研究目的
Investigating the optimization of thermal and electrical energy consumption in the ADREAM building through dynamic simulations and intelligent energy management.
研究成果
The paper concludes that combining different types of models is beneficial for optimizing the energy management of Smart Buildings. The Simulink model, in particular, offers a comprehensive approach to simulating the global functioning of the building and its systems, with potential for further exploration in combination with electrical network models.
研究不足
The study acknowledges the complexity of simulating the interaction between ANN and physical elements, which impacts the robustness of the ANN models. Future work aims to explore more efficient control strategies for energy optimization.
1：Experimental Design and Method Selection:
The study employed three modeling techniques: Dynamic Thermal Simulation (DTS) using Pleiades+Comfie, a 'Black Box' model using Artificial Neural Networks (ANN), and a physical model developed in Matlab/Simulink.
2：Sample Selection and Data Sources:
The ADREAM building, equipped with 6500 integrated sensors and a wide network of embedded systems, served as the sample. Data was collected from in situ measurements, sensor data, and occupant surveys.
3：List of Experimental Equipment and Materials:
The study utilized software tools Pleiades+Comfie and Matlab/Simulink, along with the building's HVAC systems.
4：Experimental Procedures and Operational Workflow:
The DTS was calibrated based on measurements and surveys. The ANN model was trained with past outputs to predict system parameters. The Simulink model was developed to simulate the interaction of all systems with the building.
5：Data Analysis Methods:
Statistical indicators NMBE and CVRMSE were used to validate the models.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
