Research On Key Technologies Of Intelligent Car Control Based On Traffic Image Information Collection And Processing

In the context of intelligent traffic system,the key technologies of intelligent car automatic control based on the acquisition and processing of traffic image information are researched in this paper,which lays the foundation for the autopilot of vehicles in traffic system.In order to realize the automatic control of the intelligent car,three key technologies are studied in this paer,namely(1)Tracking the desired straight path when the intelligent car goes straight.(2)Tracking the desired turning trajectory when the intelligent vehicle turns at the crossroads.(3)Collecting high-quality macroscopic traffic image information when the intelligent car driving path is optimized.Based on the above three key technologies,the main research content is as follows:The intelligent car direct control system based on lane identification technology is designed to realize the intelligent car tracking the desired straight path.The traffic image acquisition and processing module for straight path is designed firstly: the position and straight-line state(center,left or right)of the intelligent car can be acquired in real time by recognizing the lane line or crossroads dividing line using image preprocessing technique,improved probabilistic Hough transform and other lane recognition techniques.Secondly,the intelligent car straight-line control module is designed: the module develops direct control instructions(straight,correction)on basis of the results of image recognition to correct the track of the intelligent car timely and ensure its accurate straight.Finally,the intelligent vehicle driven by DC motor is used to validate the direct control system on the intelligent traffic sand table.The result shows that the intelligent car can track the desired straight path.The intelligent car crossroad turn control system based on number identification positioning is designed to guarantee the intelligent car to track the desired turning trajectory.The traffic image acquisition and processing module for turning path is designed firstly: the position of the intelligent vehicle on the turning path is determined by identifying the numbers on the turning path using improved KNN algorithm.Simultaneously,the driving state of the intelligent car can be obtained real-timely by identifying the crossroads dividing line or zebra crossing on the turning path using improved probabilistic Hough transform.Secondly,the intelligent car turning control module is designed: based on the results of the image recognition,the intelligent car control command(straight,correct and turn)is designed to track the expected turning trajectory.Finally,on basis of the verification platform of the direct control system,the turn control system is validated.The result shows that the intelligent car can track the desired turning trajectory rightly.In order to realize the intelligent car path planning system can collect high-quality macroscopic traffic image information,aiming at the problem that the aerial quadrotor UAV in the intelligent car path planning system has external disturbance,its own state delay and the actuator failure affect its aerial stability,improved guaranteed cost controller and model reference adaptive control are designed to ensure the UAV flight quality in aerial photography.Improved guaranteed cost control is designed to solve the problems of external disturbance and state delay,thus ensure the smooth flight of the quadrotor UAV in the absence of fault conditions.Model reference adaptive control is used to reconfigure the quadrotor UAV controller to compensate the actuator faults,thus ensuring the good flight quality of the quadrotor UAV under fault conditions.The simulation results show that the designed controller can guarantee the quadrotor UAV tracking performance for desired aerial attitude angles,which laies the foundation for high-quality aerial photography.