A New Algorithm Based on Immune Algorithm and Hopfield Neural Network for Multimodal Function Optimization

This paper analyzes immune theory and Hopfield Neural Network (HNN), and then proposes a new algorithm for multimodal function. This new algorithm uses the advantages of both HNN and immune algorithm, and it appears excellent characteristic in optimal problems of multimodal function. In detail, we obtain a group of solutions with variety by immune algorithm (IA) first; and then the solutions are partitioned into some clusters. Finally we take cluster centroids returned by clustering algorithm as the initial value of each HNN, and run the Hopfield neural networks to obtain all minima.