Research On The Vehicle Positioning Technology Based On Muti-source Information Fusion

Land-based platform XX XX system is the significant force to attack the enemy in modern warfare.Suddenness and complexity of modern warfare requires that land-based XXs equipment have fast response,independent and random launch capability under a wide range of maneuvering.The high precise autonomous positioning capability of land-based XX equipment was presented.The robustness and reliability of positioning technology which is based on single information source are not good.A vehicle positioning system which has multi-redundant,highly accurate and complementary advantages of information can be taken by multi-source information fusion positioning algorithm with the development of measurement technology.The positioning algorithm can be used in complex environment to complete high-precision positioning tasks autonomously and provide technical support land-based XX systems in launching target accurately.Based on the vehicle strapdown inertial navigation system(SINS),this dissertation combines multiple sources of localization information to improve the localization accuracy of land-based XX XX system and the adaptability of complex environment.The multi-source information fusion vehicle positioning technology was studied.The main contents and achievements of this research include:(1)An open multi-source information fusion algorithm architecture is proposed.The Strapdown Inertial Navigation System(SINS)is used as the main reference system which forms a sub-filter with each information source,and the output of the sub-filter is diagnosed.The result of sub-filtered whether effective or not is decided by reconstruction mechanism.The positioning information source is divided into three categories according to the signal output characteristics: the signal is sent at a fixed frequency,the signal is output asynchronously and has a short interval,the signal is output asynchronously and has a long interval.And the corresponding data fusion method is used to the information source data which have classified process information fusion.(2)The federated Kalman filter optimal fusion algorithm is used to conduct data fusion of the first type of location information sources.SINS as the main reference system form sub-filters with other sources of information.The sub-filter output is decided by fault diagnosis and system reconfiguration whether go to the main filter for information fusion.The expansion of the algorithm framework and fault tolerance performance are improved due to each sub-filter independently filtered without affecting each other.The simulation results and experiment results show that the proposed algorithm has lower error state dimension and better robustness compared with the traditional centralized kalman filtering algorithm.In the event of information source failure,the algorithm can still maintain the positioning accuracy of the system and improve the environmental adaptability of the vehicle positioning system to ensure the performance of the vehicle positioning system.(3)Aiming at the second type of information source,the asynchronous non-uniform interval information fusion algorithm based on factor graph is studied.The factor graph is used to correlate the system state model and the observation model,and the state estimation based on the maximum posterior probability is used to process the measurement information sequentially.This method effectively solves the fitting error of the traditional non-interval data fusion algorithm and plug-and-play problems.Simulation results show that the positioning error of the asynchronous non-uniform information fusion algorithm is reduced by 30% compared with the method of correcting location of SINS directly using GMNS information.In addition,experiment was conducted to further verify the positioning accuracy of the algorithm and to solve the plug-and-play problem of the vehicle positioning information source.(4)The comprehensive correction technology of SINS based on position information is used to conduct data fusion of the third type of location information sources.Aiming at the third type of information source represented by landmarks,an integrated position correction algorithm was used to generate the error transfer model of gyro drift and azimuth error by using the intermittently obtained external position information.The gyro drifts were estimated and compensated by least square method.Making full use of the third type of information source represented by landmarks,and the long-time positioning accuracy of the strapdown inertial navigation system is significantly improved.(5)Aiming at the open framework of the multi-source information fusion algorithm which include the federated Kalman filter algorithm,the asynchronous data fusion algorithm based on factor graph and the comprehensive correction technology based on position information did a simulation analysis and experiments in this paper.The simulation results show that the proposed algorithm can integrate many types of positioning information sources,and the position accuracy is consistent with the highest positioning accuracy of sub-system.Experiment results show that the algorithm significantly improves the deficiencies of error divergence which exist in the traditional integrated navigation algorithm.Therefore,the algorithm can control the error within a certain range and has a good fault diagnosis,system reconfiguration capabilities,meanwhile the plug-and-play of position information source is solved.