摘要
针对同步电机参数辨识问题,建立了精确表示非同步采样及高次谐波在内的极值优化模型,利用改进混合遗传算法对该模型进行求解,为提高同步电机瞬态和超瞬态参数的精确辨识打下了良好基础。针对普通遗传算法收敛慢和经典迭代法初始点敏感问题,该改进混合遗传算法结合了全局寻优的遗传算法和局部寻优的模式搜索方法,不需要计算矩阵导数,可实现无需指定初值的电机参数快速求解。理论和仿真实验表明,该方法所需数据窗小,能有效提高参数测量的运行效率和计算精度。
A parameter extremum optimization model of synchronous machine is developed, in which such error factors as nonsynchronous sampling and harmonics are precisely depicted. In order to realize accurate identification of transient and subtransient parameters, a new measuring method based on hybrid genetic algorithm is improved and then applied into solve this extremum optimization model. In view of that the ordinary genetic algorithm converges slowly and the classic method is sensitive to initial guess, the improved hybrid genetic algorithm combines the global optimization of genetic algorithm and the strongly local search of pattern search method, consequently, do not need to calculate matrix derivative and quickly to solve problem without specifying the initial value of the parameters. Theory and simulation experiments show that the method requires a short data window and can effectively improve the measurement of the efficiency and accuracy.
出处
《电力系统保护与控制》
EI
CSCD
北大核心
2010年第1期51-55,共5页
Power System Protection and Control
基金
南京工程学院院级科研基金项目(KXJ08107)
关键词
参数辨识
同步电机
混合遗传算法
模式搜索
parameter identification
synchronous machine
hybrid genetic algorithm
pattern search
