基于闭环自适应辨识的速度环PI参数自整定

针对永磁同步电机伺服系统速度环比例积分（PI）参数整定过程中需要反复调节、效率低等问题，提出了一种基于闭环自适应卡尔曼滤波（AKF）系统辨识的伺服系统速度环PI参数自整定方法。首先根据输入信号激励速度闭环系统，分析不同频率激励作用下闭环辨识序列的信噪比与实际输出，然后引入AKF算法辨识闭环被控对象的离散模型，最后通过遗传算法仿真搜索最优速度环PI参数。仿真与实验结果表明：该算法能有效抑制量测噪声等扰动对系统辨识精度的影响，辨识结果能够反映实际系统的动态输入输出特性，优化后的速度环具有优良的响应性能和较高的精度，便于实际工业应用。

Speed PI Parameter Auto-tuning Based on Closed-loop Adaptive Identification

In order to solve the problems of speed PI parameter tuning process needed to be adjusted repeatedly or had low efficiency for PMSM servo system, a practical method of speed PI parameter auto-tuning was proposed based on the closed-loop AKF system identification. Though speed closed-loop input excitation, the signal-to-noise ratio and actual output of closed-loop identification sequence under different frequencies were analyzed, then the discrete model of closed-loop controlled objects was identified by AKF algorithm, and finally by genetic algorithm the optimal speed PI parameters were searched through simulation. Simulation and experimental results show that the presented algorithm can effectively suppress the influences of the measurement noise disturbances on system identification accuracy, and the recognition result can reflect the dynamic input-output characteristics of the actual system, moreover, excellent response and high accuracy have appeared after speed optimization and it is convenient to the practical industrial applications.