Uncertain T-S model-based robust controller design withD-stability constraints—A simulation study of quadrotor attitude stabilization

• DOI: 10.1016/j.engappai.2017.08.021
• Publication Date: 2017-10-08T12:00:28Z
• Authors: Abdelmadjid Cherifi; Kevin Guelton; Laurent Arcese
• Keywords: 000; [MATH.MATH-DS]Mathematics [math]/Dynamical Systems [math.DS]; 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology; [SPI.AUTO]Engineering Sciences [physics]/Automatic

The problem of the attitude stabilization of a quadrotor unmanned aerial vehicle is considered in simulation. An uncertain-like TakagiSugeno model of the quadrotor, where the nonlinear entries which depend on the input motor voltages are reported in uncertainties, is proposed to avoid closed-loop algebraic loops. The design of a Parallel Distributed Compensation (PDC) control law is considered with D-stability constraints in order to improve the closed-loop transient response. In this context, new generic and relaxed Linear Matrix Inequality (LMI) conditions are proposed. These new PDC controller design conditions constitute an alternative to previously proposed D-stabilizing non-PDC controller ones. Indeed, non-PDC controllers require online inversions of time-varying matrices, which is a major drawback for embedded applications with low computational capabilities. The conservatism of the proposed LMI conditions with respect to previous results is illustrated through an academic example. Then, simulations of the quadrotor with usual prescribed LMI regions are proposed to show the effectiveness of the proposed conditions.