Spatiotemporal Distribution Patterns and Prediction of Construction Material Prices in Sichuan Province

Abstract

Material prices play a decisive role in construction projects. In order to achieve material price prediction in Sichuan, this study focuses on four types of building materials: first-class sawn timber, shale porous bricks, rebar and white cement. By collecting multi-source data such as economy, climate and geography of cities (and regions) at various levels in 2024, an adaptive neural fuzzy reasoning system (ANFIS) model was built, and the training, verification and testing of the model was completed using MATLAB. The results show that the prediction performance of first-class sawn timber and rebar is relatively good, while the prediction accuracy of shale porous bricks and white cement is low. Research shows that the ANFIS model can be used for preliminary price estimation and trend reference for building materials (such as sawn timber and rebar) affected by macro factors, and provide data support for building materials price regulation and engineering cost management in Sichuan Province.