Forecasting urban construction waste generation utilizing the TI-TSTLNGM(1,1) Model based on the COOT Algorithm.

Abstract

To address the challenges in scientifically estimating and predicting construction waste (C&W) generation trends, a novel forecasting method and an enhanced grey model are presented. By transforming the initial sequence and incorporating the three-parameter interval grey number, the model effectively resolves data volatility and incompleteness. For the first time, the combination of the three-parameter interval grey number with waste generation coefficients is introduced to determine the range of C&W generation, providing a novel and more accurate approach for prediction. The time lag coefficient is further optimized using the COOT algorithm to improve prediction accuracy. Taking Guangzhou as a case study, the model demonstrates superior performance compared to traditional grey prediction and exponential smoothing methods, providing crucial reference values for C&W disposal planning and policymaking. The empirical analysis conducted in this study validates the model's rationality, applicability, and effectiveness, while providing a scientific foundation for regional construction waste treatment, disposal, resource allocation, and comprehensive management.