基于神经网络的内置式永磁电动机弱磁区最佳电流相位曲线拟合

针对电传动系统中内置式永磁(IPM)电动机内部磁场的非线性问题，提出采用BP神经网络拟合电动机弱磁区的最佳电流相位曲线。分析了IPM电动机每安培电流最大转矩/电流控制的原理，并进一步在考虑逆变器输出电压和电流限制的基础上，指出了IPM电动机在弱磁区产生最大转矩的条件。构建了用于完成IPM电动机弱磁区转速和电流幅值与最佳相位角之间非线性映射的BP神经网络。计算和仿真结果表明，该神经网络的拟合结果与有限元方法（FEM）计算结果的最大误差不到1°.

Optimal Current Phase Curve Fitting of IPM Motors in Flux-Weakening Region Based on Neural Network

To take the nonlinear effect of interior permanent magnet(IPM) motors into account,a method of fitting the optimal phase curve in flux-weakening region by BP neural network is presented.The theory of maximum torque per ampere(MTPA) control is analyzed.Considering the voltage and current limits of power inverter,the condition for IPM motors to output maximum torque in flux-weakening region is derived.A BP neural network is constructed to map the motor speed and current amplitude to optimal phase angel in flux-weakening region.Computation and simulation results show that the maximum error between the fitting result of the constructed BP neural network and the result of finite element method(FEM) is less than 1°.