Optimering av motorstyrning och strukturdesign i robotsystem för ledmekanismer : Positionsreglering av PMSM baserad på backstepping med hänsyn till glapp och konstruktion av flerfrihetsgradsmotorer

With the rapid development of the robotics industry, higher demands are placed on joint motor position control. Permanent Magnet Synchronous Motor (PMSM), commonly used in robotic joints, are often coupled with gearboxes to enhance torque density. However, gearboxes introduce backlash gaps that degrade positioning accuracy and cause limit cycle oscillations. While advanced control strategies, including AI-based methods, have been proposed, they often face practical limitations due to complexity and robustness issues. This work aims to develop a simple yet effective control strategy to mitigate backlash effects while ensuring practical feasibility. A PMSM model incorporating a dead-zone representation of backlash was developed using parameters measured from a real gearbox system. A conventional cascaded Field-Oriented Control (FOC) structure with PID controllers was first applied to the position, speed, and current loops. The impact of backlash nonlinearity on limit cycle oscillations was analyzed using the describing function method to predict oscillation frequency and amplitude. Subsequently, a backstepping-based control approach combined with a Shaft Torque Observer (STO) was implemented. Two configurations were investigated: in the two-step scheme, the position loop used PID control, while the speed and current loops employed backstepping laws; in the three-step scheme, backstepping laws were applied across all loops. Comparative studies were conducted on three control strategies, considering the effects of backlash and feedforward compensation. The results show that both backstepping-based strategies effectively suppressed oscillations and reduced position tracking errors. Notably, the backstepping approach exhibited inherent resilience to backlash disturbances, with the three-step scheme delivering the best overall performance. These findings demonstrate a backlash-resilient PMSM control strategy that features simple parameter tuning and ease of implementation. Simulation-based comparisons validate the robustness and practical applicability of the proposed design under realistic operating conditions.