A Study On Obstacle Detection Method With Unpiloted Vehicle

With the development of modern sciences and technology and the improvement of social economy and human’s living standards. People put forward higher requirements for transportation means. Opening a precedent in history to unmanned vehicle technology, because peolpe has been gave unmanned that is the complete beauties. Obstacle detection research one of the key for unmanned vehicle.First of all, this paper detailed description of communication protocol for GPS inertial navigation, laser radar, and through the realization of the communication between IPC and the platform of Visual Basic to obtain the detection data of GPS inertial navigation, laser radar.Secondly, In order to solve the problem of obstacle motion state detection in unknown environment conditions, a detection method of independent dynamic obstacles based on laser radar is put forward in this paper. The close neighborhood method and the rules classifier algorithm are used to cluster the data which detected by the laser radar and analyze the characteristic parameters of the clustering obstacles. The types of obstacles is determined by analysing the confidence interval correlation of the clustering obstacle data, and the velocity and azimuth of the obstacles are analyzed and determined by the relative coordinate transformation.Problem of road boundary and Lateral obstacle detection, This paper uses the method of lidar detection by laser radar. By using recursive median filtering method makes the border road changing more smooth, and use of the ultrasound detection of lateral obstacles.Finally, Visual Basic software platform developed to accept data from each sensor interface, At the same time for the clustering of the barrier, Finally, Visual Basic software platform developed to accept data from each sensor interface, while the clustering obstacles, obstacles motion, road side boundary and obstacle detection experiments were made. The results show that: the ability to effectively clustering obstacles in order to achieve the same obstacle detection speed and heading; simultaneous detection of the road boundary and side obstacles Finally, Visual Basic software platform developed to accept data from each sensor interface, while the clustering obstacles, obstacles motion, road side boundary and obstacle detection experiments were made. The results show that:the ability to effectively clustering obstacles in order to achieve the same obstacle detection speed and heading; simultaneous detection of the road boundary and side obstacles experiments were done. Therefore to realize the same speed and direction of obstacles detection and the fruition of a road boundary and lateral obstacle detection.