小波神经网络在无人机电机控制的应用

Application of Wavelet Neural Network for UAV Motor Control

为提高无人机轨迹控制,研究了一种新型无人机电机调速系统。对无位置传感器永磁同步电动机进行设计,采用了一种改进算法的自适应小波神经网络控制方法。使用新型梯度下降算法(Adma)优化训练后的网络,并且利用永磁同步电动机的相电压和转子电角度之间特定的非线性关系,建立无人机电机无位置传感器的小波神经网络控制系统。仿真结果表明,该系统能够有效控制电机换相,对无人机全程速度具有良好的静、动态控制效果。

In order to improve unmannel aerial vehicle’s trajectory control,a new drone motor speed control system was studied.The design of the position sensorless permanent magnet synchronous motor was carried out,and an improved wavelet neural network control method was introduced.The new gradient descent algorithm(Adma)was used to optimize the trained network,and the wavelet system of the UAV motor without position sensor was established by using the nonlinear relationship between the interphase voltage and the rotor angle of permanent magnet synchronous motor.According to the simulation results,the system can effectively control the commutation of the UAV motor and has a good static and dynamic control effect on the full speed of the UAV.