Exclude Niche Genetic Algorithm Research And Applications

After a brief and impressive survey on the basic theory and mathematical foundations of genetic algorithms, their main implementation techniques and the development, a kind of crowding niche genetic algorithms based on Penalty Function is improved, and its searching efficiency is advanced.A new crowding niche genetic algorithms based on clustering analysis is proposed in this work through analysis of niche genetic algorithms. This kind of new algorithms combines clutering analysis and crowding techniques organically, which can search many peaks of the multimodal function space effectively, and can control the quantity of the convergent niche to avoid to find invalid peaks by adjusting minimal clutering distance. The information of the radius of the niche and the quantity of the niche must not be known by this kind of algorithms.Based on the study and comparisons with other algorithms on the several standard multimodal test problems, the supremacy of the new algorithms is demonstrated in this work, and the proposed algorithms is applied to optimize the model of the district design proposed in this work.