解图着色问题的一个新的遗传算法-计算机软件与理论专业论文.docxVIP

Abstract Genetic algorithm (GA) is a kind of globally random searching methods using natural evolution and genetic mechanism. Its application predominance lies in the ability to deal with complicated and non-linear problems which are difficult for traditional searching methods. Graph coloring problem (GCP) is one of the most widely studied combinatorial optimization problems with many applications both in theory and engineering. But it is also a well-known NP-complete problem. Therefore, designing efficient algorithm for this problem is not only of great theoretical value, but also of important application value. After analyzing the classical genetic algorithm, this paper presents a new improved genetic algorithm for the graph coloring problem. A novel scheme of producing the initial population is proposed, and it can ensure diversity of population and enhances convergence. Then an improved greedy partition crossover operator is designed and it can ensure that the big subsets of the vertices will not be destroyed. Moreover, two new mutation operators are proposed and they can increase diversity of population and improve local searching ability of the algorithm. Finally, The simulation experiments on some standard coloring instances are made and the simulation results indicate that the new improved genetic algorithm has good ability of searching optimal solution for various types of graphs. Keywords: Genetic algorithm GCP Crossover operator Mutation operator 创新性声明 秉承学校严谨的学风和优良的科学道德，本人声明所呈交的论文是我个人在导 师指导下进行的研究工作及取得的研究成果。尽我所知，除了文中特别加以标注 和致谢中所罗列的内容以外，论文中不包含其他人已经发表或撰写过的研究成果； 也不包含为获得西安电子科技大学或其它教育机构的学位或证书而使用过的材 料。与我一同工作的同志对本研究所做的任何贡献均已在论文中做了明确的说明 并表示了谢意。 申请学位论文与资料若有不实之处，本人承担一切的法律责任。 本人签名： 日期 关于论文使用授权的说明 本人完全了解西安电子科技大学有关保留和使用学位论文的规定，即：研究生 在校攻读学位期间论文工作的知识产权单位属西安电子科技大学。学校有权保留 送交论文的复印件，允许查阅和借阅论文；学校可以公布论文的全部或部分内容， 可以允许采用影印、缩印或其它复制手段保存论文。同时本人保证，毕业后结合 学位论文研究课题再攥写的文章一律署名单位为西安电子科技大学。 （保密的论文在解密后遵守此规定） 本学位论文属于保密，在 年解密后适用本授权书。 本人签名： 日期 导师签名： 日期 第一章 绪论 1 第一章 绪论 1.1 选题依据 图着色问