Research On Fault Detection And Quantitative Filtering Response For Integrated Navigation Of Aerobat

Integrated navigation system can improve the accuracy of navigation system,which is the key to ensure the accuracy of the system.However,it is particularly important to ensure the reliability of the system.Fault detection technology is an important measure to ensure the safety and reliability of navigation system.When a measurement fault occurs,the traditional fault detection method will choose to remove all the information from the observation subsystem.In this paper,a method of system global and local joint cross fault detection based on Kalman filter residual chi-square fault detection and least squares deviation fault detection is proposed,which can eliminate all fault observations,retain normal observations for integrated navigation,and avoid unnecessary waste of correct measurement information.Compared with the traditional fault detection method,this paper innovatively proposes a method to distinguish the state fault from the measurement fault,and can effectively define the state fault from and measurement fault.When the fault is less than the Minimum Detectable Biases(MDB),it will lead to over-estimation of the filter and increase the probability of failure missed detection.In this paper,we reconstruct the error covariance matrix of filter estimation by using the research method of quantizing filter response to faults.The experimental results show that the research method of quantized filter response to faults can effectively suppress the influence of filter over-estimation when faults occur so as to reduce the impact of fault omission and improve the reliability of the system.The main contents of this paper are as follows:(1)This paper introduced the basic principle of strapdown inertial navigation system,and analysed its main error sources,deduced the navigation system error equation.For integrated navigation filtering algorithm and fault detection method,this paper introduces the mathematical basis of conventional Kalman filter,extended Kalman filter,unscented Kalman filter,particle filter algorithm,residual chi-square fault detection and least square deviation fault detection in detail.(2)In order to improve the reliability of integrated navigation system,this paper presents a global and local combined cross fault detection method based on Kalman filter residual chi-square fault detection and least squares deviation fault detection,and a method to distinguish system measurement fault and state fault.The validity of this research method is verified by numerical simulation.Finally,this method is compared with the traditional residual fault detection method to verify its superiority.(3)When the measurement fault is less than MDB,it will lead to over-estimation of the filter,which will increase the probability of system failure missed detection.In this paper,the causes and effects of overestimation are analyzed,and the covariance matrix of filter estimation error is reconstructed by quantifying the filter response,so as to restrain the influence of overestimation,reduce the occurrence of failures and improve the reliability of the system.