As offshore wind farms become larger and further from the shore, there are strong economic and climate incentives to perform transfers required for operations and maintenance from floating vessels, rather than employing expensive and slow jack up rigs. However, successful transfers of heavy and sensitive equipment from a floating vessel (in all but benign sea/wind conditions) are heavily dependent on multiple degrees of freedom, high performance control. This project aims to bring a novel modelling and simulation methodology in Simulink that could be used to assess offshore wind installation and maintenance procedures. More specifically, thegoal is to demonstrate that a crane prototype assumed to be located on a floating ship can transfer loads of hundreds of tons onto a fixed platform. Furthermore, this process should be completed with good precision and minimal impact force during equipment loading onto the stand. This problem has not yet been answered in research, with the only relevant patent in the field being the Ampelmann platform, a motionless bridge allowing technicians to access the offshore turbine. The first main contribution to knowledge of this thesis was the design of a 90 m crane that could handle a 660 tons load. This thesis presents a procedure, based on both mechanical/hydraulics design as well as empirical findings, which could be re-used for scaling the crane model to a more realistic dimension. It is worth noting that the goal here was to assess whether a realistically weighing piece of equipment could be stably handled, while the actual size of the crane was deemed unimportant. Another missing gap in literature this project wanted to fill was achieving active motioncompensation for a larger scale system such as the current one. This refers to balancing out the base motions on multiple axes, so the payload can be moved on a given trajectory unaffected by them. Currently, research in the field mainly consists of crane mechanisms that feature active heave compensation, which only refers to the vertical axis. Hence, two control design methods were employed to assess the viability of heavy payload positioning from floating vessels through the development of a simulation approach using Simulink. The crane prototype was designed and modelled to operate under simulated vessel motions given by sea states with a significant wave height of 5 m and maximum wave frequency of 1 rad/s. Then, traditional control (feedback and feedforward) was designed to achieve active motion compensation with steady-state position errors under 20 cm. A second controller architecture was then designed/implemented as a comparison basis for the first one, with the aim being to find the most robust solution of the two. The nonlinear generalised minimum variance (NGMV) control algorithm was chosen for control design in this application. Due to its ability to compensate for significant system nonlinearities and the ease of implementation, NGMV was a good candidate for the task at hand. Tuning controller parameters to stabilize the system could also be based on the previously determined traditional control solutions. An investigation of controllers’ robustness against model mismatch was carried out by introducing various levels of uncertainty which influence actuators’ natural frequency to assess system sensitivity. The outcome of the investigation determined that traditional and NGMV controllers provided comparable regulating performance in terms of reference tracking and disturbancerejection, for the nominal case. This confirmed the assertion that the PID-based NGMV weightings selection is a useful starting point for controller tuning. Increasing the mismatch between the nominal system based on which the controllers’ were designed and the actual plant showed that the traditional control was marginally more robust in this application. The final contribution to knowledge this thesis aimed to bring was minimising the impact force during load placement on a fixed and rigid platform. To that end, the contact forces between the payload and a platform were first successfully modelled and measured. A switching algorithm between position and force control was then developed based on a methodology found in literature but on a microscopic scale project. To execute smooth load placement, an automated hybridforce/position control scheme was implemented. The proposed algorithm enabled position control on x and y axes, while minimising impact forces on the z-axis. Unfortunately, preliminary findings showed that there is still work to be done to claim any success in this regard. However, the author hopes this offers a good starting point for future work.
Date of Award | 27 Jul 2023 |
---|
Original language | English |
---|
Awarding Institution | - University Of Strathclyde
|
---|
Sponsors | University of Strathclyde |
---|
Supervisor | Paul Blackwell (Supervisor) & Dorothy Evans (Supervisor) |
---|