Neural Network with Migration Parallel GA for Adaptive Control of Integrated DE-PSO Parameters

Abstract

This study develops an evolutionary strategy called DEPSO-GANN, which uses an artificial neural network (ANN) based on a parallel genetic algorithm (PGA) with migration for the adaptive control of integrated differential evolution (DE) and particle swarm optimization (PSO) to solve large-scale optimization problems, reduce calculation costs, and improve the stability of convergence towards the optimal solution. This approach combines the global search ability of DE and the local search ability of adaptive system with migration parallel GA. The proposed algorithm incorporates concepts from DE, PSO, PGA and neural networks (NN) to facilitate the adaptive control of parameters. DEPSO-GANN is applied to several numerical benchmark tests with multiple dimensions to evaluate its performance, it is also compared with other evolutionary algorithms (EAs) and memetic algorithms (MAs), which is shown to be statistically significantly superior to other EAs and MAs. We confirm satisfactory performance through various benchmark tests.