Open AccessPARAMETRIC METHODS FOR ECT INVERSE PROBLEM SOLUTION IN SOLID FLOW MONITORING
Author(s) -
Andrzej Romanowski,
Krzysztof Grudzień,
Hela Garbaa,
Lidia Jackowska-Strumiłło
Publication year - 2017
Publication title -
informatyka, automatyka, pomiary w gospodarce i ochronie środowiska
Language(s) - English
Resource type - Journals
eISSN - 2391-6761
pISSN - 2083-0157
DOI - 10.5604/01.3001.0010.4582
Subject(s) - parametric statistics , markov chain monte carlo , inverse problem , computer science , monte carlo method , probabilistic logic , algorithm , bayesian probability , artificial intelligence , mathematics , statistics , mathematical analysis
The article presents the parametrisation-based methods of monitoring of the process of gravitational silo discharging with aid of capacitance tomography techniques. Proposed methods cover probabilistic Bayes’ modelling, including spatial and temporal analysis and Markov chain Monte Carlo methods as well as process parametrisation with artificial neural networks. In contrast to classical image reconstruction-based methods, parametric modelling allows to omit this stage as well as abandon the associated reconstruction errors. Parametric modelling enables the direct analysis of significant parameters of investigated process that in turn results in easier incorporation into the control feedback loop. Presented examples are given for the gravitational flow of bulk solids in silos.