The Study On Nonlinear Filter And Fusion Method For Ship Position Measure System

With the development of science and technology, People compete increasingly formarine resources. Ship Dynamic Positioning technology is an effective means for thedevelopment of offshore resources, but how to improve the DP ship job security andpositioning accuracy has always been the focus of this technology. At sea, just rely on asingle sensor to achieve precise positioning to be unsatisfactory, due to a variety ofenvironmental forces and various uncertainties. Thus, the use of multi-sensor data fusion hasbecome a hot research. This article as the main research background, on the conditions ofship nonlinear, proposed three kinds of state estimation method based on the ship positionmeasuring system to estimate the state of positon. Make use of the structure of multi-levelmulti-sensor data fusion to get the current state of motion of ship. It provides a strongguarantee for the precise positioning of the ship.Firstly, This article need to establish a modle of the system. It mainly consists of the shipmotion model and sensor observation model. Ship motion model include high and lowfrequency movement of the ship and a variety of environmental interference factors. Themodeling of the sensor observation model means to establish the modle of ship positionmeasuring system (the tension cable position measurement system, the location of theunderwater acoustic measurement system, GPS position measurement system).Secondly, It introduces briefly several common methods of state estimation. It is in thefield of nonlinear filtering. Though introduing and performance analysis to Extended Kalmanfilter、Unscent Kalman Filter、Central Difference Kalman filter and Cubrature Kalman Filter,to descript deeply the several hotspots nonlinear state estimated method. It makes goodfundation for the simulator of the next chapter.Then, on the nonlinear ship conditions of motion, It proposes three kinds of nonlinearstate estimated method based on the ship’s position measurement system. Making use of theCubrature Kalman filter (CKF), Unscented Kalman filter (UKF), Central Difference KalmanFilter (CDKF) state to estimate the state of ship’s position, It fully proved that the method isverified by simulation in this system and comparing the verification proved their performancemerits.Finally, descripted briefly of the fusion of several data structures (centralized, distributed,multi-stage), and then, based on the the position measurement system, a multi-level datafusion structure is proposed and simulated in the full feedback and no feedback cases. It has fully proved the availability of the fusion method.