NAVIGATION CONTROLLER DESIGN USING FUZZY LOGIC THEORY FOR VEHICLE PARALLEL AUTOMATIC PARKING

摘要

With the increasing development of intelligent vehicles, the automatic parking system came into being, it could help people improve the driving safety and reduce the car accidents. From this point of view, it is greatly necessary to serve the automatic parking research. The three stages of the automatic parking system include parking space detection, the kinematics model building and the navigation controller design. Firstly, the laser radar sensor is applied to obtain the distance information around the vehicle, then the median filtering algorithm and the least squares method are employed to process the laser data. The next step is to detect the parking space by using K-Means clustering and threshold segmentation method. The second stage is to build car kinematics model by employing the Arman steering theory. Thirdly, the navigation controller design and simulation are accomplished based on the fuzzy logic control theory. The parallel automatic parking simulations are shown by using the Matlab toolbox. Experimental results show that the parking space detection method has perfect performance and the navigation controller design algorithm has good feasibility.