论文部分内容阅读
Lamarck学习理论已被引入进化计算,能有效提高其局部搜索能力,逐步发展成为进化计算的新热点-Memetic计算.文中从神经系统与免疫系统在生物机体内的整合调节机理,提出了免疫Memetic计算模型,设计了模拟神经系统对免疫反应单向调节的Lamarck学习策略,并针对数值优化问题,提出了基于Lamarck学习的免疫Memetic算法.该算法结合了免疫算法和传统数学规划算法的不同特性,具有较理想的搜索性能.基于10个低维和10个高维基准测试问题的仿真结果表明,基于Lamarck学习的免疫Memetic算法与基于遗传算法的基本Memetic算法相比具有明显的优越性.
Lamarck learning theory has been introduced into evolutionary computation, which can effectively improve its local search ability and gradually evolve into a new hot spot in evolutionary computation - Memetic computation. In this paper, immune Memetic computation is proposed from the integration and adjustment mechanism of nervous system and immune system in biological organism Model, a Lamarck learning strategy is proposed to simulate the unilateral modulation of the immune response by the nervous system. In order to solve the numerical optimization problem, an immune Memetic algorithm based on Lamarck learning is proposed. This algorithm combines the different characteristics of the immune algorithm and the traditional mathematical programming algorithm, The simulation results based on 10 low-dimensional and 10 high-dimensional benchmark test problems show that the immune Memetic algorithm based on Lamarck learning has obvious advantages over the basic Memetic algorithm based on genetic algorithm.