Research On Nonlinear Filtering Method For Initial Alignment Of Integrated Navigation System

The initial alignment precision impacts directly on the inertial navigation system(INS).Initial alignment is one of the key techniques of inertial navigation system.The Kalman filtering method is the most mature and widely used method in all the filtering methods.On Kalman filtering method,the disturbance must be Gauss white noise whose variance has been known.But in most actual systems,the condition of Gauss white noise is not exist.Also most actual systems are nonlinear systems.Both the condition and the nonlinearity impact the Kalman filter’s working performance.Predictive Filtering method is a real time filtering method.The method is on actual basics.The Predictive Filtering can estimate the model error and correct the model on line.In this paper,Predictive Filtering method associating with Extended Kalman Filtering method has been taken used in the simulation of the initial alignment’s process.Result showed that the Predictive Filtering method associating with Extended Kalman Filtering method can follow the initial alignment’s process in a short time.Kernel methods were applied to the pattern analysis mostly.In this paper,the process of initial alignment was simulated by using the kernel method.The simulation results showed that the kernel method can accurately simulate the initial alignment of integrated navigation system.