摘要
提出了利用遗传算法求解泵站优化运行问题的方法。同时,引入SQP局部搜索应用于遗传算法中,对随机生成的初始解集进行搜索改进,提高了算法的稳定性。对优化变量采用整数编码,相比常用的二进制编码,缩短了编码长度,加快了算法收敛速度。经过实例计算并与同等离散情况下的动态规划法的计算结果比较,基于SQP局部搜索和整数编码的遗传算法计算结果较优,计算用时较短。这表明该方法在泵站优化运行以及相近领域有较高的实用价值。
Developed a method of solving optimal pump scheduling problem with GA.Improved initial solutions of GA with SQP local search,which can increase the stability of GA.Integer coding is used for dealing with decision variables of GA,the convergence rate of GA is accelerated for the length of integer coding is much shorter than the length of binary coding.An calculation example was conducted,the result of SQP-GA was compared with the result of dynamic programming under the same discretization of decision variables.The result of SQP-GA is better and the computing time is short which indicates that SQP-GA can provide a high application value to the field of optimal pump scheduling problem and related fields.
出处
《农机化研究》
北大核心
2013年第9期16-19,共4页
Journal of Agricultural Mechanization Research
基金
国家自然科学基金项目(51279173)
江苏省水利动力工程重点实验室开放项目(K100019)
关键词
泵站优化
遗传算法
SQP局部搜索
optimal pump scheduling problem
genetic algorithm
sequential quadratic programming
