Distributed Control of Multizone HVAC Systems Considering Indoor Air Quality

Abstract

with these challenges, this work develops a two-level distributed computation paradigm for HVAC systems based on problem structures. Specifically, the upper level control first calculates zone mass flow rates for maintaining comfortable zone temperature with minimal energy cost and then the lower level strategically regulates the computed zone mass flow rates as well as ventilation rate to satisfy IAQ while preserving the near energy saving performance of the upper level control. As both the upper and lower level calculation can be implemented in a distributed manner, the proposed method is scalable to large multi-zone deployment. The method's performance both in maintaining comfort (i.e., thermal comfort and IAQ) and energy cost saving is demonstrated via simulations in comparisons with the centralized method, the distributed token-based scheduling strategy and the demand controlled ventilation strategies.

Proximal ADMM for Nonconvex and Nonsmooth Optimization

“…Proof of Prop. 5: By examining the terms of T k+1 c defined in (7), we only require to establish the lower boundness property of…”

“…We consider a case study with N = 10 zones and the predicted horizon T = 48 time slots (a whole day with a sampling interval of 30 mins). We set the lower and upper comfortable temperature bounds as T i min = 24 • C and T i max = 26 • C. The specifications for HVAC system can refer to [6,7]. We apply the proposed proximal ADMM to solve this problem in a distributed manner.…”

“…We therefore close the proof. Recalling the definition of the Lyapunov function in (7) and invoking (10), we have…”

“…Problem (P) is directly originated from smart buildings where smart devices are empowered to make local decisions while accounting for the interactions or the shared resource limits with the other devices in the proximity (see, for examples [6,7]). Many other applications also fit into this formulation, including but not limited to smart sensing [8], electric vehicle charging management [3,9], power system control [10], wireless communication control [11].…”

A Survey of ADMM Variants for Distributed Optimization: Problems, Algorithms and Features

“…The well-known alternating direction method of multipliers (ADMM) has emerged as one of the most popular tools for distributed optimization. It has found massive applications in broad areas ranging from statistical learning [5,6], multiagent reinforcement learning [7], imaging processing [8,9], data mining [10,11], power system control [12][13][14][15], smart grid operation [16][17][18][19][20][21][22], smart building management [23][24][25], multi-robot coordination [26], wireless communication control [27,28], autonomous vehicle routing [29,30] and beyond. The popularity of ADMM can be attributed to its many distinguishing advantages, such as modular structure, superior convergence, easy implementation and high flexibility.…”