针对混合极性的并行表格技术的遗传算法

Optimization of Mixed Polarity Functions Using Genetic Algorithm with Parallel Tabular Technique

针对混合极性的最佳极性优化问题,提出一种基于并行表格技术的遗传算法.在3n混合极性搜索过程中,采用并行表格技术计算遗传算法中种群的适应度函数;并行表格技术不按变量顺序产生on-set项,克服了在传统表格技术中顺序产生相关项造成数据相关性问题,有效地提高了CPU利用率.实验结果表明文中算法在保证最优结果的同时,可平均缩短8%的处理时间.

This paper presents a genetic algorithm （GA） using parallel tabular technique for the optimization of mixed polarity functions. The algorithm is to find optimal solution among 3n different solutions for large functions. To overcome the slow convergence of GA, the calculation of the cost function is based on parallel tabular technique, in which new on-set terms are generated at one time instead of generating in sequence. As a result, the correlation between newly generated terms and previously generated terms is avoided. Experimental results show that, the proposed algorithm is efficient in terms of CPU time and achieves 8% improvement on average, without generating all the possible 3n polarities.