Three-stage vertical parking path planning for low-intelligence vehicles in standardized garages

Currently, China has 435 million vehicles, with a high proportion of low-intelligence vehicles. These vehicles have low automation, cannot perform complex driving operations, but are equipped with basic onboard sensors and V2X functionalities. With advancements in computer vision, the use of relative positioning from reverse images and acoustic radar has been effectively applied to onboard computers. This paper proposes a three-stage parking operation based on a parking selection library to address these characteristics. For known parking lot structures, a fixed-point steering model for standardized parking lots is developed. Complete parking trajectories and paths are obtained using vehicle and garage parameters to create a parking selection library. Basic sensors provide relative position parameters to match the parking selection library information. The vehicle then performs steering operations at specified points. This method effectively compensates for the low perception and weak decision-making capabilities of low-intelligence vehicles.