Energy-aware robust Model Predictive Control based on wireless sensor feedback

Abstract

Flexibility, ease of deployment and of spatial reconfiguration, and low cost make wireless sensor networks (WSNs) fundamental component of modern networked control systems. However, due to the energy-constrained nature of WSNs, the transmission rate of the sensor nodes is a critical aspect to take into account in control design. Two are the main contributions of this paper. First, a general transmission strategy for communication between controller and sensors is proposed. Then, a scenario with a controller and a wireless node providing measures is investigated, and two energy-aware control schemes based on explicit model predictive control (MPC) are presented. We consider both nominal and robust control in the presence of disturbances, and convergence properties are given for the latter. The proposed control schemes are tested and compared to traditional MPC techniques. The results show the effectiveness of the proposed energy-aware approach, which achieves a profitable trade-off between energy consumption of wireless sensors and loss in system performance.