Control of a compliant gripper via least-squares support vector regression (LS-SVR) with particle swarm optimization (PSO) algorithm

This study focuses on controlling a compliant gripper using least-squares support vector regression (LS-SVR) combined with a particle swarm optimization (PSO) algorithm. The compliant gripper is designed to grip small objects with high precision. However, repeated use can lead to reduced precision due to the hysteresis inherent in the gripper’s mechanism. To address this, an algorithm developed to mitigate the effect of hysteresis is seen to improve control accuracy. The algorithm is further designed to control the end-effector position of the gripper. Simulation results show that applying LS-SVR with the PSO algorithm enhances gripping precision. After implementing the control algorithm, gripping displacement was compared across three configurations: no controller, a conventional proportional–integral (PI) controller, and the proposed LS-SVR with PSO. The root mean square error (RMSE) decreased significantly to 5.13 × 10⁻² mm with the proposed controller, compared to 7.03 × 10⁻² mm without a controller and 6.69 × 10⁻² mm with PI control. These results demonstrate that the LS-SVR with PSO significantly improves the precision of the compliant gripper.