Research On The Key Algorithms Of Multi-source Fusion Navigation System

With the constant development of science and technology,there are more and more kinds of navigation sources.However,navigation systems are relatively independent and can not share resources.The multi-source fusion navigation system can comprehensively utilize the information of various navigation sources,combine them effectively,and complement each other to provide full-time and high-precision navigation.In this paper,the key algorithm of multi-source fusion navigation system is studied,including the design of system framework,fusional judgement,fusion algorithm and integrity monitoring algorithm.The main contents of this paper include the following items:1.According to the characteristics of navigation system,The framework of multi-source fusion navigation system is designed in this paper,dividing the multi-source fusion navigation system into four parts: information measurement,fusional judgement,information fusion and integrity monitoring,and expounds the relationship between the functions of each part and each part.2.In the study of the fusional judgement part,this paper designs the fusional judgement model based on the performance evaluation index of the multi-source fusion navigation system.This article divides the fusional judgement into two parts: the basic judgment and the promotional judgment,and expounds the relationship between them.The decision principle and decision method of reliability,continuity,precision are analyzed in detail,which lays a foundation for the study of the fusibility of multi-source fusion navigation system.3.In the research of fusion algorithm,the existing fusion algorithm is divided into three categories: centralized fusion algorithm,parallel fusion algorithm and sequential fusion algorithm.And the principle and performance of these algorithms are analyzed in detail and verified by simulation.Then the characteristics and application scenarios of the three algorithms are summarized.Aiming at the problem that existing federated filtering algorithms are insensitive to abnormal observations and have poor fault tolerance,an adaptive federated filtering algorithm based on observation innovation is proposed.The experimental results show that the adaptive federated filtering algorithm can automatically adjust the system error matrix when the navigation source is abnormal,thus effectively suppressing the error and improving the fault tolerance of the multi-source fusion navigation system.Moreover,the method is simple in form,small in computation and easy to implement in engineering.4.In the study of integrity monitoring algorithm,this paper introduces the principle of fault detection algorithm based on innovation detection and fault detection algorithm based on innovation extrapolation in detail,and analyzes and verifies their performance through simulation experiments.Then a integrity monitoring algorithm based on synchronous detection is designed.Synchronous detection algorithm combines the advantages of innovation detection method and innovation extrapolation method,can detect three types of common faults in time,and can infer the specific fault types through the detection results of two parts.In this paper,according to the differences of different fusion algorithms,centralized integrity monitoring algorithm and parallel integrity monitoring algorithm are designed,and their performance is verified by simulation.Experimental results demonstrate the availability and effectiveness of integrity checking algorithm based on synchronous detection.In summary,this paper focuses on the key algorithm of multi-source fusion navigation system,and designs and analyzes the system framework and the fusional judgement model of multi-source fusion navigation system.An adaptive federated filtering algorithm based on innovation is proposed to improve the fault tolerance of multi-source fusion navigation system.The centralized integrity monitoring algorithm and parallel integrity monitoring algorithm based on synchronous detection are designed,and their performance is verified by simulation,which provides theoretical basis and guidance method for the research and analysis of multi-source fusion navigation system in engineering practice.