COMPARISON OF GENETIC ALGORITHM, HOPFIELD AND MLP NEURAL NETWORK TECHNIQUES FOR A CONSTRAINED OPTIMIZATION PROBLEM

In recent decays, soft computing techniques such as genetic algorithm (GA) and artificial neural networks (ANN) are increasingly employed in a diverse area of applications. As optimization tools, genetic algorithm and Hopfield net are successfully applied in constrained optimization problems. In this study suitability of Multi-Layered Perceptron (MLP) for a constrained optimization problem; namely economic dispatch problem, is investigated and a comparison is carried out between GA, Hopfield and MLP. As a case study, the performance of the MLP techniques in economic dispatch problem is compared to the results given in literature. For economic dispatch problem, the MLP approach is compared with an improved Hopfield NN approach (IHN) [1], a fuzzy logic controlled genetic algorithm (FLCGA) [2], an advance engineered-conditioning genetic approach (AECGA) [3] and an advance Hopfield NN approach (AHNN) [4]. Results show that although MLP neural network is inherently not suitable for optimization task when compared to GA and Hopfield, the performance of MLP neural network may be highly improved in order to obtain comparable results to that of GA and Hopfield neural networks