Generalized efficient robust predictive control for networked interval type-2 T–S fuzzy system with adaptive event-triggered scheme
Xiaoming Tang,Jialiang Wang,Kun Zhao,Xiao Lv
DOI: https://doi.org/10.1016/j.isatra.2023.08.025
IF: 7.3
2023-08-27
ISA Transactions
Abstract:Highlights • A synthesis method of generalized efficient robust predictive control (GERPC) is proposed for IT-2 fuzzy NCSs with data dropout, adaptive event-triggered scheme (AETS) and bounded disturbance. • Design a fixed feedback gain offline, and the control perturbation introduced is calculated online. • Introduce an adaptive event-triggered scheme with dynamically adjustable threshold to avoid network communication congestion. • The quadratic bounded (QB) technique and the convex LMI reformulation method deal with bounded disturbance and the non-convex terms, respectively. • Designing controller based on GERPC can obtain good control performance, larger initial feasible region and smaller online computation. This article studies the generalized efficient robust predictive control (GERPC) for the nonlinear systems denoted by interval type-2 (IT2) Takagi–Sugeno (T–S) fuzzy model over networks with consideration of packet dropout, bounded disturbance and adaptive event-triggered scheme (AETS). First, the packet dropout is described by a Bernoulli process, and an AETS with thresholds that can be self-regulated according to different situations is employed to decide whether to release data into the communication network, so as to reduce the data transmission frequency and the network burden; Second, taking into consideration of packet dropout and AETS, a unified IT2 T-S fuzzy control model based on GERPC is established, and the stability of closed-loop system with bounded disturbance is guaranteed by quadratic boundedness (QB) technique; Third, generalized efficient robust predictive controller is designed to obtain a larger initial feasible region, smaller online computation and acceptable control performance. The unconstrained control law is designed offline, and a new variable will be optimized online to achieve good control performance, which is introduced to adjust the control quantity. Then, the convex LMI reformulation method is utilized to handle non-convex terms in optimization problem, and the effectiveness of the GERPC algorithm is proved by the matrix partition technique. At last, the effectiveness of the designed controller based on the GERPC strategy is verified by a given simulation experiment.
automation & control systems,instruments & instrumentation,engineering, multidisciplinary