A design procedure of fuzzy PD control for mechanical systems

This paper presents a design procedure of fuzzy PD controllers for trajectory control in mechanical systems, which relies on the practical success of popular fuzzy PD control and robust control design techniques. The proposed control consists of a conventional fuzzy PD and a robust compensation component aimed at compensating the hard nonlinearities common in servo mechanisms. The fuzzy PD control uses the tracking error and its derivative to generate a control action according to an expert experience. The design of this fuzzy control also gives a linear parametrization that allows the design a ldquoplug-inrdquo robust control component aimed at increasing the tracking performance. Instead of assuming a model of nonlinearities or a bound on their effects, this robust fuzzy PD control is designed relying on the fact that PD-type controls can yield an acceptable tracking performance, a compensation is required only for the purpose of improving the tracking accuracy. Simulation and experiment results are included to illustrate the proposed procedure.