CN106568462A - Multi-probe star sensor fusion attitude testing method - Google Patents

Abstract

The invention provides a multi-probe star sensor fusion attitude testing method. On the basis of error characteristics of a star sensor, corresponding attitude testing methods are decomposed, classified, and drafted, the problem that there is no rule for the tests of the fusion attitude of a multi-probe star sensor is solved, and a guide is provided for the fusion applications of the star sensor in the system level. The provided system error test method and period error test method can explore the fusion attitude ability boundary of the star sensor, further provide constrains for star sensor support, whole star thermal control, and mechanical structure design, and are benefit for the implementation of a high precision satellite plan.

Description

Multi-probe star sensor fusion attitude testing method

Technical Field

The invention relates to the technical field of star sensors, in particular to a multi-probe star sensor fusion attitude testing method which is suitable for a star sensor test for fusion attitude determination according to information of a plurality of probes, can provide guidance for system-level fusion application of the star sensor, and can also be applied to non-information fusion type star sensor tests.

Background

The star sensor is a high-precision optical attitude sensor widely applied to attitude measurement of spacecrafts at present, and takes a fixed star as a measurement target, the fixed star is imaged on a photoelectric converter through an optical system, and the orientation of an optical axis vector of the star sensor under an inertial coordinate system is determined by combining a fixed star table through star point extraction and star map identification, so that the attitude measurement is completed.

Most of the traditional star sensors are of an integrated structure, namely, the optical probes correspond to processing lines one by one, and the processing lines only calculate the inertia postures of the corresponding probes. In an actual application environment, a spacecraft selects the number of star sensors according to different requirements, designs corresponding installation orientations, and then performs System-level Attitude determination in an Attitude and Orbit Control System (AOCS) (reference [1 ]: Jie Li, Yiqing Chen. constant-gain information filter for Attitude determination of precision pointing space technology. 47th International adaptive convergence, 1996; reference [2 ]: Liu Yiwu, Chengxi. star sensor measurement model and application thereof in a satellite Attitude determination System. aerospace science report, 2003). The system-level fusion attitude determination algorithm is generally related to star sensor performance indexes and installation orientations.

In order to improve the dynamic performance and accuracy index of the Star Sensor and enhance the reliability of the product, research institutes have developed a multi-probe information fusion type Star Sensor, that is, a single processing line can process Multiple probe information and the single probe information can be used for Multiple processing lines, and in the processing lines, the probe information can be fused according to the installation position of each probe to give higher accuracy attitude measurement, such as HYDRA Star Sensor (reference: l.blank, n.permon.new Multiple Head Star Sensor (HYDRA) description and maintenance, developed by SODERN corporation, france, a high elevation autonomy, access and vertical robust system to board for the Navigation access systems, aiaa guide, Navigation, and exposure, control, 2005). The method for measuring the attitude by utilizing the information fusion indirectly enlarges the combined view field of the star sensor and increases the number of available stars. In addition, the star sensor can independently complete the functions of calibration of mounting precision among probes, low-frequency error estimation, time difference aberration compensation and the like, so that the attitude determination precision is improved, and the output attitude is called as a fusion attitude. The second-generation triaxial stable stationary orbit meteorological satellite in China adopts HYDRA as a main attitude sensor, and similar fusion algorithms are designed on the AOCS system level of the meteorological satellite as the traditional satellite.

In summary, the existing AOCS uses star sensors by means of system-level information fusion, a fusion algorithm has a certain correlation with product characteristics, ground tests do not pay attention to various errors of the star sensors, even relevant tests are not implemented, and system design cannot be effectively checked. The development and application of the information fusion type star sensor are still in a preliminary stage, certain ground test is required to guarantee how system-level fusion or star sensor-level fusion is decided on track, and in addition, the ground test can find out the fusion capability boundary of the star sensor, so that the constraint is provided for the design of a star sensor support, the integral star thermal control and a mechanical structure, and the implementation of a high-precision satellite scheme is facilitated.

Disclosure of Invention

The technical problem of the invention is solved: the method overcomes the defects of the prior art, provides a multi-probe star sensor fusion attitude test method, solves the problem that the multi-probe star sensor fusion attitude test is not applicable, and can provide reference for how to select system-level fusion or sensor-level fusion for the spacecraft on orbit.

The technical solution of the invention is as follows: a multi-probe star sensor fusion attitude testing method comprises the following steps:

(1) when a random error test is carried out, the step (2) is carried out, when a system error test is carried out, the step (4) is carried out, when a periodic error test is carried out, the step (6) is carried out, and when a time difference optical performance test is carried out, the step (7) is carried out;

(2) obtaining star sensor theoretical measurement attitude matrix C_SIStar sensor optical axis noise index sigma_xStar sensor horizontal axis noise index sigma_zAnd calculating to obtain a star sensor measurement attitude matrix tempC under random error as

In the formula, randn (σ) represents the generation of gaussian noise with mean 0 and mean square error σ, and σ is represented by σ_xOr σ_z；

(3) For random errorPerforming orthogonal normalization processing on the star sensor measurement attitude matrix tempC under the difference to obtain the actual measurement attitude matrix of the star sensorWherein,is composed ofIn the case of the 1 st row of (c),is composed ofIn the case of the 2 nd row of (c),is composed ofRow 3 of (2), norm is the norm normalization function of vector L2, × is the vector cross multiplier

Turning to the step (9);

(4) obtaining equivalent installation deviations delta x, delta y and delta z of the star sensor under a body coordinate system of an installed satellite, and calculating to obtain an installation deviation matrix delta C of the star sensor_SBIs composed of

ΔC_SB＝R_y(Δy)·R_x(Δx)·R_z(Δz)

Wherein R is_x(upsilon) is a direction cosine array R when the angle upsilon rotates around the X axis in the satellite body coordinate system_y(upsilon) is a direction cosine array R when the angle upsilon rotates around the Y axis in the satellite body coordinate system_z(v) is around the satellite bodyA direction cosine array when a Z-axis rotation angle upsilon is in a coordinate system;

(5) obtaining a theoretical installation matrix C of the star sensor_SBSatellite inertial attitude matrix C_BIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

Turning to the step (9);

(6) receiving a rotating shaft selection instruction sent from the outside, selecting a rotating shaft, and acquiring the interference amplitude A of the star sensor_STStar sensor interference period T_STWhen the selected rotating shaft is the optical axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_XIs composed of

When the selected rotating shaft is the horizontal axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_ZIs composed of

Then, according to the star sensor measurement error matrix and the theoretical measurement attitude matrix C_SICalculating the actual measurement attitude matrix of the current star sensorIs composed of

Or

Turning to the step (9);

wherein t is the star sensor measuring time; the rotating shaft selection instruction comprises an optical axis or a transverse axis; the rotating shaft comprises an optical axis and a transverse axis;

(7) when the current epoch is obtained, then calculating the age compensation matrix C of the star sensor_PRIs composed of

Therein, ζ_A、θ_APhi, phi and delta are the time difference description parameters;

(8) obtaining a theoretical measurement attitude matrix C of the star sensor_SIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

Turning to the step (9);

(9) calculating the actual measurement attitude matrix of the current star sensorCorresponding quaternionThen, acquiring a quaternion q output by the actual measurement of the star sensor_SICalculating to obtain error quaternionIs composed of

In the formula,is q_SIThe number of the conjugate quaternion of (c),is a quaternion multiplier;

then based on the error quaternionCalculating to obtain the three-axis equivalent attitude error e of the current star sensor_x、e_y、e_zIs composed of

In the formula,representing error quaternionI-1, 2, 3.

Compared with the prior art, the invention has the following advantages:

(1) on the basis of comprehensively analyzing the error characteristics of the star sensor, the testing method solves the problem that the multi-probe star sensor fusion attitude test is not applicable and recyclable, and is more favorable for understanding various performance indexes of products compared with the prior art;

(2) compared with the prior art, the error testing method is beneficial to analyzing the relationship between the attitude output effectiveness and the system deviation of the multi-probe star sensor, indirectly provides constraint for the installation deviation of the star sensor, and provides help for the mechanical structure design of the whole star;

(3) compared with the prior art, the error testing method can better analyze the relation between the fusion attitude determination precision and the period error of the multi-probe star sensor, indirectly provides constraint for the structural deformation of the star sensor mounting bracket on the premise of meeting a certain precision index, and provides reference for the mechanical structure and the thermal control design of the whole star;

(4) compared with the prior art, the error testing method is more beneficial to mining various performance indexes of products, and has a guiding function on the on-orbit sensor level/system level fusion selection.

Drawings

FIG. 1 is a block diagram of a test system according to the present invention;

FIG. 2 is a main flow chart of the test of the present invention;

FIG. 3 is a flow chart of various test modes of the present invention.

Detailed Description

The invention provides a multi-probe star sensor fusion attitude testing method aiming at the defects of the prior art. The following takes the HYDRA star sensor test developed by SODERN of France as an example, and the detailed implementation of the method of the invention is described in detail with reference to the accompanying drawings, and comprises the following steps:

(1) the star sensor ground test system is connected as shown in fig. 1 and consists of ground test equipment and satellite-borne equipment. The Ground test equipment is a dynamics computer and egse (electrical group Support equipment), the satellite-borne equipment comprises a star sensor (including a processing circuit Electronic Unit and an Optical Head) and an AOCS, and the dynamics computer runs the test programs (namely, the formula algorithm in step 2).

(2) According to the test purpose, a star sensor test mode is set, a dynamics computer selects a corresponding test sub-process according to the test mode, and fig. 2 is a test main process block diagram. When a random error test is carried out, the step (2a) is carried out, when a system error test is carried out, the step (2b) is carried out, when a periodic error test is carried out, the step (2c) is carried out, when a time difference optical performance test is carried out, the step (2d) is carried out, and fig. 3 is a sub-flow block diagram of each test program.

(2a) Random error test sub-flow: obtaining star sensor theoretical measurement attitude matrix C_SIStar sensor optical axis noise index sigma_xStar sensor horizontal axis noise index sigma_zAnd calculating to obtain a star sensor measurement attitude matrix tempC under random error as

In the formula, randn (σ) represents the generation of gaussian noise with mean 0 and mean square error σ, and σ is represented by σ_xOr σ_z；

Carrying out orthogonal normalization processing on the star sensor measurement attitude matrix tempC under random errors to obtain the actual star sensor measurement attitude matrixWherein,is composed ofIn the case of the 1 st row of (c),is composed ofIn the case of the 2 nd row of (c),is composed ofRow 3 of (2), norm is the norm normalization function of vector L2, × is the vector cross multiplier

Turning to the step (3);

(2b) a system error testing sub-process: obtaining equivalent installation deviations delta x, delta y and delta z of the star sensor under a body coordinate system of an installed satellite, and calculating to obtain an installation deviation matrix delta C of the star sensor_SBIs composed of

ΔC_SB＝R_y(Δy)·R_x(Δx)·R_z(Δz)

Wherein R is_x(upsilon) is a direction cosine array R when the angle upsilon rotates around the X axis in the satellite body coordinate system_y(upsilon) is a direction cosine array R when the angle upsilon rotates around the Y axis in the satellite body coordinate system_z(upsilon) is a direction cosine array when a rotation angle upsilon is formed around the Z axis in a satellite body coordinate system;

obtaining a theoretical installation matrix C of the star sensor_SBSatellite inertial attitude matrix C_BIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

Turning to the step (3);

(2c) periodic error testing sub-process: receiving a rotating shaft selection instruction sent from the outside, selecting a rotating shaft, and acquiring the interference amplitude A of the star sensor_STStar sensor interference period T_STWhen the selected rotating shaft is the optical axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_XIs composed of

When the selected rotating shaft is the horizontal axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_ZIs composed of

Then measuring an error matrix delta C according to the star sensor_XOr Δ C_ZTheoretical measurement attitude matrix C_SICalculating the actual measurement attitude matrix of the current star sensorIs composed of

Or

Turning to the step (3);

wherein t is the star sensor measuring time; the rotating shaft selection instruction comprises an optical axis or a transverse axis; the rotating shaft comprises an optical axis and a transverse axis;

(2d) the sub-process of the years difference optical performance test: when the current epoch is obtained, then calculating the age compensation matrix C of the star sensor_PRIs composed of

Therein, ζ_A、θ_AThe psi and delta are the time difference description parameters and are only related to epoch, and specific calculation can be found in literature (Liulin. spacecraft orbit theory. national defense science and technology university Press);

obtaining a theoretical measurement attitude matrix C of the star sensor_SIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

Turning to the step (3);

(3) calculating the actual measurement attitude matrix of the current star sensorCorresponding quaternionExciting EGSE, and obtaining quaternion q output by star sensor_SICalculating to obtain error quaternionIs composed of

In the formula,is q_SIThe number of the conjugate quaternion of (c),is a quaternion multiplier;

then based on the error quaternionCalculating to obtain the three-axis equivalent attitude error e of the current star sensor_x、e_y、e_zIs composed of

In the formula,representing error quaternionI-1, 2, 3.

The invention has not been described in detail and is within the skill of the art.

Claims (1)

1. A multi-probe star sensor fusion attitude testing method is characterized by comprising the following steps:

(1) when a random error test is carried out, the step (2) is carried out, when a system error test is carried out, the step (4) is carried out, when a periodic error test is carried out, the step (6) is carried out, and when a time difference optical performance test is carried out, the step (7) is carried out;

(2) obtaining star sensor theoretical measurement attitude matrix C_SIStar sensor optical axis noise index sigma_xStar sensor horizontal axis noise index sigma_zCalculating to obtain random errorsThe star sensor measurement attitude matrix tempC under difference is

$tempC=C_{SI}+\left[\begin{array}{ccc}randn\left({\mathrm{\σ}}_x\right)& randn\left({\mathrm{\σ}}_x\right)& randn\left({\mathrm{\σ}}_x\right)\\ 0& 0& 0\\ randn\left({\mathrm{\σ}}_z\right)& randn\left({\mathrm{\σ}}_z\right)& randn\left({\mathrm{\σ}}_z\right)\end{array}\right]$

In the formula, randn (σ) represents the generation of gaussian noise with mean 0 and mean square error σ, and σ is represented by σ_xOr σ_z；

(3) Carrying out orthogonal normalization processing on the star sensor measurement attitude matrix tempC under random errors to obtain the actual star sensor measurement attitude matrixWherein,is composed ofIn the case of the 1 st row of (c),is composed ofIn the case of the 2 nd row of (c),is composed ofRow 3 of (2), norm is the norm normalization function of vector L2, × is the vector cross multiplier

$\left\{\begin{array}{c}{\stackrel{\‾}{C}}_{SI}3=norm\left(tempC3\right)\\ {\stackrel{\‾}{C}}_{SI}2=norm(tempC3\×tempC1)\\ {\stackrel{\‾}{C}}_{SI}1=norm({\stackrel{\‾}{C}}_{SI}2\×{\stackrel{\‾}{C}}_{SI}3)\end{array}\right.;$

Turning to the step (9);

(4) obtaining equivalent installation deviations delta x, delta y and delta z of the star sensor under a body coordinate system of an installed satellite, and calculating to obtain an installation deviation matrix delta C of the star sensor_SBIs composed of

ΔC_SB＝R_y(Δy)·R_x(Δx)·R_z(Δz)

Wherein R is_x(upsilon) is a direction cosine array R when the angle upsilon rotates around the X axis in the satellite body coordinate system_y(upsilon) is a direction cosine array R when the angle upsilon rotates around the Y axis in the satellite body coordinate system_z(upsilon) is a direction cosine array when a rotation angle upsilon is formed around the Z axis in a satellite body coordinate system;

(5) obtaining a theoretical installation matrix C of the star sensor_SBSatellite inertial attitude matrix C_BIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

${\stackrel{\‾}{C}}_{SI}={\mathrm{\ΔC}}_{SB}\·C_{SB}\·C_{BI};$

Turning to the step (9);

(6) receiving a rotating shaft selection instruction sent from the outside, selecting a rotating shaft, and acquiring the interference amplitude A of the star sensor_STStar sensor interference period T_STWhen the selected rotating shaft is the optical axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_XIs composed of

${\mathrm{\ΔC}}_X=R_x\left(A_{ST}sin\frac{2\mathrm{\π}t}{T_{ST}}\right);$

When the selected rotating shaft is the horizontal axis, the star sensor measurement error matrix delta C under the influence of interference is obtained through calculation_ZIs composed of

${\mathrm{\ΔC}}_Z=R_z\left(A_{ST}sin\frac{2\mathrm{\π}t}{T_{ST}}\right);$

Then measuring an error matrix delta C according to the star sensor_XOr Δ C_ZTheoretical measurement attitude matrix C_SICalculating the actual measurement attitude matrix of the current star sensorIs composed of

Or

Turning to the step (9);

wherein t is the star sensor measuring time; the rotating shaft selection instruction comprises an optical axis or a transverse axis; the rotating shaft comprises an optical axis and a transverse axis;

(7) when the current epoch is obtained, then calculating the age compensation matrix C of the star sensor_PRIs composed of

$C_{PR}=\left[\begin{array}{ccc}1& 0& -\mathrm{\ψ}sin\mathrm{\ϵ}-{\mathrm{\θ}}_A\\ 0& 1& {\mathrm{\θ}}_A{\mathrm{\ζ}}_A-\mathrm{\Δ}\mathrm{\ϵ}\\ \mathrm{\ψ}sin\mathrm{\ϵ}+{\mathrm{\θ}}_A& -{\mathrm{\θ}}_A{\mathrm{\ζ}}_A+\mathrm{\Δ}\mathrm{\ϵ}& 1\end{array}\right]$

Therein, ζ_A、θ_APhi, phi and delta are the time difference description parameters;

(8) obtaining a theoretical measurement attitude matrix C of the star sensor_SIAnd calculating to obtain the actual measurement attitude matrix of the current star sensorIs composed of

${\stackrel{\‾}{C}}_{SI}=C_{SI}\·C_{PR}$

Turning to the step (9);

(9) calculating the actual measurement attitude matrix of the current star sensorCorresponding quaternionThen, acquiring a quaternion q output by the actual measurement of the star sensor_SICalculating to obtain error quaternionIs composed of

$\tilde{q}=q_{SI}^{-1}\⊗{\stackrel{\‾}{q}}_{SI}$

In the formula,is q_SIThe number of the conjugate quaternion of (c),is a quaternion multiplier;

then based on the error quaternionCalculating to obtain the three-axis equivalent attitude error e of the current star sensor_x、e_y、e_zIs composed of

$e_x=2\tilde{q}\left(1\right)$

$e_y=2\tilde{q}\left(2\right)$

$e_z=2\tilde{q}\left(3\right)$

In the formula,representing error quaternionI-1, 2, 3.