Optimization method for energy consumption prediction of campus buildings based on Bayesian theory

Abstract

Education rapid developing has led to an increasing scale of buildings in universities, resulting in a significant increase in energy consumption. How to predict and optimize campus building energy consumption is a hot research topic. Based on this, a Bayesian theory was used to construct an optimization model for predicting energy consumption in campus buildings. Firstly, data mining techniques were used to analyze and model campus building energy consumption data, and an energy consumption prediction model was established. Then, the practical application effect of this method was evaluated through experimental verification. Finally, the actual data are used to test and analyze model prediction and optimization performance. The average accuracy, recall, and error of this Bayesian prediction in extracting energy consumption features of campus buildings are 89.63%, 92.46%, and 13.86%, respectively, which are significantly higher than the comparison algorithms. And in office buildings, mixed buildings, and dormitory buildings, the average energy consumption prediction errors optimized by this Bayesian prediction are 10.16%, 6.95%, and 11.26%, respectively. This Bayesian prediction has high accuracy and reliability in predicting and optimizing campus building energy consumption. This research aims to provide technical support and guidance for predicting campus building energy consumption. It provides useful reference and guidance for the construction of low-carbon, green, and sustainable development.