The Study Of Robot Manipulators Fuzzy Modeling

The control problems of robot have received great attention in theoretical and engineering for many years. When the robot model is exactly known, the traditional technique in nonlinear systems can solve the problems very well. However, in fact when the robot is a complex variable, the strong coupling, nonlinear systems, its analytic modeling process needs to be done lots of simplify and assumption, so that the modeling accuracy is extremely limited. In addition to the changes in load and the impact of external disturbances, it virtually impossible obtains a precise mathematical model. In recent years the fuzzy model used to describe the highly nonlinear and uncertainty of the actual system is very successful, there has also been a lot of fuzzy modeling and applications and related research results. This makes the study that how to fuzzy modeling and robust adaptive control for robotic system has not only theoretical meaning but also practical value.In this dissertation, the robotic system with unknown model is regarded as controlled plant. It is focus on fuzzy modeling and robust adaptive control of the robotic system using the references available, specific tasks as follows:Firstly, an offline fuzzy modeling method of robotic system is firstly presented, which analyzed the clustering number and fuzzy coefficient deeply and improved the fuzzy c-means clustering algorithm, resulting in a higher precision approach to the fuzzy modeling of robotic dynamics.Secondly, against real-time control of robotic system, an online modeling method was proposed, which can be achieved the identification and updating of fuzzy model online and avoiding the complexity of the system. The capable of outside interference and other uncertainty factors included in the fuzzy model, provide a good foundation for robust adaptive control of robotic system.Thirdly, based on the online fuzzy modeling, an adaptive controller ispresented for tracking control of uncertain robot, whose fuzzy model can be identification and updating online, and after that the adaptive law is given. This method achieved a high-precision and fast-track control of uncertain robot and opened up a new approach for robust adaptive control of the robotic system.