Research On Path Planning And Control Of Driverless Logistics Train

Environmental perception,positioning,decision-making,path planning and tracking control are the key technologies for intelligent vehicles to realize driverless,and path planning and tracking control are the premise and foundation of realizing driverless.Driverless logistics transport vehicles can travel in various end distribution scenarios requiring short-distance logistics transportation,such as industrial parks and residential areas,which can effectively reduce the transportation cost and labor cost of the logistics industry and improve the transportation efficiency.Therefore,in order to further study the relationship between driverless logistics transport vehicle and path planning and path tracking technologies,in this thesis,a driverless logistics train platform with trailer is established,and the path planning and tracking control technologies are studied based on the train platform.The specific contents are as follows:The kinematics and vehicle road system model of driverless logistics train are established.Firstly,this thesis analyzes the driverless logistics train system with single and multi trailer,deduces the corresponding kinematic equation and establishes the kinematic model.Secondly,the simulation tests of single trailer train and multi trailer train are carried out to verify the accuracy of the train kinematics model.Finally,based on the distributed driving vehicle used in the driverless logistics train platform,the vehicle road system model is established,and the lateral deviation at the preview point and the directional deviation at the centroid are selected as the system control variables.A local path planning algorithm based on quintic polynomial curve is designed.Firstly,when the vehicle is driving,the coefficients of the quintic polynomial are calculated based on the information of the current point and the preview point in front of the vehicle,so as to obtain a certain number of candidate paths.Secondly,a cost function is designed,which integrates the security,smoothness,centrality and consistency of candidate paths,and the path with the lowest cost function value is selected as the final target path.Finally,the algorithm is tested by joint simulation,and the simulation results verify its effectiveness and feasibility.A speed control method based on sliding mode control and a path tracking control strategy based on feedforward control and feedback control are designed.Aiming at the intelligent tractor used in the driverless train platform,a speed control method based on sliding mode theory is designed to make the tractor run at the expected speed.In the path tracking control strategy,the feedforward control makes the vehicle obtain a certain direction angle in advance in order to overcome the curvature of the road ahead and improve the response speed of the vehicle at the same time.The feedback control is designed based on the linear quadratic regulator(LQR)control theory.The goal is to reduce the directional deviation and lateral deviation between the vehicle and the desired path.Finally,the simulation test of the control strategy is carried out.The simulation results show that vehicle can better complete the global path tracking,and verify the feasibility and effectiveness of the proposed path tracking control strategy.The real vehicle test platform of driverless logistics train is built,and the relevant test verification is carried out.Firstly,a driverless logistics train platform with trailer is built,and the main structure and intelligent equipment of the train platform are described.Secondly,the longitudinal speed control method,path tracking control strategy and path planning algorithm proposed in this thesis are tested and explored.In the longitudinal speed test,set different target speeds for the train,increase the train load by carrying car tires on the trailer,and test the longitudinal speed of the train under different loads.The results show that the train can run at a more stable speed under different loads.For the path tracking control test,the test is carried out under two classic road conditions: double lane change and splay roads to verify the path tracking control performance of the train under different road conditions and different speeds.In the path planning experiment,the experiments were carried out on the closed site and the actual road of the campus.The test results show that the train can better complete the behavior strategies of obstacle avoidance and detour,parking and waiting.Among them,when testing on the actual road of the campus,the positioning method of Simultaneous Localization and Mapping(SLAM)is combined,which effectively solves the problem of inaccurate GPS positioning of the train under the trees,and verifies the feasibility of SLAM positioning in the actual road test of the campus.