Research On Car-like Mobile Robot Formation Control And Path Planning Problem In Rough Terrain Environment

Multi-robot system has become one of the research hotspots due to its high efficiency,fault tolerance ability and flexible extensibility.Formation control is a key technology to realize the cooperation and coordination of the multi-robot system.With the expansion of robot application,flat terrain workspace,such as laboratory,factory workshop and other indoor environment,cannot satisfy the increasing need of complicated robot tasks.In numerous situations,robots have to face uneven rough terrain in a variety of outdoor natural environments,such as rescue after disaster,science investigation,mine clearance and planetary exploration.Car-like mobile robots are able to bear a certain degree of rough terrain.But the complicated environments may lead to wheel slip,wheel subsidence,robot skid and shock,which affect the control performance of robots.Based on the rough terrain robot workspace,this thesis studies the formation control methods for car-like mobile robots in rough terrain.A compensation controller design method is proposed to deal with the wheel slip and a dynamics controller is also designed.Finally,a path planning algorithm using multi-objective optimization is proposed to solve the path planning problem in rough terrain.The proposed algorithms are simulated using Microsoft Robotic Developer Studio 4,which is based on a physics engine called PhysX.The main work is as follows:(1)A kinematics formation controller is proposed according to the kinematics model of car-like mobile robots.Using the leader-follower approach and combining with the formation parameters,the formation system error state equations are obtained through coordinate transformation.Then Lyapunov function of the trajectory tracking system is constructed via the backstepping method.Simulations are designed using MRDS4 to analyze the effects of control parameters,undulation degree and formation parameters,which provide the foundation for the following research.(2)For the formation control problem of the car-like mobile robot with longitudinal slip,a formation control algorithm is proposed based on the compensation of slip ratio and fuzzy logic.Firstly,the real velocity of the robot is acquired by sensors.And the noise is decreased through filtering.Then the real velocity is compared with the input velocity signal to obtain the estimation of slip ratio,which represents the degree of slip.According to the leader-follower method,a formation control strategy with compensation of slip ratio is designed,and fuzzy logic is adapted to tune controller parameters.Finally,simulations are designed in MRDS4.A standard slope workspace and a common rough terrain workspace are designed to test the proposed controller.The simulation results show the effectiveness of the presented method in many situations where the leading or following robot slips.(3)A dynamics controller design method based on characteristic model is proposed for the formation control problem of car-like mobile robots.Only kinematics controller is not enough for sone cases such as the environment is rugged,the mass and velocity of the robot is large,and the dynamic parameters of the robot are time-varying.First,we obtain the kinematic error state equations according to the leader-follower method.A kinematics controller is designed and the stability is proved by Lyapunov theory.Then the characteristic model of the dynamics inner loop is established.A sliding mode controller is designed based on the second order discrete model,and the stability is analyzed.Finally,simulations are designed in MATLAB and Microsoft Robotics Developer Studio 4 to verify the effectiveness of the proposed method.Sinulation results show that the proposed method can improve the responding speed of the mobile robots and maintain high formation accuracy.(4)A path planning algorithm for car-like mobile robots operating on a known static rough terrain environment is proposed.The purpose of this approach is to find collision free and feasible paths with minimum length and terrain roughness.First,a new workspace modeling method is proposed to model the rough terrain envirorunent.Then,considering the nonholonomic constraints of car-like robots,a MOPSO(multi-objective particle swarm optimization)based method is used to solve the problem.In the proposed algorithm,a new updating method for particle's global best position based on crowding radius is used to increase population diversity.And to improve the algorithm efficiency,a nonuniformity factor is adopted to update the particle's position when the path collides with obstacles.Finally,two simulation tests are designed using Microsoft Robotics Developer Studio 4 and Matlab.Results show the advantages of the proposed algorithm in finding Pareto optimal paths.The effectiveness of the optimal paths is validated using MRDS4 simulations.(5)In order to verify the effectiveness of the proposed kinematics and dynamics controllers,two groups of simulation cases are designed.Firstly,combining the path found by the multi-objective path planning algorithm with the control strategy,a multi-robot formation control simulation is carried out.The advantages of the proposed inner and outer loop controllers are analyzed.Secondly,a mine clearance simulation case is built with MRDS4 to show the practicability of high accuracy formation control.Using the same formation parameters,the search coverage detected by robots using the proposed controllers are compared with the one using general controllers.Then the formation parameters are modified to guarantee complete coverage of the detection area.The search converges with the two control methods are compared to reflect the mine clearance efficiency.