基于Baldwin效应的自适应有性繁殖遗传算法及其仿真研究

为了解决传统遗传算法存在的局部搜索能力差、早熟收敛、随机漫游等问题,提出一种基于Baldwin效应的自适应有性繁殖遗传算法。该算法模拟自然界普遍存在的有性繁殖现象,引入个体亲子间的适应度信息传递,并按照个体的性别差异以及种群间竞争与协作关系,对有性遗传操作进行自适应调整。同时,借鉴生物学中的Baldwin效应理论,引导个体后天正向或反向学习。通过与标准遗传算法和自适应遗传算法的仿真实验比较,证明了这种改进遗传算法的有效性。

Abstract：

Aimed to overcome the problems of traditional Genetic Algorithms,such as poor capabilities in local search,premature convergence,and stochastic roaming,a sexual reproduction adaptive Genetic Algorithm based on Baldwin effect was presented.The proposed algorithm simulated sexual reproduction and enables the transmission of fitness information from parent individuals to the offspring.Based on competition,cooperation,and innate difference between two sexual subgroups,the proposed algorithm adaptively adjusted sexual genetic operators.Meanwhile,inspired by the acquired reinforcement learning theory based on Baldwin effect,the proposed algorithm guided individuals to forward or reverse learning to adapt individuals＇ acquired fitness.Compared with Standard Genetic Algorithm （SGA） and Adaptive Genetic Algorithm （AGA） by numerical simulations,the proposed algorithm is proved more effective and advantageous.

Sexual Reproduction Adaptive Genetic Algorithm Based on Baldwin Effect and Simulation Study

Aimed to overcome the problems of traditional Genetic Algorithms,such as poor capabilities in local search,premature convergence,and stochastic roaming,a sexual reproduction adaptive Genetic Algorithm based on Baldwin effect was presented.The proposed algorithm simulated sexual reproduction and enables the transmission of fitness information from parent individuals to the offspring.Based on competition,cooperation,and innate difference between two sexual subgroups,the proposed algorithm adaptively adjusted sexual genetic operators.Meanwhile,inspired by the acquired reinforcement learning theory based on Baldwin effect,the proposed algorithm guided individuals to forward or reverse learning to adapt individuals＇ acquired fitness.Compared with Standard Genetic Algorithm （SGA） and Adaptive Genetic Algorithm （AGA） by numerical simulations,the proposed algorithm is proved more effective and advantageous.