Multi-agent control for centralized air conditioning systems serving multi-zone buildings

Abstract

Coordinating different components in a complex air conditioning system is challenging for centralized controls due to the large number of optimization variables. In this scenario, de-centralized controls are more appropriate alternatives. This study proposes a multi-agent control methodology for the optimal control of centralized air conditioning systems that are typically adopted in multi-zone commercial buildings. A hierarchical multi-agent framework is developed in which the agents cooperate to find the optimal operating point. Two consensus-based distributed optimization algorithms are formulated for this specific type of problem, which form the underlying mechanism of intra-agent optimization and inter-agent cooperation. Finally, a 3-zone building case study is used to demonstrate the performance of the proposed approach.