Analysis of PWA control of discrete-time linear dynamics in the presence of variable input delay. (30th September 2017)

• Record Type: Journal Article
• Authors: Laraba, Mohammed-Tahar; Olaru, Sorin; Niculescu, Silviu-Iulian
• Is Part Of: IFAC journal of systems and control. Volume 1(2017)
• Page Start: 24
• Page End: 36
• Publication Date: 2017-09-30
• Subjects: Time-delay systems -- Model predictive control -- PWA controller -- Delay margins; Automatic control -- Periodicals; Relay control systems -- Periodicals; Embedded computer systems -- Periodicals; Feedback control systems -- Periodicals; Artificial intelligence -- Periodicals; Artificial intelligence; Automatic control; Embedded computer systems; Feedback control systems; Relay control systems; Electronic journals; Periodicals; 629.89
• DOI: 10.1016/j.ifacsc.2017.06.002
• Languages: English
• ISSNs: 2468-6018

Abstract

Abstract: This paper focuses on the robustness of a specific class of control laws, namely the piecewise affine (PWA) controllers, defined over a bounded region of the state-space. More precisely, we are interested in closed-loop systems emerging from linear dynamical systems controlled via feedback channels in the presence of varying transmission delays by a PWA controller defined over a polyhedral partition of the state-space. We exploit the fact that the variable delays are inducing some particular model uncertainty. Our objective is to characterize the delay invariance margins : the collection of all possible values of the time-varying delays for which the positive invariance of the corresponding region is guaranteed with respect to the closed-loop dynamics. These developments are proving to be useful for the analysis of different design methodologies and, in particular, for model predictive control (MPC) approaches. The proposed delay margin describes the admissible transmission delays for an MPC implementation. From a different perspective, the delay margin further characterizes the fragility of an embedded MPC implementation via the on-line optimization and subject to variable computational time.