感应电机参数辨识三种智能算法的比较  

Comparison of three intelligent optimization algorithms for parameter identification of induction motors

作　　者：陈振锋[1] 钟彦儒[1] 李洁[1] 

机构地区：[1]西安理工大学自动化与信息工程学院,陕西西安710048

出　　处：《电机与控制学报》2010年第11期7-12,共6页Electric Machines and Control

摘　　要：针对感应电机参数辨识,采用3种智能优化算法,即遗传算法、蚁群算法、微粒群算法。感应电机的实际输出电流和电气模型的观测电流之间的差值被作为目标函数不断对电气模型中的参数进行更新,从而辨识全部的感应电机参数。变速运行实验是在电机不带负载的情况下进行的。通过实验,对感应电机参数辨识3种智能优化算法进行比较,并得出结论。遗传算法可以得到最准确的电机参数,微粒群算法次之,蚁群算法最差。蚁群算法所需时间最短,遗传算法次之,微粒群算法最长。Genetic algorithm,ant colony optimization and particle swarm optimization were introduced and applied to the parameter identification of an induction motor for vector control.The errors between the actual stator current output of an induction motor and the stator current output of the model were used as the criterion to correct the model parameters,so as to identify all the parameters of an induction motor.Experiments were conducted on speed-varying operation with no load.The three kinds of optimization algorithms were compared with each other and conclusions were summarized.Genetic algorithm can acquire the most accurate parameters of induction motor.Particle swarm optimization takes second place,and ant colony optimization is the worst.In computing time,the ant colony optimization is the fastest.Genetic algorithm takes second place,and particle swarm optimization runs the most slowly.

关 键 词：感应电机 矢量控制 参数辨识 智能优化算法 

分 类 号：TU313[建筑科学—结构工程]