The Research And Realization Of Automatic Path-finding And Obstacle-avoiding Intelligent Vehicle System

Comparing to traditional vehicles,intelligent vehicles have a great advantage because they have the ability to think.Without manual operation from drivers,intelligent vehicles can start,accelerate,park on their own.What's more,they can plan driving route to avoid obstacles according to the surrounding environment.In the complicated and changeable environment,the "brain" of intelligent vehicles has faster response ability than that of humans.They can make the best plan in a short time.Intelligent vehicles have faster response than humans to deal with emergencies,so they can avoid some of the traffic accidents,making intelligent transportation become true.Nowadays,with the popularization of GPS,vehicle autonomous navigation system has become increasingly mature.However,because of the complexity of the actual conditions,it is still a problem for vehicles to recognize the path and avoid obstacles.It is quite necessary to study and develop path planning of intelligent vehicles system based on camera sensor.The study of path recognition and obstacle avoidance algorithm,which has important value of research,becomes a hot topic in the field of studying intelligent vehicles.In this paper,the author designs a kind of automatic path-finding and obstacle-avoiding intelligent vehicle system based on machine vision.By identifying single guide line,the previous system often leads to big errors.The author proposes,in this paper,an intelligent vehicle system that can identify bilateral guide line and avoid obstacles autonomously on the driving path.As far as hardware is concerned,Freescale 32-bit microprocessor MK60FN1M0VLQ15 is applied as main control module in the intelligent vehicle system.It first acquires image information of the track through the CMOS camera OV7620 and then processes the collected original images so as to identify the track and the obstacles on the track accurately.The right components are chosen to complete the design of various modules of intelligent vehicles,including micro controller module,the system power supply module,path perception module,debugging module.At the same time,the mechanical structure of intelligent vehicles has also been designed so as to make sure that the intelligent vehicles are able to run smoothly.As to software,this paper introduces the popular guide line recognition methods used by intelligent vehicles,including the method of threshold segmentation and edge testing method.By comparison and validation,edge testing method is finally used as the way to recognize the guide line of intelligent vehicles.For obstacle-avoiding path planning used by intelligent vehicles,this paper has adopted gravitational field function as the artificial potential field method in intelligent vehicles.In view of the fact that the intelligent vehicle runs in a dynamic environment,therefore,the distance between the obstacle and the intelligent car,the velocity and the acceleration information are added into the artificial potential field algorithm.By adding "escape power",we can make the intelligent vehicle get rid of the limit of local minimum.Fuzzy control algorithm is used as the strategy of steering controlling direction of the intelligent vehicle steering control.From fuzzy control strategy through simulation,steering gear makes the intelligent vehicle have faster response speed,and has thus improved the stability of the system.Incremental PID control strategy is used to control the speed of the intelligent vehicle.Compared with traditional PID control strategy,the incremental PID control strategy can make the system overshoot volume decrease,so the controller can adjust the speed better.