A powerful modified Genetic Algorithm for multimodal function optimization

In this paper, we describe an efficient approach for multimodal function optimization using Genetic Algorithms. Through analyzing the main mechanism leading GAs to be premature, we find that the phenomenon called Nepotism causes the confliction between accuracy and speed. To realize the twin goals to satisfy precision requirement and improve running speed of the GA, we proposed a GA based on heuristic mutation with final-zero-rate (HMGA). In order to check the optimization ability of this new approaches, we also apply it into several difficult optimization problems, selected from the literature. The results produced by this new approaches, proving that this technique generates better genetic algorithm can be used as a trade-offs and that the g highly-efficient, highly-precise and reliable optimization tool.