一种改进遗传算法在超临界电厂动态负荷优化分配中的应用研究  

Application of Accelerating Genetic Algorithm in Dynamic Load Optimal Distribution for Supercritical Power Plant

作　　者：李云娟[1] 张立明[1] 黄钺[1] 

机构地区：[1]昆明学院自动控制与机械工程系,云南昆明650118

出　　处：《现代电子技术》2010年第21期147-150,共4页Modern Electronics Technique

摘　　要：考虑到超临界电厂机组实际运行情况及实际约束条件,建立了机组多时段动态负荷优化分配的数学模型。在传统遗传算法(SGA)的基础上对其进行改进,首次提出了基于实数编码、双倍初始种群生成法、变参数法和压缩空间法的加速遗传算法(AGA)。实例仿真分析说明,AGA能稳定地收敛到全局最优解,时间短,效率高,较SGA性能有了显著的提高,在一定程度上能够满足动态负荷优化分配对算法运算速度的要求。Taking into account of the practical operation and practical constraint of generating units in supercritical power plants,an algorithm model of multiperiod dynamic load optimal distribution of power generating units is established.Based on the modification of the simple genetic algorithm（SGA）,the accelerating genetic algorithm（AGA） is proposed according to real coding,double initial population generation method,variation method of parameters and method of space compression.The simulation of an example shows that the AGA can converge stably to the global optimal solution.The convergence time is short and its efficiency is high.In comparison with the performance of SGA,that of AGA was improved significantly,and can meet the requirement of the dynamic load optimal distribution to the algorithm operation speed to some extent.

关 键 词：动态负荷优化分配 加速遗传算法 实数编码 变参数法 

分 类 号：TN919-34[电子电信—通信与信息系统]