building envelope

building energy

green roof

thermal comfort

air temperature

heating cooling

severe cold

Renewable Energy

The use of phase change materials (PCM) and its efficient integration can significantly lower the carbon emissions of building sector. However, optimal integration of thermal energy storages, including PCM, in the buildings is a challenging problem. Digital engineering offers a reliable mechanism to optimize such integration. In order to realize the usefulness of digital resources, especially on the global scale, a methodological approach is required based on validated and data-driven model(s) supported by optimal solution selection in the set of localized as well as external variables. This recommendation system, for the optimal solution, is exemplified here as a novel approach offering solution for an integrated solar chimney (naturally ventilated solar façade) for a building with passive energy storage based on PCM. The complete novel methodology is demonstrated for the sake of illustrating the procedure, followed by regional solutions and their impacts on a global scale. While the solution and results have significance in terms of their quantification; the real purpose and novelty of data engineering mechanism is the robustness and procedural illustration of the digital twin generation from the perspective of energy systems and building performances. The sequential procedure is summarized here with the case study of a solar chimney that has PCM storage in a building. The building is simulated for the four climatic zones covering a major chunk of the world map. The influencing variables are selected along with viable ranges for the prediction of optimal configurations. The multi-objective optimization leads to the region-specific characteristics of the geometry as well as phase-changing material (PCM) selections. While the response parameters have their own projections in their own specific climates. The results has indicated that multilayer perceptron artificial neural network model trained with Levenberg-Marquardt and Bayesian regularization algorithm having topology architecture 10-[10]-1 and 10-[40]-1 can ...

Data engineering for digital twining and optimization of naturally ventilated solar façade with phase changing material under global projection scenarios