Model modification and state estimation for an experimental planar two-link flexible manipulator by using an efficient approach

In this paper, an efficient approach of model modification and state estimation for an experimental planar two-link manipulator possessing significant structural flexibility is presented. The dynamic model of the manipulator is derived by employing the Lagrangian approach in conjunction with the assumed modes method. This strongly nonlinear dynamic model is then linearised around a set of general operating points [2]. In order to represent the un-modelled dynamic and compensate inaccuracy of linearised model over the whole operating area, an extra input for each link is introduced, and an augmented model is obtained. Based on this model, all states and the two extra inputs are estimated by using the "strong tracking filter" (STF) [1], Experimental results demonstrate the effectiveness of the proposed approach.