Based On The Visual Feedback System Of The Smart Car Research

With the development of science and technology, the developing trend for machine system is to improve the intelligent and visual capability. Mobile robot is a high intelligent machine system constituted by sensing module, planning and decision module, control and executing module and so on. It is playing important roles in industry, battlefield, homes and hospitals. This thesis takes the intelligent car that is a mobile robot with simple function and structure as research object, and focuses on the navigation and control problems of intelligent car system.Firstly, the intelligent car system platform is designed. And this thesis introduces the constitute part and function of intelligent car system based on visual feedback.Secondly, the navigation approach that uses monocular vision to detect road leading line is researched. A novel approach that uses monocular vision to get image deep information is presented. At the same time, the detection algorithm of road leading line is also introduced. The experiment results show that proposed navigation approach is effective and image processing algorithm is quick and effective.Thirdly, this thesis researches the control strategy of intelligent car system. Because vision navigation can obtain much more useful information, by imitating the preview driving behavior of human, a preview PID control strategy for path tracking is developed. Considering the fact that driving behavior of human is fuzzy control process, an effective fuzzy control method for path tracking is developed. The experiment results show that the fuzzy control approach can not only make the intelligent car follow various reference paths, but also make the control process smoothly. In short, the fuzzy control approach is a ideal control strategy for intelligent car system.Finally, this thesis introduces the hardware and software design of intelligent car system. And the obtained research results are summarized and future work is addressed.