【摘要】 在借鉴生命科学中免疫概念与理论的基础上 ,提出了一种新的算法——免疫规划算法 .该算法的核心在于免疫算子的构造 ,而免疫算子又是通过接种疫苗和免疫选择两个步骤来完成的 .理论证明免疫规划是收敛的 .文中给出了免疫疫苗的一种自适应选取与构造方法 .仿真结果表明该算法不仅可以有效地抑制原算法在进化过程中所出现的退化现象 ,而且在搜索能力和自适应性能方面有较大改善 ,进而提高收敛速度 .更多还原

【Abstract】 A novel algorithm, the immune programming——IP, is proposed based on the immune theory in biology, which constructs an immune operator accomplished by two steps, a vaccination and an immune selection. In IP, the local information is employed to intervene in the globally searching process to a certain extent in order to restrain or avoid repetitive and useless work in the operating course, which can overcome the blindness of the original operators in EP, refrain the degenerative of evolutionary process and increase the fitness of population steadily. The immunity possesses memory property that can identify a population or an individual that has appeared previously. IP can be proved to be convergent with the theory of random process and a method of adaptively selecting and constructing a vaccine is also given in this paper. The simulation test shows that IP can not only evidently alleviate the undulate phenomenon during the evolutionary process, but also improve the searching ability and the adaptability greatly, therefore increases the convergent speed.更多还原