基于RBF神经网络的无刷直流电动机速度无关控制法

为提高无刷直流电动机无位置传感器控制精度,提出了一种基于RBF神经网络的无刷直流电动机速度无关控制新策略。该策略主要包含两个部分:一方面,利用RBF神经网络的自适应、非线性控制等优良性能,结合电机运行状态,修正神经元之间的连接权值,从而可以克服由于无刷直流电动机系统的非线性和部分参数不确定性造成精度下降的负面影响。另一方面,神经网络的输出经过滤波处理后,采用速度无关位置函数法(函数法)输出电机换相信号。该方法在转子转速由近零到高速变化的过程中,都能够对转子位置进行检测并给出换相时刻。仿真和实验表明,该策略具有优良的控制性能。

A New Speed-Independent Method of Position Function Control Based on RBF Neural Network for Brushless DC Motor

A new speed-independent position function control method based on RBF neural network for brushless DC motor(BLDCM) was proposed in order to improve the performance of the BLDCM position sensor less control.This method contained two parts.On the one hand,taking advantages of RBF neural network’s self-adaptation and nonlinear control,together with operation situation of the motor,the connection weight of neural network was revised to overcome negative effects,due to nonlinear and part parameters’ uncertainty of the BLDCM system.On the other hand,after neural network’s output was filtered,it was converted to commutation signals by the speed-independent position function.When the rotor speed was changed from almost zero to high,this function can detect the rotor position and then give commutation signals.Simulation and experimental results indicate that the proposed method has good performance in control of BLDCM.