A method for space-time alignment of high dynamic target fast acquisition and steady tracking

Abstract

The application discloses a kind of high dynamic target fast acquisition and steady-state tracking space-time alignment method, through gyroscope information is recursively to star sensitive exposure time and star sensitive information are filtered and fused to realize satellite platform high-precision attitude determination, again, satellite attitude determined by star sensitive exposure time is recursively to miss distance quantity time stamp, satellite attitude corresponding to miss distance quantity time stamp is obtained, and the time alignment of satellite attitude information and target observation information is completed;Again, the satellite attitude of miss distance quantity time stamp is used to transform the target line-of-sight vector under satellite camera coordinate system into the target line-of-sight vector of inertial system, the inertial position and velocity of miss distance quantity time stamp target are estimated by Kalman filtering, and then the inertial position of miss distance quantity time stamp target and the inertial position of satellite are recursively to control implementation time, and the space alignment of satellite and target is completed.The application effectively solves the problem of target pointing control misalignment caused by the inconsistency of multi-information space-time, and realizes the fast acquisition and steady-state tracking of high dynamic target.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of satellite control, in particular to a space-time alignment method for fast acquisition and steady tracking of high dynamic targets. BACKGROUND

[0002] The attitude and orbit control subsystem of a satellite is used to implement the attitude and orbit control of the satellite, and is one of the most important subsystems of the satellite platform, and plays a vital role in the successful completion of the satellite flight mission. The attitude and orbit control subsystem mainly consists of sensors, actuators and controllers. Commonly used sensors include star sensors and gyroscopes. The gyroscope is an angular rate sensor of a spacecraft, which can directly measure the three-axis inertial angular rate of the satellite. The star sensor is an angle sensor, which directly measures the three-axis inertial attitude of the satellite. Both of them are the core units for the satellite to complete the on-orbit attitude determination, stable control and attitude maneuvering function. In addition, in order to capture and track high dynamic targets, the satellite usually carries a camera (belonging to other subsystems other than the attitude and orbit control subsystem) to provide real-time target observation information (such as off-target amount information).

[0003] There is a time difference between the timing of the internal components of the attitude and orbit control subsystem and the timing between the external camera. For general targets, the existing technology mainly solves the different step of the star sensor and the gyroscope, improves the accuracy of the attitude determination, and does not introduce real-time target observation information in the calculation of the platform attitude pointing, and the sensor processing is not coupled with the external target observation information.

[0004] For high dynamic targets, compared with general targets, high coupling between the attitude and orbit control subsystem and the camera is required. In the case of information misalignment, high-precision satellite position and attitude determination and fast attitude maneuvering are completed, so as to realize fast acquisition and steady tracking of the target. The higher the target dynamic is, the greater the pointing control deviation caused by information misalignment is. Therefore, in order to realize fast acquisition and steady tracking of high dynamic targets, this problem must be solved, so as to accurately and quickly calculate the attitude adjustment control quantity required by the attitude and orbit control subsystem according to the target observation information provided by the camera. SUMMARY

[0005] The technical problem solved by the present application is that, in view of the technical difficulties in the prior art, a space-time alignment method for fast acquisition and steady tracking of high dynamic targets is provided, which ensures the fast synchronization of the processing timing of each link inside the attitude and orbit control subsystem and between the attitude and orbit control subsystem and other external subsystems, so as to obtain high-precision satellite position and target position in space, thereby realizing fast and accurate satellite attitude adjustment and achieving the purpose of fast acquisition and steady tracking of high dynamic targets.

[0006] The technical solution of the application is a space-time alignment method for high-dynamic target fast acquisition and steady tracking, which is realized by a star sensor, a gyroscope of an on-board attitude and orbit control system, and a space-borne camera, and comprises the following steps:

[0007] S1, obtaining satellite inertial attitude quaternion estimation at star sensor exposure time by adopting star sensor and gyroscope joint filtering;

[0008] S2, estimating target position and velocity corresponding to miss distance time stamp, wherein the miss distance time stamp is the time corresponding to the target image captured by the space-borne camera;

[0009] S21, converting the satellite inertial attitude quaternion estimation at star sensor exposure time calculated in S1 to the miss distance time stamp to obtain satellite inertial attitude quaternion corresponding to the miss distance time stamp;

[0010] S22, calculating satellite inertial position corresponding to the miss distance time stamp;

[0011] S23, calculating target line-of-sight vector in the camera coordinate system and then transforming the target line-of-sight vector to the geocentric inertial system;

[0012] S24, estimating target inertial position and velocity corresponding to the miss distance time stamp based on the calculation results of S21, S22 and S23;

[0013] S3, recursively calculating target inertial position and velocity corresponding to the miss distance time stamp to the control implementation time to obtain target inertial position at the control implementation time;

[0014] S4, calculating satellite inertial position at the control implementation time; and obtaining the guide quantity required for the satellite to capture and track the target by the satellite inertial position at the control implementation time and the target inertial position, thereby completing the space-time alignment between the satellite and the target.

[0015] Further, in S24, the target inertial position and velocity corresponding to the miss distance time stamp are estimated in the following manner: the Kalman filtering is adopted to estimate the state quantity X k which is set as the target inertial position and velocity corresponding to the miss distance time stamp to be estimated; and the kth filtering process is as follows:

[0016] (1) system matrix updating;

[0017]

[0018] wherein μ is the earth gravity constant, ΔT pix is the time difference between the current miss distance time stamp and the previous miss distance time stamp, r t,k-1 is the target inertial position corresponding to the state quantity of the previous system, I 3×3 is a 3x3 unit matrix.

[0019] (2) State one-step prediction;

[0020] X k|k-1 = A k X k-1

[0021] wherein X k-1 is the state quantity of the last frame, and X k|k-1 is the state quantity of one-step prediction;

[0022] (3) Measurement matrix update and observation update;

[0023] The measurement matrix is:

[0024]

[0025] wherein C1 is the star vector r st,k corresponding to the Jacobian matrix, r st,k = r t,k|k-1 -r pix,k ; 0 3×3 is a 3x3 zero matrix; r t,k|k-1 is the target position vector of state one-step prediction, and r pix,k is the target position vector calculated according to the miss distance;

[0026] The observation is:

[0027]

[0028] wherein ΔS i,k is the target line-of-sight vector increment in the inertial system, and the calculation method is:

[0029]

[0030] wherein S i,k is the target line-of-sight vector in the current frame in the inertial system, and S i,k-1 is the target line-of-sight vector in the previous frame in the inertial system;

[0031] (4) Residual calculation;

[0032] Δγ k = y k -C k X k|k-1

[0033] wherein Δγ k is the residual;

[0034] (5) One-step prediction of error covariance matrix;

[0035]

[0036] wherein P k|k-1 is the error covariance matrix of one-step prediction, P k-1 is the error covariance matrix of the last step, and Q is the noise covariance matrix.

[0037] (6) Gain matrix calculation;

[0038]

[0039] wherein K k is the gain matrix, δ0 is the amplification parameter, and R is the measurement noise covariance matrix.

[0040] (7) State quantity estimation;

[0041] X k = X k|k-1 + K k Δγ k

[0042] (8) Error covariance matrix update;

[0043]

[0044] The state quantity X k obtained in step (7) is the target inertial position and velocity corresponding to the time stamp of the miss distance quantity to be calculated.

[0045] Further, in S1, a star sensor and a gyroscope are combined to filter to obtain the satellite inertial attitude quaternion estimation at the star sensor exposure time, and the specific manner is as follows:

[0046] Based on the attitude quaternion of the star sensor relative to the geocentric inertial coordinate system obtained by the star sensor, the first satellite inertial attitude quaternion is calculated.

[0047] Based on the satellite inertial angular velocity obtained by the gyroscope, the second satellite inertial attitude quaternion is calculated.

[0048] The first satellite inertial attitude quaternion and the second satellite inertial attitude quaternion are compared to obtain the attitude deviation quaternion; the Kalman filtering method is used to eliminate the high-frequency items in the attitude deviation quaternion to estimate the gyroscope constant drift and the attitude deviation; the satellite inertial angular velocity output by the gyroscope is corrected by the gyroscope constant drift, and the second satellite inertial attitude quaternion is corrected by the attitude deviation; the corrected second satellite inertial attitude quaternion is the satellite inertial attitude quaternion estimation at the star sensor exposure time.

[0049] Further, the first satellite inertial attitude quaternion and the second satellite inertial attitude quaternion are both the attitude quaternions of the satellite body system relative to the geocentric inertial system.

[0050] Further, in S21, the satellite inertial attitude quaternion estimation conversion mode is:

[0051] The satellite inertial attitude quaternion estimation at the star exposure time is integrated into the miss distance timestamp by using the satellite inertial angular velocity, so as to obtain the satellite inertial attitude quaternion corresponding to the miss distance timestamp.

[0052] Further, in S22, the calculation mode of the satellite inertial position corresponding to the miss distance timestamp is any calculation mode based on GNSS data calculation or based on the injection parameter calculation.

[0053] Further, in S23, the target line-of-sight vector in the geocentric inertial system is calculated, and the specific mode is:

[0054] First, the target line-of-sight vector in the camera coordinate system is calculated; the target line-of-sight vector in the camera coordinate system is converted into the satellite body system by using the installation matrix of the camera coordinate system relative to the satellite body system; the target line-of-sight vector in the satellite body system is converted into the geocentric inertial system by using the satellite inertial attitude quaternion corresponding to the miss distance timestamp obtained in S21, so as to obtain the target line-of-sight vector in the geocentric inertial system.

[0055] Further, in S3, the target position at the control implementation time is obtained by using the unpowered recursion mode.

[0056] Further, in S4, the direction cosine matrix A ti which is required for the satellite to capture and track the target, is calculated. ti The direction cosine matrix A

[0057]

[0058] A ti =[OY*OZ OY OZ] T

[0059] wherein, is the target inertial position vector at the control implementation time; r exc is the satellite inertial position vector at the control implementation time.

[0060] Compared with the prior art, the present application has the following advantages:

[0061] (1) The present application firstly realizes high-precision satellite platform attitude determination by filtering and fusing star sensor information and star sensor exposure time point information, and then recursively determines the satellite attitude at the target observation information time point (miss distance time stamp) to obtain the satellite attitude corresponding to the miss distance time stamp. The above two steps complete the time alignment of the satellite attitude information and the target observation information. Further, the target line-of-sight vector in the satellite camera coordinate system is transformed into the target line-of-sight vector in the inertial system using the satellite attitude at the miss distance time stamp, and the inertial system position vector and velocity vector of the target at the miss distance time stamp are estimated using the target line-of-sight vector in the inertial system as the observation quantity. Further, the inertial system position vector of the target at the miss distance time stamp and the inertial system position vector of the satellite platform are recursively determined at the control implementation time of the satellite platform actuator, and the above two steps complete the spatial alignment of the satellite and the target.

[0062] (2) The present application provides a specific implementation method for estimating the inertial system position vector and velocity vector of the target at the miss distance time stamp by Kalman filtering. The inertial system position and velocity of the target at the miss distance time stamp are estimated using the target line-of-sight vector in the inertial system as the observation quantity.

[0063] (3) The present application effectively solves the problem of inaccurate target observation reference calculation caused by the space-time inconsistency of multiple input information when the satellite platform attitude is coupled to target observation estimation without increasing redundant products, effectively improves the dynamic adaptability and robustness of target observation, and realizes reliable high-dynamic target rapid capture and steady tracking. BRIEF DESCRIPTION OF DRAWINGS

[0064] Figure 1 The present application relates to a method process schematic diagram. DETAILED DESCRIPTION

[0065] In order to better understand the technical scheme of the present application, the embodiments of the present application will be specifically described below with reference to the drawings.

[0066] The present application is aimed at the attitude sensor (star sensor, gyro) of the satellite platform attitude and orbit control subsystem and the target observation information (miss distance information) obtained from the outside of the attitude and orbit control subsystem. The information time sequence is inconsistent, which leads to different information times, and the corresponding time of the inertial position vector of the target to be observed and the satellite platform is inconsistent. In order to realize real-time target capture and tracking, the above problems of different times and space inconsistency are aligned in time and space by the method of the present application, so as to ensure high-precision high-dynamic target pointing control.

[0067] Firstly, the time relationship and the meaning of different time points involved in the present method are described, Figure 1In this context, ETR refers to the periodic time reference generated by external hardware and input into the attitude and orbit control subsystem; the satellite angular velocity data output by the gyroscope at ETR k is denoted as gyroscope data k; the satellite attitude data output by the star sensor at ETR k is denoted as star sensor data k, and the corresponding star sensor exposure time is T. k-2 +Ams; The miss distance information output by the camera at time ETR k is denoted as miss distance information k, and the corresponding camera detector exposure time is T. k-2 +Bms; A represents the deviation of each beat star-sensor exposure time from the ETR time, and B represents the deviation of the off-target timestamp from the ETR time.

[0068] Taking a certain in-orbit model as an example, this method includes the following process:

[0069] Step 1: Using a combined star-sensor and gyroscope filtering method, obtain the satellite inertial attitude quaternion estimate at the star-sensor exposure time.

[0070] Specifically as follows:

[0071] (1) The attitude quaternion q of the star-sensor relative to the geocentric inertial coordinate system is obtained based on star-sensor measurements. si Based on the onboard calibration information, the star-sensor installation and thermal deformation were corrected. Then, after coordinate transformation and fusion of the star positions of the two star-sensor heads, the attitude quaternion of the satellite's main system relative to the geocentric inertial frame was calculated. Its error includes star sensor measurement noise;

[0072] (2) Obtaining the satellite's inertial angular velocity ω based on a gyroscope m Using the attitude kinematic equations based on the given initial values, the attitude quaternion of the satellite's main system relative to the geocentric inertial frame at the current moment is calculated. Its error includes gyroscope constant drift and measurement noise;

[0073] (3) Compare attitude quaternions and Obtain the attitude deviation quaternion (Contains gyroscope constant drift, measurement noise, and star sensor measurement noise). An error model is established based on the principles of the star sensor and gyroscope, and Kalman filtering is used to eliminate these noises. The high-frequency terms (noise from the star sensor and gyroscope) are estimated to obtain the gyroscope constant drift b and attitude deviation q. e Use b to correct the satellite inertial angular velocity output of the gyroscope, and use q. e Correct the attitude quaternion calculated by the gyroscope. Revised That is, the quaternion estimate of the satellite's inertial attitude is obtained.

[0074] Step 2, estimate the target position and velocity corresponding to the miss distance timestamp

[0075] Considering that the sensor information, miss distance information, and orbit information have different update periods, they are all unified to the control implementation time, as shown in Figure 1 . Among them, the miss distance information is provided by the camera to the attitude and orbit control subsystem, including the miss distance obtained by the camera shooting target, and the miss distance timestamp, that is, the shooting time of this shooting target.

[0076] In order to realize the estimation of the target position and velocity corresponding to the miss distance timestamp, the satellite inertial attitude quaternion at the same time needs to be obtained pix , combined with the target line-of-sight vector in the inertial system to solve.

[0077] (1) Calculate the satellite inertial attitude quaternion corresponding to the miss distance timestamp

[0078] Let the time corresponding to the kth ETR be T k , and there is a certain time difference between the miss distance timestamp and the beat of the gyroscope obtaining the satellite inertial angular velocity. The relationship can be expressed as: T k-2 ±Bms, where B represents the deviation of the miss distance timestamp relative to the time T k-2 of the previous 2 ETRs in each beat. Due to the change of satellite platform timing, B is different in each beat, but it is usually controlled within 50ms.

[0079] Based on the inertial angular velocity ω bi,k-1 obtained by the previous beat (i.e. T k-1 ) gyroscope, the satellite inertial attitude quaternion estimate obtained in step 1 is integrated to the miss distance timestamp, to obtain the satellite inertial attitude quaternion corresponding to the miss distance timestamp

[0080] (2) Calculate the satellite inertial position r pix

[0081] The on-orbit orbit calculation method is divided into GNSS data-based calculation method or on-annotation parameter-based calculation method, and any method can be selected as needed. According to the current satellite orbit calculation method, the satellite inertial position r pix corresponding to the miss distance timestamp is obtained.

[0082] (3) Calculate the target line-of-sight vector S i

[0083] The camera image plane center coordinates are (C x , C y ), and the camera coordinate system Oc X c Y c Z c , the origin is at the center of the image plane, Z c axis is perpendicular to the image plane and outward, X c Y c plane is defined in accordance with the definition of the image plane coordinate axis.

[0084] First, the target line-of-sight vector S c in the camera coordinate system is calculated. bc The installation matrix Aof the camera coordinate system relative to the satellite body coordinate system is obtained, and the target line-of-sight vector in the body coordinate system is obtained, and the satellite inertial attitude quaternion corresponding to the miss distance timestamp is obtained. The target line-of-sight vector is converted to the inertial system S i .

[0085] (4) Estimate the target position and velocity corresponding to the miss distance timestamp

[0086] After the target line-of-sight vector in the inertial system is solved, according to the dynamics of the target, the Kalman filter is used to estimate the target inertial position and velocity corresponding to the miss distance timestamp

[0087] The state quantity X k is set as the target inertial position and velocity corresponding to the miss distance timestamp to be estimated; the k-th filtering process is:

[0088] (1) System matrix update;

[0089]

[0090] Where μ is the earth's gravitational constant, the value is 3.986005e14, ΔT pix is the time difference between the current miss distance timestamp and the miss distance timestamp of the previous shot, r t,k-1 is the target inertial system position corresponding to the system state quantity of the previous shot, I 3×3 is a 3x3 unit matrix;

[0091] (2) State one-step prediction;

[0092] X k|k-1 = A k X k-1

[0093] Where X k-1 is the state quantity of the previous shot, and X k|k-1 is the prediction of the state quantity of the current shot;

[0094] (3) Measurement matrix update, observation update;

[0095] The measurement matrix is:

[0096]

[0097] wherein C1 is the star vector r st,k The corresponding Jacobian matrix, r st,k = r t,k|k-1 - r pix,k ; 0 3×3 is a 3x3 zero matrix; r t,k|k-1 is r t,k|k-1 is the target position vector of one-step prediction of the state:

[0098] r t,k|k-1 = X k|k-1 (1:3)

[0099] r pix,k is the target position vector constructed according to the miss distance;

[0100] The observation is:

[0101]

[0102] wherein ΔS i,k is the target line-of-sight vector increment in the inertial system, and is calculated as:

[0103]

[0104] wherein S i,k is the target line-of-sight vector in the current frame of inertia, and S i,k-1 is the target line-of-sight vector in the previous frame of inertia, both of which are calculated from the target miss distance;

[0105] (4) Residual calculation;

[0106] Δγ k = y k - C k X k|k-1

[0107] wherein Δγ k is the residual;

[0108] (5) One-step prediction of error covariance matrix;

[0109]

[0110] wherein P k|k-1 is the one-step predicted error covariance matrix, P k-1 is the error covariance matrix of the previous frame; and Q is the noise covariance matrix;

[0111] (6) Gain matrix calculation;

[0112]

[0113] wherein, K k is the gain matrix, δ0 is the amplification parameter, which can be taken as δ0 = 10 14 , and R is the measurement noise covariance matrix.

[0114] (7) State quantity estimation;

[0115] X k = X k|k-1 + K k Δγ k

[0116] (8) Error covariance matrix update;

[0117]

[0118] The state quantity X k obtained in step (7) is the target inertial position corresponding to the miss distance timestamp to be calculated velocity Specifically,

[0119]

[0120] Step 3, recursively calculating the target position at the control implementation time

[0121] The control implementation time is an artificially set time, and the target inertial position and velocity corresponding to the miss distance timestamp have been estimated through Kalman filtering. In order to calculate the control attitude angle at the control implementation time, the target inertial position at the control implementation time is needed. Therefore, the target position and velocity corresponding to the miss distance timestamp are recursively calculated to the target position at the control implementation time; preferably, the calculation can be performed in a non-powered recursive manner.

[0122] Step 4, calculating the guidance information

[0123] First, the satellite inertial position at the control implementation time is calculated, and the calculation method is similar to (2) in step 2;

[0124] Then, according to the target position and satellite inertial position at the control implementation time, the target-geocentric double vector reference is adopted to obtain the direction cosine matrix A ti from the inertial system to the tracking reference system, i.e. the guidance information is obtained.

[0125] wherein, the direction cosine matrix A ti from the inertial system to the reference system is obtained

[0126]

[0127] A ti = [OY x OZ OY OZ] T

[0128] Preferably, the direction cosine matrix can be converted into a quaternion q ti , so as to calculate the guide attitude angle, angular velocity and angular acceleration.

[0129] It can be understood that the present application is described through embodiments, and those skilled in the art know that various changes or equivalent replacements can be made to the features and embodiments without departing from the spirit and scope of the present application. In addition, the features and embodiments can be modified to adapt to specific conditions under the guidance of the present application without departing from the spirit and scope of the present application. Therefore, the present application is not limited by the specific embodiments disclosed herein, and the embodiments falling within the scope of the claims of the present application are within the scope of protection of the present application.

[0130] The contents not described in detail in the specification of the present application are the known technology of those skilled in the art.

Claims

1. A method for space-time alignment of high dynamic target fast acquisition and steady tracking, characterized in that: The star sensor, the gyroscope and the satellite camera are jointly implemented based on an on-board attitude and orbit control subsystem, and the implementation includes: S1, a satellite inertial attitude quaternion value at a star sensor exposure time is obtained by using a star sensor and a gyroscope joint filtering method; S2, a target position and a target speed corresponding to a miss distance time stamp are estimated, the miss distance time stamp being a time corresponding to a target image captured by the satellite camera; S21, the satellite inertial attitude quaternion value at the star sensor exposure time calculated in S1 is converted to the miss distance time stamp, so as to obtain a satellite inertial attitude quaternion corresponding to the miss distance time stamp; S22, a satellite inertial position corresponding to the miss distance time stamp is calculated; S23, a target line-of-sight vector in a camera coordinate system is calculated, and a target line-of-sight vector in an earth-centered inertial system is further obtained by transformation; S24, based on the calculation results of S21, S22 and S23, a target inertial position and a target speed corresponding to the miss distance time stamp are estimated; S3, the target inertial position and the target speed corresponding to the miss distance time stamp are recursively calculated to a control implementation time, so as to obtain a target inertial position at the control implementation time; S4, a satellite inertial position at the control implementation time is calculated, and a guide quantity required for the satellite to capture and track the target is obtained by using the satellite inertial position at the control implementation time and the target inertial position, so as to complete the space-time alignment between the satellite and the target.

2. The spatio-temporal registration method of high dynamic target fast acquisition and steady tracking according to claim 1, characterized in that: In S24, the target inertial position and speed corresponding to the off-target timestamp are estimated in the following manner: the Kalman filter is used for estimation, and the state quantity X k The target inertial position and speed corresponding to the off-target timestamp to be estimated are set; and the kth filtering process is as follows: (1) system matrix updating; where μ is the earth gravitational constant, ΔT pix is the time difference between the current miss distance timestamp and the previous miss distance timestamp, r t,k-1 is the target inertial frame position corresponding to the system state quantity of the previous shot, I 3×3 is a 3x3 identity matrix; (2) one-step state prediction; X k|k-1 = A k X k-1 where X k-1 is the state quantity of the previous frame, X k|k-1 is the state quantity of one-step prediction; (3) measurement matrix updating and observation updating; the measurement matrix is: where C1is the star vector r st,k The corresponding Jacobian matrix, r st,k = r t,k|k-1 - r pix,k ; 0 3×3 is a 3 x 3 zero matrix; r t,k|k-1 is the target position vector predicted one step ahead of the state, r pix,k is the target position vector calculated from the current miss distance; the observation is: where ΔS i,k is the target line-of-sight vector increment in the inertial frame, calculated as Wherein, S i,k is the target line-of-sight vector in the current frame of inertia, S i,k-1 is the target line-of-sight vector in the previous frame of inertia; (4) residual error calculation; Δγ k = y k - C k X k|k-1 where Δγ k is the residual error; (5) one-step error covariance matrix prediction; where P k|k-1 is the error covariance matrix of one-step prediction, P k-1 is the error covariance matrix of the last frame; Q is the noise covariance matrix; (6) gain matrix calculation; where K k is the gain matrix, δ0is the amplification parameter, and R is the measurement noise covariance matrix. (7) state quantity estimation; X k = X k|k-1 + K k Δγ k (8) error covariance matrix updating; The state quantity X obtained in step (7) k The target inertial position and velocity corresponding to the off-target quantity timestamp required for the calculation.

3. The method of claim 1, wherein: In S1, the satellite inertial attitude quaternion value at the star sensor exposure time is obtained by using the star sensor and the gyroscope joint filtering method, and the specific method is as follows: a first satellite inertial attitude quaternion is calculated based on an attitude quaternion of the star sensor relative to an earth-centered inertial coordinate system; a second satellite inertial attitude quaternion is calculated based on a satellite inertial angular velocity obtained by the gyroscope; a comparison between the first satellite inertial attitude quaternion and the second satellite inertial attitude quaternion is performed to obtain an attitude deviation quaternion; a Kalman filtering method is used to eliminate high-frequency items in the attitude deviation quaternion, so as to estimate a gyroscope constant drift and an attitude deviation; the satellite inertial angular velocity output by the gyroscope is corrected by using the gyroscope constant drift, and the second satellite inertial attitude quaternion is corrected by using the attitude deviation, and the corrected second satellite inertial attitude quaternion is the satellite inertial attitude quaternion value at the star sensor exposure time.

4. The spatio-temporal registration method of high dynamic target fast acquisition and steady tracking according to claim 3, characterized in that: Both the first satellite inertial attitude quaternion and the second satellite inertial attitude quaternion are attitude quaternions of a satellite body system relative to the earth-centered inertial system.

5. The method of claim 1, wherein: In S21, the satellite inertial attitude quaternion value conversion method is as follows: the satellite inertial attitude quaternion value at the star sensor exposure time is integrated to the miss distance time stamp by using the satellite inertial angular velocity, so as to obtain the satellite inertial attitude quaternion corresponding to the miss distance time stamp.

6. The method of spatio-temporal registration of high dynamic target fast acquisition and steady tracking according to claim 1, characterized in that: In S22, the calculation method of the satellite inertial position corresponding to the miss distance time stamp is any calculation method based on GNSS data calculation or based on an uploaded parameter calculation.

7. The method of claim 1, wherein: In S23, the target line-of-sight vector in the earth-centered inertial system is calculated, and the specific method is as follows: First, the target line-of-sight vector in the camera coordinate system is calculated; the target line-of-sight vector in the satellite coordinate system is obtained by using the installation matrix of the camera coordinate system relative to the satellite coordinate system; the target line-of-sight vector in the earth-centered inertial coordinate system is obtained by using the satellite inertial attitude quaternion corresponding to the time stamp of the miss distance obtained by S21, and the target line-of-sight vector in the satellite coordinate system is converted to the earth-centered inertial coordinate system.

8. The method of claim 1, wherein: In S3, the target position at the control implementation time is obtained by using the unpowered recursive method.

9. The space-time registration method for high dynamic target fast acquisition and steady tracking according to claim 1, characterized in that: In S4, the direction cosine matrix A from the inertial frame to the tracking reference frame ti As a guidance quantity required for the satellite to capture and track the target, A ti is calculated by the following equation: A ti = [OY x OZ OY OZ] T wherein, r is the target inertial position vector at the time of control implementation; exc r is the satellite inertial position vector at the time of control implementation.

Images