演化求解高维函数优化问题的独立编码策略

演化求解高维函数优化问题的独立编码策略 THE INDEPENDENT ENCODING STRATEGY FOR EVOLUTIONARY OPTIMIZING OF HIGH D IMENSIONAL FUNCTION 徐静雯 Xu Jingwen 1 2 first-author 李元香 Li Yuan-clang 1 2 陈子仪 Chen Ziyi 1 2 武汉大学软件工程国家重点实验室,武汉 武汉大学软件工程国家重点实验室,武汉 State Key Laboratory of Software Engineering, Wuhan University State Key Laboratory of Software Engineering, Wuhan University ,430072 ,430072 在用演化算法解高维函数优化问题时，一般采用传统的二进制整体编码策略．本文将函数的自变量独立分组，分别进行编码和演化操作，以突破传统二进制整体编码策略的束缚，将这种分组编码策略称为独立编码策略．同时，讨论了独立编码串行实现的灵活性，提出了一个基于变量分组的并行演化计算模型．大量的数值实验结果表明：在问题求解的精确度和求解速度方面，独立编码策略都优于传统的二进制整体编码策略． in this paper, we propose an independent encoding strategy which makes a breakthrough of the traditional binary encoding strategy in evolutionary optimizing of high--dimensional functions. By this new strategy, variables are divided into independent variable groups, then in each group variables are encoded and operated. Evolutionary algorithms based on this strategy can be flexibly implemented on serial machine, furthermore this strategy suggests a new parallel model of evolutionary computation. Numerical experimental results show that the independent encoding strategy is superior to ordinary binary encoding strategy not only in accuracy but in time as well. 独立编码 二进制整体编码 高维函数优化 演化算法 并行处理 independent encoding binary integral encoding high-dimensional function optimization evolutionary computation parallel algorithm O224 国家自然科学基金(61672102) 1998-05-01 2021-04-01 5