{"id":161,"date":"2019-11-13T15:30:17","date_gmt":"2019-11-13T14:30:17","guid":{"rendered":"http:\/\/rwajman.iis.p.lodz.pl\/wordpress\/?page_id=161"},"modified":"2019-11-18T12:28:18","modified_gmt":"2019-11-18T11:28:18","slug":"scientific-profile","status":"publish","type":"page","link":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/scientific-profile\/","title":{"rendered":"Scientific profile"},"content":{"rendered":"\n<p>Updated: 1.09.2019<br><a href=\"http:\/\/rwajman.iis.p.lodz.pl\/wordpress\/profil-naukowy\/\"><span style=\"text-decoration: underline;\">wersja w j\u0119zyku polskim<\/span><\/a><\/p>\n\n\n\n<p>Among the area of my research, introduced in a synthetic way in the Summary of Professional Accomplishments (SPA), as the most important and the main research current I considered the achievements in the domain of the computer methods for&nbsp;non-invasive three-dimensional tomographic diagnostic and fuzzy control dedicated to the two-phase flow processes.<\/p>\n\n\n\n<p>Hence, in a frame of the achievement &nbsp;the mono-thematic series of research papers which consists of eight articles. They have been listed in table 1 in chronological order <\/p>\n\n\n\n<p><em><strong>Table 1: <\/strong>List of the mono-thematic series of papers according to the article 16 paragraph 2 of the Law on&nbsp;Academic Degrees and Title and Degrees and Title in the Arts <\/em><\/p>\n\n\n\n<table class=\"wp-block-table\"><tbody>\n<tr><td width=\"55px\">\n  <strong>Ref number<\/strong>\n  <\/td><td>\n  <strong>Article<\/strong>\n  <\/td><\/tr><tr><td>\n  I.B.1.\n  <\/td><td>   <strong>R. Wajman,<\/strong> R. Banasiak, \u0141. Mazurkiewicz, T.   Dyakowski, D. Sankowski \u201e<em>Spatial   imaging with 3D capacitance measurements<\/em>\u201d; Meas. Sci. Technol. 17 (July   2006) pp. 2113-2118   &nbsp; <br>DOI: 10.1088\/0957-0233\/17\/8\/009   <\/td><\/tr><tr><td>\n  I.B.2.\n  <\/td><td>   <strong>R. Wajman<\/strong>, R. Banasiak, \u0141. Mazurkiewicz, D.   Sankowski \u201e<em>Reply to comments on   &#8216;Spatial imaging with 3D capacitance measurements&#8217;<\/em>\u201d; Meas. Sci. Technol.   18 No 11 (November 2007) pp. 3668-3670   &nbsp;    <br>DOI: 10.1088\/0957-0233\/18\/11\/N02   <\/td><\/tr><tr><td>\n  I.B.3.\n  <\/td><td>   M. R. Rz\u0105sa, <strong>R. Wajman<\/strong> \u201e<em>Dob\u00f3r metody   wyznaczania mapy czu\u0142o\u015bci dla tomografu pojemno\u015bciowego o zwi\u0119kszonej   czu\u0142o\u015bci przy \u015bciance<\/em>\u201d ; Automatyka 13\/3 (2009), AGH, Krak\u00f3w, 1361-1368   &nbsp;   <\/td><\/tr><tr><td>\n  I.B.4.\n  <\/td><td>   <strong>R. Wajman<\/strong>, P. Fiderek, H. Fidos, T.   Jaworski, J. Nowakowski, D. Sankowski and R. Banasiak \u201c<em>Metrological evaluation of a 3D electrical capacitance tomography   measurement system for two-phase flow fraction determination<\/em>\u201d; Meas. Sci.   Technol., (2013) Vol. 24 No. 065302   &nbsp;   <br>DOI: 10.1088\/0957-0233\/24\/6\/065302   <\/td><\/tr><tr><td><br>\n  \n  I.B.5.\n  <\/td><td>   R. Banasiak, R. <strong>Wajman<\/strong>, T.   Jaworski, P. Fiderek, H. Fidos, J. Nowakowski, D. Sankowski \u201e<em>Study on two-phase flow regime visualisation and identification using 3D   electrical capacitance tomography and fuzzy-logic classification<\/em>\u201d   International Journal of Multiphase Flow, Vol. 58, (January 2014), pp. 1-14   &nbsp;   <br>DOI:   10.1016\/j.ijmultiphaseflow.2013.07.003   <\/td><\/tr><tr><td>\n  I.B.6.\n  <\/td><td>   <strong>R. Wajman<\/strong>, R. Banasiak &#8220;<em>Tunnel-based method of sensitivity matrix   calculation for 3D-ECT imaging<\/em>&#8220;, Sensor Review, Vol. 34 Iss: 3,   (2014), pp.273 \u2013 283   &nbsp;   <br>DOI: 10.1108\/SR-06-2013-692   <\/td><\/tr><tr><td>\n  I.B.7.\n  <\/td><td>   P. Fiderek, J. Kucharski, <strong>R.   Wajman<\/strong> \u201c<em>Fuzzy inference for   two-phase gas-liquid flow type evaluation based on raw 3D ECT measurement   data<\/em>\u201d, Flow Measurement and Instrumentation, Vol. 54, April 2017,   pp.88\u201396   &nbsp;   <br>DOI:   10.1016\/j.flowmeasinst.2016.12.010   <\/td><\/tr><tr><td>\n  I.B.8.\n  <\/td><td>   P. Fiderek, T. Jaworski, R. Banasiak, J. Nowakowski, J. Kucharski, <strong>R. Wajman<\/strong> \u201cIntelligent system for the   two-phase flows diagnosis and control on the   basis of raw 3D ECT data\u201d, IAPGOS, Vol. 7, No. 1, 2017, pp. 17-23,   ISSN 2083-0157   &nbsp;   <br>DOI: 10.5604\/01.3001.0010.4576    <\/td><\/tr><\/tbody><\/table>\n\n\n\n<hr class=\"wp-block-separator\"\/>\n\n\n\n<p>It is worth to note that my research achievements are the outcomes of my active participation in the international project titled&nbsp; \u201e<em>Development of Excellence in Non-Invasive Diagnostic System for Industrials and Scientific Applications<\/em>\u201d (acronym DENIDIA) (2006-2010). This project was hosted by the Institute of Applied Computer Science (IIS) of Lodz University of Technology (LUT) and was in a frame of the 6. Frame Program Mobility &#8211; Marie Curie Host Fellowships for the Transfer of Knowledge. Additionally, I was the leader of one research project in Computer Science (panel ST6 of MNiSW) and the main contractor in four other projects funded by the National Science Centre. Various publications are the&nbsp;consequence of my research works done in&nbsp;a&nbsp;frame of the mentioned projects. After PhD graduation, I am a co-author of:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>10\narticles published in journals indexed in the JCR database and in the list A\nof&nbsp;MNiSW,<\/li><li>4\nchapters in English-language books,<\/li><li>1\nEuropean patent,<\/li><li>24\narticles published in journals indexed in the list B of MNiSW,<\/li><li>7\nprojects which were awarded or honoured\n16 times in the international inventions exhibitions.<\/li><\/ul>\n\n\n\n<p>Each time I worked (once as a leader) together with some interdisciplinary research teams together with the scientists of different research areas like computer science, process engineering, automatic, mechanic as well as metrology. Such way of activity attracts the co\u2011authorship of research publications. Nevertheless, this SPA document explains only my research achievements which state about my contribution to the applications of computer science into the process tomography.<\/p>\n\n\n\n<h4 class=\"wp-block-heading\"><a>Discussion of the research goals and the obtained results along with the discussion of their utilitarian aspects<\/a><\/h4>\n\n\n\n<h5 class=\"wp-block-heading\">Preliminary remarks<\/h5>\n\n\n\n<p>The\nmain research area of my interest dedicated to\nthe industrial applications are <strong>the&nbsp;computer methods for non-invasive diagnosis and\nregulation of two-phase gas-liquid flow processes<\/strong>. The main research\nachievements and the obtained results are discussed here in the form of thematically related articles\u2019\nset submitted for evaluation. <\/p>\n\n\n\n<p>The\ntwo-phase flow (TPF) processes belong to the most rapidly growing trend in\nfluid mechanics research. In recent years there has been a significant, but\nstill insufficient progress in the development of knowledge about these\nindustrial processes (Fang\nDong et al., 2012). TPFs arouse growing interest\nbecause of their great practical significance. They are closely related to the\nrapidly developing field of research in bioprocess engineering, biotechnology,\nenvironmental engineering, energy, and many other related branches.<\/p>\n\n\n\n<p>One\nof the fundamental problems, where knowledge of which is necessary to describe\nthe hydrodynamics of TPF mixtures, include: determination of the mixtures of TPF\u2019s\npatterns, determination of the void fraction in the flowing mixture and the\nflow resistance of mixtures (Cheng\net al., 2008; Dziubinski et al., 2004; Ruspini et al., 2014). Besides,\nfor purposes of description of mass transfer in TPF systems we need to&nbsp;know the interfacial surface area,\ncoalescence of gas bubbles and mass transfer coefficient. So far, none of these\nissues has been satisfactorily presented in the literature. This is due to&nbsp;the complicated mechanism\nof flow dynamics, often connected with difficulties in its description from the\nmathematical point of view and it is related to the complicated measurement\nmethods typically used in the TPFs studies.<\/p>\n\n\n\n<p>The\ntwo-phase gas-liquid flows are a very\nimportant component of many industrial processes (Mokhatab,\n2008). One of the numerous examples is the\naeration processes (Jothiprakash\net al., 2015; Teng et al., 2016) in chemical reactors (Kiambi\net al., 2011), in flotation processes (Vadlakonda\nand Mangadoddy, 2017), in water and sewage aeration\nsystems (Guo\net al., 2013). The main task of aeration systems\nis a production of a proper fraction of\naerated liquid and oxygen. Oxygen\ninjection process is important due to the fact\nthat liquid circulation has to be achieved.\nIt helps to intensify a mass transfer process. One of the fundamental problems\nis a proper evaluation of the inter-phase surface \u2013 this is an important\nparameter from the mass transfer point of view. The example of water and sewage\naeration may be biological sewage treatment plants (Beux\net al., 2007). An\nimportant role in these processes plays\naerobic bacteria, which grow only in the presence of free oxygen from the\natmosphere or dissolved in water. The level of aeration must be within the specified\nrange, which depends on water temperature. One way of aeration is to inject\ninto the system, compressed air, making the need for freedom of gas bubbles\nmovement in the liquid column up.<\/p>\n\n\n\n<p>The\nTPF processes also occur in the bubble columns (Abdulmouti,\n2015; Delnoij et al., 2000; Mewes and Wiemann, 2003). Their purpose is to implement the\nvarious physical and chemical processes. Controlling the size of the interface\noften determines the progress intensity of these processes. For example, in\nair-lift columns (Akita\net al., 1988; Bla\u017eej et al., 2004; Kassab et al., 2009) and ejectors (Balamurugan\net al., 2007) the movement of the liquid stream\nis forced by the gas stream. Such devices are\ncommonly used in the extractive industries (e.g. flotation processes) (Zargaran\net al., 2016) or to precipitation of some\nfraction of the liquid in the sedimentation processes (Malijevsky\nand Archer, 2013), such as decreasing where the size of bubbles is significant.<\/p>\n\n\n\n<p>There\nis also a separate group of industrial processes in which gas bubbles may be formed in the liquid as a result of chemical\nreactions. It can occur for example in chemical reactors or in the process of\nelectrolysis (Lafmejani\net al., 2017; Olesen et al., 2016), where the gas phase is a product\n(often kind of a by-product) of a chemical reaction. Then the appearance\nof&nbsp;bubbles indicates the quality of ongoing changes, and the measurement\nof bubble size provides information about the process. There are industrial\nprocesses as well in which the occurrence of bubbles is undesirable, such as\nheat exchangers or heating devices. In these systems,\nthe appearance of unwanted gas gives evidence of boiling liquid phenomenon (Solotych\net al., 2016), and enforces signalling state of\nemergency. This occurs similarly to\ncavitation phenomena in rotational pumps (Liu\net al., 2015; Zhu et al., 2015), which are caused by a rapid\ndecrease in pressure below the pressure of the boiling liquid. This phenomenon\nis undesirable because it leads to erosion of the pump blades. This phenomenon\ncould also be the evidence of system leakage.<\/p>\n\n\n\n<p>The\nmentioned above the <strong>growing needs of industry for simple,\nversatile, relatively inexpensive, non-invasive and rapid method of process\ndiagnosis and control<\/strong> for TPFs in horizontal and vertical pipelines justify\nthe importance of my research topic. The knowledge of the characteristics and of the gas-liquid flow type is very important\nfor the design and implementation of industrial-scale research facilities as\nwell as for the process of numerical\nmodelling. The continuous monitoring and diagnosis of any abnormalities can\nprovide valuable information about their\ndynamic state and allow for continuous and automatic control. <\/p>\n\n\n\n<p>Nowadays,\nthe applicability of similar technologies in modern production systems is the\nmain trend of development and technological progress in many industrial sectors\ndetermining thereby the energy-saving and quality enhancing trends. Wherever in\nthe production process the phase mixture\nis transported and it is not optimal or\nnot economical, there is a need\nto&nbsp;develop a system which would be able to prevent crashes, unexpected\nproduction line hold\u2011ups or situations where for reasons of bad flow\nparameters, the final product is defective. Such solution could also be\nirreplaceable when a flow process requires constant supervision, or when the\nwork environment would be a danger to the safety or cause loss of either\nemployees&#8217; health or life and simultaneously it is required continuous, automated, non-invasive and\nefficient monitoring of inaccessible parts of pipelines. <\/p>\n\n\n\n<h5 class=\"wp-block-heading\">Computer methods for\nthree-dimensional tomographic diagnosis<\/h5>\n\n\n\n<p>The\nmain aspect of many identification and control systems is a diagnosis of\nindustrial processes. Since many years the research conducted on the two-phase\ngas-liquid flows still does not deliver\nconsistent answers to many questions according to this phenomenon (Fang\nDong et al., 2012). It is because of its stochastic\nnature as well as its dynamics but also\nit is related to&nbsp;the research capabilities. In the case of these flow processes,\nthe diagnostic methods developed so far are\nbased on usage of most sophisticated measurement techniques (Brebbia\nand Mammol, 2011). These methods either did not allow\nto obtain repeatable results (in most cases the description of the phenomena\nwas ambiguous) (Abbagoni\nand Yeung, 2016; Arvoh et al., 2012; Bertola, 2003; Ozbayoglu and Ozbayoglu,\n2009) or was actually precise enough but\ninterfered with changing of the process features (Venkata\nand Roy, 2012; Xie et al., 2004).<\/p>\n\n\n\n<p>Nevertheless,\none of the computer measurement non-invasive methods for the dynamic processes\ndiagnosis is the Electrical Capacitance Tomography (ECT). This technique delivers\nthe two- and three-dimensional imaging in\nthe basis of dielectric features of the\nprocess\u2019 components, e.g. flow. The first\nECT systems (Pl\u0105skowski\nA., Beck M.S., Thorn R., 1995; Reinecke and Mewes, 1996) allowed to obtain only the rough\nevaluation of the process state because the information encoded in measurement\ndata represents merely the fragment of the process which additionally was approximated into the cross-section surface\nthrough the sensor (2D ECT) (Isaksen,\n1996; Yang and Peng, 2003). Unfortunately, this kind of\nimaging occurred to be insufficient from the process control point of view. The\ncross-sectional image does not reflect enough the spatial phase distribution or\nflow structure in a measurement volume. It is because the image is generated according to the approximated\nmeasurement values from the whole electrodes surfaces. In the case of long electrodes,\nthe high level of approximation prevents the precise measurements. Moreover, the spatial electrostatic field\ndistribution is neglected. Therefore, the\nECT diagnosis was developed in the direction of&nbsp;other data processing\nmethods like cross-correlation (Mosorov\net al., 2002), image processing and analysis,\nmulti-layered tomography (Gadd\net al., 1992; Holder, 2004; Metherall et al., 1996; Wang et al., 2003) etc. Nevertheless, independently of\nthe mentioned extensions still the classic 2D ECT measurement suffered from the\nlimitations of the cross-sectional approximation. Many of the industrial\nprocesses are characterized by the\nspatial features and their reduction to the planar solution results in\nundesirable simplifications.<\/p>\n\n\n\n<p>The\nliterature studies, I have conducted concerning existing research publications\nin the field of electrical capacitance tomography in 2006, allowed me to\nidentify some unsolved problems and\nmotivated me to the further exploration of this issue. Most of the research\nworks based on the ECT technique have implemented the image reconstruction\nmethods which suffer from the poor final\nquality. In the images (tomograms) it was impossible to distinguish the clear\nborders between the process components (phases) (Polydorides\nand Lionheart, 2002; Warsito and Fan, 2005, 2003; York et al., 2003). Nonetheless, in many industrial\napplications supported by the ECT diagnosis beside\nthe accurate measurement devices a key\nrole plays the data processing and image\nreconstruction methods. From these methods,\nit is expected the high image quality together with the short processing time. These&nbsp;challenges\nconcern the essential computer issues like the image resolution enhancement\nsaving quality and processing time what definitely\nresults in the&nbsp;computation complexity, computer modelling of measurement\nsensors, numerical errors and finally in the necessity of development of new\nalgorithms for efficient management\nof&nbsp;resources as well as computation power.<\/p>\n\n\n\n<p>The\nissues mentioned above caused that my research area was mainly focused on the <strong>development\nof new computer methods for\nthree-dimensional tomographic data processing and visualisation dedicated to non-invasive diagnosis and regulation\nof&nbsp;industrial flow processes<\/strong>. The results of conducted research, as\nwell as the developed methods, help me to contribute the visualisation quality,\nfastness and the diagnosis precision enhancement to the computer science\ndiscipline and the ECT domain. The scope of my research covered the\ndevelopment, implementation and verification of:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>raw\ntomographic measurement data processing algorithms in the context of TPF diagnosis;<\/li><li>computer\nmethods for spatial ECT sensor modelling and designing;<\/li><li>fuzzy\ninference algorithms for the TPFs type identification and regulation;<\/li><li>software deploying the developed algorithms and methods for the purpose of real flow processes monitoring and regulation.<\/li><\/ul>\n\n\n\n<p>After\nobtaining my PhD degree, the main streams\nof my research I have conducted at the Institute of Applied Computer Science (IIS)\nat LUT have been as follows:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>algorithms\nfor computer modelling and sensitivity analysis of three-dimensional ECT\nsensors for tomographic visualisation;<\/li><li>computer methods for supporting the 3D ECT sensor designing\nprocess;<\/li><li>identification\nand regulation of TPF type in the basis of\nfuzzy inference and tomographic diagnosis.<\/li><\/ul>\n\n\n\n<figure class=\"wp-block-image is-resized\"><img loading=\"lazy\" decoding=\"async\" src=\"http:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-content\/uploads\/2019\/11\/Schemat-EN-1024x758.png\" alt=\"\" class=\"wp-image-169\" width=\"768\" height=\"480\"\/><figcaption> <strong>Figure 1<\/strong>. Author\u2019s algorithms (after PhD degree) for tomographic diagnosis and regulation of industrial two-phase gas-liquid flows in the context of their usefulness  within the most important research projects <\/figcaption><\/figure>\n\n\n\n<p>In\nfigure 1 the algorithms, I designed after PhD degree, for the tomographic\ndiagnosis and regulation of industrial two-phase gas-liquid flows were summarised in context of their usefulness within the most important research\nprojects I take part in. The algorithms\ntogether with the references to the articles from the cycle were grouped regarding the main research streams. <\/p>\n\n\n\n<p>Next,\nthe discussion on the research results of my algorithms and computer methods is provided and it is divided into sections according to the main\nresearch streams listed above.<\/p>\n\n\n\n<h5 class=\"wp-block-heading\">Algorithms for computer\nmodelling and sensitivity analysis of three-dimensional ECT sensors for\ntomographic visualisation<\/h5>\n\n\n\n<p>In\n2006 after obtaining my PhD degree I continued the research on computer methods\nfor ECT sensor modelling. I finished then the project concerning the numerical\nimplementation for 16 electrodes 3D ECT system together with the complete\nsoftware for 3D ECT sensor simulation, 3D image reconstruction and\nvisualisation. The obtained results I published in July 2006 in [Att. 3 pos. I.B.1]. <\/p>\n\n\n\n<p>My\nfirst research task was to develop and verify <strong>the algorithms for the construction\nof&nbsp;the 3D&nbsp;ECT sensor computer model and forward problem solution<\/strong>. This stage is responsible for the calculation of the electric field distribution\npreconditioned by the initial permittivity distribution inside the sensor and\nnext for the inter-electrodes\ncapacitances values calculation. Next, I developed two algorithms. The first one determines <strong>the spatial sensitivity matrices determination for 3D ECT sensor<\/strong> in\nthe basis of obtained electric potentials values in the grid nodes (voxels).\nThe <strong>algorithm for calculation of the\ninverse problem for three-dimensional\nimage reconstruction<\/strong>, in turn, I developed in\nthe basis of crucial modifications\nof&nbsp;2D iterative algorithms. All mentioned algorithms I verified\nnumerically and together with the results I published in [Att. 3 pos. I.B.1] and implemented using WINAPI C++\nplatform. The built software called \u201c<em>WinRECO<\/em>\u201d\nprovided the tool for complete 3D ECT image reconstruction and was shared in the Process Tomography Laboratory\nat Institute of Applied Computer Science (IIS) for next research purposes where\nthe 3D ECT system was used as\na&nbsp;diagnostic module. More details concerning\nthe practical application of my algorithms may be found in one of the last sub-chapter titled \u201c<em>Usefulness of the conducted research<\/em>\u201d.\nThe&nbsp;diagnostic knowledge obtained thanks to the phase distribution visualisation\nin the form of 3D images had expanded the\npossibilities of applying the ECT systems in industrial processes.<\/p>\n\n\n\n<p>Furthermore,\nin a frame of my early achievements in the ECT domain I developed the&nbsp;specialised sensitivity model for\ntwo-dimensional ECT sensor. In 2007 as the main\nexecutor I contributed to the research project MNiSW no&nbsp;PB\u20111318\/T10\/2005\/28 titled \u201e<em>Development of&nbsp;the measurement method\nfor flow structures identification<\/em>\u201d supervised by Mariusz R.&nbsp;Rz\u0105sa\nPhD, DSc from Opole University of Technology. My task within this project was\nto&nbsp;develop a new image reconstruction method for the dedicated ECT system\nto enhance the&nbsp;spatial resolution and the detection abilities in the\nneighbourhood of pipe walls. The&nbsp;proposed solution assumed the new\nstructure of the capacitance sensor which allowed the&nbsp;measurement of flow\ntypes characterised with the homogenous\nphase distribution along the pipeline and very small objects like bubbles or\ndrops, e.g. churn, annular, slugs or\nplugs flows. The sensor allowed the diagnosis of the thin liquid layer in the\narea next to the pipe wall. The main characteristic feature of the sensor has\nbeen the <em>extra<\/em> electrodes with the\nwidth many times smaller than the measurement electrodes mounted between them.\nDuring the measurement, the <em>extra<\/em> electrodes adjacent to the\nmeasurement electrode have assigned the electric potential the same as the\nopposite electrode which participates in the measurement. As a result (Fig.\n2b in [Att. 3 pos.&nbsp;I.B.3]) the measured capacitance value has been equal to\nthe sum of three component capacitances. Because\nthe <em>extra <\/em>and measurement electrodes have\nbeen next to each other, then considering the\nelectric field distribution the highest\nsensitivity of the set has been in the area located near the pipe wall.\nAdjusting the width of the <em>extra<\/em>\nelectrodes it has been possible to achieve the optimal attitude between the\nsensitivity in the sensor centre and in the neighbourhood of the wall. Next, having\nthe designed sensor\u2019s structure, I\ndeveloped and implemented the <strong>image\nreconstruction method that in forward as well as in inverse problem considers\nthe new sensor geometry<\/strong>. As a first step, I\ndeveloped <strong>the algorithm for the sensor\narea discretisation<\/strong>. It considers the new sensor\u2019s elements and also\nadjusts the grid density increasing the numerical accuracy only in the area next\nto the electrodes but simultaneously decreasing the influence of\nunder-determining aspect of image reconstruction process (i.e. significantly\nlarger number of image points (variables) than the number of measurement data).\nAfter that, I developed <strong>the algorithms for numerical calculations\nof the electric field potentials distribution and sensitivity matrices for the\nsensor with the greater detection abilities in the\nneighbourhood of pipe walls<\/strong>. In this case, the algorithms consider not only the\nsensor geometry but also the <em>extra <\/em>electrodes\ninfluence on the measurement. Therefore,\nfor each measurement pair, the algorithm assigns\nthe electric potential value to two associated <em>extra <\/em>electrodes. This feature limited the total number of\nmeasurement electrodes in the design tomographic device excluding thereby the\npossibility of spatial (3D) measurement. Nevertheless, as it was shown in [Att. 3 pos.&nbsp;I.B.3], the developed measurement system occurred to have the\nstrengthened abilities to detect the structures of&nbsp;counterflow and thereby\nto give a more accurate tool for the\ndescription of this dynamic phenomena especially in the area of the pipeline\nwall.<strong><\/strong><\/p>\n\n\n\n<p>One\nof the stages of the image reconstruction process for the 3D ECT is the\nsensitivity matrix determination. This matrix is the most important factor\ndeciding about the final image quality. With the issue of sensitivity matrix for\nECT I am conversant since my PhD\nresearch. Then, but for the 2D case, I developed the innovative algorithm for\nsensitivity matrix calculation along the electrical field lines and image\nreconstruction method which were the basis of the research described in the\narticle (Loser, Wajman,\nMewes, 2001). The work still enjoys the wide interest from the\ninternational research teams what may be proved by the total citations count\nequals to 69 (from Scopus base excluding self-citations) and within the last 4\nyears 13 times. The results of the research works I performed during my\nscientific internship in&nbsp;Hannover (Germany) I considered in my PhD thesis.<\/p>\n\n\n\n<p>In work\n[Att. 3 pos. I.B.1], in turn, my contribution was to\ndevelop the method of&nbsp;the&nbsp;sensitivity matrix determination for the\n3D&nbsp;ECT case. This method I developed and implemented mainly on the basis of energy field dependencies. It\ndelivers, in most cases, approvable reconstruction results and also gives the possibility to iterate the\nprocess considering in each step the sensitivity matrix update adjusting it to\nthe changing electrical field. However, in 2014 taking into account the still\ngrooving quality diagnostic demands I designed the substantial modification of\nthe method considering the electrical field distribution inside the sensor. <strong>Developed and implemented the tunnel-based\nalgorithm <\/strong>determines the sensitivity matrix in each step of the image reconstruction\nprocess along the electrical field lines using the mathematical and physical dependences\nderived from the electrical field energy analysis. In [Att. 3 pos. I.B.6] it was shown that the new sensitivity maps more\nprecisely reflect the non-linearity of the electrical field phenomena.\nAccording to the characteristic geometrical shape of the resulting maps, the method is called as tunnel-based. It assumes\nin fact that most of the deterministic image reconstruction methods, which use\nthe sensitivity matrix, are derived from the linear tomography represented by\nthe X-ray, gamma-ray or optical tomography. The applicability of&nbsp;the\nlinear image reconstruction methods for\nthe electrical tomography is conditioned by a suitable sensitivity matrix which\nthe main task is to approximate the phenomena of the electrical field inside\nthe ECT sensor. In the linear tomography, the sensitivity analysis is based on the calculation of the ray\nprojection absorption factors in each point of the mesh. The electrical field\nlines, likewise, can be considered as\nrays projected in linear tomography. Following these lines in the space of 3D ECT sensor, the sensitivity tunnels can be constructed which reflect\nthe voxel participation in the projection (i.e.&nbsp;electrical field lines are\nconsidering the permittivity [material density] distribution).<\/p>\n\n\n\n<p>Furthermore, in work [Att. 3 pos. I.B.6] the results of my experiments may be found. There are both simulations as well as\nexperiments conducted on the TPF facility in Tom Dyakowski Process Tomography\nResearch Laboratory in IIS at LUT. To&nbsp;evaluate\nthe measurement resolution and image reconstruction quality of 3D ECT sensors, I developed the complex numerical method.\nIn a&nbsp;frame of it, I implemented the <strong>algorithm for metrological analysis\nof&nbsp;sensitivity model<\/strong>. For the\nimage error estimation, I applied the\nfollowing criteria: the&nbsp;normalised\nmean square error, the Pearson\u2019s linear correlation coefficient. In each\nexperiment, the achieved results proved the adequateness of assumptive solution\nin comparison to the traditional methods (based on the electrical field\nenergy). The improvement in the convergence for the tunnel-based sensitivity\nmatrix approach is noticeable. This feature is\nespecially emphasised while increasing the belligerence (fastness) of\nthe reconstruction process by increasing the relaxation factor. It is worth to\nmention that for the unduly high value of this factor the traditional method\nultimately lost his convergence and the image errors kept rising in each step. The\nusage of the tunnel-based method is fruitful even in this case mapping the TPF\nstructures on the tomograms in a correct way. Furthermore,\nthis algorithm was successfully applied in a non-linear image reconstruction\nmethod. It was implemented in the diagnostic module for the two-phase\ngas-liquid flow identification system developed in frame of the research\nproject MNiSW no 4664\/B\/T02\/2010\/38 titled \u201c<em>Application\nof three-dimensional electrical capacitance tomography for phase distribution\nand structure identification in gas-liquid flows in horizontal and vertical\npipelines<\/em>\u201d supervised by Robert Banasiak PhD, DSc in the years 2010-2012.<\/p>\n\n\n\n<h5 class=\"wp-block-heading\">The computer methods\nsupporting 3D ECT sensor designing process<\/h5>\n\n\n\n<p>The\ntask of design and fabrication the most efficient sensor structure from the mechanical\nand electrical point of view is relatively difficult (Zhang\net al., 2014). The electrodes layout, their\nshapes and geometrical dimensions, horizontal and vertical gaps, the width of\nthe inner-electrodes rings, boundary screens and finally the thickness of the\nisolation need to be determined according\nto the measuring sensitivity abilities of the measurement hardware unit and to\ndielectric properties of the diagnosed medium. Sensitivity\nof the ECT system requires the measured values and their changes to be above a\ncertain minimum level. It enforces the use of electrodes with a sufficiently\nlarge surface area. In the case of ECT\nimaging technique measured capacitance range can varying from tens to hundreds of picofarads and a difference of capacitance caused by a change of dielectric\npermittivity distribution can be a&nbsp;range of femtofarads. Taking the above\ninto consideration (whereas, e.g. the\nhomogeneous sensitivity in the whole sensor volume is expected) the most\nimportant is to develop the&nbsp;mechanism for supporting the time-consuming\ntask of sensor design under given conditions of&nbsp;its deployment for\npurposes of the industrial process diagnosis. These sensor\u2019s parameters to be\ndetermined include a surface area of the\nelectrodes and their shape, the inter-electrode horizontal and vertical\nspacing, width of the inter-electrode and boundary screen, also the thickness\nof protective insulation. Moreover, the\ndevelopment of analogous sensors\u2019 concepts\non pipes with different profile\u2019s diameters and additionally to study liquids\nwith different dielectric permittivity values is not a matter of simple\ngeometry &#8220;rescaling&#8221;. The electric field is known to be highly\nnon-linear, and the choice of&nbsp;ECT sensor geometry parameters must be also optimised\nexperimentally by using ultra-precise LCR meter. During these experimental work, the inner space of the test sensor should be filled by the water (a\nmedium of high dielectric constant) and the air (the medium of low dielectric\nconstant). Then a static capacitances measurement by the RLC meter should be\nperformed. The analysis of experimentally collected capacitance data allows for\nadjusting optimal key parameters of sensor design to an ECT tomography sensor\nmeasuring range.<\/p>\n\n\n\n<p>One\nof my first research issue regarding the 3D ECT sensor design was the dead\nzones identification. In 2007 (Warsito\net al., 2007) in a context of the article [Att. 3\npos. I.B.1] and the polemic with the research group from Ohio State University\n(Prof. Liang-Shih Fan) I&nbsp;developed the <strong>algorithm for the analysis of the 3D ECT sensor dead zones in the inter\u2011electrodes areas<\/strong>. The results were published in [Att. 3 pos.&nbsp;I.B.2]. It was pointed out that in published sensor\nstructure the dead zones were significantly reduced and also, this sensor was characterised\nby the increased measurement resolution in comparison to the sensor offered by\nthe Prof. Fan\u2019s team (Warsito\nand Fan, 2003). The conclusions published together\nwith the results of the experiments let to explain the doubts finally.<\/p>\n\n\n\n<p>The\nmethods developed in a frame of the mentioned issue allows me to develop <strong>the<\/strong>&nbsp;<strong>computer method for geometrical and mechanical 3D ECT sensors\u2019\nproperties determination<\/strong> in the context\nof various applications. As part of this method,\nI developed a series of algorithms and implemented them in the <em>WinRECO<\/em> software, i.e. algorithms for 3D ECT model mesh\ngeneration, the algorithm for electrical\nfield simulation, algorithms for metrological and sensitivity analysis of the\nsensor\u2019s dead zones. Moreover, the method was\nused within the experiments carried out for two articles [Att. 3 pos.\nII.A.2] and&nbsp;[Att. 3 pos. II.A.3] purposes. I&nbsp;was there responsible\nfor the ECT sensor design and fabrication as well as the ECT experimental setup\nimplementation. Each time the application conditions were different therefore\nit was a need to evolve various sensors\u2019 conceptions.<\/p>\n\n\n\n<p>The\nresearch task described in [Att. 3 pos. II.A.2] aimed to evaluate the\napplicability level of the ECT systems to investigate the defects inside the\ndielectric materials. The mentioned 3D ECT sensor structure consisted of four\nelectrodes\u2019 rings. The electrodes in the following ring were shifted to these in the previous ring about 5<sup>o<\/sup>.\nMoreover, the central rings contained four electrodes more. Such approach brought\nabout alignment in the homogeneity of the\neffective imaging area within the whole\nsensor volume and simultaneously allowed to reduce the number of electrical\nfield intensity overloads next to the external edges of the sensor.<\/p>\n\n\n\n<p>In\na frame of the research described in [Att. 3 pos. II.A.3], which deals with the\ndevelopment of the image reconstruction algorithms for 3D ECT, the direct\ntemporal image reconstruction method was applied\nas an alternative to the common 4D image reconstruction in the basis of the images sequence (the fourth\ndimension is the experiment time). The sensor construction contained the two\ncentral rings narrower than the external ones. It was because using my previous\nsensors the measured capacitances between electrodes from the central rings\nwere significantly greater than others. For the dynamic processes, which are\nadditionally indicated by the temporary and simultaneously huge disparities\nbetween the permittivity values for given fractions, it could be observed that the capacitances values within\nthe measurement set frequently exceeded the level determined for the sensor\ntotally filled with the fraction of the maximal permittivity value. This phenomenon causes that the applied ECT system produced\nthe measurements values greater than the converter range distorting thereby the results of the experiments. Finally, the\ndeveloped sensor structure under my concept allowed to collect the balanced measurement\nvalues set even for dynamic processes.<\/p>\n\n\n\n<p>The\nindustrial process, which I diagnosed, were characterised\nby the minor discrepancy in electrical\npermittivity between used materials. Together with air I applied the plastic\ngranules (e.g. pcv, erthalon) as well as the rise or chipping (sand) [Att. 3 pos. II.L.9]\nand&nbsp;[Att. 3 pos.&nbsp;II.L.11]. However, quite distinct conditions I met\nconstructing the system for the mentioned\nbefore research project 4664\/B\/T02\/2010\/38 where the two-phase gas-liquid\nmixtures were diagnosed. In this case, as a liquid,\nI use polypropylene glycol, carboxymethylcellulose as well as water which are\nindicated by significantly higher (even 80 times) relative electrical\npermittivity value in an attitude to air. Within many simulations I&nbsp;perform\nfor the sensor I noticed the negligible\npenetration of the electric field in its centre especially in the case when the sensor volume with electrodes mounted\nexternally on&nbsp;a&nbsp;pipe wall was\ncompletely filled with the liquid. Moreover, the analysis of the\nequipotential lines indicated that the electric field was unduly concentrated\nin the pipe\u2019s wall made from PMMA with the relative permittivity significantly\nlower than the relative permittivity value of the used liquid. This phenomenon (which can be seen in figure 5a in\n[Att.5 pos. I.B.4]) did not occur in&nbsp;the case of the sensor with internal\nelectrodes I&nbsp;developed for [Att. 3 pos. I.B.5] and [Att. 3 pos. II.E.16]. The\nexperiments based on the sensitivity analysis proved, that the new construction\nfor the same liquids allowed to achieve the effect of deeper penetration of the\nelectric field throughout the process. In a frame of this task, I developed and implemented <strong>the algorithm for computer determination of\nthe geometrical sensor parameters to reduce the inhomogeneity in the\nsensitivity distribution<\/strong> which effectively allowed to adjust the proper\nsensor geometry together with the spatial electrodes layout under the process\nconditions (extreme big relative electrical permittivity values). In the basis\nof the method, I was able to design a suitable sensor. The experiments results published in [Att. 3 pos. I.B.5] proved\nthe better applicability of the sensor for liquid flow processes.<\/p>\n\n\n\n<p>In\nthe IIS for the 3D ECT research purposes,\nthe TomoKIS team (i.e. the group of researchers in the Institute dealing with\nthe tomography diagnosis) uses the ET3 tomography device fabricated by the team\nof Dr Roman Szabatin from the Institute of\nRadioelectonics at Warsaw University of\nTechnology (Brzeski\net al., 2003). From the industry point of view, the most desirable feature of this device\nis the possibility of freely adjusting the gains of the measurements circuits\nfor each measurement electrodes pair separately. Therefore, this device is\nuniversal of use and can be applied for\ndiagnosis of various processes. In case of&nbsp;32\u2011electrodes sensor, there are 496 measurement chains\ncontrolled by four parameters which prejudging about the stability as well as\nthe amplitude of the signal what finally determines the SNR (<em>signal-to-noise ratio<\/em>) of the\nmeasurements. There are two amplifiers (4&nbsp;values of setting for each) and\nfeedback resistors and capacitors (it gives together 12 values of settings).\nSuch diversity of settings caused that the manual tuning of the device is very\ntime-consuming. Simultaneously, there is very important to tune the system as precisely\nas possible remembering that despite the big and different distances between\nelectrodes and the different sensor geometry the small measurement capacitances\nshould oscillate in the measurement range of the device and additionally should\nbe sensitive enough to the changes of the permittivity distribution between\nfull and empty sensor. Taking the above into consideration,\nI developed and implemented <strong>the&nbsp;method\nwhich automatically sets the gains for all the measurement channels<\/strong> based\non the user conditions and calibrates the tomography device to any sensor\ngeometry [Att. 3 pos. II.L.17]. The method was\nverified by the TomoKIS team by numerous experiments adjusting the ET3\nto the different industrial processes and sensors (i.a. [Att. 3 pos. II.L.11],\n[Att. 3 pos. II.L.13] and [Att. 3 pos. II.L.14]).<\/p>\n\n\n\n<p>During\nmy multiannual research works, I had the possibility to meet various conditions\nof&nbsp;the 3D ECT application for industrial processes. Each time the most\ndesirable issue was to&nbsp;construct the special sensor structure to increase\nthe measurement resolution as well as to&nbsp;amplify its sensitivity, especially in its specified areas. To validate the parameters, which unambiguously\nmay decide about the usefulness of the sensor to the given process, in 2013 in\na&nbsp;frame of the research project (4664\/B\/T02\/2010\/38) I developed the <strong>set of&nbsp;algorithms for metrological and\nsensitivity analysis of the ECT sensors<\/strong> in context of&nbsp;gas-liquid flows\ndiagnosis [Att. 3 pos. I.B.4]. According to this task, I performed the numerical analysis of the sensitivity\ndistribution of the designed 3D ECT sensors to evaluate their measurement\nresolution. It is worth to note that the simulation effectively supported the\nprocess of optimal sensor geometry construction under the homogeneity\nmeasurement condition for flows with the liquids of high permittivity value.<\/p>\n\n\n\n<h5 class=\"wp-block-heading\">Identification and regulation\nof two-phase flow type in the basis of\nfuzzy inference and tomographic diagnosis<\/h5>\n\n\n\n<p>In\nmany industrial applications for TPFs it\nis important to provide the diagnostic signals which simultaneously allow to\nidentify the materials behaviour (qualitative) and to measure the\ncharacteristic features of the dynamic flow (quantitative) in real time. The TPFs\n(liquid\/liquid, gas\/liquid) needs and demands include efficient monitoring as well as an automatic regulation. The computer\nmeasurement systems based on&nbsp;video signals analysis and processing in\ncomparison to the classic measurement devices are featured by two main\nadvantages which allow to face up to these\ndemands. Definitely in a&nbsp;better way they can illustrate the\nphysical and chemical phenomena (in space and in time) in the industrial\nprocess what next allows to develop more efficient methods for control and\ndiagnosis.<\/p>\n\n\n\n<p>My\ncontribution to this area was to develop and verify the <strong>algorithm for non-invasive phase distribution detection in the basis of three-dimensional tomographic images<\/strong>\n[Att. 3 pos. II.E.13]. This algorithm was an innovative solution in case of regulation\nand diagnosis of&nbsp;two-phase gas-liquid mixtures flows. The concept based on\n3D flows tomograms analysis and processing like image segmentation (not yet artificial intelligence techniques) and as a&nbsp;result,\nthe percentage value of phase distribution in the sensor volume was determined. Apart from this, the developed software was constructing\nthe 3D images distinguishing the liquid fraction. Afterwards, I developed the <strong>algorithm for error analysis in the context of phase distribution evaluation <\/strong>and\npublished it in [Att. 3 pos. I.B.4]. To\nverify the results, I&nbsp;also designed the <strong>algorithm for percentage phase distribution in the basis of flow images acquired form CCD camera <\/strong>[Att.\n3 pos. II.L.15].<\/p>\n\n\n\n<p>The\nachievement within the mentioned research became an inspiration for me to\nsearch for new solutions in respect to an\nimprovement in flow diagnosis. The\ngrowing needs of industry for a versatile,\nrelatively inexpensive, non-invasive and rapid method of TPFs regulation and\ndiagnosis justify the importance of the continued research topic. In addition, the literature studies, I have\nconducted concerning existing (in 2010) research publications in this field,\nlet me state that there was a lack of the intelligent diagnosis and regulation\nmethods dedicated for the two-phase flows supported with the 3D ECT non-invasive\nmonitoring.<\/p>\n\n\n\n<p>In\nthe years 2011 \u2013 2014 <strong>I supervised<\/strong>\nthe research project SONATA no&nbsp;2011\/01\/D\/ST6\/07209 titled \u201c<em>The intelligent system for two-phase flow\ncontrol in basis of capacitance\ntomography diagnosis<\/em>\u201d founded by National Science Centre which has been positively evaluated. The main goal was to\ndevelop the intelligent diagnostic and control system for monitoring and\nautomatic control the two-phase gas-liquid mixtures flows processes in the\nhorizontal and vertical pipelines. The diagnostic signals obtained in a frame\nof&nbsp;fuzzy inference allowed me to design the methods for the spatial flow\ncharacteristic identification, flow maps determination and analysis. This, in turn, was the input signal to another algorithm for flow process regulation\nand provided the complex knowledge about the current state of this process. I\napplied the fuzzy inference and as the input, I used the raw measurement data. Thanks\nto the 3D ECT it has been possible the real-time, non-invasive diagnosis and\nmonitoring of the pipeline interior. <strong>The\nessence of my research task was to&nbsp;develop the methods based on fuzzy\nclusterisation and on spatial relations\nanalysis for&nbsp;the flow structures shapes and localisation determination directly from raw data (not images) and\nfor regulation of this flow process<\/strong>. It has been an innovative approach not encountered before in scientific resources.\nUltimately, the developed algorithms and methods were deployed to&nbsp;the two-phase\ngas-liquid mixtures flow facility of a half-industrial\nscale and closed in a&nbsp;feedback loop.<\/p>\n\n\n\n<p>The\nculmination of my project is the intelligent diagnostic and regulation system\nnamed <em>intelliFlowControl<\/em> able to the\ncomprehensive evaluation of the flow behaviour in the basis of&nbsp;3D tomography measurements and able to produce the\ncontrol signals to maintain the process under given conditions. The system\nsupports the research facility entirely\nin Tom Dyakowski Process Tomography Laboratory in the IIS at LUT. Furthermore,\nin September of 2015, I&nbsp;applied to\nthe Polish Patent Office for the protection of the invention titled \u201c<em>Fuzzy regulator for two-phase flow type\ncontrol<\/em>\u201d. The application numbered P.413804 has just been published. The added value of the system\nis its commercialisation potential. The\ninvention was doubly awarded on International Inventions Shows and presented as\nwell for wide industrial representants group (Azoty W\u0142oc\u0142awek, Orlen S.A.,\nNational Instruments, etc.) and is the\npart of the industrial offer of Lodz University of Technology (LUT).<\/p>\n\n\n\n<p>The\ncore of the <em>intelliFlowControl<\/em> system\nconsists of two algorithms:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>for\nTPF structure identification in basis on raw tomographic measurement data and\nfuzzy inference,<\/li><li>of\nfuzzy regulator for the intelligent TPFs type\nregulation closed in the feedback loop together with the identification\nalgorithms.<\/li><\/ul>\n\n\n\n<p>The\nfirst stage of <strong>the fuzzy identification\nalgorithm<\/strong> [Att. 3 pos. I.B.7] is the fuzzy clusterisation FCE (Fuzzy c-elliptotypes) responsible for a description of\nthe typical flow structures like slugs and plugs. The results of the\nclusterisation block together with the results of the raw data analysis are input\nto the main flow identification block, which uses the algorithm for the current\nflow state determination in the basis of\nboth inter-object spatial relations\u2019 map as well as the fuzzy evaluation of these relations (Zadeh,\n1999). The obtained map, which considers\nthe mutual location, size and shape of the structures, is finally applied in\nthe matching process of the patterns as a&nbsp;vector of characteristic\nfeatures of the flow. These patterns are previously determined experimentally.\nFurthermore, this map, in turn, is useful for producing other flow maps\ndependent on the pipeline and the medium. The determined flow maps are\nimportant as a pattern for control relations and current flow state.<\/p>\n\n\n\n<p>My\nconcept for the identification algorithm has been\nbased on row (not images) 3D&nbsp;tomographic measurement data obtained\nfrom the sensor located on the examined part of&nbsp;the pipeline and on this direct\ndata analysis as well as on the identification of the flow structures\u2019\nshapes. Next, the result of identification is\nused to conclude on flow type [Att. 3 pos.&nbsp;I.B.7]. The usage of raw data\nin case of flow structures recognition becomes an effective (qualitative but also\nin the meaning of processing time) modification\nof the method developed in 2012 in a frame of the research project no\n4664\/B\/T02\/2010\/38 supervised by dr&nbsp;hab. in\u017c.\nRobert Banasiak, I took part in. This method based on reconstructed three\u2011dimensional tomographic images and on sophisticated parallel GPU computing\nalgorithms. Despite acceptable results I noticed that during its work there was\na risk to discount (not to detect) some characteristic flow features like slugs\nor plugs which (highly dynamic) are essential in the TPF identification\nprocess. The significant part of diagnostic information is able to be directly obtained on the stage of raw data analysis just before the image reconstruction.\nSuch analysis decreases the risk of missing the crucial knowledge like gentle\nbubbles which become invisible after image reconstruction. The raw data\nprocessing is related to the significant increase of total measurements samples\nbecause the real-time image\nreconstruction process is time-consuming.\nTherefore, the flow type evaluation is possible in a real-time mode without the need of&nbsp;application\nof expensive computer hardware equipped with numerous efficient graphics cards and parallel GPU computing\nalgorithms. Moreover, the processing time is limited only by the measurements\ncapability of ECT system hardware (the mentioned ET3 device can provide 12 complete measurements frames per\nsecond for 32 electrode sensor). The solution I evolved simplifies the\nimplementation and finally mitigates the total system costs.<\/p>\n\n\n\n<p>The\nresearch on the fuzzy identification algorithm I conducted as a supervisor of\nthe research team I established within my project. The PhD student mgr in\u017c.\nPawe\u0142 Fiderek was one of the team members. I act as an assistant supervisor of\nhis PhD thesis (supervisor dr hab. in\u017c. Jacek Kucharski, prof. of LUT). Together\nwith PhD student, I developed the fuzzy\ninference algorithm which is based on the statistical flow features and\nneglects thereby the patterns\u2019 set and refuses the requirement of the knowledge\nbase determination. It is worth to note that the innovative shape of membership\nfunctions was introduced. The sections of\nthe trapezoids functions\u2019 sides (legs) are\nconstructed in the basis of\nexponential functions. Thanks to this the non-crisp nature of the border\nbetween any of the TPFs types is reflected more accurately and is better\ncustomizable (only by a&nbsp;single parameter) to the subjective assessment of\nthe user.<\/p>\n\n\n\n<p>The\nidentification algorithm allows the non-intrusive and non-invasive flow\nmonitoring in&nbsp;a real-time mode\nproviding the accurate flow dynamic analysis, determining the flow types and\nwarning against undesirable states. Moreover, one of the algorithm\u2019s attributes\nis its usefulness for the new flow maps determination. One of the examined\nliquid was the propylene glycol. It is a highly\nhygroscopic liquid which attracting and holding water molecules from the\nsurrounding environment changes its physical parameters. Furthermore, because\nof its low value of surface tension, glycol is a liquid vulnerable to foaming\nduring the flow what affects the instability of flow structures and the lack of\nreproducibility of flow character under the same conditions and in the short\ntime intervals. Admittedly, this feature is important in case of various\napplications where there is a need either to reclaim the moisture excess or to heat\nexchange in heating devices. But additionally,\nthis feature makes that this liquid suffers from lack of unambiguous flow maps (Parsi\net al., 2015; van Nimwegen et al., 2015). The mentioned identification\nalgorithms occurred to be helpful by construction such maps i.a. for facilities in the Process Tomography\nLaboratory. These maps consider the acceptable measurement range of gas and\nliquid streams powering the flow rig. This\ntranspired to be important especially in the case\nof&nbsp;horizontal sections of the pipeline. The determined flow maps differ\nsignificantly from these commonly used so far e.g.&nbsp;in\nthe border area between a slug and plug\nflow [fig 1. in the article (Att. 3 pos.\nI.B.7)].<\/p>\n\n\n\n<p>Furthermore,\nfrom the industrial point of view the diagnostic task is crucial because, in&nbsp;spite of the appropriate flow type identification, the algorithm\nshares the mechanism of&nbsp;malfunctioning detection which could and\nfrequently does affect the process hold-up, worse quality of the final product\nor even the emergency states like accidents or disasters. An algorithm is thus an efficient tool for\nundesirable states detection.<\/p>\n\n\n\n<p>The algorithm for two-phase flow facility control described in [Att. 3 pos. I.B.8] uses\nin turn knowledge about the objects and their spatial relations to work in a\nclosed feedback loop. This module allows to set the flow type in a special part\nof the pipeline and to evaluate the impact of the noisy factors (i.e. rapid\npressure changes, emergency) on the process behaviour. The&nbsp;algorithm is based on the fuzzy inference techniques,\ntherefore, makes a decision about the control\nstrategy. The fundamental analysis of the world literature, performed by me in\n2010, let me state that at the time there was a lack of similar computer methods\nwhich would be able to regulate the flow type in the basis of the tomographic diagnosis and fuzzy inference.<\/p>\n\n\n\n<p>In\nthe algorithm, the safety technique was\nimplemented. It has two tasks. First, it conducts the observation of the\ncurrent state of the flow to detect undesirable states. Second, the analysis of\nthe current diagnostic signals to predict the emergency and to avoid such\nsituations.<\/p>\n\n\n\n<p>The\nalgorithm for two-phase flow facility control is characterised by the short response time but also by its\nuniversality of application. It is dedicated\nfor various TPF processes which require precise and effective regulation. The\nmentioned precision was achieved thanks to the decomposition of the input\nsignals into two classes. The classic fuzzyfication\nwas not applied here because the input consists of&nbsp;two fuzzified relations\n(i.e. current and given flow type) and two fuzzified variables (i.e. values\nof&nbsp;gas and liquid streams). Similar to the identification algorithms the\nsame membership functions were applied to more precisely reflect the non-crisp\nborders between flow types. Therefore, one of the features is the possibility\nof changing the required flow type changing only the value of one parameter. This is performed within the most effective\nrange of the powering devices avoiding their overloads, saving lifespan and\ndecreasing any additional maintenance costs. Moreover, this functionality is\nresponsible for the regulator protection against\nthe over-loop state in the case when the\nalgorithm would try to reach the unsupported or&nbsp;undesirable flow type.<\/p>\n\n\n\n<p>Finally,\nit is worth to mention the simple deployment. This\nmay be performed by&nbsp;the maintenance staff of the flow rig and is guaranteed by\nsharing the individual sets of parameters and the intuitive inference rules. The\nmain research task was intensified\ntowards the automatic adjustment of the shape and number of membership\nfunctions of the regulator both input and output. Therefore, it is not required\nto determine any of complicated computer models of the flow rig, what is unfortunately necessary when using the classic regulators. The developed solutions\nallowed me to design the intelligent Black-Box type regulator dedicated to regulate the two-phase gas-liquid mixtures\nflows.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Usefulness of the\nconducted research<\/h3>\n\n\n\n<p>The\ncrucial aspect of the mentioned research area is the practical deployment of my\nnovel computer methods and algorithms\ninto the real processes. The first issue is developed in 2009 the complex\nsoftware for monitoring the TPF rig. This research I conducted in&nbsp;cooperation\nwith the team from the Opole University\nof Technology consisted of&nbsp;increasing the abilities of the ECT system\ndiagnosis by the sensor wall for vertical counter flows \u2013 the research project\n1318\/T10\/2005\/28 titled \u201c<em>Development of\nthe measurement method for two-phase gas-liquid flow structures identifications<\/em>\u201d.\nIn a frame of this project, I modified\nand shared the <em>WinRECO<\/em> computer\nprogram implementing the specialised\nmethods for raw measurement data and cross-sectional images analysis.<\/p>\n\n\n\n<p>In\nthe years 2006 \u2013 2010 I took part as research\nas well as an administrative fellow in the\ninternational 6FP Marie-Curie scientific-research project DENIDIA. In addition, I was responsible for the\npreparation of the proposal as well as further maintenance and report\npreparation for other domestic project founded by Polish MNiSW which aimed to\nsupport DENIDIA. I developed the computer methods which contributed to&nbsp;the\nachievement of the main DENIDIA project\u2019s objectives (actions), i.e.&nbsp;improvement of&nbsp;ECT systems spatial\nresolution, measurement data acquisition, image reconstruction and\nvisualisation saving the main advantage of ECT \u2013 measurement speed. The\nalgorithms I&nbsp;developed supported: the control and synchronisation of the rotatable ECT sensor [Att. 3 pos. II.L.14],\nthe image reconstruction process in dual-mode capacitance &#8211; resistance [Att. 3 pos.\nII.L.8] for the sensor dedicated for multi-phase flow processes visualisation [Att.\n3 pos. II.E.14] [Att. 3 pos. II.E.15]. My contribution to this second task\nallowed to develop the&nbsp;multimodal measurement system (Gamma, ECT, ERT).\nThe innovative solutions applied in the sensor are\nnow protected with the European patent I am a co-author. <\/p>\n\n\n\n<p>It\nis also important to point out that one of the major achievements of the DENIDIA\nproject is the Tom Dyakowski Process Tomography Laboratory at IIS at LUT. I was\nactively involved in the design, organisational and finally in constructional\ntasks. The laboratory consists of three facilities for multiphase flow\nprocesses non-invasive examination. In order to\nensure the similar conditions as in case of real industrial processes, the laboratory stands are built in\na&nbsp;semi-industrial scale.<\/p>\n\n\n\n<p>In\n2011 together with dr hab. in\u017c. Robert Banasiak\nI completed development of the software name <em>TomoKISStudio<\/em> [Att. 3 pos. II.L.17]. The built platform is a\ncomplex tool dedicated to electrical\nprocess tomography systems and makes the basic diagnostic software in the\nprocess tomography laboratory. Within this software I am the only author of\nalgorithms responsible for:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>the\ntomographic devices communication and configuration,<\/li><li>the\ncomputer method of automatic calibration of the tomography configuration\nparameters to adjust the 3D sensor to the process conditions considering the\nexamined medium, various sensor geometry and the measurement values to fit them\ninto the measurement range of the device and simultaneously to produce the\nsignificant difference in measurement for the empty\nand full sensor, <\/li><li>3D\nECT and ERT computer model determination including the generators of finite elements\nmesh, 3D electrodes layout, spatial electrical potentials and sensitivity\ndistribution [Att. 3 pos. II.E.12],<\/li><li>the\nsensitivity tunnel-based model determination,<\/li><li>the\ntomographic images reconstruction in dual (capacitance-resistance) mode\n(algorithm developed for DENIDIA purposes),<\/li><li>the\ncontrol and synchronisation of the\nrotatable capacitance sensor (algorithm developed for DENIDIA purposes),<\/li><li>the\ntomographic measurement data transmission via the Internet network for the\nremote diagnostic in the different (often distant) measurement sections,<\/li><li>the flow analysis in the basis of\nreconstructed images to produce the percentage phase distribution [Att. 3 pos. II.E.13],<\/li><li>the\ngraphical user interface, multithreaded and multimodal software structure.<\/li><\/ul>\n\n\n\n<p>The\ndeveloped <em>TomoKISStudio <\/em>software\nsupported with my algorithms for tunnel-based sensitivity modelling and maps\ndetermination was applied i.a. for\ndiagnostic of silos discharging process in a frame of the research project no\nPB-3687\/B\/T02\/2009\/37 titles \u201c<em>Non-invasive\nmethod of measurement of the dynamics of industrial processes of gravitational\nflow of bulk materials<\/em>\u201d supervised by dr in\u017c. Krzysztof Grudzie\u0144. Results\nobtained from the software allow determining\nthe sensitivity matrices necessary for 3D imaging of the TPF of silo\ndischarging. The results were helpful in the further research on particle flow\nmonitoring [Att. 3 pos. II.L.11]. The research was conduct in&nbsp;cooperation\nwith dr hab. in\u017c. Maciej Niedostatkiewicz from the Gda\u0144sk University of&nbsp;Technology. The verification was made on a real silos model, built in the\nlaboratory of the IIS at LUT. Additionally, I am a co-author of the article which in the basis of&nbsp;my algorithms and <em>TomoKISStudio\n<\/em>describes the developed and implemented solutions for the online three-dimensional imaging of the\ngravitational flow in silo both statically and during discharging [Att. 3 pos. II.E.9].\nThese solutions were exhibited\non&nbsp;international inventions shows winning numerous prizes and medals.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Summary<\/h3>\n\n\n\n<p>The main\nareas of the research I have conducted in the IIS at LUT after obtaining a PhD\ndegree focused on the <strong>development of computer methods for three-dimensional\ntomographic data visualisation and processing for purposes of industrial flow\nprocesses non-invasive diagnosis and regulation<\/strong>. My most important\nachievements have been listed below which\nprovide my contribution to the computer science research area:<\/p>\n\n\n\n<ul class=\"wp-block-list\"><li>the algorithm for image reconstruction process for the three-dimensional capacitance\ntomography including the 3D sensitivity matrix calculation algorithm, iterative\n3D image reconstruction algorithm [Att.\n3 pos. I.B.1],<\/li><li>the\nalgorithm for identification of the dead zones in the 3D capacitance sensor\ninter-electrodes areas [Att. 3 pos. I.B.2],<\/li><li>the\nalgorithm for geometrical and mechanical properties determination for 3D ECT\nsensors to reduce the inhomogeneity in the sensitivity distribution [Att. 3 pos.\nI.B.3], [Att. 3 pos. I.B.4], [Att. 3 pos. I.B.5],<\/li><li>the\nalgorithm for the sensitivity matrix determination dedicated to counter flows diagnosis and for the ECT\nsensors characterized by the greater\ndetection abilities in&nbsp;the neighbourhood of pipe walls [Att. 3 pos. I.B.3],<\/li><li>the\nalgorithm for tunnel-based sensitivity matrix calculation for the 3D ECT sensor\n[Att. 3 pos. I.B.6],<\/li><li>the\nalgorithm for metrological analysis of the sensitivity model [Att. 3 pos. I.B.4],<\/li><li>the\nalgorithm for percentage phases distribution\ncalculation of TPF in the basis\nof&nbsp;3D ECT images [Att. 3 pos. II.E.13],<\/li><li>the\nalgorithm for two-phase flow structures identification in the basis of raw tomographic data and fuzzy\ninference [Att. 3 pos. I.B.7],<\/li><li>the\nalgorithm for two-phase flow type control closed in&nbsp;the&nbsp;feedback loop\nwith the identification algorithm [Att. 3 pos. I.B.8].<\/li><\/ul>\n\n\n\n<p>The\nmethods I developed are characterised by\nsignificant usefulness and commercial potential. The algorithms designed to the\n3D ECT systems expand this diagnostic technique\u2019s abilities through increasing\nits application scope for industrial processes. The inference methods, in turn,\ndesigned to the purposes of TPFs diagnosis and control supported by the ECT\nsystem open the new possibilities for industrial solutions. Simultaneously, the\nwide area of flows processes\u2019 exploitation causes that my computer methods can\nbe used for optimisation of many\nindustrial processes as well as to warn against unexpected failures or even\nindustrial disasters. A crucial element of my research was an&nbsp;experimental\nverification of the methods and solutions proposed on the basis of semi-industrial installations of pneumatic\ntransport, a TPF and a gravitational flow while silo\ndischarging.<\/p>\n\n\n\n<p>The\nresearch tasks I conducted are strictly IT solutions and contribute to the\ncomputer science area. The algorithms and methods pertain to the computer\ntechniques of modelling, image reconstruction, visualisation, analysis and\nprocessing. The scope of my last research was related to artificial intelligence techniques. I developed\nnew solutions in the area of&nbsp;fuzzy clustering, of defining fuzzy\nlinguistic variables for TPFs purposes. This,\nadmittedly, extends the world state of the art\nin case of AI industrial applicability. Moreover,\nthis research tasks on intelligent techniques had an effect of new statistical\ndata and signal processing methods and in the case\nof algorithmics upgraded the electrical field simulation for 3D&nbsp;ECT using\nfinite elements method. The innovative intelligent fuzzy regulator shares\nthe&nbsp;contribution into the industrial\nautomatic control and computer sciences discipline. The usage of the\nregulator allows to&nbsp;automate the\nproduction line providing the mechanism of&nbsp;failure prevention or safety.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a>Other scientific and research achievements<\/a><\/h3>\n\n\n\n<p>Since\n2006 I took part in various research works which were not connected with the presented in the previous section my\nacademic achievement. <\/p>\n\n\n\n<p>In\nthe years 2008 \u2013 2011 I took part as a research\nfellow in a scientific-research project entitled \u201c<em>An autonomous military robot designed for reconnaissance and mine\ndetection<\/em>\u201d no.&nbsp;0010\/1\/R\/T00\/2008\/05. The aim of the research,\nconducted in cooperation with the Prexer\ncompany was to test the possibilities of detection of explosives in the ground by means of an electrical capacitance\ntomography system. Within the framework of the research with the use of\nreference materials (sand and wood), I\nundertook, along with the TomoKIS team, an&nbsp;attempt to assess a relative\nvalue of electrical permittivity of the pyrotechnic materials studied. The conducted\nstudies have demonstrated that it is possible to detect a specific class of\nexplosive materials, non-sensitive to an electric field environment, using electrical capacitance tomography.<\/p>\n\n\n\n<p>In\na frame of the DENIDIA project in IIS in 2010,\nI developed the method and the device for fractions measurement of the\nmultiphase flow especially fractions of water, oil and gas of&nbsp;flows in the\noil platform pipelines [Att. 3 pos. II.E.14] i\n[Att. 3 pos. II.E.15]. The essence of this innovation protected since\n27.02.2015 by the European patent no EP-2416127 is the application of three tomographic modules (EIT\/ECT\/GRT) within one\nsensor and development of the computer\nmethods for data aggregation. My contribution in it was to&nbsp;develop and\nimplement the algorithms for electrical tomography diagnosis i.a. the algorithm for capacitance and resistance\ntomography synchronisation and calibration, the algorithm of image\nreconstruction of capacitance and resistance modules as well as these images\naggregation algorithm.<\/p>\n\n\n\n<p>I\nam also an inventor of two components of <em>intelliFlowControl<\/em> system, not mentioned so\nfar, which <em>de facto<\/em> do not belong to\nmy academic achievements but admittedly extend the system industrial proposal.\nFirst of them is the algorithm for timestamp distribution service\nto&nbsp;synchronise the measurement data obtained from various devices with\ndifferent acquisition frequency [Att. 3 pos.\nII.E.19]. Two capacitances tomography systems, CCD camera as well as numerous\nmeasurement devices in the flow rig like gasometers, flow meters and barometers\nwere synchronised. Each of the programs for collecting the&nbsp;measurement\ndata received the timestamps provided in the basis\nof time determined by one of the computers and assigned them to every\nmeasurement value. This solution was definitely\nhelpful while further analysis as well as while metrological validation of the\ncomplete system. The second component is the mobile touchable panel for system\nmodel visualisation, its current state monitoring supported by the graphical\nuser interface for programming own flow rig models. I&nbsp;developed the method\nfor presenting the current state of all\nstrategic parts (elements) of the flow rig. Using the \u201ctouchable gesture\u201d\n(common mobile devices feature), it is\npossible to&nbsp;regulate the process from the tablet or phone. It is possible\nto change the pipeline, flow type or separate set of power devices. Furthermore, this solution thanks to the shared\nself-programmable features allows creating\nsome new extended elements for visualisation, monitoring and control. These\nissues, mentioned above, make an alternative to expensive commercial industrial visualisation systems.<\/p>\n\n\n\n<p>My\nexperience in the field of 3D ECT systems occurred to be helpful in the case of two different research tasks published\nin the journals from the Thomson Reuters JCR list: [Att. 3 pos. I.A.1] and [Att. 3 pos. I.A.4] which I am a co-author. In case of both articles, I was responsible for the design of the experimental setup which\nconsisted of the measurement 3D ECT system and measurement data acquisition\nmodule. My duties included as well the experiments conducting.<\/p>\n\n\n\n<p>Finally,\nin 2017 I was invited to be the co-supervisor of one of the PhD student in a frame\nof the Marie Sk\u0142odowska-Curie European Innovative Network (HORYZONT 2020) \u201c<em>Smart tomographic sensors for advanced\nindustrial process control (TOMOCON)<\/em>\u201d which joins 12&nbsp;international\nacademic institutions and 15 industry partners, who work together in the\nemerging field of industrial process control using smart tomographic sensors. Since\n2017 I am responsible for one of 15 research tasks titled \u201c<em>ERT tomography for measuring the&nbsp;crystallization\nprogress in a batch reactor<\/em>\u201d, and I\nam an assistant supervisor of one PhD student Guruprasad Rao. <\/p>\n\n\n\n<p>Additionally,\nmy experience and research achievements in the field of&nbsp;electrical process\ntomography computer methods appeared to be useful for two research projects\u2019\npurposes I&nbsp;am actively involved in.\nBoth projects started in 2017 and are funded\nby polish National Centre of&nbsp;Research and Development institution and will be conducted in cooperation with the\nNETRIX research and development company from Lublin. The first project (acronym LETS) deals with the problem of innovative\napplication of&nbsp;process tomography systems for\nthe purpose of an area imaging and monitoring using potentials nodes\nmap. Next project aims to&nbsp;develop \u201c<em>The\nnew generation of industrial tomography platform for diagnostics and process\ncontrol<\/em>\u201d &#8211; PLATOM.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\"><a>Bibliography<\/a><\/h3>\n\n\n\n<ul class=\"wp-block-list\"><li>Abbagoni, B.M., Yeung, H., 2016. Non-invasive classification of gas-liquid two-phase horizontal flow regimes using an ultrasonic Doppler sensor and a neural network. Meas. Sci. Technol. 27, 84002. doi:10.1088\/0957-0233\/27\/8\/084002<\/li><li>Abdulmouti, H., 2015. Bubbly Two-Phase Flow: Part II- Characteristics and Parameters. Am. J. Fluid Dyn. 4, 115\u2013180. doi:10.5923\/j.ajfd.20140404.01<\/li><li>Akita, K., Okazaki, T., Koyama, H., 1988. Gas holdups and friction factors of gas-liquid two-phase flow in an air-lift bubble column. J. Chem. Eng. Japan 21, 476\u2013482. doi:10.1252\/jcej.21.476<\/li><li>Arvoh, B.K., Hoffmann, R., Halstensen, M., 2012. Estimation of volume fractions and flow regime identification in multiphase flow based on gamma measurements and multivariate calibration. Flow Meas. Instrum. 23, 56\u201365. doi:10.1016\/j.flowmeasinst.2011.11.002<\/li><li>Balamurugan, S., Lad, M.D., Gaikar, V.G., Patwardhan, A.W., 2007. Hydrodynamics and mass transfer characteristics of gas\u2013liquid ejectors. Chem. Eng. J. 131, 83\u2013103. doi:10.1016\/j.cej.2006.12.026<\/li><li>Bertola, V., 2003. Two-Phase Flow Measurement Techniques. Springer Vienna, pp. 281\u2013323. doi:10.1007\/978-3-7091-2538-0_6<\/li><li>Beux, S., Nunes, E., Barana, A.C., 2007. Effect of temperature on two-phase anaerobic reactors treating slaughterhouse wastewater. Brazilian Arch. Biol. Technol. 50, 1061\u20131072. doi:10.1590\/S1516-89132007000700017<\/li><li>Bla\u017eej, M., Cartland Glover, G.M., Generalis, S.C., Marko\u0161, J., 2004. Gas\u2013liquid simulation of an airlift bubble column reactor. Chem. Eng. Process. Process Intensif. 43, 137\u2013144. doi:10.1016\/S0255-2701(03)00010-2<\/li><li>Brebbia, C. a., Mammol,&nbsp; a. a., 2011. Computational Methods in Multiphase Flow VI.<\/li><li>Brzeski, P., Mirkowski, J., Olszewski, T., Pl\u0105sowski, A., Smolik, W., Szabatin, R., 2003. Multichannel capacitance tomograph for dynamic process imaging. Opto-Electronics Rev. 11, 175\u2013180.<\/li><li>Cheng, L., Ribatski, G., Thome, J.R., 2008. Two-Phase Flow Patterns and Flow-Pattern Maps: Fundamentals and Applications. Appl. Mech. Rev. 61, 50802. doi:10.1115\/1.2955990<\/li><li>Delnoij, E., Kuipers, J.A.., van Swaaij, W.P.., Westerweel, J., 2000. Measurement of gas?liquid two-phase flow in bubble columns using ensemble correlation PIV. Chem. Eng. Sci. 55, 3385\u20133395. doi:10.1016\/S0009-2509(99)00595-3<\/li><li>Dziubinski, M., Fidos, H., Sosno, M., 2004. The flow pattern map of a two-phase non-Newtonian liquid\u2013gas flow in the vertical pipe. Int. J. Multiph. Flow 30, 551\u2013563. doi:10.1016\/j.ijmultiphaseflow.2004.04.005<\/li><li>Fang Dong, Li-De Fang, Hong-Lian Li, Yan Zhu, 2012. Characterization of gas-liquid two-phase flow pattern based on complexity measures analysis, in: 2012 International Conference on Machine Learning and Cybernetics. IEEE, pp. 996\u20131000. doi:10.1109\/ICMLC.2012.6359490<\/li><li>Gadd, R., Vinther, F., Record, P.M., Rolfe, P., 1992. Reconstruction of three-dimensional data for electrical impedance tomography. Electron. Lett. 28, 974\u2013976. doi:10.1049\/el:19920619<\/li><li>Guo, X., Zhou, X., Chen, Q., Liu, J., 2013. Flow field and dissolved oxygen distributions in the outer channel of the Orbal oxidation ditch by monitor and CFD simulation. J. Environ. Sci. (China) 25, 645\u201351.<\/li><li>Holder, D.S., 2004. Electrical Impedance Tomography: Methods, History and Applications. CRC Press. doi:10.1118\/1.1995712<\/li><li>Isaksen, \u00d8., 1996. A review of reconstruction techniques for capacitance tomography. Meas. Sci. Technol. 7, 325\u2013337. doi:10.1088\/0957-0233\/7\/3\/013<\/li><li>Jothiprakash, V., Bhosekar, V.V., Deolalikar, P.B., 2015. Flow characteristics of orifice spillway aerator: numerical model studies. ISH J. Hydraul. Eng. 21, 216\u2013230. doi:10.1080\/09715010.2015.1007093<\/li><li>Kassab, S.Z., Kandil, H.A., Warda, H.A., Ahmed, W.H., 2009. Air-lift pumps characteristics under two-phase flow conditions. Int. J. Heat Fluid Flow 30, 88\u201398. doi:10.1016\/j.ijheatfluidflow.2008.09.002<\/li><li>Kiambi, S.L., Kiriamiti, H.K., Kumar, A., 2011. Characterization of two phase flows in chemical engineering reactors. Flow Meas. Instrum. 22, 265\u2013271. doi:10.1016\/j.flowmeasinst.2011.03.006<\/li><li>Lafmejani, S.S., Olesen, A.C., K\u00e6r, S.K., 2017. VOF modelling of gas\u2013liquid flow in PEM water electrolysis cell micro-channels. Int. J. Hydrogen Energy. doi:10.1016\/j.ijhydene.2017.05.079<\/li><li>Liu, Y., Wang, L., Zhu, Z., 2015. Numerical study on flow characteristics of rotor pumps including cavitation. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 229, 2626\u20132638. doi:10.1177\/0954406214562634<\/li><li>Loser, T., Wajman, R., Mewes, D., 2001. Electrical capacitance tomography: image reconstruction along electrical field lines. Meas. Sci. Technol. 12, 1083. doi:10.1088\/0957-0233\/12\/8\/314<\/li><li>Malijevsky, A., Archer, A.J., 2013. Sedimentation of a two-dimensional colloidal mixture exhibiting liquid-liquid and gas-liquid phase separation: A dynamical density functional theory study. J. Chem. Phys. 139, 144901. doi:10.1063\/1.4823768<\/li><li>Metherall, P., Barber, D.C., Smallwood, R.H., Brown, B.H., 1996. Three-dimensional electrical impedance tomography. Nature 380, 509\u2013512. doi:10.1038\/380509a0<\/li><li>Mewes, D., Wiemann, D., 2003. Two-Phase Flow with Mass Transfer in Bubble Columns. Chem. Eng. Technol. 26, 862\u2013868. doi:10.1002\/ceat.200300009<\/li><li>Mokhatab, S., 2008. Two-Phase Flow: Theory and Applications, 1st Edition Clement Kleinstreuer Taylor &amp; Francis Group, Published May 2003, 454 pages ISBN: 1591690005. Can. J. Chem. Eng. 84, 260\u2013261. doi:10.1002\/cjce.5450840214<\/li><li>Mosorov, V., Sankowski, D., Mazurkiewicz, ?, Dyakowski, T., 2002. The \u201cbest-correlated pixels\u201d method for solid mass flow measurements using electrical capacitance tomography. Meas. Sci. Technol. 13, 1810\u20131814. doi:10.1088\/0957-0233\/13\/12\/302<\/li><li>Olesen, A.C., Romer, C., Kaer, S.K., 2016. A numerical study of the gas-liquid, two-phase flow maldistribution in the anode of a high pressure PEM water electrolysis cell. Int. J. Hydrogen Energy 41, 52\u201368. doi:10.1016\/j.ijhydene.2015.09.140<\/li><li>Ozbayoglu, E.M., Ozbayoglu, M.A., 2009. Estimating Flow Patterns and Frictional Pressure Losses of Two-Phase Fluids in Horizontal Wellbores Using Artificial Neural Networks. Pet. Sci. Technol. 27, 135\u2013149. doi:10.1080\/10916460701700203<\/li><li>Parsi, M., Vieira, R.E., Torres, C.F., Kesana, N.R., McLaury, B.S., Shirazi, S.A., Schleicher, E., Hampel, U., 2015. On the effect of liquid viscosity on interfacial structures within churn flow: Experimental study using wire mesh sensor. Chem. Eng. Sci. 130, 221\u2013238. doi:10.1016\/J.CES.2015.03.033<\/li><li>Pl\u0105skowski A., Beck M.S., Thorn R., D.T., 1995. Imaging Industrial Flows \u2013 Applications of Electrical Process Tomography. IOP Publishing, Bristol.<\/li><li>Polydorides, N., Lionheart, W.R.B., 2002. A Matlab toolkit for three-dimensional electrical impedance tomography: a contribution to the Electrical Impedance and Diffuse Optical Reconstruction Software project. Meas. Sci. Technol. 13, 1871\u20131883. doi:10.1088\/0957-0233\/13\/12\/310<\/li><li>Reinecke, N., Mewes, D., 1996. Recent developments and industrial\/research applications of capacitance tomography. Meas. Sci. Technol. 7, 233\u2013246. doi:10.1088\/0957-0233\/7\/3\/004<\/li><li>Ruspini, L.C., Marcel, C.P., Clausse, A., 2014. Two-phase flow instabilities: A review. Int. J. Heat Mass Transf. 71, 521\u2013548. doi:10.1016\/j.ijheatmasstransfer.2013.12.047<\/li><li>Solotych, V., Lee, D., Kim, J., Amalfi, R.L., Thome, J.R., 2016. Boiling heat transfer and two-phase pressure drops within compact plate heat exchangers: Experiments and flow visualizations. Int. J. Heat Mass Transf. 94, 239\u2013253. doi:10.1016\/j.ijheatmasstransfer.2015.11.037<\/li><li>Teng, P., Yang, J., Pfister, M., 2016. Studies of Two-Phase Flow at a Chute Aerator with Experiments and CFD Modelling. Model. Simul. Eng. 2016, 1\u201311. doi:10.1155\/2016\/4729128<\/li><li>Vadlakonda, B., Mangadoddy, N., 2017. Hydrodynamic study of two phase flow of column flotation using electrical resistance tomography and pressure probe techniques. Sep. Purif. Technol. 184, 168\u2013187. doi:10.1016\/j.seppur.2017.04.029<\/li><li>van Nimwegen, A.T., Portela, L.M., Henkes, R.A.W.M., 2015. The effect of surfactants on air\u2013water annular and churn flow in vertical pipes. Part 1: Morphology of the air\u2013water interface. Int. J. Multiph. Flow 71, 133\u2013145. doi:10.1016\/J.IJMULTIPHASEFLOW.2014.03.008<\/li><li>Venkata, S.K., Roy, B.K., 2012. An Intelligent Flow Measurement Scheme using Ultrasonic Flow Meter, in: Modelling, Identification and Control \/ 770: Advances in Computer Science and Engineering. ACTAPRESS, Calgary,AB,Canada. doi:10.2316\/P.2012.769-033<\/li><li>Wang, H., Liu, S., Fan, J., Yang, W., 2003. 3D Presentation of Images with Capacitance Tomography, in: 3rd World Congress on Industrial Process Tomography. Banff, Canada, pp. 331\u2013336.<\/li><li>Warsito, W., Fan, L.-S., 2005. Dynamics of spiral bubble plume motion in the entrance region of bubble columns and three-phase fluidized beds using 3D ECT. Chem. Eng. Sci. 60, 6073\u20136084. doi:10.1016\/j.ces.2005.01.033<\/li><li>Warsito, W., Fan, L.-S., 2003. Development of three-dimensional electrical capacitance tomography, in: Proc. 3rd World Congress on Industrial Process Tomography (Banff). p. 391.<\/li><li>Warsito, W., Marashdeh, Q., Fan, L.S., 2007. Some comments on \u201cSpatial imaging with 3D capacitance measurements.\u201d Meas. Sci. Technol. 18, 3665\u20133667. doi:10.1088\/0957-0233\/18\/11\/N01<\/li><li>Xie, T., Ghiaasiaan, S.M., Karrila, S., 2004. Artificial neural network approach for flow regime classification in gas\u2013liquid\u2013fiber flows based on frequency domain analysis of pressure signals. Chem. Eng. Sci. 59, 2241\u20132251. doi:10.1016\/j.ces.2004.02.017<\/li><li>Yang, W.Q., Peng, L., 2003. Image reconstruction algorithms for electrical capacitance tomography. Meas. Sci. Technol. 14, R1\u2013R13. doi:10.1088\/0957-0233\/14\/1\/201<\/li><li>York, T., Mazurkiewicz, L., Polydorides, N., Mann, R., Grieve, B., 2003. Image Reconstruction for Pressure Filtration Using EIDORS 3D, in: 3rd World Congress on Industrial Process Tomography. p. 559.<\/li><li>Zadeh, L.A., 1999. Fuzzy sets as a basis for a theory of possibility. Fuzzy Sets Syst. 100, 9\u201334. doi:10.1016\/S0165-0114(99)80004-9<\/li><li>Zargaran, A., Mozaffari, E., Giddings, D., 2016. Gas-liquid slip velocity determination in co-current column flotation. Sep. Purif. Technol. 169, 179\u2013186. doi:10.1016\/j.seppur.2016.05.018<\/li><li>Zhang, W., Wang, C., Yang, W., Wang, C.-H., 2014. Application of electrical capacitance tomography in particulate process measurement \u2013 A review. Adv. Powder Technol. 25, 174\u2013188. doi:10.1016\/j.apt.2013.12.003<\/li><li>Zhu, Z.T., Wang, Y., Zhao, L.F., Ning, C., Xie, S.F., Liu, Z.C., 2015. The effect of gas fraction on centrifugal pump. IOP Conf. Ser. Mater. Sci. Eng. 72, 32022. doi:10.1088\/1757-899X\/72\/3\/032022<\/li><\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Updated: 1.09.2019wersja w j\u0119zyku polskim Among the area of my research, introduced in a synthetic way in the Summary of Professional Accomplishments (SPA), as the most important and the main research current I considered the achievements in the domain of &hellip; <a href=\"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/scientific-profile\/\">Continue reading <span class=\"meta-nav\">&rarr;<\/span><\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-161","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/pages\/161","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/comments?post=161"}],"version-history":[{"count":23,"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/pages\/161\/revisions"}],"predecessor-version":[{"id":200,"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/pages\/161\/revisions\/200"}],"wp:attachment":[{"href":"https:\/\/rwajman.iis.p.lodz.pl\/wordpress\/wp-json\/wp\/v2\/media?parent=161"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}