Robust Networked Control Adopting Prediction–Compensation Mechanism for IoT-Based Building Field-Level Control Concerning Network Uncertainties

Abstract

Internet of Things (IoT) technologies offer great potential benefits to the development of smart buildings. However, the functionalities of IoT applications in buildings, especially those involving time-critical control tasks, are still limited due to the strict real-time and reliability requirements. These tasks could be easily affected by network uncertainties in the IoT environment. Current optimization methods aimed at mitigating network impacts have limitations in their applications and often overlook the impacts in real engineering cases. This study, therefore, proposes a robust networked control adopting the prediction-compensation mechanism to improve the robustness of building field-level controls implemented in the IoT-enabled building automation system. The control mainly consists of a predictor to estimate the controlled variable, and a compensator to evaluate the uncertainties. To assess the performance and the improvement on control robustness, a typical time-critical building field-level control task is implemented in a networked building field-level control simulation platform, considering network uncertainties. The proposed robust control is adopted for implementing the control task. The results show that the proposed robust networked control is a promising option due to its significant improvement in the control robustness when affected by network constraints, especially in critical conditions of the control process.