Title
A novel dynamic progress forecasting approach for construction projects
Abstract
In this paper, we propose a novel construction project progress forecasting approach which combines the grey dynamic prediction model and the residual modified model to forecast the current progress during the construction phase. Firstly, four typical S-curves simplified from various sigmoid curves are proposed and fitted to the grey dynamic prediction model. For higher prediction accuracy, three different residual modified models are taken to amend the initial prediction value which was derived from the above step. The mean absolute percentage error (MAPE) and standard deviation of the estimate of Y (SDY) are used to assess the accuracy of the composite results. The better residual modified prediction model is adopted to combine the grey dynamic prediction model to form the novel progress forecasting approach. Then, practical completed construction cases are provided for testing the prediction ability of the proposed progress forecasting approach. Results show that the forecasting approach proposed to forecast construction progress during construction phase is able to get better prediction accuracy almost within 10% whether typical S-curves or practical cases. The new approach relatively provides an accurate, simple and stable method for predicting construction progress in comparison with the previous traditional forecasting methods.
Year
DOI
Venue
2012
10.1016/j.eswa.2011.07.093
Expert Syst. Appl.
Keywords
Field
DocType
higher prediction accuracy,typical s-curves,construction phase,better prediction accuracy,initial prediction value,prediction ability,grey dynamic prediction model,forecasting approach,construction progress,construction project,residual modified prediction model,novel dynamic progress forecasting,polynomial,forecast
Mean absolute percentage error,Residual,Data mining,Polynomial,Computer science,Artificial intelligence,Dynamic prediction,Standard deviation,Machine learning,Sigmoid function
Journal
Volume
Issue
ISSN
39
3
0957-4174
Citations 
PageRank 
References 
2
0.42
1
Authors
4
Name
Order
Citations
PageRank
M. Chiao Lin120.42
H. Ping Tserng250.83
S. Ping Ho320.42
D. L. Young4164.59