Joint attitude determination method and satellite attitude control system based on unscented Kalman filter

Abstract

The invention belongs to the technical field of satellite attitude determination, and discloses a joint attitude determination method based on an unscented Kalman filter and a satellite attitude control system; a joint attitude determination algorithm based on a UKF star sensor and a gyroscope, and constructs a differential nonlinear state equation , and the fourth-order Runge-Kutta integral method is used to solve the time update of the state variable, which avoids the discretization process of the nonlinear differential state equation; the UKF algorithm is used to filter and correct the quaternion of the star sensor and the angular velocity of the gyro. The present invention performs time update through the fourth-order Runge-Kutta integration method, uses quaternion multiplicative errors to calculate the weighted mean and covariance, adopts the unscented Kalman filter algorithm, introduces the observation value of the star sensor for filter update, and finally The pose data is corrected using the estimated error quaternion. It has good attitude determination performance when the measurement error of the attitude sensor is large, and has higher estimation accuracy than the attitude determination algorithm based on the extended Kalman filter.

Description

technical field [0001] The invention belongs to the technical field of satellite attitude determination, in particular to a joint attitude determination method based on an unscented Kalman filter and a satellite attitude control system. Background technique [0002] At present, the closest existing technology: attitude determination technology is the process of processing the measured value of the attitude sensor with error through a certain algorithm, reducing the error, and correcting the measured value. It is widely used in satellite navigation and positioning and other fields. For surveying and mapping satellites, the accuracy of attitude determination is an important prerequisite for high-precision positioning. At present, satellites mainly use the combination of star sensor and gyroscope, and use dynamic estimation method for filtering. The dynamic estimation method uses a series of continuous time measurement information, adopts attitude kinematics models and other t...

AI Technical Summary

Problems solved by technology

However, when the UKF algorithm solves the nonlinear attitude determination problem where the state equation is a continuous nonlinear ordinary differential equation system, it usually discretizes the state equation. On the one hand, the derivation of the discretization process is complicated, and on the other hand, discretization errors will occur.

In addition, when the traditional UKF solves the mean and covariance of the state variables, the quaternion is equivalent to the general vector solution, which cannot meet the normalization condition of the quaternion

[0003] To sum up, the problems existing in the existing technology are: the existing algorithm linearizes and discretizes the state equation and the measurement equation, which inevitably produces linearization and discretization errors, and ignores the quaternion operation characteristics and Physical meaning, leading to filter performance degradation

[0005] The difficulty of the above technical problems lies in: 1) how to avoid linearization and discretization processing, and effectively improve the accuracy of attitude estimation; 2) how to apply the operational characteristics of quaternions to the process of solving the mean value and covariance matrix

Images