| Automatic parking system is a kind of automotive electronics system that enables a car to get into the parking space automatically. It finds the right parking spaces depending on surroundings information gotten by sensors, and measures the relative position between the car and the parking spaces then controls the car's speed and steering with a certain method, finally, completes the parking. Comparing with manual parking which is difficult to operate, but gets a higher accident rate and the low intelligent parking with a traditional reversing radar, automatic parking system provides a convenient parking function, reduces the difficulty of parking, and also improves the intelligence level and safety of the car.In this paper, three automatic vertical parking methods have been developed and designed. They have been verified through the computer simulation.First of all, according to Ackermann steering geometry, the kinematics model is created from the simplified model of a vehicle that is at a low speed. Next, the parking space is identified by perceptual system built using a variety of sensors. Then, based on the law that the parking process is reciprocal with the outbound one, and in reference to the national standard, three different vertical parking paths are designed, considering obstacle avoidance requirement. Finally, use Bezier curves to deal with continuous driving problems.In simulation, a simulation system is built up using Matlab's GUI, and a variety of path planning methods are simulated with it. With this simulation system, we study the car's movement law in the automatic vertical parking; analyze the impact of parking trajectory caused by the obstacle avoidance constraints, and find out how to solve the problem of turning mutations, which introduced a method of continuous driving.The automatic vertical parking methods designed in this paper can complete the automatic parking smoothly, but there are also many problems which can be improved, follow-up studies are expected. |
PDF Full Text Request