摘要
针对标准遗传算法容易"早熟"的缺陷,提出聚类遗传算法;改进了选择算子和交叉算子,并利用Shubert多峰测试函数验证了聚类遗传算法的优势.引入水轮机组运行效率梯度变化因素提出改进变异算子,弥补了变异搜索过随机的缺陷.最后,将改进方式应用于三峡-葛洲坝梯级水利枢纽短期电力调度优化研究中,提出和构建了相应的优化模型以及机组组合启停和运行效率同步实现策略.实例优化结果表明:聚类遗传算法和改进变异算子能有效弥补"早熟"的缺陷,并能显著提高优化搜索效率,适用于梯级电站电力调度优化问题.优化得出的梯级电力调度方案可以满足设定目标和约束,并提高了梯级的发电效率.
In view of the premature weakness of standard genetic algorithm,a cluster genetic algorithm has been put forward to improve the selection and crossover operators,and its advantages has been testified by Shubert multimodal function. Based on the change gradient of unit＇s operation efficiency,improved mutation operator has been proposed to deal with excessive randomization. All the above improvements have been applied to short term hydropower dispatching optimization of Three Gorges and Gezhouba cascaded hydropower junctions and the optimization model and synchronization optimization strategies have been developed. Real case optimization results show that cluster genetic algorithm and improved mutation operator can reduce the premature possibility and enhance the optimization efficiency. Hydropower dispatching solutions based on such improvements can meet all the objectives and constraints,and also increase the power generation efficiency.
出处
《天津大学学报》
EI
CAS
CSCD
北大核心
2010年第1期1-8,共8页
Journal of Tianjin University(Science and Technology)
基金
"十一五"国家科技支撑计划资助项目(2008BAB29B09)
国家杰出青年科学基金资助项目(50725929)
关键词
聚类遗传算法
机组运行效率变化梯度
改进变异算子
短期电力优化
三峡-葛洲坝梯级水利枢纽
cluster genetic algorithm
change gradient of unit＇s operation efficiency
improved mutation operator
short term hydropower dispatching optimization
Three Gorges and Gezhouba cascaded hydropower junctions
