KR101294623B1 - Enhanced ground based precise attitude determination method of imaging satellite - Google Patents

Abstract

PURPOSE: An enhanced ground based precise posture determination method of a low-orbit imaging satellite is provided to accurately determine the posture of the satellite by dually using gyro measurement data and star tracker measurement data. CONSTITUTION: Gyro measurement data and star tracker measurement data are obtained in different intervals through a gyro sensor and a star tracker using a dual MIL-STD-1553 bus communication device (S100). The gyro measurement data and the star tracker measurement data are stored and transmitted to a ground station (S200). A posture quaternion estimated value is repeatedly predicted until the star tracker measurement data is received after calculating an angular speed by restoring the transmitted gyro measurement data again (S300). A modified value is calculated by using the posture quaternion estimated value and the star tracker measurement data, and an error posture quaternion is applied as a corrected value by using the modified value (S400). [Reference numerals] (S100) Measured data obtaining step; (S200) Measured data transmitting step; (S300) Angular speed calculating and predicting step; (S400) Correcting step

Description

Improved Ground Based Precise Attitude Determination Method of Imaging Satellite

The present invention relates to a precision posture determination method of a low orbit image photographing satellite, and more particularly, an onboard computer equipped with a satellite in which the fine orbital image positioning satellite of a low orbit image photographing satellite transmitting high-resolution images to a ground station while driving a low orbit image is transmitted. Instead, the present invention relates to a terrestrial precision posture determination method of a low orbit imaging satellite, which can be performed on the ground to improve the performance of attitude determination in a moving region of the low orbit imaging satellite.

A low-orbit imaging satellite, which uses a high-resolution imaging device mounted on a low-orbit solar orbit for several days and transmits it to the ground by photographing objects or areas on the ground, and these imaging satellites are sophisticated and accurate. In order to acquire an image, the attitude of the satellite must be precisely controlled. Therefore, precise attitude determination of the satellite is required. To this end, a high-performance sensor such as a star tracker and a gyro is installed in the imaging satellite, Precise attitudes to the satellite are made.

However, these sensors may change the installation state due to external force, space effect due to satellite launch, and in this case, an error may occur in the captured image. Relative correction for geometric errors such as change, misalignment of gyro due to influence of space environment, scale factor error, and misalignment between star trackers is performed first. After the relative correction is completed, absolute correction is performed using the image reference point information. The absolute correction is performed using attitude information determined by a computer mounted on a satellite.

On the other hand, when determining the attitude of a satellite in a computer mounted on a satellite, the measurement values obtained from a high performance star tracker and a gyro sensor are used, and it takes a long time to perform the image processing obtained from the star tracker. In this case, since the processor throughput is limited, the onboard computer determines the attitude of the satellite by acquiring the measurement data with the gyro measurement period set according to the star tracker measurement period (typically 10 Hz). In this case, the angular velocity of the satellite is used as the average angular velocity for the unit time calculated by Equation 1 below, in which case the measurement noise and additional errors are added to increase the uncertainty in the prediction of the process noise (process noise), This lowers the accuracy of attitude determination.

는 각속도,

는 각도이고,

는 측정시간이다.
here,

Is the angular velocity,

Is an angle,

Is the measurement time.

In addition, when the satellite computer is used to determine the attitude of the satellite, the measurements obtained from the high-performance star tracker and the gyro sensor are input to a filter such as a Kalman filter, and the filter parameters are relaxed to increase the robustness of the attitude determination. Because it is set, it takes a long time for the attitude value to converge and it is difficult to apply it to the maneuvering section of the satellite.

For this reason, attempts have been made to improve the performance by performing the attitude determination using the star tracker measurements and the gyro measurements collected from satellites on the ground without limitation in processing capacity, and also by adjusting the Kalman gain. Although it helps to improve the convergence performance of the attitude determination, the problem of deteriorating the accuracy of the angular velocity prediction remains because it is performed by setting the star tracker measurement value and the gyro measurement value acquisition period in the same manner as in the prior art.

Therefore, the present invention has been made to solve the problems of the conventional satellite attitude determination method as described above, by using dual MIL-STD-1553 bus communication to obtain the gyro measurement data and star tracker measurement data at different periods, The purpose of the present invention is to provide a positioning method of an imaging satellite, which dualizes the obtained gyro measurement data and star tracker measurement data to determine the attitude of the satellite. There is this.

The object of the present invention as described above is to measure the gyro measurement data and the star tracker at different cycles using the dual MIL-STD-1553 bus communication device provided in the computer mounted on the satellite, the terrestrial precision posture determination method of the imaging satellite A measurement data acquisition step of acquiring data respectively; A measurement data transmission step of storing the gyro measurement data and the star tracker measurement data and transmitting the same to a ground station; A prediction step of reconstructing the gyro measurement data transmitted in a relatively short period at the ground station to calculate the posture quaternion value until the star tracker measurement data is available; When the star tracker measurement data is available, the correction value is calculated using the posture quaternion prediction value and the star tracker measurement data, and then the error posture quaternion and gyro drifting error are calculated using the correction value, and the error posture quaternion By applying as a correction value, it is achieved by constructing a correction step of updating the estimated attitude of the satellite.

In this case, it is preferable that the Extended Kalman Filter (EKF) is used as the attitude determination algorithm.

In addition, the gyro sensor obtains the measured value data at a period of 50Hz, the star tracker is more preferably to obtain the star tracker measurement data at a 10Hz period.

According to the present invention, a dual MIL-STD-1553 bus communication device is provided in a satellite-mounted computer, and the gyro measurement data and the star tracker measurement data are dualized through a separate bus, acquired, stored, and transmitted to the ground station. The angular velocity is calculated based on the gyro measurement data, which has a relatively short measurement period, and used to determine the attitude and drifting error of the satellite.At the same time, the gyro drifting error estimation performance is improved. Performance is also improved in the section where starting and disturbance interference occurs.

In addition, the present invention is continuously updated and used Kalmandeuk used to determine the attitude of the satellite, which improves the convergence speed of the attitude determination of the filter, so that the geometric error of the satellite in a section in which the high-speed motion is performed to several shooting targets in a short time Can be reduced.

1 is a configuration diagram showing an example of an apparatus for performing a method for terrestrial precision posture determination of a low orbit image photographing satellite according to the present invention;
2 is a flowchart showing a method for terrestrial precision posture determination of a low-orbit imaging satellite according to the present invention;
3 is a flowchart showing a specific procedure for the method for terrestrial precision posture determination of low-orbit imaging satellite according to the present invention.

Hereinafter, the configuration of the present invention through the accompanying drawings showing a preferred embodiment will be described in more detail.

The present invention relates to a terrestrial precision posture determination method of a low-orbit satellite photographing satellite, and the precision posture determination method of the present invention includes a measurement data acquisition step (S100) and a measurement data transmission step (S200) as shown in FIGS. 2 and 3. , Angular velocity calculation and prediction step (S300) and correction step (S400).

(1) Measurement data acquisition step (S100)

This step is to acquire the basic data for determining the attitude of the satellite, for this purpose, the satellite is provided with a gyro 10 and a star tracker 20, wherein the gyro 10 and the three axis direction of the satellite The rotational angular velocity is measured and transmitted, and the star tracker 20 transmits star track measurement data.

At this time, the onboard computer is equipped with a separate dual MIL-STD-1553 bus communication device as shown in FIG. 1 so that the gyro 10 acquires measurement data at a period of 50 Hz, and the star tracker 20 measures data at a period of 10 Hz. Acquire it.

(2) Measurement data transmission step (S200)

In this step, when the gyro measurement data and the star tracker measurement data are acquired at different periods by the measurement data acquisition step (S100), the obtained gyro measurement data and the star tracker measurement data are stored in a binary cycle and transmitted to the ground station. Receiving.

(3) angular velocity calculation and prediction step (S300)

In this step, when the relevant data is transmitted to the ground station by the measurement data transmission step (S200), the ground station calculates the angular velocity using the gyro data transmitted in a short period until the star tracker measurement time, that is, until the star tracker measurement data is available. Prediction quaternion and gyro drifting error values are predicted. Hereinafter, a method of predicting the change of satellite posture by the Extended Kalman Filter (EKF) using the transmitted gyro measurement data will be described.

)를 계산한다.
① First, using the transmitted gyro measurement data using the following equation (2) the fuselage angular velocity corresponding to the average angular velocity of the fuselage during the unit time (

).

은 동체 각속도,

는 자이로에 의해 측정된 회전각이다.
here,

Silver fuselage angular velocity,

Is the rotation angle measured by the gyro.

)가 계산되면, 아래의 수학식 3에서와 같이 동체 각속도(

)에서 자이로 표류오차 추정값(

)을 빼 동체 각속도 추정값(

)을 구한다.
② Body angular velocity (Equation 2 above)

) Is calculated, the fuselage angular velocity (

) Estimates the gyro drifting error (

Subtract the) to estimate the fuselage angular velocity (

).

는 동체 각속도 추정값,

은 동체 각속도 측정값,

는 자이로 표류오차 추정값이다. 자이로 표류오차 추정값은 별추적기 측정 정보를 이용하는 보정단계에서 결정된다.
here,

Is an estimate of the fuselage angular velocity,

Is the fuselage angular velocity measurement,

Is the gyro drifting error estimate. The gyro drifting error estimate is determined in the calibration step using the star tracker measurement information.

)을 이용하여 아래의 수학식 4에 의해 이전 주기의 자세정보의 변화량 예측값(

)을 구한 다음, 이전 주기에서 예측 또는 갱신된 위성의 위성자세 추정값(

)과 아래의 수학식 4에서와 같이 결합하여 현재의 위성자세 예측값(

)을 구하고, 아울러 수학식 5에 의해 자이로 표류오차 추정값(

)을 예측한다.
③ Estimated body angular velocity obtained from Equation 3 above (

Equation (4) using the following equation to estimate the change amount of the posture information of the previous period (

) And then estimate the satellite attitude of the satellite predicted or updated in the previous period (

) As shown in Equation 4 below

) And the gyro drift error estimate (Equation 5)

).

는 이전 주기의 자세정보의 변화량 추정값,

는 주기(0.02초),

는 이전 주기의 동체 각속도 추정값이다.
here,

Is an estimated amount of change in posture information of the previous cycle,

Is a period (0.02 seconds),

Is an estimate of the fuselage angular velocity of the previous period.

는 자세 쿼터니언 예측값,

는 이전 주기의 자세정보의 변화량 추정값,

는 주기,

는 이전 주기의 동체 각속도 추정값,

은 이전 주기의 자세 쿼터니언 추정 혹은 예측 값이다.
here,

Is a posture quaternion prediction,

Is an estimated amount of change in posture information of the previous cycle,

Cycle,

Is an estimate of fuselage angular velocity from the previous period,

Is the attitude quaternion estimate or prediction value of the previous period.

는 자이로 표류오차 추정값으로써 별추적기 측정 값을 이용한 보정단계를 통해 갱신되고, 예측단계에서는 동일한 값으로 설정된다.
here,

Is updated through a correction step using a star tracker measurement value as an gyro drifting error estimate value, and is set to the same value in the prediction step.

)을 계산하기 위해 아래의 수학식 7에서와 같이 상호분산 행렬의 예측 값도 구해준다.
④ After the current satellite posture is predicted by Equation 5 as above,

In order to calculate), the prediction value of the covariance matrix is also obtained as shown in Equation 7 below.

는 상태오차의 상호분산 행렬이고,

는 과정잡음의 상호분산 행렬이며,

는 상태천이 행렬이다.here,

Is a covariance matrix of state errors,

Is the covariance matrix of process noise,

Is the state transition matrix.

이고,

이다.Each definition is

ego,

to be.

,

,

은 자이로 측정값의 잡음특성으로써 각각 AWN(Angular White Noise), ARW(Angular Random Walk), RRW(Rate Random Walk)의 분산값을 나타내고,

는 단위행렬을 나타낸다.
here,

,

,

Are noise characteristics of the gyro measurement values, and represent variance values of Angular White Noise (AWN), Angular Random Walk (ARW) and Rate Random Walk (RWR), respectively.

Denotes a unit matrix.

)과 상태오차의 상호분산(

)을 예측 또는 계산하는 ② 내지 ④의 과정은 별추적기 측정 데이터가 이용 가능할 때까지 반복(5회)되는데, 이때 상기 과정에 의해 추정된 값들은 다시 이전 주기의 값으로 사용되고, 이에 의해 자세정보의 추정값(

)과 상태오차의 상호분산(

) 등이 계속 갱신된다.
⑤ The estimated value of the current posture information from the gyro measurement data as described above (

) And variance of state error (

The process of ② to ④ of predicting or calculating) is repeated (5 times) until the star tracker measurement data is available, wherein the values estimated by the process are used as the values of the previous periods, thereby Estimate

) And variance of state error (

) Is constantly updated.

(4) correction step (S400)

This step is a correction value using the posture quaternion estimation value and the star tracker measurement data updated by the prediction step after the prediction step S300 is finished at the time when the star tracker measurement data acquired at a relatively long period is available. Next, the error attitude quaternion and gyro drifting errors are calculated using the correction value, and the correction value is determined.

In the present invention, an attitude determination algorithm (filter) is used to determine the precise satellite attitude of the satellite, wherein the attitude determination algorithm used is a prediction and correction method such as EKF, UKF, Particle Filter, Quest, and Predictive Filter. Either one can be used and the following describes a method of correcting by the Extended Kalman Filter (EKF) using the star tracker measurement data.

를 이용하여 아래의 수학식 8에 의해 칼만이득(

)을 계산한다.
① First, when the star tracker measurement data is input, the star tracker measurement data and obtained in the prediction step (S300)

By using Kalman gain (8)

).

는 칼만이득,

는 상태오차의 상호분산,

는 관측행렬,

은 별추적기 측정잡음의 상호분산,

와

는 동체좌표계에서 각각의 별추적기 기준좌표계로의 변환행렬이고, 이때 관측행렬

이다.

와

는 최초 지상에서 측정된 값으로 주어진 후 궤도상에서 수행하는 자세제어계 보정단계를 거쳐 갱신해 준다.
here,

Is full of swords,

Is the variance of state error,

Is the observation matrix,

Is the covariance of the star tracker measurement noise,

Wow

Is the transformation matrix from the fuselage coordinate system to the reference coordinate system of each star tracker.

to be.

Wow

Is given as the value measured at the ground first and then updated through the calibration process of the attitude control system performed on track.

)을 계산하여 측정값(

)과 아래의 수학식 9에서와 같이 결합하여 별추적기 기준좌표계에서의 자세 오차(

)를 구한 다음, 아래의 수학식 9에 의해 별추적기에서 관측된 자세 오차에 대한 관측벡터(

)를 생성한다.
② Next, the posture quaternion estimated value from the Earth Centered Inertia (ECI) to the star tracker reference coordinate system (

) And the measured value (

) And posture error in the star tracker reference coordinate system by combining

), And then the observation vector for the attitude error observed in the star tracker

).

은 별추적기 기준좌표계에서의 자세 오차,

와

는 각각 지구중심관성좌표계에서 별추적기 기준좌표계로의 변환 자세 쿼터니언 추정값과 측정값이다.
here,

Is the posture error in the star tracker reference coordinate system,

Wow

Are the estimated posture quaternion estimates and measurements from the geocentric inertial coordinate system to the star tracker reference coordinate system, respectively.

는 관측벡터,

와

는 각 별추적기 측정값과 비교한 오차 자세 쿼터니언이며, (1:3)은 쿼터니언의 벡터 부분을 나타낸다.
here,

Is an observation vector,

Wow

Is the error attitude quaternion compared with each star tracker measurement, and (1: 3) represents the vector portion of the quaternion.

)과 관측벡터(

)를 이용하여 아래의 수학식 11에 의해 수정값(

, correction)을 계산한다.
③ Kalman gained by the above process (

) And the observation vector (

By using the following equation (11)

, correction).

는 수정값 벡터,

는 칼만이득이고,

는 관측벡터이다.
here,

Is the correction vector,

Is a sword full,

Is the observation vector.

)이 계산되면, 이를 아래의 수학식 12에 대입하여 보정할 자세 오차 쿼터니언(

)을 구한 다음, 이를 자세 쿼터니언 예측값(

)과 결합하여 자세 쿼터니언 추정(

)을 구함으로써 그 결과 나타나는 값을 정밀위성자세로 결정하고, 아울러 이래의 수학식 13에 의해 자이로 표류오차 추정값(

)을 갱신한다.
④ Corrected value (

) Is calculated, the attitude error quaternion (

), Then calculate the attitude quaternion prediction (

Combined with) to estimate posture quaternion (

), The resulting value is determined by the precise satellite posture, and the gyro drifting error estimated value (

).

는 결정된 정밀자세에 해당하는 쿼터니언,

는 자세 오차 수정량에 해당하는 오차 쿼터니언,

는 자세 쿼터니언 예측값이고, 이때

이다.
here,

Is the quaternion corresponding to the determined precision posture,

Is the error quaternion corresponding to the posture error correction amount,

Is a posture quaternion prediction,

to be.

는 자이로 표류오차 추정값,

은 수정값 벡터의 표류오차 보정 부분이다.
here,

Gyro drifting error estimate,

Is the drift error correction part of the correction vector.

)을 갱신한다.
⑤ Finally, the covariance matrix (

).

는 상태오차의 상호분산행렬,

는 단위행렬,

는 칼만이득행렬,

는 관측행렬,

은 별추적기 측정 잡음의 상호분산행렬이다.
here,

Is the covariance matrix of state errors,

Is a unit matrix,

Is a kalman,

Is the observation matrix,

Is the covariance matrix of the star tracker measurement noise.

)를 계산하는 과정부터 보정단계(S400)의 상호분산행렬을 갱신하는 과정까지의 전체 과정이 반복해서 수행된다.
(6) After the covariance matrix of the state error is updated, it returns to the above prediction step (S300) and the fuselage angular velocity (

), The entire process from the process of calculating to the process of updating the covariance matrix of the correction step (S400) is repeatedly performed.

As described above, the present invention acquires, stores and transmits the gyro measurement data and the star tracker measurement data having different measurement cycles by using the dual MIL-STD-1553 bus communication device, and then transmits the star tracker measurement data on the ground. The angular velocity is calculated using the gyro measurement data, which has a relatively short measurement cycle until input, and used to predict posture, thereby reducing the influence of uncertainty of process noise. do.

Claims (4)

In the method of determining the attitude of the low-orbit imaging satellite from the ground using a gyro sensor and a star tracker,
Measurement data acquisition step of acquiring gyro measurement data and star tracker measurement data at different cycles through a gyro sensor and a star tracker using a dual MIL-STD-1553 bus communication device provided in the computer mounted on the satellite (S100). Wow;
A measurement data transmission step of storing the gyro measurement data and the star tracker measurement data and transmitting them to a ground station (S200);
An angular velocity calculation and prediction step (S300) of reconstructing the gyro measurement data transmitted in a relatively short period at the ground station to calculate the angular velocity and then repeatedly predicting the posture quaternion estimate until the star tracker measurement data is received;
When the star tracker measurement data is available, the correction value is calculated using the posture quaternion estimate value and the star tracker measurement data, and then the error posture quaternion is obtained using the correction value and applied as the correction value. Comprised of a correction step (S400) to determine the posture,
The gyro sensor acquires measurement data at a period of 50 Hz, and the star tracker acquires by dividing the star tracker measurement data at a period of 10 Hz.
The method according to claim 1,
The attitude determination algorithm of a satellite is any one selected from among an extended Kalman filter (EKF), an even Kalman filter (UKF), a particle filter, a particle filter, a Quest, and a predictive filter. Ground precision posture determination method.
The method according to claim 1,
The attitude determination algorithm is an extended Kalman filter (EKF), and the Kalman gain applied to the extended Kalman filter (EKF) is continuously updated. delete

Images