Research On The Detection Of Preceding Vehicle And Safe Distance Control Based On Fuzzy Logic

With the increasing of vehicle's ownership and enhancement of vehicle's performance,the problems of traffic congestion and traffic accident has been at a high level.Due to the shift from the concern about dynamic performance to the concern about vehicles ' safety performance,intelligent transportation system is emerging.Intelligent transportation system constructs the traffic,the vehicle and the driver to one integrate via information technology,sensor technology,automatic control theory and artificial intelligence et al.A transportation system with safety,high level of efficiency,energy conservation is the trend of vehicle and transport development.The research of vehicle active safety technology started late in our country,and has a long way to go to catch up with other countries.With the increasing alcohol-related and fatigue-related traffic accidents,the problem of safe distance control needs to be studied and resolved in the further research.In recent years,fuzzy theory has been researched by domestic and overseas scholars,and been applied to many fields including image processing and automatic control.Due to the outstanding performance of robustness and completeness,data mining has become one of effective methods in noise recognition,image segmentation and vehicle detection.The paper has researched visual-based vehicle recognition and measurement,and designed an active brake control system based on vision distance measurement and velocity estimation.The main ideals are shown as follows:(1)The image preprocessing methods based on fuzzy logic have been researched:The image noise is identified by using data mining method and fuzzy weighted mean filter is used to deal with mixed noise;The fuzzy-theory-based gray level transformation has been adopted to enhance the filtered images.(2)Feature extraction and recognition methods of front vehicle have been researched:The area of interest is extracted from preprocessed image after using data-mining-based image segmentation;Data-mining-based learning method has been adopted to identify vehicles by using features of Hu moment,contour moment and rectangle degree;And BP neural net work has been adopted to verify the vehicles by using parameters of edge feature,symmetry and texture.(3)Vehicle locating and tracking methods of front vehicle have been researched:The relationship between vertical coordinate and actual distance has been confirmed by calibrating camera internal parameters;The vehicle locating method using edge extraction has been studied,and been used to track and update vehicle margin information;The bottom margin of the vehicle in sequence image has been obtained to estimate relative velocity.(4)The safe-distance-based vehicle active braking control method has been researched:Safe distance model has been established using vehicle speed and safe distance;The expert PID and fuzzy PID ABS control strategy have been compared based on single wheel model,and fuzzy PID ABS control is selected;The safe distance control system have been build based on real-time safe distance measurement and ABS control strategy.The front vehicle detection and locating and safe-distance-based vehicle brake control have been particularly researched in the paper based on fuzzy theory.The simulation experiment has verified the effectiveness and practicability of the brake control system under different circumstances.