Research On System Architecture And Navigation Strategy Of Intelligent Vehicle

Mobile robot is an important branch of robotics, and the autonomous navigation is a hot researching question of mobile robot, but environment perception and path planning are two key technologies to achieve autonomous navigation. In the paper, the model of mobile robot--- intelligent vehicle is taken as the research object to mainly research architecture and navigation strategy of intelligent vehicle.At the aspect of architecture, according to the actual navigation requirements of intelligent vehicle, autonomous and hybrid architecture is adopted as experimental platform for autonomous navigation of intelligent vehicle by analyzing and comparing several architecture's characteristics. The double CPU is used to simultaneously process environment information which is captured by a CCD image sensor and multi-ultrasonic sensors.At the aspect of processing multi-sources information, in order to recognize target location and pits or hollow obstacles on the road, the Sobel template is used to process target image which is captured by CCD, as well as, dynamic threshold method is adopted to extract characteristics from road images. When multi-ultrasonic sensor information are being processed, a algorithm of avoiding obstacles that is based on fuzzy comprehensive evaluation is proposed, and experiments show that it is effective to achieve the autonomous avoiding obstacles of intelligent vehicle by using the algorithm.At the aspect of path planning, a algorithm of path choices based on fuzzy logic inference is proposed, and according to fusion result of the surface of road and the obstacle information, intelligent vehicle's optimal path choice and the autonomous navigation are implemented by using the fuzzy inference rules in the algorithm. The simulations and results of experiment confirm that autonomous navigation of intelligent vehicle under complex environment can be achieved by using the algorithm proposed.