Duty-Based Dumping Control for Underground Load- Haul-Dump Vehicle Based on Reinforcement Learning
Abstract
Underground load-haul-dump (LHD) vehicles, with the excellent maneuverability and efficiency, are widely used to the materials removal in underground mines and constructions.Owing to the harsh operation environment, the autonomous functions of LHDs have been urgently needed for the consideration of safety, efficiency, and cost.Dumping is one of the core functions for underground load-haul-dump (LHD) vehicles.Typical underground environment including narrow space and varying dumping locations restricted the unconstrained autonomous dumping.This paper probes into the autonomous dumping control of LHDs under narrow and constrained environment based on the reinforcement learning (RL) frame.Firstly, the kinematic and dynamic models of a LHD are established considering the data-dependency of the RL method.The duty-based RL control method, is proposed to realize the coordinative control of the driving system and the working system in dumping duty, in which the safety in avoiding collision and working efficiency were considered.A test platform was established with the LHD to validate the methods.For the RL-based method, the performance of different algorithms was evaluated.A typical dump duty was performed to validate the method.The results shown that the method could finish the dumping operation autonomously under constrained environment.The duty-based RL control method shown great coordinative characteristics and environmental adaptability with the RL frame under constrained environment.
Authors 4
-
University of Science and Technology Beijing
Affiliation as printed
School of Mechanical Engineering , University of Science & Technology Beijing , Beijing , 100083 , China
-
Affiliation as printed
ifas , RWTH Aachen University , Campus-Boulevard 30 , 52074 Aachen , Germany
-
Affiliation as printed
ifas , RWTH Aachen University , Campus-Boulevard 30 , 52074 Aachen , Germany
Cited by 0 stored of 0
No patents citing this paper on Lens.org (checked 2026-10-06).