Intelligent Operation & Maintenance Regulation of Shopping Mall Air Conditioning Embedded with Policy Constraints ——Balancing Compliance, Energy Efficiency and Customer Thermal Comfort

Hong Bin, Lei Jian, Sun Fan, Lu Xiyao, Yu Lisha^*

Lingnan University

Abstract：

A zero-sum game dilemma exists among policy compliance constraints, energy conservation and customer thermal comfort for shopping mall HVAC systems. This paper constructs a three-way tension model of "Policy–Energy Efficiency–Customer Experience" and develops a five-layer intelligent regulation architecture integrating LSTM load forecasting, MPC rolling optimization and SAC reinforcement learning. Combined with on-site quasi-experiments and calibrated simulation verifications, the results demonstrate that the proposed method fully satisfies summer air temperature control policies. Compared with conventional constant-temperature control, the system achieves an energy-saving rate of 15%-30%, restricts indoor temperature fluctuation within ±1.0°C, and improves overall thermal comfort performance by 21.7%. This research provides effective technical support for low-carbon, policy-compliant and customer-friendly intelligent operation & maintenance of commercial buildings.Keywords: Shopping mall HVAC; Intelligent operation and maintenance; Policy constraints; Three-way tension model; LSTM-MPC reinforcement learning; Quasi-experiment verification.

Key Words：

shopping mall HVAC; intelligent operation and maintenance; policy constraints; three-way tension model; LSTM-MPC reinforcement learning; quasi-experiment verification