Abstract
This study addresses the common issues of slow dynamic response and significant hysteresis in nonlinear electro-hydraulic systems, with a particular focus on optimizing the control performance of shotcrete robots operating under complex working conditions. The electro-hydraulic proportional control system was first designed and mathematically modeled. Based on input-output data collected under actual operating conditions of the shotcrete manipulator, signal preprocessing was performed using a wavelet soft-threshold denoising algorithm. Subsequently, system parameters were identified using a particle swarm optimization (PSO) algorithm enhanced by least squares estimation, resulting in a high-accuracy system transfer function. To overcome the limited robustness of traditional PID controllers under strong nonlinearities, as well as the real-time computational burden of standalone model predictive control (MPC), a dual-loop control strategy was proposed-employing PID feedback as the inner loop and MPC as the outer loop-to optimize and simulate the electro-hydraulic system. Experimental validation was conducted on a six-degree-of-freedom shotcrete robot platform through extension and rotational motion control tests of the robotic boom. Results show that, compared to conventional PID control, the proposed PID-MPC approach significantly improved system responsiveness and tracking accuracy. Specifically, in the extension test, the maximum tracking error decreased from 0.13 m/s to 0.04 m/s, and the maximum settling time was reduced from 3.0 s to 0.45 s; in the rotational test, the maximum tracking error dropped from 8°/s to 4°/s, and the maximum settling time shortened from 2.6 s to 0.8 s. This study offers a practical and effective solution for accurate modeling and high-performance control of complex electro-hydraulic systems, providing a solid theoretical foundation for the development and engineering application of intelligent and efficient shotcrete equipment.