Predictive modeling of carbon emissions in Jiangsu Province's construction industry: An MEA-BP approach

Understanding and predicting carbon emissions in the construction industry is crucial for sustainable development planning. In this study, we employed the GM(1,1) model and the Mind Evolution Algorithm (MEA) to optimize the BP neural network model and forecast carbon emissions in Jiangsu Province from 2022 to 2026. Using data related to energy consumption and building material usage in the Jiangsu construction industry from 1995 to 2021, we estimated carbon emissions using the life cycle assessment method and IPCC carbon emission coefficient method. Additionally, we utilized the STIRPAT model and Principal Component Analysis (PCA) to select and reduce important variables, identifying three principal components. Subsequently, the MEA-BP model was employed to provide detailed forecasts of carbon emissions over the next five years, revealing a declining trend in carbon emissions from the construction industry. The study thoroughly analyzed the direct and indirect impacts of population growth, economic development, and industrial restructuring on carbon emissions, proposing targeted policy recommendations including sustainable urban planning and green building design. Future research could enhance the applicability and accuracy of the model by expanding the dataset and considering more uncertainty factors. This research offers valuable insights into the trends and influencing factors of carbon emissions in the construction industry of Jiangsu Province, providing guidance for policymakers and future research directions. •Combining the STIRPAT model with PCA principal component analysis for factor selection, three principal components were ultimately chosen.•Using the Mind Evolution Algorithm (MEA) to optimize the BP neural network, achieving good performance in testing.•Utilizing GM(1,1) for forecasting the input variables, followed by principal component analysis, and finally integrating the MEA-BP model to achieve a favorable predictive outcome.