Control System Based On Adaptive Filtering Techniques

The goal of control theory is to make a given dynamical system (the "plant") behave in a user-specified manner as accurately as possible. This objective may be broken down into two separate tasks: control of plant dynamics; and control of plant disturbance. Conventionally, one uses feedback to treat all two problems simultaneously. Compromises are necessary to achieve good solutions.Adaptive inverse control is a method to treat the two control tasks separately. First, the plant is controlled using a feedforward controller; secondly, a disturbance canceller is used to reject plant disturbances. Adaptive filters are used as controller and disturbance canceller, and algorithms adapt the transfer functions of the filters to achieve excellent control.This dissertation studies the adaptive inverse control for linear and nonlinear plants. An important part of this work is the development of an ameliorative algorithm of normalized variable stepsize for adaptive filtering (ANVS). Based on the improvement of MNVS, ANVS has high convergent speed and fast tracking ability. In chapter 4 an improved adaptive control of plants disturbance is proposed. By predicting the disturbance in the circuit, the delay of the noise that caused by the digital feedback circuit can be compensated efficiently. Then the ability of the control of plant disturbance is improved. In chapter 5 a method of designing state feedback system based on Chebyshev I filter principle is proposed. This method avoids the complicated calculations of linear quadratic regulators.To the nonlinear plants, the shortage of normal inverse control method for nonlinear plants is discussed in chapter 6. ANFIS is effectively applied to nonlinear inverse dynamic control system. A project about how to design nonlinear inverse dynamic controller is also discussed here. It effectively avoid the inherent imperfects in the normal inverse control methods. These methods show a great promise.Simulation results are presented to verify the analysis and synthesis methods. Overall, excellent results are obtained.