基于BP神经网络的开关磁阻电机直接转矩控制系统及实现

发布时间：2018-05-12 02:24

本文选题：开关磁阻电机 + 反向传播神经网络　；参考：《电力系统及其自动化学报》2017年01期

【摘要】：直接转矩控制可有效抑制开关磁阻电机(SRM)转矩脉动。由于开关磁阻电机双凸结构和磁路的严重饱和,造成其转矩是关于电流和转子位置的严重非线性函数,转矩计算非常困难。针对这一问题,本文提出一种采用基于BP神经网络建立开关磁阻电机转矩模型的方法。利用有限元仿真得到的转矩样本对BP神经网络经行进行离线训练,完成电流、位置角度到转矩的非线性映射,构造出基于BP神经网络的转矩观测器。再将构造好的转矩观测器应用于电机直接转矩控制系统中,对电机的转矩经行进行实时在线估算。最后,将估算转矩经行反馈,完成电机的直接转矩控制。该控制方法利用了BP神经网络泛化、逼近能力强的优点,同时控制过程简单,无需在线训练。实验结果表明,所提方法转矩计算速度快、计算精度高,可以满足实时控制的要求,有效地减小了电机的转矩脉动。
[Abstract]:Direct torque control (DTC) can effectively suppress the torque ripple of SRM. Due to the severe saturation of the double convex structure and magnetic circuit of switched reluctance motor, the torque is a serious nonlinear function of current and rotor position, so the calculation of torque is very difficult. In order to solve this problem, a method based on BP neural network is proposed to establish the torque model of switched reluctance motor (SRM). The torque sample obtained by finite element simulation is used to train the BP neural network off-line, and the nonlinear mapping of current, position angle to torque is completed, and the torque observer based on BP neural network is constructed. Then the torque observer is applied to the direct torque control system of the motor, and the torque line of the motor is estimated in real time and online. Finally, the torque feedback is estimated to complete the direct torque control of the motor. This control method takes advantage of the generalization of BP neural network and the advantage of strong approximation ability. At the same time, the control process is simple and there is no need for online training. The experimental results show that the proposed method has the advantages of fast torque calculation speed and high accuracy, which can meet the requirements of real-time control and effectively reduce the torque ripple of the motor.
【作者单位】：湖南大学电气与信息工程学院;湘潭市电机车厂集团有限公司;
【基金】：湖南省战略新兴产业科技相关类项目(2014GK1036)
【分类号】：TM352;TP273

【相似文献】