Research On Automatic Driving Technology Of Vehicle Adapted To The Condition Of Curved Ramp In Underground Parking Garage

The development of automated valet parking technology can provide convenient and efficient parking services for the driver,and effectively reduce the accident rate in the parking garage,and reduce energy consumption.The curved ramps,which are typical entrances to underground parking garage,often cause car collisions.It is great significance to study how automatic driving technology of vehicle adapted to the curved ramp in underground parking garage.In the existing automated valet parking technology,the environmental sensing system requires expensive infrastructure,and there is no curved ramp condition in the the collision avoidance scenario,and the path tracking control law is based on time.This paper studies how automatic driving technology of vehicle adaptes to the condition of curved ramp in underground parking garage,so that the intelligent vehicle can safely enter the parking garage through the curved ramp in the process of the automated valet parking,This paper focuses on the research of environment-sensing system,collision avoidance control strategy and path tracking control law.The effectiveness of the proposed collision avoidance control strategy and path tracking control law were verified by simulation.The actual vehicle tests verify the effectiveness of the path tracking control law.First,the ultrasonic radar-based environment-sensing system was built and the physical characteristics of ultrasonic radar were studied.Through the analysis of radar echoes,echoes of useful objects were screened out.The location of objects was identified by adjacent radar.The early-warning areas and warning-levels were designed so that the designated environment-sensing area could completely include the front end and the rear end of the vehicle.Using Canoe?PVC pipe?grid cloth and other tools,the detection range and early warning level of the environment-sensing system were tested.The results show that the designed environment-sensing system has enough detection range and the warning level distribution was reasonable,and the warning information is clear.Secondly,the collision avoidance control strategy of the intelligent vehicle driving in the curved ramp was studied.Based on the early-warning information of the environment-sensing system,a kind of collision avoidance control method based on fuzzy algorithm is proposed.The collision scenes in the same lane and adjacent lanes within the curved ramp are taken into account.The fuzzy collision avoidance controller controls the the speed of the intelligent vehicle to avoid collision or reduce the damage.Through co-simulation of Carmaker and Simulink,the designed fuzzy collision avoidance controller was verified.The results show that the fuzzy collision avoidance control strategy can provide the intelligent vehicle with an effective collision avoidance in the curved ramp.Thirdly,the path tracking control technology of the intelligent vehicle driving in the curved ramp was studied.Aiming at the problem of the three-dimensional position parameters tracking in the process of the curved ramp path tracking in the underground parking garage,considering that the time-variable-based path-tracking control law is greatly affected by vehicle speed and time,the control error would be increased by time when the driving process of intelligent vehicle is stopped due to collision avoidance.By using the Z-axis position variable as a non-time-referenced depend variable,the three-dimensional space position parameters is converted into two-dimensional position parameters to solve the problem that the control error is influenced by time.a non-time-based path tracking control law is proposed.At the same time,the complexity of the path tracking control strategy has been reduced and the real-time control performance is improved.The simulation results show that the designed path tracking controller has good path tracking effect in the curved ramp.On this basis,an automatic driving system simulation model,which applied to the curved ramp,was constructed within the automated valet parking technology.The simulation results show that even if there was the parking condition in the path tracking process,the actual path could still substantially coincide with the target path.System control deviation is not affected by time.Finally,a real vehicle test for path tracking of the intelligent vehicle driving in the curved ramp was carried out.The effectiveness of the underlying electrical system was tested,and the acquisition of the target path was completed through the inertial navigation technology.The validity of non-time-based path tracking control was verified.The path tracking test results show that the system control error is not affected by vehicle speed and time,and the designed path tracking controller can accurately track the target path.