Critical Analysis of Optimization Techniques for a MRPID Thermal System Controller

Abstract

In this paper, critical analysis of performance comparison of various optimization techniques for a MRPID (Multi Resolution Proportional Integral Derivative) thermal system controller is presented. This controller controls the uncertainty or mismatch between real and reference temperature. Wavelet coefficients of error between the two temperatures and their corresponding gains are added to produce a control signal for the thermal system. The main purpose of the proposed method is to optimize the tuning parameters of a wavelet-based MRPID controller to regulate the switching pulse of the thermal system. Modeling and simulations are carried out using MATLAB/Simulink@2015. Figures-of-merit such as disturbance rejection, system stability and transient response are also compared using various techniques like Particle swarm optimization (PSO)-MRPID, Genetic Algorithm (GA)-MRPID (existing techniques), Firefly-MRPID controllers and Elephant herding optimization (EHO) based Recurrent neural network (RNN)-MRPID algorithm. Firefly-MRPID and EHO with RNN-based MRPID are proposed for determining unique (best) operating conditions of the thermal system in this paper.

KEYWORDS:

• EHO algorithm
• Firefly algorithms
• GA
• MRPID controller
• PSO
• Thermal system

Additional information

Notes on contributors

Abhas Kanungo

Abhas Kanungo obtained his bachelor's degree in instrumentation and control engineering from RGTU, Bhopal in 2008. Then he obtained his master's degree in control systems from the NIT, Kurukshetra, Haryana India in 2010 and now doing PhD from the NIT, Kurukshetra, Haryana, India. His current research interest is wavelet-based control signal processing. He has been an assistant professor in EIE Dept, KIET Group of Institutions, Ghaziabad for the last 7.5 years.

Monika Mittal

Monika Mittal completed her graduation in electrical engineering from M M M Institute of Technology, Gorakhpur, India in 1992 and completed her post-graduation from NIT Kurukshetra, India in 1994 with specialization in control systems. She completed her doctorate from Electrical Engineering Dept., NIT Kurukshetra in 2013. She has a teaching experience of about 23 years and authored about 50 research papers in international and national journals and conferences. Presently, she is working in the areas of signal processing applications in control systems, computational algorithms and wavelets in control as an associate professor in Electrical Engineering Department, NIT Kurukshetra. Email: [email protected]

Lillie Dewan

Lillie Dewan has completed her graduation in electrical engineering from Punjab University in 1983 and completed her post-graduation from Kurkshetra University, Haryana, India in 1987 with specialization in controls. She completed her doctorate in 2001. She has a teaching experience of about 31 years. She authored about 130 research papers in international and national journals and conferences. Presently, she is working in the areas of control, instrumentation and biomedical signal processing, as professor in Electrical Engineering Department, NIT Kurukshetra. Email: [email protected]