Research On The Thrust Allocation For Dynamic Positioning System

With the rapid development of modern industry, and the raising of living standards of the humanity, people’s demands on all kinds of resources are increasing day by day.Recently, along with the drying up of onshore or offshore resources, exploitation of deep sea is becoming more and more frequent. Since anchors are not suitable for positioning in deep sea, more and more drilling rigs or vessels have adopted dynamic positioning system(DPS).Ships will be positioned at a desired position or tracking a desired path with DPS, by generating forces to counter-balance external forces, based on its own thrust system. As an important component of DPS, the function of thrust allocation is to calculate the optimized thrust of each thruster according to the total thrust of the DPS, and hence improve the performance of the DPS.The study subject of this thesis is the “NEW HAIHU9”, which adopts two propeller/rudder pairs as its main thrusters. Therefore, main attentions of this thesis are focused on the thrust allocation problem for ships with propeller/rudder pairs as their main thrusters. This problem is more difficult than its Z-propeller counterpart.According to the propeller/rudder pair which is the research object, build the corresponding objective function and constraints. The objective function take less energy consumption, less mechanical wear of thrusters, minimization of thrusters’ maximum into account. The constraints consider several aspects such as thrust region of thrusters, rudder angle rate of change, thrust rate of change, etc.Two optimization algorithms are adopted in this paper, such as Sequential Quadratic Programming(SQP)and Adaptive Chaotic Particle Swarm Optimization(ACPSO) to handle thrust allocation. These two algorithms have their own advantages and disadvantages, for example, SQP is stable and fast, as well as local convergence, ACPSO is an intellective bionic algorithm and it has a structure, the global convergence and it is easy to implement,but it has a long running time.Applying the above algorithm into the thrust allocation problem, under the different test environment conditions, given the appropriate parameters for the simulation test, then compared with the results obtained from the two algorithms after several simulations.Experiments show that the above algorithms can achieve the desired effect. ACPSO hasmore accurate optimization results compared with SQP, but some improvements must be done to reduce operational time in the application of ACPSO.