Research On Automatic Parking Path Planning And Tracking Control Strategy For Articulated Vehicles

With the rapid development of economy and society,articulated vehicles with long size and large load capacity have become more and more important in the cargo transportation industry.Therefore,developing a container traction-semi-trailer with automatic parking function will greatly improve the transportation efficiency of vehicles in port terminals and logistics parks,which not only reduces the driving cost of drivers,but also improves the efficiency of transportation.Compared to a bicycle,the articulated vehicle has a semi-trailer constrained by a traction pin,which makes the vehicle system steadier when driving forward.Due to the restricted field of vision and perception of the driver and the agglomeration divergence,the vehicle is prone to folding instability when the vehicle moves in the reverse direction.It is difficult for even an experienced driver to park the vehicle in the warehouse at one time.The automatic parking technology of passenger cars has matured at present.Many companies have already installed this technology on mass-produced mid-to-high-end cars,which has brought great convenience to some novice drivers and owners who are fear of parking.However,universities and research institutions at home and abroad have slightly studied on commercial vehicle automatic parking.Based on the above reasons,it is necessary to carry out research on the path planning and tracking control strategy of articulated vehicles.The specific research contents of this article are as follows:(1)The vertical parking and parallel parking scenes of articulated vehicles are set up using TruckSim,and the reversing motion trajectory of the articulated vehicles is analyzed using the TruckSim ’s vehicle model.The traction-semi-trailer model with axle arrangement of "S_SS + SSS" was selected.Simplify two types of vehicle models,which one is the traction pin placed at the front of midpoint of the rear axle of the tractor and the other is the traction pin placed at the midpoint of the rear axle of the tractor,deduce the kinematic models,give the model parameters and simulation parameters of the two models,and build the two motion models of vehicles in Simulink.At the same time,dynamic models of the same vehicles are built in TruckSim.Under the same input conditions,compare the simulation output of the kinematic model derived in Simulink with the vehicle model built in TruckSim.(2)The A * algorithm is introduced,the calculation methods of its three heuristic functions are given,and the pseudo code of the algorithm is described in detail.By setting the starting point,target points and obstacles in the plane,A * algorithm is simulated and verified in a Matlab script file.The vehicle kinematics model is added on the basis of the traditional A * algorithm to form a Hybrid A * algorithm that planned the path nodes to meet the vehicle’s non-holonomic constraint.The extended method of hybrid A * algorithm grid node,the composition form of heuristic function,and the definition of the function in algorithm pseudo code are given.The algorithm pseudo code is described in detail based on the flowchart of hybrid A * algorithm.By adding Reeds-Sheep curve search strategy in the grid node expansion,the algorithm’s solving efficiency is improved.Giving calculation methods of 48 combinations of 9 types of curves,write the curve solving algorithm in Matlab and verify by simulation.By adding collision constraints,articulation angle constraints,and cost constraints,the vertical parking and parallel parking path planning simulations of the tractor and two articulated vehicles have built can be performed.(3)Considering the semi-trailer when reversing as a virtual tractor,the semi-trailer is the driving unit and the tractor is the follower unit.The inverse kinematics model of a simplified articulated monorail model is derived by taking the midpoint of semi-trailer axle as the reference point,and the conversion relationship between the steering angle of the virtual steering shaft and the actual front wheel angle of the tractor is obtained.Calculate the horizontal and vertical coordinates of the preview point by using the heading angle vector of the trailer and the preview distance,and find the path node closest to the preview point by solving the distance between the adjacent reference path node and the preview point,then the position deviation between the lookahead point and the path node is converted into a virtual steering angle and brought into the inverse kinematics conversion relation to obtain the actually required front wheel turning angle of the tractor.By designing path-tracking horizontal and vertical joint controllers,the vertical parking path tracking and parallel parking path tracking are simulated and verified with the traction pin placed at the front of midpoint of the rear axle of the tractor based on the TruckSim and Simulink joint simulation platform.