Integrated Navigation Method Using Marine Inertial Navigation System and Star Sensor Based on Model Predictive Filtering
Wang qiuying*, Zhang minghui, Guo Zheng,College of Information and Communication Engineering, Harbin Engineering University, China; Wu Hui, Dalian Shipbuilding Industry Offshore Co., Ltd, China
Location: Big Sur
The inertial navigation system (INS) is a common navigation equipment for surface ship, but the error of INS accumulates with time. It is difficult to meet the requirements of the ship’s navigation accuracy, so the other navigation equipment is needed to correct the error parameters of inertial navigation system. Traditionally, the integrated navigation with the global position system (GPS) and INS is often used. However, in the electronic environment of modern battlefield, the GPS signal cannot be used because it is easily interfered or deceived. The star sensor can output high precision navigation information. More importantly, its measurement results do not have the problem of error accumulation in INS, and there is no interference of GPS signals. Thus, in order to improve the accuracy of surface ship navigation system, the fully autonomous navigation technique for the integrated navigation using the marine inertial navigation system and star sensor is proposed. The integrated navigation method employs navigation information from a star sensor as a benchmark to perform periodic recalibration of the divergent positioning error of the inertial navigation system, so as to improve the accuracy of the integrated navigation system. However, the position calculation of star sensor needs horizontal attitude information provided by inertial navigation system. Therefore, the inertial horizontal attitude error is coupled to the star sensor calculation, which results in inaccurate description of the system model. Aiming at the problem that star sensor depends on the level information of inertial navigation system, the mathematical model between the star positioning error and the horizontal attitude error of inertial navigation system is established. At the same time, the random error cannot be accurately modeled in the system. Therefore, the data of inertial navigation system and star sensor are fused by model predictive filtering (MPF) based on the position information difference between inertial navigation system and star sensor. The MPF is a filter that uses predictive output tracking measurements output. This method can estimate any unknown model errors, and then modify the system model. Thus, this can achieve the error suppression of star sensor position information. The estimated position error is fed back to star sensor to obtain accurate location information. Then, the accurate positioning information is used as the reference, and then fed back to the integrated navigation system to obtain high-precision navigation information. This can improve the precision of integrated navigation system, and meet the requirements of full autonomy, high precision and long endurance voyage of ship. In additions, during the experiment, the star sensor and inertial navigation system is fixedly connected. Then the INS/star sensor which appears to be a single unit is fixed on the ship’s deck to remove measurement errors due to the relative movement of the two devices in the experimental course. The laptop is used to acquire the data, in order to compare and analyze the effects and accuracy of the systems. The vehicle was navigating on the lake, and the navigation trajectory is assigned. Finally, the method is verified by simulation, and the effectiveness of the method is verified by the real ship test on the lake surface.