Research On Hybrid Information Fusion Methods Of The Reference-system For Dynamic Positioning

With the increasing development of world economic, ocean resource become more important to human, and there are more and more ocean engineering project such as constructing subsea tunnel, exploiting subsea oil, laying subsea cable. So it is high request to the technical equipment of ship equip in resource development and assignment engineer project. The application of dynamic positioning technique on the ship makes people use and exploit ocean resource effectively. It is a key technology in the study of dynamic positioning system that how to get exact location information from measurement of multi-location reference system. In this article, the object of study is the reference system of dynamic positioning, aiming at the uncertainty of system model and noise that caused by external disturbance, using the way of hybrid fusion based on self-adaption robust SRCKF algorithm to achieve multiple sensor fusion, in order to improve the positional accuracy and whole performance of dynamic positioning system.Firstly, according to the motion situation of dynamic positioning, the useful coordinate system and the way of coordinate transformation are introduced and ship motion model is built; Considering the environmental disturbance, the environmental disturbance model is built; Aiming at the dynamic positioning system, there are three dynamic positioning position reference system mathematical models including: DGPS position reference system,hydroacoustic position reference system and tensioning cable position reference system.Secondly, the sampling frequency of each position reference system is different and the sampled data in reference system are also based on coordinate of themselves. The disturbance of environment could cause some problem like freeze, outliers, float will turn up to data that is measured, those problems will influence data accuracy and veracity. In order to make sure the fusion reliability, it is necessary to make time registration, spare registration and quality detection, that measured data of every location reference system.Then, in the real ship assignment, the system model is usually uncertain and the statistical property of noise is probably unknown. In this article, by analyzing CKF, SRCKF successively, aiming at the uncertain system model, the robust SRCKF algorithm is proposed, and based on robust SRCKF, aiming at the uncertain statistical property of noise,the adaptive robust SRCKF algorithm is proposed. In the end, the simulation comparison is operated.Finally, based on reset federation fusion method, using the advantage of centralization fusion structural, the hybrid fusion method is proposed. And associated with SRCKF, the multi-sensors of dynamic positioning reference system is best to fusion. By making compare with the fusion way of hybrid and reset-style in simulation, we get the different property on them, that make sure the effectiveness of hybrid information fusion based on SRCKF.