Application of cloud model - Bayesian network and genetic algorithm for assessment and optimization of building energy consumption

Building energy consumption (BEC) accounts for a large proportion of total global energy consumption, and has a huge prospect of energy conservation. In order to reduce building energy consumption, it is necessary to effectively evaluate the energy consumption level and optimize it. Due to the mutual influence of multiple factors on building energy consumption, there are often complex interactions and relationships between these factors, making the evaluation and optimization of building energy consumption a complex and uncertain problem. Therefore this paper aims to combine the cloud model (CM)-Bayesian network (BN) and genetic algorithm (GA) to build a comprehensive assessment and optimization model to solve this problem. First a comprehensive and applicable assessment indicator system and rating criteria for BEC is constructed. Second, the advantages of CM in reasoning multi-index complex uncertainty problems are utilized, and the quantitative value is converted into qualitative probability by BN to better solve the randomness and fuzziness of evaluation. Subsequently, the results of Bayesian inference analysis are used to determine the BEC optimization indexes and targets, and creatively inverse the index target values through the probability distribution of the index target level. Then, we selected a teaching building of a university in Wuhan for case validation and successfully assessment the BEC level using the model. Finally, based on the rating results and key indicators, GA is applied to quantitatively optimize the BEC indicators. Our models contribute to the accurate assessment and efficient optimization of BEC, thereby accelerating energy savings in the building sector.