Linear and Nonlinear System Modeling

6Development of New Bioinspired Hybrid Algorithms for Parameter Modeling of Photovoltaic Panels

Souvik Ganguli*, Shilpy Goyal and Parag Nijhawan

Department of Electrical and Instrumentation Engineering, Thapar Institute of Engineering and Technology, Patiala, Punjab, India

Abstract

Three hybrid bioinspired techniques are developed in this chapter. Two popular nature-dependent methods, viz., the firefly method, abbreviated as firefly algorithms (FA) and the bacterial foraging algorithm (BFA) were integrated utilizing the benefits of each. Firefly algorithms is employed for exploring the entire search domain while the three improved BFA schemes proposed by Supriyono have been utilized to search locally in the global optimization methods. Mainly, the step size of chemotaxis in bacterial foraging technique is modified with linear, quadratic, and exponential adaptations to improve upon the basic BFA technique. The bioinspired techniques not only outperform their parent methods, but they also surpass some of the most cutting-edge metaheuristic strategies. A bunch of unconstrained standard functions are taken up in this work to test the efficiency of the suggested approaches. The statistical analysis of the optimal values as well their non-parametric test also validate this fact. Further, these hybrid algorithms are also successfully used for the evaluation of two-diode model parameters using a novel objective function coined as weighted sum of square error (WSSE). The test results ...