基于遗传算法优化的BP网络速度观测器的研究

根据三相异步电机的数学模型，提出了一种基于智能算法优化的速度观测器，以实现无速度传感器在直接转矩控制系统中的速度闭环控制。在通过BP神经网络训练得到的DTC系统的速度观测器的基础上，针对BP神经网络寻优参数多、易陷于局部极值以及初始设置对训练结果影响大等的不足，采用遗传算法对其进行优化设计。由仿真结果可知，用遗传算法优化后的BP神经网络较单纯的BP神经网络速度观测器具有更高的精度。

Research of speed observer based on BP neural network optimized by Genetic Algorithm

The new concept of speed identification based on intelligent algorithm has been put forward according to the model of triple-phase asynchronous motor in this paper, which is used to accomplish the closed-loop control of speed for sensorless in the system of direct torque control. Based on the speed identification of DTC system generated from the training of BP neural network, the bug of the overfull of optimization methodology of BP neural network parameter, getting into local extremum easily and slight effect for training result from initial setup, the BP neural network has been optimized through the genetic algorithm. It is acknowledged that the BP neural network optimized by genetic algorithm has a higher precision compared with the normal BP neural network.