The Research Of Open Architecture Robot Controller

The robot is the most complicated machine, also is the human most eager to early manufactured robot friend. However, to produce a humanoid robot is not easy, just let the machine walking motion simulation of humans, scientists have to pay tens of or even hundreds of years of hard work. Developed robots do not have the communication interface compatible, so the robot development has been slow, taking "fly think of Carle" National University intelligent car contest as the background, the paper discussed the design of open architecture robot controller, the main structures of a set of open controller, the controller through the host computer can complete the simple configuration different groups of the game task, so that beginners can directly start easier entry, relieve the developer, to complete the controller, this thesis takes eighth contest rules custom controller, optimization design of carbody mechanical structure including the smart car model, hardware platform, program frame construction, algorithm, and the final debugging, and ultimately a controller compatible with all vehicles.The fly think of Carle cup smart car competition based on the study of the competition, the smart car model characteristics, design and adjust its mechanical structure. Made strict limits in the PCB shape and size, the original layout of integrated models considering structural.In the research of identification of track, this paper analyzes and compares two kinds of two value algorithm and jumping two value algorithms in the literature. And put forward a kind of dynamic jumping algorithm based on this, the algorithm can also solve the problem of remote image gray gradient and the light intensity change problem.Developed a smart car debugging platform, the platform is mainly used for debugging image processing algorithm and control algorithm, in order to improve the efficiency of intelligent vehicle development. Making the selection, hardware platform, software writing and other aspects have been described in detail in this paper. In the production of the hardware circuits and driver software debugging after the completion of the work, the platform is used in practice. After the competition, the smart car model which was made to obtain two prize in the fourth "fly think of Carle" Cup National Smart car contest. In a number of MATLAB simulation experiments, the dynamic jumping algorithm has better treatment effect. In testing, the smart car debugging platform to some extent improved the debugging development efficiency, reduced labor intensity of the developer.