Application Control Research On Autonomous Guidance Of Region Auto Valet Parking

Countries around the world have taken intelligent vehicles as an important part of the development of future cities and transportation,and many enterprises and institutions have comprehensively promoted the development and implementation of intelligent vehicle technology.Fully autonomous driving is very complexity,but at this stage,scenario-based,lowspeed,and functionalization are the main development direction of intelligent vehicle technology,and they are also the only way to achieve autonomous driving.Low-speed & scenebased auto valet parking system,as a fusion product of intelligent vehicle technology,is the focus in exploring the development of autonomous driving.This subject comes from a school-enterprise cooperative research that want to achieve automatic valet parking,and this paper focuses on the application of global path planning and path tracking control in autonomous guided driving vehicle.In order to realize the automatic driving control of intelligent vehicle,this paper establishes a vehicle kinematics model that meets the driving control requirements,based on the project system requirements and existing experimental vehicle platform.And this paper constructs a discrete state space equations that describe the vehicle's motion through linearization and discretization processing.These equations are theoretical models of state prediction and control of vehicle.Simultaneously,this paper describes the precise positional relationship among driving roads,parking spaces,trees,and obstacles by mapping the location outdoor parking lot.And combined the road driving rules and Open DRIVE road network technology,this paper establishes HD Map of the outdoor parking lot.The map provides location data for path planning and tracking control.Then based on the graph and the heuristic path search theorys,this paper designs the hybrid path planning algorithm and maneuverability strategy based on Floyd & A*,combined with the established HD Map that establishes waypoints.The hybrid algorithm can realize the online dynamic planning of reference path that is safety and optimal driving route for intelligent vehicle.In order to ensure that the intelligent vehicle can follow the reference path accurately,this paper constructs a path tracking controller based on the model predictive control algorithm that meets the requirements of the intelligent vehicle driving control.And the objective function of the controller is transformed into a quadratic programming problem.Based on the QP method and a dual algorithm,this paper proposes a new QPKWIK solver through decomposing and inverting the Hessian matrix.By analyzing the simulation data in Matlab/Simulink software,we get what QPKWIK solver is fast and effective.Finally,the actual driving data is analyzed to verify the effectiveness and feasibility of the path planning algorithm,path tracking controller,and autonomous guidance system in this paper through the experimental vehicle.Combined with autonomous guidance system and automatic parking system,this paper realizes the application of auto valet parking.