Constrained Optimal Thrust Allocation for C/S Inocean Cat I Drillship

Abstract

It is essential to have a good thruster system when it comes to offshore operations and dynamic positioning. Not only can this save fuel, but also make the vessel and operations safer. A good thruster system consists essentially of two components. First is a good thruster control system which controls the motors using either speed, torque or power control. The second is an optimal thrust allocation which determines how the necessary forces are distributed to each of the thrusters.

This thesis describes how modeling and implementation of the thruster system onto NTNU's latest model C/S Inocean Cat I Drillship is done. The model is a 1:90 scaled model based on an Arctic drillship called Inocean Cat I Drillship. The goal is to create a thruster system which is optimal and based upon the parameters from the real vessel.

Creating the thruster system similar is done by using information given by Inocean with scaling laws and experimental tests to find parameters. Model experiments,included bollard pull and drag tests were conducted in NTNU's Marine Cybernetics Laboratory. The remaining parameters are provided by theory or through simulation.

The various thruster controllers are simulated and tested on the model to see the differences and similarities they have.

The thrust allocation finally implemented on the vessel consists of a pseudoinverse algorithm which is limited by forbidden zones and singularities that may arise. It also takes into account the optimal angle path and restrictions on speed. A simulation model in which a quadratic regulator algorithm is used to optimize the allocation is created. Implementation of the quadratic regulator proved to be difficult because of compatibility and is therefore not yet fitted to the model. Potential methods for resolving this problem are given later in the thesis along with one of the tried methods.