Stochastic Optimal Control of HVAC system for Energy-efficient Buildings
This paper aims to develop an agile, adaptive and energy-efficient method for HVAC control via Markov decision process (MDP). Our main contributions are outlined First, we formulate the problem as a MDP, which incorporates i) the multiple uncertainties resulting from the weather and occupancy, ii) the elaborate Predicted Mean Vote (PMV) thermal comfort model. Second, to cope with the computational challenges, we propose a gradient-based policy iteration (GBPI) method to learn the policies based on the performance gradients. Thrid, we theoretically prove that the method can converge to an optimal policy of the formulated MDP. The advantages of the proposed method are that: i) it uses off-line computation to learn control policies thus reducing on-line computation burden, and ii) it handles the non-convex and nonlinear system dynamics efficiently and can accommodate the non-analytical thermal comfort models (e.g., PMV) in the literature. The favorable performance of the policies yield by the GBPI is demonstrated through comparisons with the optimal solution obtained by assuming all the information is available before the planning in several case studies.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
MILP-based Imitation Learning for HVAC control
To optimize the operation of a HVAC system with advanced techniques such as artificial neural network, previous studies usually need forecast information in their method. However, the forecast information inevitably cont…
Imitation LearningHVAC-DPT: A Decision Pretrained Transformer for HVAC Control
Building operations consume approximately 40% of global energy, with Heating, Ventilation, and Air Conditioning (HVAC) systems responsible for up to 50% of this consumption. As HVAC energy demands are expected to rise, o…
In-Context Reinforcement LearningReinforcement Learning (RL)LSTM-based Space Occupancy Prediction towards Efficient Building Energy Management
Energy consumed in buildings takes significant portions of the total global energy usage. A large amount of building energy is used for heating, cooling, ventilation, and air-conditioning (HVAC). However, compared to its…
energy managementManagementPredictionTime Series AnalysisLaxity-Aware Scalable Reinforcement Learning for HVAC Control
Demand flexibility plays a vital role in maintaining grid balance, reducing peak demand, and saving customers' energy bills. Given their highly shiftable load and significant contribution to a building's energy consumpti…
reinforcement-learningReinforcement LearningReinforcement Learning (RL)Go Beyond Black-box Policies: Rethinking the Design of Learning Agent for Interpretable and Verifiable HVAC Control
Recent research has shown the potential of Model-based Reinforcement Learning (MBRL) to enhance energy efficiency of Heating, Ventilation, and Air Conditioning (HVAC) systems. However, existing methods rely on black-box …
Model-based Reinforcement Learning