Predictive Analysis of Integrated Operational Energy Consumption in Public Buildings Based on GAWOA-BiLSTM Neural Networks

As an important part of urban energy consumption, the study of energy consumption in public buildings is of great significance for optimising energy management, reducing carbon emissions and achieving sustainable development goals. In this paper, a comprehensive operational energy consumption prediction model for public buildings based on GAWOA-BiLSTM neural network is constructed for the long-term dependence and non-linear characteristics of public building energy consumption data. The model improves the traditional Whale Optimisation Algorithm (WOA) by modifying the population initialisation method and introducing corresponding optimisation measures for the weight coefficients, so as to obtain the Global Adaptive Whale Optimisation Algorithm (GAWOA) with higher optimal parameter search performance. And combined with the ability of BiLSTM in time series data processing, it shows high accuracy and stability in public building energy consumption prediction, providing accurate prediction support for energy management and load regulation of buildings.