The Research Of An Autonomous Vehicle Based On Integration Of Function And Behavior

In recent years, many people have researched on Autonomous Vehicle in many aspects. The emphasis of their research is control of body of Autonomous Vehicle and path planning. Some aims are achieved. Now main problem is the absence of flexibility and self-determination. Aimed at the drawbacks, I take the car-like Autonomous Vehicle as my research object, and propose some control methods such as fuzzy control, FNN(Fuzzy Neural Net)control and so on. And these methods all have good performances through simulation study. A novel track control method based on Regress Algorithm is present in this paper. Simulations illustrate the effectiveness of the proposed method. We developed a three-wheel autonomous vehicle which has a metal frame and two driven wheels. And we can go further study on it such as integration of navigation by vision. The autonomous vehicle can walk in a line and turn under the control of fuzzy control techniques. The experiment results do show self-study character to some extent. And at last an new control architecture based on integration of function an behavior is proposed in this paper. This architecture has the virtues of both two methods. This architecture endows the Autonomous Vehicle with the ability of learning , adaptation, quick reactive speed ,also with the ability of achieving giving task.