Information Reliability Estimate And Fault Diagnosis Method For Inertial Navigation System

With rapid development of aerospace, marine, aviation and other areas,modernnavigation, guidance and control system has become an important and indispensable devicefor aircraft, ships and missiles, navigation system errors and failures may cause catastrophicconsequences. As a core component of the navigation system, so the safety and reliability ofinertial navigation system is very important. Therefore, carry out research on inertialnavigation system reliability and fault diagnosis method has great significance.At present,the research about system reliability and fault diagnosis focused on the failurecaused by component failure.But when the Inertial navigation system running, there are manyinterferences from the internal and external environment affect sensitive devices, besides theinertial navigation system has the defect of error accumulation, so the system outputparameters always have larger errors when it working with long time, although the system cancontinue running with larger errors,but it has lost its meaning, which we called informationfailure.So how to diagnose information faults and evaluate system reliability of outputparameters is particularly important.In this paper, platform inertial navigation system was selectted for the study, Based onthe basic principle of inertial navigation system established the errors model of the inertialnavigation system, and analyzed error features of inertial navigation system under thedeterministic and stochastic error sources. Established the state equation and observationequation for fault diagnosis based on the error model of inertial navigation system.Based on the error characteristics of inertial navigation system and statistical analysis ofthe system parameters error, combined with fuzzy reliability theory,derived fuzzy reliabilityanalysis model used for inertial navigation system output parameters.According to the praticalexperimental data did simulation analysis,simulation results show that the fuzzy reliability ofthe calculation method is in line with engineering practice, and can estimate the informationreliability of the inertial navigation system in the case of no failure data.Based on the reason of information failure, a new detection method that based on themulti-Gaussian approximation of posterior probability density is proposed for the informationfailure. The algorithm is divided into two parts. First is the information interference detectionand estimation, followed by statistical decision, determine the type of interferenced andwhether there is information failure. Suppose there are multiple interferences that can cuasesystem failure and each interference corresponding to a fault state,and then accumulate residuals with Kalrman filter and calculate the posterior probability for each failure andestimate the interference.At the Stage of decision-making the criteria of maximum posteriorprobability is selecetted to determine the type of interference that cause informationfailure,then compared the detected interference with the given threshold, if the interferencelager than the given value, the system will Malfunctioning, Simulation results show that thealgorithm can detect and diagnose information failure with very good effect.