Automated Control of Transactive HVACs in Energy Distribution Systems

Heating, Ventilation, and Air Conditioning (HVAC) systems contribute significantly to a building's energy consumption. In the recent years, there is an increased interest in developing transactive approaches which could enable automated flexible scheduling of HVAC based on customer demand electricity prices decided by suppliers. Flexible make it prime source for participation residential response or systems. Therefore, significant identify optimal strategy control this article, reducing cost while keeping comfort level acceptable users, we argue that such should consider both user simultaneously. Accordingly, develop through solution optimization problem balances between consumer's dissatisfaction. This enables us solve decision-making first price prediction then choosing temperature settings throughout day predicted price, history settings, outside temperature. More specifically, formulate design as Markov decision process (MDP) using deep neural networks use Deep Deterministic Policy Gradients (DDPG)-based reinforcement learning algorithm find comfort.