Enhancement In Control Method for Switched Reluctance Motors through GA-BPNN

Non-linear demonstration of the flux linkages characteristic is principal to the control and execution assessment of SRM. BPNN modelling method has disadvantage marks of local minimum of BPNN which fundamentally shows the convergence rate. The estimation of GA is considered to solve BPNN's local minima. Genetic algorithm is utilized for optimizing the weights and bias of a backpropagation neural network. The goal of the genetic algorithm is to calculate the relationship between the flux and torque of the switched reluctance motor based on the current in the stator and the rotor's position. Utilizing flux linkage characteristics from FEM analysis of an 8/6 SRM, an advanced model, using GA-BPNN, is created to represent non-linear flux connections. The results of the simulation indicate that the model displays quick convergence, precise findings, and robust generalization abilities.