Productivity analysis of trailing suction hopper dredgers using stacking strategy

摘要

Trailing Suction Hopper Dredger (TSHD) is commonly used in dredging construction operations. Accurate productivity estimation and optimization are important for an efficient and smooth operation of TSHDs. An intelligent approach using stacking strategy for estimating TSHD productivity is proposed in this study. The proposed method involves two major modeling stages. First, the ReliefF-Granger algorithm is used to analyze data collected during real-world operation in order to extract the key factors influencing productivity from a vast amount of monitoring data. Several Artificial Intelligence (AI) algorithms were subsequently applied to fuse the extracted factors in order to achieve better generalization capability. In the second stage, a variety of heterogeneous AI models were adopted for mixed feature training prediction, and an optimal combined model was obtained through a grid search algorithm. To verify its accuracy and applicability, the proposed method was applied to a channel deepening project as a case study.