Abstract:

An identification method for physical parameters of joint of robot is presented based on fuzzy automatic reasoning method. The method is a kind of on line adaptive fuzzy reasoning which is deduced based on fuzzy clustering method. The method can be used in parameters identification of time-varying system. Also a 3-DOF single arm robot is used as research object, and a dynamic model of the moving robot is built. The adaptive fuzzy reasoning method is used to identify the time-varying physical parameters of the robot joint. The computing error can be reduced because the local estimating method is applied. Experiment result shows that the method has better computing precision and identification speed, and has application value in practice engineering.

Key words: robot, non-linear parameter identification, recursive fuzzy clustering, dynamics model