车载飞轮电池用无轴承永磁同步电动机神经网络控制

Neural Network Control of Bearingless Permanent Magnet Synchronous Motors for Flywheel Battery of Vehicles

针对电动汽车磁悬浮飞轮储能电池用无轴承永磁同步电动机(BPMSM)的非线性、强耦合、参数时变等特性,将单神经元PID和径向基函数神经网络(RBFNN)相结合,提出一种新型车载飞轮电池用BPMSM神经网络控制策略。采用RBFNN辨识BPMSM输入输出变量之间的关系,构建BPMSM系统的参考控制模型,通过MATLAB仿真软件对其进行仿真验证,结果表明,该控制方法可显著提高系统的静态和动态性能。

Aiming at the characteristics of bearingless permanent magnet synchronous motor （BPMSM） for magnetic suspended flywheel energy storage battery of electric vehicles, such as nonlinearity, strong coupling and time - varying parameters, a novel neural network control strategy which consists of single neuron PID and radial basis function neural network （RBFNN） is presented for BPMSM for flywheel battery of vehicles. The relation between input and output vari- ables for BPMSM is identified by using RBFNN, and the reference control model is constructed for BPMSM system. The simulation verification is carried out by MATLAB simulation software. The results show that the control method is able to significantly improve static and dynamic performance of system.