基于RBF终端滑模观测器的电机转子位置估计

无刷直流电机常采用位置传感器来检测转子位置,这会影响系统的可靠性,增加电机体积和成本。采用无位置传感器控制技术:引入终端滑模面,其具有快速收敛性和良好观测精度,可减少相位滞后问题;采用RBF神经网络来设计观测器的控制策略,将滑模变量作为神经网络输入,输出即为控制策略,简化控制结构。RBF终端滑模观测器将RBF控制与终端滑模控制的优点紧密结合,优化了控制信号,削弱了抖振现象。仿真结果表明,该观测器能快速准确地估计电机的线反电势及电机转速,系统具有良好性能,满足无刷直流电机的工作要求。

ROTOR POSITION ESTIMATION BASED ON RBF TERMINAL SLIDING MODE OBSERVER

Brushless DC motor often uses position sensor to detect the rotor position,which will affect the reliability of the system and increase the volume and cost of the motor.Therefore,this paper adopts the sensorless control technology.We introduced the terminal sliding surface,which had fast convergence and good observation accuracy,and reduced the phase lag problem.RBF neural network was used to design the control strategy of the observer.The sliding mode variable was the input of the neural network and the output was the control strategy,which simplified the control structure.RBF terminal sliding mode observer combined the advantages of RBF control and terminal sliding mode control,optimized the control signal and weakened the chattering phenomenon.The simulation results show that the designed observer can quickly and accurately estimate the linear back-EMF and motor speed.The system has good performance and meets the requirements of brushless DC motor.