Modeling and control of longitudinal motion for unmanned ground vehicle in complex environment

Unmanned ground vehicles(UGV) modeling with traditional force analysis may cause mismatching between the established model and actual system due to parameter uncertainty and inconsistency, so the simulation of traditional model makes no sense while applied to actual use. In this paper we establish a function relationship among variables based on force analysis, then consider the inherent delay characteristics of the unmanned ground vehicle and add the road slope as disturbance. Thus, by means of learning algorithm to fit parameters, the optimized model in complex environment is built. The fitted parameters and simulation results show that the established optimization model is consistent with actual system. A split-phase controller is designed for vehicle longitudinal speed control and a selection strategy of throttle threshold and brake threshold in phase controller is proposed. Simulation result in MATLAB/simulink shows that the proposed threshold selection strategy can make the system response fast and stably without frequently switching the controller.