Feedback switching adaptive control for trajectory tracking in walking piezoelectric actuator

The nonlinear characteristics and continuous foot switching in Walking Piezoelectric Actuator (WPA) necessitate an applicable control scheme for practical application. This paper proposes a Feedback Switching Adaptive Control (FSAC) based on Adaptive Proportional-Integral-Derivative (APID) and aligned with the WPA driving principles for trajectory tracking control. A Quantum Particle Swarm Optimization-based Neural Network (QPSNN) is employed for online tuning of APID control gains. The APID based on QPSNN eliminates the need to model nonlinearity and foot switching, dynamically adjusts control parameters in real time. The FSAC incorporates WPA driving principles with APID to enable precise tracking control under foot-switching conditions. The stability of the proposed APID is theoretically analyzed using a Lyapunov framework. Comparative experiments under different trajectories and varying loads were conducted on the self-designed WPA stage. The proposed FSAC is verified the superior tracking performance and strong variable load adaptability in experiment.