Adaptive Learning-Based Optimal Prescribed Performance Tracking Error Measurement of Hydraulic Support Initial Support Force

This article addresses the challenge of measuring and controlling the initial support force of hydraulic supports, which can be complicated by pressure fluctuations and losses in the hydraulic supply system. To this end, an online tracking error measurement method is presented based on optimal control, which involves developing a fluid supply system model and integrating the tracking error with the reference trajectory to streamline controller design. Our approach uniquely combines adaptive dynamic programming (ADP)-based optimal control with a prescribed performance function (PPF), effectively constraining tracking error and adapting to system uncertainties in real-time. A novel online adaptive learning law is proposed to estimate the unknown critic neural network (NN) weights, thereby eliminating the need for an actor NN and enhancing control performance. Simulation results demonstrate the developed approach achieves superior accuracy and faster response times compared to existing methods.