Intelligent Control Method and Self-position-azimuth Correction Method for Autonomous Vehicle

This paper proposes an intelligent control system using fuzzy-neural network (FNN) for a utonomous vehicle .The autonomous vehicle will appropriately and automatically recognize and judg e the running environment and run along a given orbit by using FNN. In order to acquire the training data of FNN, the driving knowledge is extracted from the human driving data. Then, the driving data is processed and normalized by rough set and defuzzy method. Moreover, a self-position- azimuth method is proposed to detect and correct the position and azimuth of the vehicle by outside- world information from stereo camera and laser sensor. The effectivity of the methods has been verified by tests using a model vehicle.