Prediction of Dredging Productivity Using a Rock and Soil Classification Model

Abstract

The prediction of dredging productivity plays an important role in controlling the costs and optimizing the scheduling of the dredging process. However, it is very difficult to achieve satisfactory forecasting efficiency because there are so many uncertain variables, including the rock, soil, and water properties; the main performance index of the dredger; and environmental restrictions. In this paper, the authors have developed a quantitative classification model for dredging materials under complex conditions, in which the weight of the condition attributes is calculated using rough-set theory and conditional entropy. Then, by considering the effect of the rock and soil conditions, the main performance index of the dredger, and the influence of the underwater environment, a prediction model for dredger production efficiency was developed. This approach is applied to a land reclamation project in Tianjin, China. In the study case, the efficiency of the dredger production is predicted fairly well when compared with the realistic conditions, so this model provides a feasible method and a technical tool for dredging cost control and schedule planning.