Ant colony optimization algorithm and fuzzy logic for switched reluctance generator control

Abstract

<abstract> <p>This article discusses two methods to control the output voltage of switched reluctance generators (SRGs) used in wind generator systems. To reduce the ripple of the SRG output voltage, a closed-loop voltage control technique has been designed. In the first method, a proportional-integral (PI) controller is used. The parameters of the PI controller are tuned based on the voltage variation. The SRG is generally characterized by strong nonlinearities. However, finding appropriate values for the PI controller is not an easy task. To overcome this problem and simplify the process of tuning the PI controller parameters, a solution based on the ant colony optimization algorithm (ACO) was developed. To settle the PI parameters, several cost functions are used in the implementation of the ACO algorithm. To control the SRG output voltage, a second method was developed based on the fuzzy logic controller. Unlike several previous works, the proposed methods, ACO and fuzzy logic control, are easy to implement and can solve numerous optimization problems. To check the best approach, a comparison between the two methods was performed. Finally, to show the effectiveness of this study, we present examples of simulations that entail the use of a three-phase SRG with a 12/8 structure and SIMULINK tools.</p> </abstract>

Speed control of switched reluctance motors based on fuzzy logic controller and MATLAB/Simulink

“…Adaptive control technique, and artificial neural network (ANN) [14]- [16] those advanced algorithms that necessitate complex processes and professional users during the design process, which limits their practical applications. Another type is optimization methods, such as the algorithm of ants colony, particle swarm optimization (PSO), or some optimal control techniques [17]- [20] that use the performance index for a given speed response as the optimal control purpose, establishes restriction situations, following that enhances the performance of given speed response. Even so, there is no assumption that the controlling function of the optimization techniques necessitates a significant amount of computing and processing of data, which places a larger market on the controller's arithmetic abilities in real applications.…”