Research On Algorithm Of Intelligent Vechicle's Motion Control And Obstacle Avoidance In Urban Environment

With the development of artificial intelligence technology and the introduction of intelligent transportation system concept, the technology of intelligent vehicle has attracted much research attention. Moreover, complicated urban environment requires higher standards for the intelligent vehicle technology. To address this need, this article mainly studies the following three topics: path planning, motion control, and autonomous obstacle avoidance of the intelligent vehicle in urban environment.1) In order to solve the problem of path planning in the area without perception map of intelligent vehicle, a method of path planning based on Google Map was proposed. First,Design a webpage using Google Maps API interface functions and JavaScript scripting language to load Google Map; Then, Implement the path planning between the starting point and the destination point using the path planning class provided by Google; Finally, Imbed the webpage into the main program which is based on MFC, the intelligent vehicle could get the path planning information of Google Map.2) In order to improve intelligent vehicle's accuracy of decision-making and precision of motion control, a motion controller based on Fuzzy Control and BP Network was designed.At first, formulate fuzzy rules based on artificial driving experience; Latter, quantify fuzzy rules into training samples; At last train the samples by BP network. The motion controller combined the advantages of Fuzzy Control and Neural Network Control, input the relative position information between vehicle and the desired trajectory, output the front-wheel turning angle and velocity of the intelligent vehicle. Experiments show that the method has higher control accuracy and better robustness3) In order to solve the problem of collision between the rear part of intelligent vehicle and the obstacle during the autonomous obstacle avoidance, the Secure Massage Method was improved. The core concept is to calculate the relative positional relationship between the smart car's rear and the obstacle at every moment. In addition, to solve the problem that the intelligent vehicle needs to change its trajectory according to the guidance of the conical barrel in the urban construction environment, a laser radar data processing algorithm was designed which could get the virtual boundary of the conical barrel region. Repeatedexperiments show that the obstacle avoidance algorithm is feasible and effective...