Research On SINS/RCNS Integrated Navigation Algorithm Based On Narrow FOV Star Sensor

Integrated navigation theory is an important research direction in navigation applications.The traditional integrated navigation mode based on inertial error model has many approximate calculations in the modeling process,and the resulting model error will lead to the reduction of system solution accuracy.In the integrated navigation system combining strapdown inertial navigation system and celestial navigation,the navigation parameters output by the two navigation systems are used for error compensation,which can effectively improve the system accuracy.The star sensor can be divided into two types according to the size of the field of view.Compared with the large field of view,the narrow field star sensor is smaller in size,lighter in weight,and has the anti-interference ability of the narrow field of view star sensor in the midlow altitude flight altitude.Stronger.At the same time,due to the narrow field of view,measurement loss,delay and system noise,it is easier to affect the positioning accuracy of the narrow field of view star sensor integrated navigation system.In the celestial navigation system,the commonly used positioning methods include orbital dynamics-based navigation methods,direct sensitive flat navigation methods,starlight refraction indirect sensitive navigation methods,and the algorithms that introduce starlight refraction into integrated navigation systems in recent years are becoming more and more having attention.Based on the National Natural Science Foundation of China(61573113),this paper studies the filtering algorithm and integrated navigation algorithm for the inertial/celestial integrated navigation system under the narrow field of view star sensor.The specific work is as follows:The filtering algorithm commonly used in nonlinear systems is elaborated.The problem of the motion state of the carrier is described for the quaternion.Three representative filtering algorithms are selected for simulation comparison.Under different initial conditions,the performance of each filtering algorithm is compared and analyzed stability.According to the nonlinear system model of inertial/celestial integrated navigation,the quaternion orbital dynamics equation is discretized to obtain the equation of state with coupled noise terms.Analyze the mean and variance of the coupled noise and its effect on the error of the nonlinear system.In the narrow field of view star sensor environment,the filter algorithm system model is designed for the coupled noise term in the equation of state.An improved Gaussian Cubature Kalman filter algorithm is proposed.The filter information is updated by the posterior probability density and can be simulated and analyzed.Effectively suppress the influence of coupled noise on the accuracy of the integrated navigation system.The SINS/RCNS integrated navigation system model is established by the method of starlight refraction indirect sensitive flatness.The traditional inertial/celestial integrated navigation method is limited by the accuracy of the horizontal reference,and the positioning accuracy is not high.At the same time,the traditional integrated navigation method can not accurately estimate the drift of the accelerometer,resulting in the system accumulating speed and longitude error over time.In this paper,an improved inertial/celestial integrated navigation model is proposed based on the theory of starlight refraction.The quaternion error,position error and refraction star height error are used as observation measurements.The star sensor output information is used to correct the gyro drift of the inertial navigation system.Acceleration error,combined with atmospheric refraction model,obtains the corrected position error of the refracting star height.Through simulation experiments,the superiority of the traditional algorithm is verified,and the effectiveness of the integrated navigation system model when using different numbers of refraction stars is analyzed.