基于实验和遗传算法的无刷双馈电机参数估算

Parameter Estimation of Brushless Double Fed Machine Based on Experiments and Genetic Algorithm

等效电路是分析无刷双馈电机(brushless double fed machine,BDFM)同步运行性能的有效工具,但是缺乏准确的参数测量方法。在等效电路的基础上将电压和电流向量分解为实部和虚部,建立起无刷双馈电机同步运行时的非线性数学模型,并在此基础上提出一种将实验数据和遗传算法相结合的参数估算方法。将等效电路中的参数作为一组决策变量,求解不同运行状态下的数学方程,将计算值与测量值之差作为目标函数,通过遗传算法,对决策变量进行优化,使计算值与测量值之差最小。给出了参数估算的流程并编程实现,计算结果表明算法合理。

The equivalent circuit of brushless double fed machine（BDFM） is an effective tool in the analysis of synchronous operation performance, but lacking accurate parameter measurement method. A nonlinear mathematical model based on equivalent circuit was established by decomposing voltage and current vector into real and imaginary parts, and a method of parameter estimation combing experiment data and genetic algorithm was presented. The parameters of equivalent circuit were treated as a group of decision variables to solve the mathematical equation in different operating conditions. The difference between calculated and measured values was considered as a target function. The genetic algorithm was employed to optimize the decision-making variables and minimize the target function. The estimation processes were given and programmed. The calculated results show that the method is feasible.