Energy-aware robust Model Predictive Control with feedback from multiple noisy wireless sensors

Wireless Sensor Networks (WSNs) are becoming fundamental components of modern control systems. Although WSNs provide tremendous advantages in versatility, their use poses new issues in the design of the control system, in particular the discharge of batteries of sensor nodes, which is mainly due to radio communications, must be taken into account. In a previous work, for the case of a single wireless measurement device and no measurement noise, we have provided a general transmission strategy for communication between controller and sensors and an energy-aware robust Model Predictive Control (MPC) algorithm that achieve a profitable trade-off between transmission rate (battery energy savings) and loss of closed-loop system performance. In this paper we extend the approach by taking into account unknown but bounded noise on state measurements, and by considering a local area network of multiple wireless sensors measuring the state vector for disturbance rejection.