Abstract: In order to improve the performance of genetic algorithm，an improved genetic algorithm based on cultural algorithm framework is developed to be applied to the function optimization.The improved genetic algorithms are embedded into cultural algorithm framework and compose population space and belief space.In population space，it introduces a random population to extend search area and parts of the worst individuals are organized to crossover with part of best individuals that the belief space provides in probability.A dissipative system theory is used in genetic algorithm so as to tapping the high guidance of the whole population and regulation potential of self-organizing to enhance accuracy and efficiency.The performance of the proposed improved generic algorithm is evaluated by a number of test functions.Experimental results show that the algorithm can be efficiently applied to the function optimization.

Key words: cultural framework, genetic algorithm, dissipative structure

摘要： 为了提高遗传算法的性能，将遗传算法纳入到文化算法框架中组成群体空间和信念空间，提出一种新的优化算法。在群体空间的遗传进化过程中引入随机种群来增加算法的勘探能力，并组织较差个体依概率与信念空间中更新后的优秀个体进行交叉操作；在信念空间充分利用对优秀个体所包含信息的开采能力并采用耗散结构来提高整个空间的自组织能力，更新优秀个体，在很大程度上提高了算法的速度和效率。实验结果表明，新算法能有效地应用于函数优化。

关键词: 文化框架, 遗传算法, 耗散结构