State Estimation Based Model Predictive Control for LHD Vehicles

LHD (load-haul-dump) vehicles are used extensively in underground mining operations for ore transporting, primarily in tunnels where access is difficult or dangerous. To ensure underground efficient and safe LHD’s performance, a robust feedback control strategy is needed. A state estimation based MPC scheme was designed for control purposes, and evaluated by simulation. The state estimator was developed by testing four approaches, in order to select the optimal one: the extended Kalman Filter, Particle Filter, Moving Horizon Estimator and a genetic algorithm based Moving Horizon Estimatior. The simulation shows that non-linear MPC performs better than linear MPC for path tracking.