改进型遗传算法在机组负荷优化组合中的应用

Optimization of unit commitment of marine power system using improved genetic algorithm

中文关键词  遗传算法  优化分配  惩罚函数

英文关键词  genetic algorithm  unit commitment  penalty function

中文摘要

      针对某舰用电站机组负荷组合的特点, 提出了一种基于浮点数和二进制数统一编码的改进遗传算法, 并进一步对算法中的编码解码方式、初始种群生成、约束条件处理、遗传算子映射及控制参数调节等作了改进, 解决了机组优化组合的0--1混合整数非线性规划问题. 改进后的算法不仅较好地处理了机组优化组合中的各种约束条件, 同时改善了算法的收敛性. 优化结果表明机组油耗率降幅最大可达2%, 效果显著.

英文摘要

      In accordance with the characteristics of the unit commitment problem in marine power system, we propose an improved genetic algorithm with both float-point coding and binary coding. To apply the nonlinear 0-1 mixed integer programming to the optimization of the unit commitment, other items in the algorithm are modified accordingly, including the coding and decoding modes, initial population creation, constraint conditions disposal, fitness function selection, genetic operators and the control parameters modulation. In the application of this improved algorithm, not only the constraint conditions can be handled more readily, but the convergence speed and the solution precision are also improved. The application advantage is demonstrated by a 2% reduction in average oil consumption rate.