论文部分内容阅读
针对传统克隆选择算法在变异时存在的盲目性和随机性而导致的退化现象以及容易陷入局部最优等问题,本文以生物免疫机制中抗体是由稳定区和可变区构成为理论支撑,利用粗糙集中的核值概念,提出了一种基于粗糙集核值的克隆选择算法.该算法用核值构造稳定区,使每一代优秀抗体的核信息在成熟的过程中保持不变,只对可变区进行变异操作,使得变异具有稳定性和向最优解靠近的方向性.实验部分采用经典的测试函数对两种算法的性能进行测试对比.实验结果表明该算法在收敛速度、抗体多样性以及避免早熟等方面均比传统克隆选择算法具有更好的效果,且算法在迭代后期还具有较强的局部搜索能力.
Aiming at the problems of the blindness and randomness of the traditional clonal selection algorithm in mutation and the problem of being easily trapped in the local optimality, this dissertation is based on the theory that the antibody is composed of stable region and variable region in biological immune mechanism. This paper proposes a clone selection algorithm based on the kernel value of rough set, which uses the kernel value to construct the stable region so that the nuclear information of each generation of excellent antibody remains unchanged in the mature process, only variable Region to make the mutation have the stability and direction to the optimal solution.The experimental part uses the classic test function to test the performance of the two algorithms.Experimental results show that the algorithm in the convergence rate, antibody diversity and Avoid precocious and so on than the traditional clonal selection algorithm has better results, and the algorithm also has strong local search ability in the late iteration.