Related Experiment Videos
Construction project risk prediction model based on EW-FAHP and one dimensional convolution neural network
Yawen Zhong1, Hailing Li2, Leilei Chen3
1School of Engineering, Southwest Petroleum University, Nanchong, China.
Plos One
|February 9, 2021
Summary
This study introduces an advanced construction project risk prediction model using Entropy Weight Method-Fuzzy Analytic Hierarchy Process (EW-FAHP) and One Dimensional Convolutional Neural Network (1D-CNN). The model achieves high accuracy, with average absolute errors below 0.1% for predicting construction period and cost risks.
Area of Science:
- Construction Management
- Artificial Intelligence
- Risk Analysis
Background:
- Traditional construction project risk prediction models often suffer from low accuracy.
- Accurate prediction of construction period and cost risks is crucial for project success.
Purpose of the Study:
- To develop a novel, highly accurate risk prediction model for construction projects.
- To improve the precision of predicting construction period and cost risks.
Main Methods:
- Literature analysis to select risk evaluation indices.
- Combining Entropy Weight Method (EW) and Fuzzy Analytic Hierarchy Process (FAHP) to determine comprehensive risk index weights.
- Utilizing a One Dimensional Convolutional Neural Network (1D-CNN) for risk prediction.
Main Results:
- The proposed EW-FAHP and 1D-CNN model demonstrates significantly improved accuracy in risk prediction.
- Experimental results show average absolute errors below 0.1% for construction period and cost risk predictions.
- The model effectively addresses the limitations of traditional risk assessment methods.
Conclusions:
- The developed model offers a robust solution for accurate construction project risk prediction.
- The integration of EW-FAHP and 1D-CNN provides a powerful tool for enhancing project management and mitigating financial and temporal risks.