Authors: Aseel Hussein, Abid Abu- Tair and Halim Bousabaine
Modelling the environmental impacts on build environment assets still have a lot of setbacks mainly due to the complexity of ecological system and the stochastic nature of the variables. Therefore, Artificial Intelligence technique, Neural Network is proposed. Through the learning mode of ANN, environmental risk analysis can be established for each project. The variables records are gathered from historical data. Backpropagation technique was used to compensate for the missing data. Designing Neural Networks model differ from project to another, so this research would explore the effectiveness of using Neural Networks in modelling the environmental risk impact. Data were mainly divided into three models based on the environmental condition. The analysis results indicate Neural Network Model 1 (Mild condition) has the best prediction for service life by achieving the highest regression rate. The outcome shows the importance of data type on Neural Network learning and predicting assets service life.
Keywords: Environmental Risks, Service Life, Bridge Deterioration, Artificial Neural Networks