Research On Path Tracking And Obstacles Avoidance For Autonomous Underground Mining Articulated Vehicles

Articulated vehicles are important equipment in underground mining industry because of its flexible,which are widely used in the process of loading and transport process of underground ore.Implementation of mental non-mental mine"Unmanned and mechanized transportation" is China’s development strategy in mining industry.Path tracking and intelligent obstacle avoidance for articulated vehicles are the key technology to realize the unmanned mining equipment.This paper focus on path tracking and obstacle avoidance control for articulated vehicles,in order to provide technical support to achieve self-driving of articulated vehicles.First of all,the dynamics characteristics of the articulated vehicles are analyzed,and the articulated vehicle dynamics model is built using modular modeling method,which includes tire model,wheel model,auxiliary computing module and vehicle body model.The auxiliary computing module is used to estimate the driving state of the vehicle,which contains the tire slip rate,the tire slip angle,the vertical load and coordinate transformation of the tire.The 4 DOF vehicle body model is built based on the D’ Alembert principle.And the vehicle dynamics model are integrated with the above modules,which are used to validate the proposed control method as a simulation platform.Then,a speed planning method and an integrated control method is proposed for the path tracking control of articulated vehicles.The kinematics characteristics of the vehicle is analyzed before speed planning,through which the relationship between the curvature of the path and the vehicle speed is built to maintain the lateral stability of the vehicle.The function of speed planning is to obtain the vehicle speed constraints under different path conditions.The sliding mode control(SMC)and weighted adaptive PI control are applied to control the vehicle speed and steering respectively,which improve the adaptability and robustness of the controller.Then,the model predictive controller(MPC)is designed using the vehicle dynamics model to achieve the path tracking for the articulated vehicles.The vehicle dynamics model is linearized using the Jacobian linearization method,and the iterative equation for the model predictive control.The output constraints are added in the cost function,referring to the output of the real vehicle,which makes the output of the controller more realistic.The proposed MPC controller is validated through simulation,whose results satisfy the performance requirements.It also provide a way for the application of MPC to the nonlinear systems.The relation space method(RSM)is proposed to plan the obstacles avoidance path,and the obstacles avoidance controller is designed based on fuzzy logic.The optional driving space can be got using self-organizing competitive neural network to classify the space data.The space cost function is proposed to evaluate the optional space,and the vehicle driving direction can found based on the space geometry relationship.This method is validated by experiments,it shows that the proposed method can find the vehicle optimum driving direction in different conditions.The safety prediction model is proposed to decide the vehicle autonomous driving mode.The membership function and fuzzy response surface is built.A simulation software is developped using C++ to validate the proposed obstacles avoidance method.The simulation and test result show advantage of the proposed path planning and vehicle control method.Finally,a small scale model prototype of the articulated vehicle is built,whose control system is similar to the real vehicle.Meanwhile,the electrification of the real vehicle is upgraded to meet the requirements of autonomous driving.The experiments is conducted in Shaoguan Fankou Leas-zinc mine,to validate some of the proposed control methods.