Design of Optimum SIMO-PID Neural Voltage-Tracking Controller for Non-Linear Fuel Cell System Based on a Comparative Study Of Various Intelligent Swarm Optimization Algorithms

This paper presents the enhancement of the output performance of a non-linear fuel cell (FC) system using a new design that comprises an adaptive SIMO-PID neural controller with different types of online swarm optimization algorithms. The work focuses on improving the use of single-input multi-output (SIMO) PID neural networks to control the non-linear FC system. The goal of the proposed adaptive SIMO-PID neural voltage-tracking controller is to rapidly and precisely identify the optimal hydrogen flow rate and oxygen flow rate control actions that are used to control the (FC) stack terminal output voltage. Three swarm optimization algorithms are used to find and tune the weights of the SIMO-PID neural controller: the Firefly algorithm, chaotic particle swarm optimization algorithm, and proposed hybrid Firefly-chaotic particle swarm optimization (F-CPSO) algorithm. Numerical simulation results show that the proposed controller using the (F-CPSO) algorithm is more accurate than with the FA or CPSO; the proposed SIMO-PID neural controller parameters are obtained more rapidly there is a high reduction in the number of function evolutions. Furthermore, the proposed controller's ability with the F-CPSO algorithm to generate a smooth flow rate control response for the non-linear (PEMFC) system without voltage oscillation in the output is determined by investigations under load variations.