Construction cost prediction model for agricultural water conservancy engineering based on BIM and neural network

Due to the complex construction conditions, long work cycles, and high uncertainty inherent in agricultural water conservancy projects, accurate construction cost prediction is crucial for investment decisions. This study presents an innovative cost prediction model for these projects, integrating BIM with neural networks. Firstly, BIM technology is utilized to digitize and visualize engineering-related information. Subsequently, a prediction model based on SSA optimized PGNN is constructed. The digital data obtained from BIM is subsequently integrated with the prediction model to estimate the construction costs of agricultural water conservancy projects. In this study, actual engineering projects are selected as case studies, utilizing material price data from January 2016 to February 2021 in Liaoning Province, along with real project data for modeling purposes. The results indicate that the maximum relative error between the predicted and actual values of the combined model is only 2.99%. Furthermore, the RMSE and R² of the simulated prediction results are 0.1358 and 0.9819, respectively. The proposed model demonstrates higher prediction accuracy and efficiency. Compared with the PGNN model, the RMSE is reduced by 33%, and R² is increased by 6%. These findings suggest that the BIM-SSA-PGNN prediction model provides more accurate and efficient construction cost predictions for agricultural water conservancy projects, promoting technological integration and innovation while optimizing construction project costs. This study provides a scientific basis for management to promote the transformation of the industry towards digital and intelligent sustainable development.