Research On Thrust Optimum Allocation For Dynamic Positioning System Of Ship

The exploration and exploitation of ocean resource such as oil and gas accelerates the development of dynamic positioning system (DPS). As an important part of the dynamic positioning system, thrust optimal allocation is responsible for making a set of thrusters jointly produce total forces and moment commanded by the higher level controller, in addition to considering power consumption, thruster-thruster interactions and thruster’s tear and wear, as well as meeting the physical constraints of the thruster. This paper presents a method of solving thrust optimal allocation to reduce power consumption of dynamic positioning ships.Firstly, the thrust optimum allocation problem of marine vessels and its basic knowledge are discussed. Subsequently, considering a supply ship with two stern pods and one bow azimuth, we propose the cost function which contains power consumption term, error between the commanded force and the allocated force term, rate constraints on the azimuth/rudder angles term, singularity avoidance term. Then the thrust optimal allocation’s constraints are determined.Secondly, based on the mathematical formulation of the thrust allocation problem, three optimization algorithms, including sequential quadratic programming (SQP), genetic algorithm (GA) and estimation of distribution algorithm (EDA), are used to solve the thrust allocation problem. Then, computer numerical simulations of all three optimization algorithms on a typical supply ship model are carried out. At last, we analyze simulation results by comparison.Lastly, considering the effects of thruster-thruster interactions, the thrust sectors are obtained. Then, based on the results of three thrust optimization allocation, the supervisor is used to monitor these sectors to obtain the power consumption and the error between the commanded force and the allocated force corresponding to each sectors. The switching in the supervisor will exploit each sectors potential to reduce the total consumed power among the thrusters. Ultimately, simulation results have confirmed that the supervisory thrust allocation method is much more power-efficient compared with a explicit thrust allocation method.