Model-based control of hopper dredgers

Model-based control of hopper dredgers

Braaksma, J.

Babuska, R. (promotor)
Klaassens, J.B. (promotor)

Mechanical Maritime and Materials Engineering

2008-02-04

The modern trailing suction hopper dredgers are advanced ships that are equipped with many automation systems that can be controlled with integrated computer systems from the bridge. From the operators it is expected that they generate the right set-points for all these systems. The latest ships are designed such that one operator is responsible for controlling the complete ship. This can only be achieved with a high degree of automation. The determination of the set-points is not unambiguous, since vital information such as the soil type, which is necessary to make a correct decision, is not always at hand. The subject of this thesis is the development of a control system that strives to optimise the dredging process operation. This control system must take into account all the aspects that influence the performance, such as the excavation process, the pumping process and the sand sedimentation process in the hopper. Whereas the consequences of the control actions are only observable in the performance at the end of the cycle, a control strategy has been chosen which is based on model predictive control (MPC). The results are based on numerical simulations and compared with measured data and simulated scenarios. These scenarios are comparable with the work method in practice. The improvement compared with the scenarios varies between 2% for coarse sand and 22% for fine sand, calculated in cubic metre sand per second for a complete dredging cycle. An important conclusion is that the sand sedimentation process in the hopper plays in important role when determining the optimal set-points for sand with a grain size smaller than 0.3 mm. Since in practice the focus of operation is on the excavation production, an increase up to 22% can be achieved by taking the sedimentation process into account.

optimisation
control
model predictive control
patternsearch
particle filter
industrial application

http://resolver.tudelft.nl/uuid:c39f031e-4b1a-448f-b33a-41a2679fd15c

978-90-9022693-4

Institutional Repository

doctoral thesis

(c) 2008 J. Braaksma