Prediction of Road Construction Costs Using an ANN Algorithm

초록

This paper proposes a construction cost prediction model through the application of Artificial Neural Networks (ANN) based on data available in the planning stage of road construction. To build the optimal ANN model, the predicted values of various input variables were compared with the actual construction costs using correlation coefficients and RMSE. The results indicate that the model utilizing total road length, road width, and design speed exhibited the best performance. An analysis of the relative importance of input variables on the predicted values showed that design speed had the most significant impact at 60.47%, while total road length had the lowest impact at 17.88%. This suggests that as design standards become more complex, construction costs increase accordingly. Thus, incorporating ANN-based predictions into future construction cost forecasting will enable efficient and reliable predictions that can account for increasingly complex design criteria.

키워드