Research On The Intelligent Control Method Of The Standard Additive Model Applied To Ship Dynamic Positioning

In recent years, with the development of offshore oil development and maritime operations, as the key technology of Marine operation system, the research on dynamic positioning system of ship has been developing rapidly abroad, which are widely used in civil and military ships. Compared with the developed countries, we have a big gap to close on the ability of designing and producing of the DPS (dynamic positioning system). Therefore the research on practical DPS with high precision and adaptability for our country’s Marine development is extremely important and strategically significant.Among the multiple types of the Additive Fuzzy System, the Standard Additive Model (or some certain type of SAM) is used by most Fuzzy Engineers in the actual utilization, because almost all the Fuzzy System could be interpreted by it. Based on which, this paper did the research on the Intelligent Control Methodology of the Standard Additive Model applied to Ship Dynamic Positioning System, and was mainly composed of the following contents.First of all, in order to realize the forecast to the Marine interference environment and build the foundation of the designation of Ship DPS controller, this paper combined the Neural Network and SAM system, and built a Fuzzy Neural Network (SAM-NET) structure based on SAM system, which testified the global approximation property of the SAM-NET Structure. It also proposed an applicable training algorithm to learn SAM-NET, which made the subordinate degree function much more reasonable. SAM-NET was used to forecast interference factors and perform the deviation analysis, and the simulation result showed that the SAM-NET was accurate and practicable to forecast the wind and wave.Secondly, the accurate information of ship position is very important during the process of the ship dynamic positioning. In order to obtain the accurate ship position information, this paper expanded the definition of "point, line, surface, object, chain" in the self-similar information system matrix of the integrated enterprise information model, so thus realized the information resource sharing and complementary among the composed parts of the DPS. A new data fusion method of multiple sensors is also put forward for the Ship Dynamic Position information on the basis of SAM model. This method combined the scoring method of competitive sports and 3 a criterion to eliminate the outliers, and chose the Gauss-type membership functions, and each triggered rule has the same weighted value. The experiment result showed that the fusion result of this data fusion method was much closer to the real position of the ship than the metrology result calculated from single position sensor.The third, DPS can’t accurately describe the mathematical model of Ship due to the changing Marine interference environment which leads to the worse controlling effect than expected. For the sake of resolving the problem mentioned above, this paper designed a new Ship Dynamic Position Controller based on improved algorithm of SAM. Using AHP (analytic hierarchy process) it removed the redundant or conflicting rules from the initial fuzzy rules of SAM, and then calculated the weighted value of each rule by Ant Algorithm. The simulation experiment showed that the Ship Dynamic Positioning controller based on the improved algorithm of SAM designed in this paper had a shorter responding time and the controlling effect was stable and accurate.The last, this paper, in accordance with the multi-mode controlling, designed the Multi-Mode Controlling System of Ship Dynamic Position based on the improved algorithm of SAM. This system, utilized the SAM-NET model to forecast the Marine environment and filtered the Marine interference using the filter, decided the controlling mode as Manual or PID controlling or SAM improved algorithm controller, realized the stable and smooth switch between the different mode using mixed modes controlling. By adopting the method of average resident time proved that the multimodal stability of switched systems.The simulation showed that the controlled ship using DPS could arrive at the designated position in a short time and kept stable, and the system possessed the excellent stability, robustness and accuracy.In summary, this paper carried out the research on the intelligent controlling mode of standardized additive model applied to the Ship Dynamic Position and gained some positive result, while we still need perform the further research and improvement.