MaOC: Model-Assisted Optimal Control for Ductless-Split Cooling Systems in Building Environments

Abstract

Demand for energy in buildings is growing exponentially, and cooling systems contribute to more than50% of buildings' energy consumption. With global warming, we need to save energy on the cooling systems. This paper targets spaces where multiple ductless-split cooling systems are deployed, commonly known as split ACs. Unlike ducted centralized cooling systems, they do not have central sensing and control. To optimize the energy consumption of the ductless-split cooling system, we propose a Model-assisted Optimal Control (MaOC) algorithm that observes the thermal environment of the room, measures efficiencies of the ACs in cooling the room, and generates an optimal execution schedule for the cooling system. We observe that the mathematical model generated for cooling systems follows the properties of the convex function. We define a MAXMIN problem to minimize energy consumption and maximize efficiency. We use the statistical distribution of cooling systems' efficiency to generate a long-term control trajectory. We evaluate MaOC for ductless-split cooling systems in a real-world environment and simulation. We compare it with solutions based on the greedy technique and Reinforcement Learning. It consumes 17% of energy compared to the one using a greedy technique and takes 34% less time to reach the desired temperature. We observe that MaOC consumes almost the same energy as a more complex one that uses Reinforcement Learning but takes less time to cool the room. In the simulation, we find that the energy consumption of MaOS is nearer to the optimum.