Multiple model-based control of multi variable continuous microbial fuel cell (CMFC) using machine learning approaches. (2nd September 2020)

• Record Type: Journal Article
• Authors: Yewale, Ashish; Methekar, Ravi; Agrawal, Shailesh
• Is Part Of: Computers & chemical engineering. Volume 140(2020)
• Publication Date: 2020-09-02
• Subjects: Microbial fuel cell -- Temperature model -- MIMO -- Machine learning -- MMB controller; Chemical engineering -- Data processing -- Periodicals; 660.0285
• DOI: 10.1016/j.compchemeng.2020.106884
• Languages: English
• ISSNs: 0098-1354

Abstract

Abstract: Microbial fuel cells (MFCs) are promising technology to simultaneously harvest energy and produce clean water from wastewater or organic matter. However, MFC is a complex, time-variant, and nonlinear system facing significant challenges in modeling and control development. Some of the challenges can be addressed by a deeper understanding of the system through mathematical modeling. A temperature based mathematical model of the MFC is developed and qualitatively validated with the experimental data from the literature. The nonlinearity present in the MFC is addressed by using multiple models-based control strategy. Various machine learning approaches are used for control integration in the developed control strategy. The performance of the developed multiple models-based control strategy shows a reduction of 65% in the average settling time of the controller when compared with the single model linear controller. Among various machine learning approaches used, WkNN model switching criteria provides the best performance. Graphical abstract: Image, graphical abstract