Model Predictive Control for Optimizing the Overall Dredging Performance of a Trailing Suction Hopper Dredger
Type:
Presented during:
WODCON XVIII - 2007 Orlando, Florida, USA
Authors:
Braaksma, et al
Abstract: Trailing suction hopper dredgers are widely used to excavate sand from the sea bottom. One or two operators control the ship who usually aim to maximize the production of sand. This production depends on the incoming sand mass flow and on the losses during the overflow phase. The incoming flow-rate, the density of the incoming mixture and the grain-size and distribution all influence the sedimentation rate. Currently, the operators’ strategy is to maximize the incoming production from the drag head, while little attention is paid to the sedimentation process. This article describes an improved control strategy based on model predictive control to optimize the total process. A predictive control strategy is necessary because the results of the control actions are only observed at the end of the dredging process. The control inputs are the pitch of the screw blades to control the ship’s speed, the diesel engine speed to control the mixture flow-rate and the drag head visor angle to control the incoming density in the drag head. The optimization objective is to maximize the sand content in the hopper at the end of the cycle, the tons of dry solids (TDS). As the dredging cycle also includes sailing and discharging, the optimal dredging strategy depends on the total cycle time. Therefore the degrees of freedom to achieve the optimization objective are the control inputs as well as the dredging time. Our simulation results show that the production-oriented strategy, currently used by operators, is suboptimal. The achievable improvement depends on the type of sand and ranges between 5 % and 11 % for coarse and fine sand.
Keywords: -