Neural-Based Adaptive Event-Triggered Tracking Control for Series Elastic Actuator with Input Dead Zones

This paper proposes a novel adaptive neural network control scheme to address the tracking control problem for series elastic actuators (SEAs) subject to dead-zone constraints in stochastic noise environments. Specifically, the controlled system investigated in this work is a single-input single-output (SISO) stochastic nonlinear system. Employing the traditional backstepping design approach to study such a system would significantly increase the computational burden. To resolve this issue, the command-filtered technique is integrated into the adaptive neural network design framework. Furthermore, a relative threshold event-triggered strategy is adopted to reduce the network communication burden to a certain extent. The proposed method ensures that the tracking error converges to a small neighborhood of the origin. Simulation results demonstrate the effectiveness of the proposed algorithm.