Dynamics Identification of a Flexible Robot using Multichannel Nonlinear Systems

This paper presents a method of a link displacement identification in flexible robots. The displacement caused by the payload and torques applied at the joints causes serious problems for position/force control. Displacement identification substantially enhances control of flexible manipulators, and it has been implemented by using nonparametric identification algorithm for nonlinear, multichannel systems. A basic multichannel system applicable two one-link flexible manipulator consists of two parallel subsystems: a nonlinear, memoryless subsystem and a dynamic, linear subsystem. We identifiy parameters of the linear subsystem by the correlation method. The nonlinearity is recovered. by the nonparametric estimate based on the Hermite series expansion. The identification algorithm is very powerful: it recovers nonlinearities regardless of their functional forms. Both local and global convergence of the algorithm is obtained for all input densities. Its rates are also investigated for nonlinearities of Lipschitz type. The method is currently being verified in simulation experiments.