A method for establishing a highly stable inertial reference for a satellite star sensor

Abstract

The application relates to a method for establishing a high-stability inertial reference of a satellite star sensor, belonging to the technical field of spacecraft control, which comprises the following steps: integrating the star sensor and a load accelerometer; using a separated two-stage light shield to suppress the stray light of the star sensor in the cabin; using a sectional heat control coating outside the two-stage light shield, and performing anti-stray-light treatment on the mounting surface of the light shield; calculating the inertial attitude of the load accelerometer by using the star sensor output, and establishing an inertial attitude model of the load accelerometer; calibrating the installation matrix of the star sensor based on the load accelerometer, and establishing a unified inertial measurement reference between the star sensor and the load accelerometer; and establishing a unified inertial measurement reference between multiple star sensors.

Description

Technical Field

[0001] This invention relates to a method for establishing a highly stable inertial reference for a satellite star sensor, and relates to the field of spacecraft control technology. Background Technology

[0002] Spacecraft typically utilize star sensors for attitude determination. The accuracy of this determination depends not only on the star sensor's own measurement precision but also on factors such as external heat flux radiation, installation accuracy, and stray light. Furthermore, star sensors serve as part of the satellite's payload, providing high-precision inertial attitude data to other payloads. Therefore, the inertial attitude data from star sensors not only affects satellite attitude determination but also influences satellite data products. This is particularly true for star sensors installed inside spacecraft, where the operating environment is more complex than that of externally installed sensors, making the factors affecting attitude determination accuracy more intricate. Additionally, on-orbit spacecraft constantly experience periodic temperature changes and relative structural deformations, preventing star sensors from providing a continuous and highly stable inertial attitude reference for the spacecraft platform and its payloads. This is especially true for spacecraft equipped with inertial acceleration sensor payloads, where existing technologies lack a unified inertial measurement reference, failing to provide a long-term, highly stable inertial reference for the inertial acceleration sensor payload. This invention proposes a method for establishing a highly stable inertial reference for satellite star sensors, providing a long-term, highly stable inertial measurement reference for spacecraft. Summary of the Invention

[0003] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a long-term, highly stable inertial measurement reference for spacecraft, ensuring high-precision attitude determination and uniformity of inertial measurement reference for spacecraft with star sensors installed in the cabin.

[0004] The objective of this invention is achieved through the following technical solutions:

[0005] A method for establishing a highly stable inertial reference for a satellite star sensor includes:

[0006] The star sensor and the load accelerometer are integrated into one unit;

[0007] A separate two-stage light shield is used to suppress stray light from the star sensors inside the cabin.

[0008] A segmented heat-controlled covering is used on the outside of the two-stage sunshade, and the mounting surface of the sunshade is treated to prevent stray light.

[0009] The inertial attitude of the load accelerometer is calculated using the output of the star sensor, and an inertial attitude model of the load accelerometer is established. Based on the calibration of the star sensor mounting matrix of the load accelerometer, a unified inertial measurement benchmark is established between the star sensor and the load accelerometer.

[0010] For multiple star sensors, a unified inertial measurement reference is established among the multiple star sensors.

[0011] In one embodiment of the present invention, the star sensor and the load accelerometer are integrated into a base, and the mounting parts of the star sensor and the base are made of the same material; the base and the whole star structure are flexibly connected to achieve thermal deformation decoupling.

[0012] In one embodiment of the present invention, the two-stage light shield includes a primary light shield and a secondary light shield. The primary light shield is fixedly connected to the star sensor body, and the secondary light shield is installed on the secondary mounting surface outside the satellite cabin. The two-stage light shields do not contact each other.

[0013] In one embodiment of the present invention, a thermal control multilayer with a thickness of N1 is wrapped around the light outlet end face of the primary light shield by a certain distance inward. The remaining part of the primary light shield is wrapped with a thermal control multilayer with a thickness of N2 according to the thermal design results, and N2>N1+the wall thickness of the secondary light shield. The thermal control multilayer on the outer surface of the secondary light shield is wrapped according to the thermal design results.

[0014] In one embodiment of the present invention, the mounting surface of the secondary light shield is blackened.

[0015] In one embodiment of the present invention, the inertial attitude model of the load accelerometer is as follows:

[0016] q acc =C b ·C s ·q m

[0017] Where q acc C represents the measured value of the load accelerometer relative to the inertial frame. b C is the matrix relating the optical reference axis of the star sensor to the coordinate system of the load accelerometer. s Let q be the matrix relating the star sensor's measurement axis to its own optical reference axis. m Let C be the measurement value of the star sensor measurement frame relative to the inertial frame; bs =C b ·C s C represents the initial mounting matrix of the star sensor relative to the load accelerometer. s C b Obtained through precise ground-based measurements.

[0018] In one embodiment of the present invention, establishing a unified inertial measurement reference between the star sensor and the load accelerometer based on the star sensor mounting matrix of the load accelerometer calibration includes:

[0019] Using the spacecraft's actuators, attitude maneuvers with fixed angular acceleration are performed in a specific sequence. The installation deviation between the star sensor and the payload accelerometer under constant acceleration is calculated using their outputs. The quaternion representation of the star sensor's measurement frame relative to the inertial frame is q. m =[q 1t ,q 2t ,q 3t ,q 4t ], where q 1t ,q 2t ,q 3t For the measured value vector part, q 4t For the scalar part, the subscript t indicates the sampling time; for two consecutive q... m By performing two difference calculations, the angular acceleration of the star sensor relative to the inertial frame is obtained:

[0020] aqdw = [aqdw] x aqdw y aqdw z ]

[0021] Among them aqdw x 、aqdw y 、aqdw z This represents the component of the star sensor's angular acceleration relative to the inertial frame in the star sensor's measurement frame.

[0022] Simultaneously, the triaxial angular acceleration Accdw = [Accdw] output from the load accelerometer is collected. x Accdw y Accdw z ]; where Accdw x Accdw y Accdw z Let Accdw be the three-axis component, and let Accdw be the theoretical satellite angular velocity;

[0023] The formula for calibrating the star sensor mounting matrix is ​​as follows:

[0024] Accdw = C as ·aqdw

[0025] Where C as To determine the true mounting matrix of the star sensor relative to the load accelerometer coordinate system, during the aforementioned attitude maneuver, real-time measurement data from the star sensor and load accelerometer were acquired and processed. The processed data was stored in a sliding window sequence. The length of the sliding window sequence was defined as 9 based on the number of parameters to be estimated. Each time the sliding window sequence data was updated, the least squares method was used to estimate C. as Perform a real-time online recursive estimation;

[0026] Based on the recursive estimation results, using C as Assigning and updating the initial installation matrix C bs This establishes a unified inertial measurement benchmark between the star sensor and the load accelerometer.

[0027] In one embodiment of the present invention, a unified inertial measurement reference is established among multiple star sensors. Taking two star sensors as an example, the method includes: first, eliminating the errors caused by vibration and thermal alternation in the orbital insertion phase between the two star sensors; then, using one of the star sensors as a reference, using measurement quaternions to perform real-time calibration and correction of the installation matrix between the two star sensors; and dynamically adjusting the calibration reference according to the angle between the solar vector direction and the optical axis of the star sensor.

[0028] In one embodiment of the present invention, real-time calibration and correction of the mounting matrix between two star sensors using measurement quaternions includes:

[0029] Define the quaternion output by star sensor 1 as q Am1 q Am2 q Am3 q Am4 The star sensor 2 outputs a quaternion q. Bm1 q Bm2 q Bm3 q Bm4 The initial installation matrix of star sensor 1 relative to the satellite is C. Abs =[X Abs Y Abs Z Abs ], where X Abs Y Abs Z Abs Let C be the direction cosine of the measurement coordinate system of star sensor 1 relative to the satellite's body, and C be the initial installation matrix of star sensor 2 relative to the satellite. Bbs =[X Bbs Y Bbs Z Bbs ], where X Bbs Y Bbs Z Bbs Let C be the direction cosine of the measurement coordinate system of star sensor 2 in the satellite's own system; using the quaternions output by the star sensor, establish the real-time attitude of the star sensor relative to the J2000 inertial coordinate system, and obtain the three-axis inertial space attitude matrix C of the measurement systems of star sensor 1 and star sensor 2. Abm and C Bbm :

[0030]

[0031] Using star sensor 1 as the calibration reference, a reference attitude of the integrated installation frame coordinate system relative to inertial space is constructed. Then, the actual on-orbit installation matrix C of star sensor 2 is calculated using the reference attitude and the real-time attitude measurement of the star sensor. nBbs :

[0032]

[0033] Next, calculate the error matrix C between the initial ground installation matrix and the actual on-orbit installation matrix of star sensor 2. ssd The formula is as follows:

[0034] C ssd =C nBbs ·C Bbs

[0035] Calculate the three-axis installation deviation angle of the star sensor according to the XYZ rotation sequence. ψ ssd :

[0036]

[0037] θ ssd =asin(C ssd [2][0])

[0038]

[0039] right θ ssd ψ ssd After filtering, the result is Constructing the installation correction matrix C Δds The details are as follows:

[0040]

[0041] Perform initial installation of star sensor 2 matrix C Bbs The corrections and updates are as follows:

[0042] C Bbs =C Δssd ·C Bbs .

[0043] In one embodiment of the present invention, the dynamic adjustment of the star sensor reference includes: when star sensor 1 is the calibration reference, calculating the angle Z1S between the optical axis ZAbs of star sensor 1 and the solar vector Si, Z1S = ZAbs·Si; if Z1S is greater than SLA, then star sensor 2 is calibrated with star sensor 1 as the reference, where SLA is the stray light suppression angle of the star sensor; otherwise, if the angle Z2S between the optical axis of star sensor 2 and the solar vector is greater than SLA, then the calibration reference is set to star sensor 2, and star sensor 1 is calibrated with star sensor 2 as the reference.

[0044] Compared with the prior art, the present invention has the following advantages:

[0045] (1) The method of the present invention can be applied to a type of spacecraft system in which the star sensor and the payload accelerometer inertial sensor are integrated and installed in the cabin. It adopts a structural integrated installation design, thermal stability and stray light protection design, calibration of the payload accelerometer and the star sensor, calibration between star sensors, and dynamic adjustment of calibration reference. This ensures that the star sensor involved in attitude determination always maintains a unified inertial reference, and comprehensively realizes the establishment of a long-term high-stability inertial measurement reference for the satellite, meeting the usage requirements of the satellite platform and the payload accelerometer inertial sensor.

[0046] (2) The present invention adopts the same material and type for the integrated installation of the star sensor and the load accelerometer inertial sensor, and the load accelerometer mounting base is installed with the flexible structure of the whole star structure, which effectively reduces the installation deviation caused by structural deformation.

[0047] (3) The present invention adopts a two-stage split-shading design method to meet the stray light suppression requirements of star sensors installed inside spacecraft under complex conditions, while also isolating the influence of external heat flow radiation on the optical system structure of star sensors, thus solving the problems of stray light suppression and heat source disturbance isolation of star sensors installed inside spacecraft cabins.

[0048] (4) This invention uses spacecraft attitude maneuvering as calibration excitation, and uses the output of the load accelerometer and star sensor during the maneuver to establish calibration equations. Using the load accelerometer as a reference, the installation deviation of the star sensor relative to the load accelerometer is calibrated online in real time, so as to achieve the unification with the inertial measurement reference of the load accelerometer. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of the installation of the two-stage light shield of the present invention.

[0050] Figure 2 This is a schematic diagram of the two-stage light shield for heat and stray light protection of the present invention. Detailed Implementation

[0051] To make the objectives, technical solutions, and advantages of the present invention clearer, the embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.

[0052] A method for establishing a highly stable inertial reference for a satellite star sensor includes:

[0053] (1) Integrated installation of star sensor and load accelerometer: The star sensor and load accelerometer are integrated through a base with extremely high thermal stability. The main structural material of the star sensor optical system and the base support are both made of aluminum-based silicon carbide with a small coefficient of thermal expansion. The same material ensures good thermal sensitivity consistency. In addition, the base is made of titanium alloy flexible corner strips and is flexibly connected to the whole star structure, realizing the thermal deformation decoupling of the base and the whole star structure, further reducing the thermal sensitivity of the structure.

[0054] (2) Separate Two-Stage Sunshield: To address the stray light suppression requirements of the star sensor inside the satellite module, a two-stage sunshield is designed based on the structural characteristics. The first-stage sunshield is fixedly connected to the star sensor body, similar in form to a traditional star sensor sunshield. The second-stage sunshield is independent and installed on the secondary mounting surface outside the satellite module. In the satellite layout and installation, the two-stage sunshield adopts a non-contact installation method, which can avoid the influence of external heat flow on the optical system of the star sensor body and isolate the heat transfer path. A schematic diagram of the two-stage sunshield installation is shown below. Figure 1 As shown.

[0055] (3) Segmented thermal control coating and stray light protection: To prevent stray light from entering the star sensor through slits inside the spacecraft cabin and to ensure a good temperature gradient effect on the star sensor's optical system and main structure, a ring-shaped thermal control multilayer with a certain unit thickness is coated on the outer surface of the two-stage shields. A thermal control multilayer of thickness N1 is coated at a distance 2mm inward from the light outlet end face of the first-stage shield. The remaining part of the first-stage shield is coated with a thermal control multilayer of thickness N2 according to the thermal design. Here, it is necessary to ensure that N2 > N1 + the wall thickness of the second-stage shield. The thermal control multilayer on the outer surface of the second-stage shield is implemented according to the specific thermal design. The mounting surface of the second-stage shield is blackened to prevent stray light interference from multiple reflections. A schematic diagram of the heat flow and stray light protection of the two-stage shields is shown below. Figure 2 As shown ( Figure 2 (The thermal control multilayer on the outer surface of the secondary light shield is not shown).

[0056] (4) Calculate the inertial attitude of the load accelerometer using the star sensor output, and establish the inertial attitude model q of the load accelerometer. acc =C b ·C s ·q m , where q acc C represents the measured value of the load accelerometer relative to the inertial frame. b C is the matrix relating the optical reference axis of the star sensor to the coordinate system of the load accelerometer. s Let q be the matrix relating the star sensor's measurement axis to its own optical reference axis. m Let C be the measurement value of the star sensor relative to the inertial frame. bs =C b ·C sC represents the initial mounting matrix of the star sensor relative to the load accelerometer. s C b Obtained through precise ground-based measurements.

[0057] (5) Calibration of the star sensor installation matrix based on the load accelerometer: Using the spacecraft actuators, attitude maneuvers with fixed angular acceleration are performed in a certain sequence. The installation deviation between the star sensor and the load accelerometer under constant acceleration is calculated. The quaternion representation of the star sensor measurement frame relative to the inertial frame is q. m =[q 1t ,q 2t ,q 3t ,q 4t ], where q 1t ,q 2t ,q 3t For the measured value vector part, q 4t For the scalar part, the subscript t indicates the sampling time; for two consecutive q... m By performing two difference calculations, the angular acceleration of the star sensor relative to the inertial frame can be obtained:

[0058] aqdw = [aqdw] x aqdw y aqdw z ], where aqdw x 、aqdw y 、aqdw z This represents the component of the star sensor's angular acceleration relative to the inertial frame in the star sensor's measurement frame; simultaneously, it controls the computer to communicate and acquire the triaxial angular acceleration Accdw output from the load accelerometer. x Accdw y Accdw z ]; where Accdw x Accdw y Accdw z Let Accdw be the three-axis component, and let Accdw be the theoretical satellite angular velocity.

[0059] The formula for calibrating the star sensor mounting matrix is ​​as follows:

[0060] Accdw = C as ·aqdw

[0061] Where C asTo determine the true mounting matrix of the star sensor relative to the load accelerometer coordinate system, during the aforementioned attitude maneuver, real-time measurement data from the star sensor and load accelerometer were acquired and processed. The processed data was stored in a sliding window sequence. The length of the sliding window sequence was defined as 9 based on the number of parameters to be estimated. Each time the sliding window sequence data was updated, the least squares method was used to estimate C. as Perform a real-time online recursive estimation.

[0062] Based on the recursive estimation results, using C as In the assignment update step (4), C bs An initial installation matrix is ​​used to establish a unified inertial measurement reference between the star sensor and the load accelerometer.

[0063] (6) Based on the calibration in step (5), a unified inertial measurement reference is established for the multi-star sensors and the load accelerometer. The dynamic adjustment of the mutual calibration of the multi-star sensors includes: first, establishing a unified inertial reference, eliminating errors caused by vibration and thermal alternation in the orbital insertion section between the two star sensors, using one of the star sensors as the reference, and using the measurement quaternion to perform real-time calibration and correction of the installation matrix between the two star sensors. The calibration reference is dynamically adjusted according to the angle between the solar vector direction and the optical axis of the star sensor. Define the output quaternion of star sensor 1 as q. Am1 q Am2 q Am3 q Am4 (Scalar), the quaternion output by star sensor 2 is q. Bm1 q Bm2 q Bm3 q Bm4 (Scalar), the initial installation matrix of star sensor 1 relative to the satellite is C. Abs =[X Abs Y Abs Z Abs ], where X Abs Y Abs Z Abs Let C be the direction cosine of the measurement coordinate system of star sensor 1 relative to the satellite's body, and C be the initial installation matrix of star sensor 2 relative to the satellite. Bbs =[X Bbs Y Bbs Z Bbs ], where X Bbs Y Bbs Z Bbs Let C be the direction cosines of the measurement coordinate system of star sensor 2 within the satellite's own system. Using the quaternions output by the star sensors, the real-time attitude of the star sensors relative to the J2000 inertial coordinate system is established, yielding the three-axis inertial space attitude matrix C of the measurement systems of star sensors 1 and 2. Abm and C Bbm :

[0064]

[0065] Using star sensor 1 as the calibration reference, a reference attitude of the integrated installation frame coordinate system relative to inertial space is constructed. Then, the actual on-orbit installation matrix C of star sensor 2 is calculated using the reference attitude and the real-time attitude measurement of the star sensor. nBbs ,

[0066]

[0067] Next, calculate the error matrix C between the initial ground installation matrix and the actual on-orbit installation matrix of star sensor 2. ssd The formula is as follows:

[0068] C ssd =C nBbs ·C Bbs

[0069] Generally, the error angle of a star sensor's actual installation relative to its initial installation is a small change. The three-axis installation deviation angle of the star sensor is calculated according to the XYZ rotation sequence. θ ssd ψ ssd :

[0070]

[0071] θ ssd =asin(C ssd [2][0])

[0072]

[0073] right θ ssd ψ ssd After filtering, the result is Constructing the installation correction matrix C Δds The details are as follows:

[0074]

[0075] Perform initial installation of star sensor 2 matrix C Bbs The corrections and updates are as follows:

[0076] C Bbs =C Δssd |C Bbs

[0077] The dynamic adjustment steps for star sensor references are as follows: When star sensor 1 is used as the calibration reference, calculate the angle Z1S between the optical axis ZAbs of star sensor 1 and the solar vector Si, where Z1S = ZAbs·Si. If Z1S is greater than SLA (where SLA is the stray light suppression angle of the star sensor), then other star sensors are calibrated using star sensor 1 as the reference. Otherwise, if the angle Z2S between the optical axis of star sensor 2 and the solar vector is greater than SLA, then the calibration reference is set to star sensor 2, and other star sensors are calibrated using star sensor 2 as the reference. This process is repeated for multiple star sensors. This dynamic adjustment method ensures that the measurement inertial reference of the star sensors involved in attitude determination remains consistent.

[0078] The contents not described in detail in this specification are common knowledge to those skilled in the art.

[0079] Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make possible changes and modifications to the technical solutions of the present invention by utilizing the methods and techniques disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the protection scope of the technical solutions of the present invention.

Claims

1. A method for establishing a high-stability inertial reference for a satellite star sensor, characterized in that, The application relates to a star sensor and a load accelerometer integratedly mounted. The star sensor in the cabin is subjected to stray light suppression by adopting a two-stage light shield. A segmented heat control coating is arranged outside the two-stage light shield, and the mounting surface of the light shield is subjected to anti-stray light treatment. An inertial attitude model of the load accelerometer is established by using the star sensor output. The star sensor installation matrix is calibrated based on the load accelerometer, and a unified inertial measurement reference between the star sensor and the load accelerometer is established. For multiple star sensors, a unified inertial measurement reference between the multiple star sensors is established. The star sensor and the load accelerometer are integratedly mounted on a base, and the star sensor and the mounting part of the base are made of the same material.

2. The method for establishing a high-stability inertial reference of a satellite star sensor according to claim 1, characterized in that, The base and the whole star structure are flexibly connected to realize thermal deformation decoupling. The two-stage light shield comprises a first-stage light shield and a second-stage light shield, the first-stage light shield is fixedly connected with the star sensor body, the second-stage light shield is mounted on a second-stage mounting surface outside the satellite cabin, and the two-stage light shield does not contact.

3. The method of claim 1, wherein the satellite star sensor high-stability inertial reference is established by: A heat control multilayer of N1 thickness is arranged inside the light outlet end surface of the first-stage light shield by a certain distance, the rest part of the first-stage light shield is coated with a heat control multilayer of N2 thickness according to the thermal design result, and N2>N1+the wall thickness of the second-stage light shield.

4. The method of claim 3, wherein, The outer surface of the second-stage light shield is coated with a heat control multilayer according to the thermal design result. The mounting surface of the second-stage light shield is subjected to blackening treatment.

5. The method of claim 3, wherein the satellite star sensor high-stability inertial reference is established by: The inertial attitude model of the load accelerometer is as follows:

6. The method of claim 1, wherein the satellite star sensor high-stability inertial reference is established by: The star sensor installation matrix is calibrated based on the load accelerometer, and a unified inertial measurement reference between the star sensor and the load accelerometer is established. wherein is the measured value of the payload accelerometer with respect to the inertial frame, is the relationship matrix of the optical reference axis of the star sensor and the payload accelerometer coordinate system, is the relationship matrix of the star sensor measurement axis and its own optical reference axis, is the measured value of the star sensor measurement system with respect to the inertial frame; and represents the initial installation matrix of the star sensor with respect to the payload accelerometer, , is obtained by ground fine measurement.

7. The method of claim 1, wherein the satellite star sensor high-stability inertial reference is established by: The star sensor installation matrix calibration formula is as follows: Using the spacecraft actuators, a fixed angular acceleration attitude maneuver is performed in a certain order, and the installation bias between the star sensor and the payload accelerometer is calculated using the outputs of the star sensor and the payload accelerometer under constant acceleration: the measurement value quaternion of the star sensor measurement system relative to the inertial system is expressed as where is the vector part of the measurement value, is the scalar part, and the subscript t represents the sampling time; twice difference calculation is performed on the adjacent two times to obtain the angular acceleration of the star sensor relative to the inertial system: wherein is the component of the angular acceleration of the star sensor with respect to the inertial system in the star sensor measurement system; Simultaneously collect three-axis angular acceleration of load accelerometer output ; wherein , , is three-axis component, take as the theoretical satellite angular angular velocity; For multiple star sensors, a unified inertial measurement reference between the multiple star sensors is established, and two star sensors are taken as examples, which include the following steps. wherein is the real installation matrix of the star sensor relative to the payload accelerometer coordinate system to be estimated, the star sensor and payload accelerometer measurement data are collected and processed in real time during the above-mentioned attitude maneuver, and the processed data are stored in a sliding window sequence. The length of the sliding window sequence is defined as 9 according to the number of parameters to be estimated. Each time the sliding window sequence data is updated, the least square method is used to perform a real-time online recursive estimation on . According to the recursive estimation result, the following is used Assignment update , so as to establish a unified inertial measurement reference between the star sensor and the payload accelerometer.

8. The method of claim 7, wherein the satellite star sensor high-stability inertial reference is established by: The installation matrix between the two star sensors is calibrated and corrected in real time by using the measurement quaternion, and the calibration reference is dynamically adjusted according to the angle relationship between the sun vector direction and the optical axis of the star sensor.

9. The method of claim 8, wherein the satellite star sensor high-stability inertial reference is established by, The installation matrix between the two star sensors is calibrated and corrected in real time by using the measurement quaternion, and the calibration reference is dynamically adjusted according to the angle relationship between the sun vector direction and the optical axis of the star sensor. The output quaternion of the star sensor 1 is defined as , , , The output quaternion of the star sensor 2 is , , , The initial installation matrix of the star sensor 1 relative to the satellite is wherein , , are the direction cosines of the measurement coordinate system of the star sensor 1 in the satellite body system respectively; the initial installation matrix of the star sensor 2 relative to the satellite is wherein , , are the direction cosines of the measurement coordinate system of the star sensor 2 in the satellite body system respectively; the real-time attitude of the star sensor relative to the J2000 inertial coordinate system is established by using the output quaternion of the star sensor, so that the three-axis inertial space attitude matrices of the measurement systems of the star sensor 1 and the star sensor 2 are obtained as follows: and ​ The star sensor 1 is taken as a calibration reference, a reference attitude of an integrated mounting frame coordinate system relative to inertial space is constructed, and a real on-orbit mounting matrix of the star sensor 2 is calculated by using the reference attitude and a real-time measured attitude of the star sensor : The error matrix of the ground initial installation matrix of the star sensor 2 and the actual on-orbit installation matrix is recalculated , and the formula is as follows: Method for calculating three-axis installation deviation angle of star sensor according to X-Y-Z rotation sequence 、 、 : right , , After filtering, the result is , , Construct the installation correction matrix The details are as follows: Perform initial installation of star sensor 2 matrix The corrections and updates are as follows: 。 10. The method of claim 9, wherein the satellite star sensor high-stability inertial reference is established by: The star sensor reference dynamic adjustment comprises: when the star sensor 1 is a calibration reference, calculating the angle between the optical axis of the star sensor 1 and the sun vector , , , If the angle between the optical axis of the star sensor 2 and the sun vector is greater than , , then the star sensor 2 is calibrated with the star sensor 1 as the reference, wherein SLA is the stray light suppression angle of the star sensor; otherwise, if the angle between the optical axis of the star sensor 2 and the sun vector is greater than , , then the calibration reference is set to the star sensor 2, and the star sensor 1 is calibrated with the star sensor 2 as the reference.

Images