This paper investigates the accelerated distributed model predictive control (MPC) strategy for the heating, ventilation and air conditioning (HVAC) systems with local and global power input constraints. The problems are firstly formulated in the distributed MPC framework and then the constrained optimization is converted into a quadratic programming problem. In the problem formulation, the thermal couplings between immediate neighboring zones are considered while designing the distributed controller, and the unknown thermal disturbances are incorporated by the robust optimization scheme. Then, using the accelerated dual gradient-projection method, a distributed fast MPC protocol is designed for HVAC systems considering both the electricity cost and occupant comforts. A distributed stopping criterion based on the distributed average consensus algorithm is utilized. Finally, numerical simulations are used to demonstrate the effectiveness of the proposed distributed MPC algorithm, and its computational advantages comparing with an existing distributed method and a centralized algorithm.

本文研究了具有局部和全局电源输入约束的供热，通风和空调（HVAC）系统的加速分布式模型预测控制（MPC）策略。首先在分布式MPC框架中提出问题，然后将约束优化转化为二次规划问题。在问题表述中，在设计分布式控制器时考虑了相邻区域之间的热耦合，并且通过稳健的优化方案合并了未知的热干扰。然后，使用加速的双梯度投影方法，为HVAC系统设计了一种分布式快速MPC协议，同时考虑了电费和乘员舒适性。利用了基于分布式平均共识算法的分布式停止准则。最后，通过数值仿真证明了所提出的分布式MPC算法的有效性，以及与现有的分布式方法和集中式算法相比的计算优势。