PSO-NMPC control strategy based path tracking control of mining LHD (scraper)

Abstract

The automation of underground articulated vehicles is a critical step in advancing digital and smart mining. Current nonlinear model predictive control (NMPC) controllers face challenges such as delays in turning on large curvature paths and correction lags during the control of underground the Load-Haul-Dump (LHD). To address these issues, this paper proposes a PSO-NMPC control strategy that integrates a particle swarm optimization algorithm (PSO) into the NMPC controller to enhance path tracking for LHDs. To verify the effectiveness of the proposed PSO-NMPC control strategy, the local path of the tunnels is selected as the simulation path, comparing it with the pure NMPC controller based on the path characteristics of the actual tunnel. The results demonstrate that the improved NMPC controller significantly enhances the trajectory tracking performance of the LHD, with maximum absolute lateral deviations for experimental paths 2, 3, and 5 improved by 89.7%, 72.2%, and 68.9%, respectively. Additionally, the improved NMPC controller exhibits superior performance in paths with large curvature compared to those with very small curvature and straight-line paths, effectively addressing the challenges of turn delay and backward lag in LHD operation, thus providing practical significance.