Energy Efficiency Analysis and Prediction of Trailing Suction Dredger Mud Pump Based on Grey Correlation Analysis and Stacking Strategy

To optimize the construction parameters of trailing suction dredger, as well as reduce the energy consumption and carbon emission of dredging operation, the energy efficiency analysis and prediction of trailing suction dredger mud pump are carried out. Firstly, the energy consumption per unit productivity (
$EC_{\text{pup}}$
) of dredging and its six influencing factors are analyzed by grey correlation analysis, and the importance of various factors affecting energy efficiency is obtained. Then, a learner based on stacking strategy (LBSS) composed of support vector regression (SVR), random forest regression (RFR) and light gradient boosting machine (LightGBM) models is used to predict the energy efficiency of the trailing suction dredger mud pump. The results show that the importance of the factors that affecting the energy efficiency of the mud pump are as follows: mud pump vacuum > mud flow velocity > mud pump pressure > mud density> mud concentration> mud pump rotation speed. The prediction performance of LBSS model is better than SVR, RFR and LightGBM models, and higher accuracy is achieved in predicting the energy efficiency of mud pump. During the dredging operation of the dredger, the operator can optimize the construction parameters according to the energy efficiency prediction results and the importance of various factors. The study can provide guidance and reference for real-time optimization, energy conservation and emission reduction in dredging construction of the same type dredger.