ABSTRACT Aiming at the inherent contradiction that fast dynamic response, steady‐state accuracy, and robustness are difficult to balance in speed control of permanent magnet synchronous motors (PMSM), this paper addresses the problem that fast convergence, steady‐state error elimination, and chattering suppression cannot be simultaneously achieved in PMSM speed control. An improved speed control strategy combining Improved Terminal Integral Sliding Mode Control with Adaptive Exponential Reaching Law (ITISMC‐ADERL) is proposed. The designed sliding surface couples the high‐order convergent term of terminal sliding mode with a smoothed nonlinear integral term: In the large‐error region, it is dominated by the high‐order terminal term to achieve finite‐time fast approximation; in the small‐error region, cumulative compensation of the smoothed integral term plays a leading role, which eliminates steady‐state errors while suppressing integral saturation and high‐frequency chattering. The proposed reaching law adaptively adjusts the power exponent and gain according to the error magnitude and its rate of change, realizing dynamic balance between “far from sliding surface—fast convergence” and “close to sliding surface—low chattering” through power exponent switching. Theoretically, the global stability of the closed‐loop system is proven based on the quadratic Lyapunov function, and the global and local upper bounds of finite‐time convergence are derived, providing a theoretical basis for parameter tuning. To verify the effectiveness of the method, validation experiments are carried out on a DSP platform with a real surface‐mounted PMSM under complex operating conditions including no‐load start‐up, speed up/down, sudden load application/removal, and flux linkage disturbance. The results show that the proposed method significantly outperforms traditional sliding mode and improved methods in key indicators such as dynamic response, steady‐state accuracy, robustness, and disturbance rejection. It integrates theoretical completeness and engineering practicability, making it suitable for high‐precision servo drive scenarios.
He et al. (Tue,) studied this question.