Research On Error Mitigation Technology Of Marine Passive Navigation Signal Fusion

Due to the rapid development demand of surface ship navigation technology,new navigational requirements such as high navigation accuracy,strong stability and full autonomy are proposed for navigational navigation equipment.At present,the information fusion of the passive navigation signals output by the strapdown inertial navigation system(SINS)and the star sensor is the research direction of the marine navigation system.The fusion method not only reduces the dependence on satellite navigation,but also ensures the navigation accuracy of the system.However,there are some problems in marine star sensor,such as the statistical changes in the measurement noise of star sensor and the introduction of SINS error when positioning relative to navigation system.Furthermore,the error is introduced into the fusion process of marine navigation information,and it will spread over time,which reduces the navigation accuracy of the system.Therefore,considering the above problems,this paper makes a deep research on SINS/star sensor information fusion system based on ship application background.The purpose is to discuss how to suppress the information fusion error of marine passive navigation system,meet the different requirements of the ship navigation system,and improve the information fusion scheme.The on-line calibration technology for the star sensor assisted gyroscope can improve the accuracy of the marine navigation system.However,due to external factors,the statistical characteristics of star sensor measurement noise are unknown,which makes the standard kalman filter divergence unable to effectively calibrate the gyroscope online.In order to solve this problem,this paper proposes a gyroscope on-line calibration method assisted by star sensor based on forgetting factor selection of a Sage-Husa filter.By using this algorithm,the measurement noise parameters of the filter are estimated and corrected continuously,and the gyro device error can be estimated effectively when the sea surface environment is complex.The accuracy of the filter is improved,and the accuracy and environmental applicability of the navigation system for ships are also increased.However,the strapdown inertial navigation/star sensor online calibration system suppresses some navigation errors,so it is necessary to further improve the accuracy of navigation system.The SINS navigation error can be corrected by using the position information of star sensor output,which can improve the navigation accuracy of the system.However,when the star sensor determines the positioning information of the ship in the navigation coordinate system,it will introduce the horizontal attitude error of the SINS,which reduces the accuracy of the star sensor.Therefore,the solution of the star sensor position information introduces the horizontal attitude error of the strapdown inertial navigation,which reduces the accuracy of the star sensor.In order to solve this problem,this paper proposes a strapdown inertial navigation/star sensor information fusion algorithm based on model predictive filtering-kalman filtering.The mathematical model between the positioning error of star sensor and the horizontal attitude error is established.The model predictive filtering(MPF)can be used to estimate the characteristics of any unknown form model errors online,and the system model can be modified to improve the star sensor positioning accuracy.Based on the corrected star sensor position information,the SINS error is estimated by kalman filter,and the divergence of navigation system error is restrained,which meets the long-term stability requirement of ship information fusion system.Finally,the online calibration algorithm based on forgetting factor adaptive selection sage-husa filter and the information fusion scheme based on model predictive filter-kalman filter are validated by simulation.The algorithm is analyzed by the actual data.And the simulation results show that the error of the marine navigation system is suppressed,which can meet the long-haul navigation requirements of the ship.