Development of stochastic deep learning model for the prediction of construction noise

Abstract

Construction noise is an occupational noise that is potentially harmful, and it usually originates from machinery in construction sites. The impact of construction noise on the health and safety of construction workers is one of the main concerns in the industry. Prolonged noise exposure duration will cause physiological, physical, psychological damage, and negatively affect the working performance of the workers as well. Furthermore, the adverse impacts arising from construction noise may jeopardize public welfare, particularly for those who live nearby the construction site. Therefore, this research aims to develop a Stochastic Deep Learning (SDL) noise prediction model by integrating stochastic modelling and deep learning techniques. Stochastic modelling is applied in this study to generate a set of simulated randomized data based on several parameters as the input for the deep learning model. The deep learning model is trained with the input from stochastic modelling to predict the noise levels emitted from the construction site. The predictive performance of the deep learning model will be assessed with several statistical measures such as absolute difference, mean absolute difference and root mean square deviation. Ten case studies are conducted to validate the reliability and accuracy of the SDL noise prediction model. The SDL model showed high accuracy of prediction results with an average absolute difference of less than 1.2 A-weighted decibels (dBA) among the case studies as compared to the measurement. The reliability of the results from the prediction model is high. In conclusion, the SDL model is established and provided a promising outcome with satisfactory predictive performance. Finally, further development of the model would be worthwhile to fully exploit the potential of the SDL noise prediction model in construction industries as a planning, managerial and monitoring tool.

Keywords

• Artificial Neural Network
• Construction Noise
• Deep Learning
• Noise Pollution
• Noise Prediction
• Stochastic Modelling