基于NNBPF-EKF的内置式永磁同步电机死区补偿方法

IPMSM Dead Time Compensation Method Based on NNBPF-EKF

针对电动汽车用内置式永磁同步电机控制方案中,由死区效应引起的电驱动系统电流谐波增加及电压损失率高等问题,提出一种基于神经网络带通滤波器和扩展卡尔曼滤波器相结合的死区补偿方法。首先,使用扩展卡尔曼滤波器对交直轴电流滤波,改善传统死区补偿算法中零电流处电流极性判别困难的情况。其次,利用基于神经网络的带通滤波器提取交直轴电流中的谐波成分,并将提取到的谐波调节后进行死区补偿,进一步提高死区补偿控制的性能。最后,通过与无死区补偿和传统死区补偿方法的对比,验证所提死区补偿方法的正确性和有效性。实验结果表明,提出的死区补偿方法能有效地抑制电流谐波、减少电流畸变,并将电压损失率降低至0.04%,进而提升电动汽车的控制性能,使其续航里程得到保证。

Aiming at the problems of increased current harmonics and high voltage loss rate of the electric drive system caused by the dead time effect in the control scheme of the interior permanent magnet synchronous motor(IPMSM)for electric vehicles,a neural network bandpass filter(NNBPF)and extended Kalman filter(EKF)were proposed.First,the extended Kalman filter was utilized to filter the direct and quadrature current to improve the difficulty of judging the polarity of the current near-zero current in the traditional dead time compensation algorithm.Second,a neural network based on a bandpass filter was used to extract the harmonic components of the direct and quadrature current.Dead time compensation was performed after the extracted harmonics were adjusted to further improve the performance of the dead time compensation control.Finally,the correctness and effectiveness of the proposed dead time compensation method were verified by comparison with no dead time compensation and traditional dead time compensation methods.The experimental results show that the proposed dead time compensation method can effectively suppress current harmonics,reduce current distortion,and decrease the voltage loss rate to 0.04%,thereby improving the control performance of electric vehicles and ensuring their mileage.