A Novel Nonlinear Model Predictive Control Method Based on LMI and Feedback linearization.pdfVIP

A Novel Nonlinear Model Predictive Control Method Based on LMI and Feedback Linearization 1 1 Rui Pang , Zhongke Shi 1.School of automation, Northwestern Polytechnical University, Xi’an, Shaanxi E-mail: pangrui517@ Abstract: Nonlinear model predictive control NMPC mainly deals with complex dynamic systems with different kinds of nonlinearities, which is much more difficult to solve than normal linear model predictive. The non-convex optimization method inside MPC is its main challenging problem. A LMI based NMPC method is proposed combined with nonlinear feedback linearization, which can efficiently convert complex non-convex optimization problem to convex. The method minimizes the upper bound of cost function instead of itself, which can stabilize the whole system and achieve the control performance. A high angle of attack HAOA aircraft controller design example is illustrated in the end to demonstrate the validity of proposed method. Key Words: NMPC, LMI, convex optimization, feedback linearization Recently, nonlinear model predictive control NMPC is the 1 Introduction major area of MPC theory research[10]-[11]. In [12], the Model Predictive Control, also known as Receding scientists proposed a method which designed a Horizon Control, is a modern feedback control methodology state-feedback control law and minimize an infinite horizon based on optimization theory. Its research and applications cost function within the framework of linear matrix on Linear Time Invariant System LTIS were studied from inequalities. This method successfully converts a 1970s and have gained much progress, while the research on non-convex quadratic optimization problem to a convex nonlinear system is much more complicated and is quadratic optimization program. developing at recent years. The paper is structured as follows: First we introduce the MPC is one kind of tracking control, which