CN105300406B - A kind of Gyro fault diagnosis method based on equilibrium equation equivalence - Google Patents

Abstract

The invention discloses a gyroscope fault diagnosis method based on the equivalence of balance equations, and discloses a gyroscope fault diagnosis method suitable for spacecraft. First, 5 gyroscopes with redundant measurement are combined with four gyroscopes to form 5 gyroscopes. group, and arbitrarily select a group of balance equations in each gyro group to calculate its balance equation coefficients and balance equation errors; then, for each group of gyroscopes, calculate parameters related to the configuration according to its balance equation coefficients, and compare the parameters with the design The product of the predetermined fault threshold is used as the fault judgment basis value; finally, the balance equation error of each group of gyroscopes is compared with the calculated fault judgment basis value, and the faulty gyroscope is located according to the comparison results of all gyro groups. The present invention is based on the equivalence conclusion between the balance equations of the gyroscope group, realizes the accurate diagnosis effect of the gyroscope fault on the basis of the minimum calculation amount of the algorithm, and provides a clear selection principle for the fault judgment threshold, and has strong engineering operability and realizability.

Description

A Gyroscope Fault Diagnosis Method Based on Balance Equation Equivalence

technical field

The invention relates to the field of spacecraft attitude control, in particular to a gyroscope fault diagnosis method based on the equivalence of balance equations, which is used for fault diagnosis of gyroscopes configured on spacecraft.

Background technique

In the satellite control system, the measurement of the attitude measurement sensor gyroscope is the key component to obtain the angular velocity of the star to realize the effective control of the star attitude, and it is the key factor that affects the stable operation of the spacecraft and restricts the life of the spacecraft on orbit. Although the spacecraft system can realize the three-axis angular velocity measurement of the star with only three gyroscopes with non-coplanar measurement axes, in order to ensure the long life and high reliability and stable operation of the spacecraft, the control system is generally equipped with more than three redundant gyroscopes .

Spacecraft gyroscope fault diagnosis is generally based on the error of the balance equation formed by the four gyroscope measurements in the redundant gyroscope to judge the gyroscope fault, that is, the star angular velocity calculated from the three gyroscope measurement values is projected in the direction of other gyroscope measurement axes and The measured deviation is used as the basis for fault judgment. In order to locate the faulty gyroscope, the number of gyroscopes participating in the gyroscope fault diagnosis of the balance equation should not be less than 5. Any combination of four gyroscopes in the 5 gyroscopes may be recorded as {i, j, k, l}, and the measured output Δg _i , Δg _j , Δg _k , Δg _l and their installation unit vectors VG _i , VG _j , VG _k , VG _l establish the following four different forms of balance equations:

ε _ijkl ＝|k _ijkl,i Δg _i +ki _jkl,j Δg _j +k _ijkl,k Δg _k -Δg _l | (1a)

ε _ijlk ＝|k _ijlk,i Δg _i +k _ijlk,j Δg _j +k _ijlk,l Δg _l -Δg _k | (1b)

ε _ilkj ＝|k _ilkj,i Δg _i +k _ilkj,l Δg _l +k _ilkj,k Δg _k -Δg _j | (1c)

ε _jkli ＝|k _jkl,l Δg _l +k _jkli,j Δg _j +k _jkli,k Δg _k -Δg _i | (1d)

Among them, ε _ijkl , ε _ijlk , ε _ilkj , and ε _jkli are the errors of the corresponding balance equations, and the three coefficients in the four balance equations are the first three elements corresponding to the four vectors as follows, namely

The current gyroscope fault method based on the balance equation generally judges the five balance errors obtained by five gyroscopes according to the specific gyroscope sequence and the fixed fault judgment threshold selected in advance. However, since the error values of ε _ijkl , ε _ijlk , ε _ilkj , and ε _jkli obtained by any combination {i, j, k, l} under the same gyroscope measurement value may be different, even for the same combination, it may be due to the balance equation Different selections lead to different judgment results, resulting in inaccurate judgment of gyro failures. In addition, the fixed thresholds set in advance may not be fully adaptable to other different gyro combinations. In order to solve the above method problems, another commonly used method is to use all the balance equation errors of the gyroscope for gyroscope fault judgment, and design a fault judgment threshold for each balance equation. Since there are 4 gyro groups for 5 gyroscopes and 4 balance equations for each gyro group, it is necessary to calculate the error values of 20 balance equations and design 20 fault judgment thresholds. Although this method can improve the accuracy of gyroscope fault diagnosis, it is obtained at the expense of gyroscope fault diagnosis calculation and system design.

Based on the theoretical analysis, the equivalence conclusions of the four balance equations in formulas (1a) to (1d) for the same gyro combination are obtained, that is, for any gyro combination, the coefficient of the gyro measurement value in any balance equation is There is a fixed proportional relationship between the ratio and the ratio of coefficients arranged in the same gyroscope order in other balance equations in the combination. The present invention proposes a gyroscope fault diagnosis method based on the equivalence of balance equations, and realizes only five balance equations The calculation amount of the fixed threshold fault method and the diagnosis effect of the fault diagnosis method based on all balance equations are adopted, and a reasonable fault threshold selection method is given.

Contents of the invention

The technical problem to be solved by the present invention is: to provide a gyroscope fault diagnosis method based on the equivalence of balance equations, using five gyroscopes to form a balance equation for accurate diagnosis of gyroscope faults, and to overcome the problem of balance equations in the prior art judgment method Inaccurate judgments lead to misdiagnosis or missed diagnosis, or other existing gyroscope fault diagnosis methods have too much calculation and too complicated design.

Technical scheme of the present invention is:

A gyroscope fault diagnosis method based on equivalence of balance equations, comprising,

S1, number the five tops as i, j, k, l, m, and divide them into five groups in the form of four tops: {i, j, k, l}, {i, j, k , m}, {i, j, l, m}, {i, k, l, m}, {j, k, l, m} carry out the fault detection of the balance equation respectively;

S2, for the gyro group {i, j, k, l}, the installation unit vectors VG _i , VG _j , VG _k , VG _l of the gyro i, j, k, l measuring axes on the star body and their gyro corresponding measurement output Δg _i , Δg _j , Δg _k , Δg _l calculate the balance equation coefficient vector k _ijkl and the balance equation error ε _ijkl :

Calculate k _ijkl and ε _ijkl according to the above gyro group {i, j, k, l}, for the other four gyro groups {i, j, k, m}, {i, j, l, m}, {i , k, l, m}, {j, k, l, m} to obtain their balance equation coefficient vectors k _ijkm , k _ijlm , k _iklm , k _jklm and their balance equation errors ε _ijkm , ε _ijlm , ε _iklm , ε _jklm .

S3, for the gyro group {i, j, k, l}, according to the sum variable _{kijkl_sum of the absolute value of each element of the vector k ijkl and} the set gyro bonus point judgment threshold δ _gelim1 and deduction point judgment threshold δ _gelim2 to combine the gyro Each gyroscope in the center carries out the addition and subtraction processing of the same score:

If ε _ijkl is less than k _{ijkl_sum} × δ _gelim1 , add points to the scores of gyroscopes i, j, k, and l;

If ε _ijkl is greater than k _{ijkl_sum} × δ _gelim2 , the scores of gyroscopes i, j, k, and l are all subtracted. According to the same process of gyro group {i, j, k, l}, for the other four gyro groups {i, j, k, m}, {i, j, l, m}, {i, k, l, m }, {j, k, l, m}, respectively calculate the sum of the absolute values of the coefficient vector elements of each balance equation k _{ijkm_sum} , k _{ijlm_sum} , k _{iklm_sum} , k _{jklm_sum} , and combine the set thresholds δ _gelim1 , δ _gelim2 to The gyroscopes in each gyroscope group are added and subtracted.

S4. Faulty gyro determination: after the above S2 and S3 processes are completed and there has been a situation where the score of the gyro has been reduced, the gyro i, j, k, l, m with the most reduced score is judged as the faulty gyro.

Further, it includes setting thresholds δ _gelim1 and δ _gelim2 according to the configuration determined by the gyro installation unit vector, and satisfying δ _gelim2 >δ _gelim1 ;

In order to avoid misdiagnosis as a fault when the gyro measurement is normal, the threshold value δ _gelim1 is selected to be greater than the normal measurement error index range b _norm given by the gyro;

In order to avoid misdiagnosis caused by multiple gyroscopes with the same score drop when a gyroscope fails, the selection of δ _gelim2 generally satisfies δ _gelim2 >2δ _gelim1 .

The advantage of the present invention compared with prior art is:

(1) The present invention proposes based on the equivalence principle of balance equations, avoids the possibility that the faulty gyroscope diagnosis results are different due to the selection of different balance equations during diagnosis in the prior art, and improves the diagnosis results on the basis of keeping the minimum diagnosis algorithm calculation amount accuracy and consistency of diagnostic results.

(2) The present invention provides a clear fault diagnosis threshold selection principle, which can effectively avoid misdiagnosis caused by unreasonable selection of diagnostic thresholds in the prior art, and greatly simplifies the complexity of system parameter design.

Description of drawings

Fig. 1 is the flow chart of gyroscope fault diagnosis based on balance equation equivalence of the present invention;

Detailed ways

As shown in Figure 1, a kind of gyroscope fault diagnosis method based on balance equation equivalence of the present invention, the steps are as follows:

(1) Use five tops i, j, k, l, m to form five groups: {i, j, k, l}, {i, j, k, m}, {i, j, l, m}, {i, k, l, m}, {j, k, l, m} balance equations for fault detection.

(2) Calculate the equilibrium equation coefficient vectors k _ijkl , k _ijkm , k _ijlm from the installation unit vectors VG _i , VG _j , VG _k , VG _l , VG _k of the gyroscopes i, j, k, l, and k measuring axes on the star , k _iklm , k _jklm :

in

Combined with the calculation of the balance equation coefficient vector above, calculate the balance equation errors ε _ijkl , ε _ijkm , ε _ijlm , ε _iklm , ε _jklm from the gyroscope’s corresponding measurement outputs Δg _i , Δg _j , Δg _k , Δg _l , Δg _m :

ε _ijkl ＝k _ijkl ×[Δg _i Δg _j Δg _k Δg _l ] ^T

ε _ijkm ＝k _ijkm ×[Δg _i Δg _j Δg _k Δg _m ] ^T

ε _ijlm ＝k _ijlm ×[Δg _i Δg _j Δg _l Δg _m ] ^T

ε _iklm =k _iklm ×[Δg _i Δg _l Δg _l Δg _m ] ^T

ε _jklm ＝k _jklm ×[Δg _j Δg _k Δg _l Δg _m ] ^T

Wherein, the superscript "-1" on the right of each of the above variables indicates matrix inverse operation, and the superscript "T" on the right indicates transposition operation of vector (matrix).

(3) Take the absolute values of the coefficients of the five gyro balance equations and sum them up to obtain variables k _{ijkl_sum} , k _{ijkm_sum} , k _{ijlm_sum} , k _{iklm_sum} , k _{jklm_sum} , namely

k _{ijkl_sum} =|k _ijkl,i |+|k _ijkl,j |+|k _ijkl,k |+|k _ijkl,l |

k _{ijkm_sum} =|k _ijkm,i |+|k _ijkm,j |+|k _ijkm,k |+|k _ijkm,m |

k _{ijlm_sum} =|k _ijlm,i |+|k _ijlm,j |+|k _ijlm,l |+|k _ijlm,m |

k _{iklm_sum} ＝|k _iklm,i |+|k _iklm,k |+|k _iklm,l |+|k _iklm,m |

k _{jklm_sum} =|k _jklm,i |+|k _jklm,k |+|k _jklm,l |+|k _jklm,m |

And the product obtained by multiplying k _{ijkl_sum} , k _{ijkm_sum} , k _{ijlm_sum} , k _{iklm_sum} , k _{jklm_sum} obtained from the above calculation with the gyro plus point judgment threshold δ _gelim1 and the deduction judgment threshold δ _gelim2 is used as the gyro score addition and subtraction of each group Basis, take the gyro group {i, j, k, l} as an example:

If ε _ijkl is less than k _{ijkl_sum} × δ _gelim1 , add 1 point to the scoring of top i, j, k, l;

If ε _ijkl is greater than k _{ijkl_sum} × δ _gelim2 , the scores for gyro i, j, k and l will be reduced by 1 point.

Do the same addition to the other four gyro groups {i, j, k, m}, {i, j, l, m}, {i, k, l, m}, {j, k, l, m} , Point reduction processing.

(4) Judgment of faulty gyroscopes: After completing the processing of adding and subtracting points for each gyroscope group, if there is a gyroscope score for deduction, the total score of gyroscopes i, j, k, l, m will be reduced to The one with the most is judged as a faulty gyro.

In the above steps, the selection of the threshold needs to meet the following conditions:

Select the threshold δ _gelim2 > δ _gelim1 ;

Select the threshold δ _gelim1 greater than the normal measurement error index range b _norm of the gyroscope;

The threshold δ _gelim2 needs to be selected according to the configuration of the gyroscope combination, and generally δ _gelim2 >2δ _gelim1 is desirable.

Embodiment 1: five gyroscopes i, j, k, l, m are working, and gyroscope fault diagnosis is performed based on equivalent balance equations.

(1) Use five gyroscopes i, j, k, l, m to form five groups {i, j, k, l}, {i, j, k, m}, {i, j, l, m}, { i, k, l, m}, {j, k, l, m} balance equations for fault detection.

The installation unit vector of 5 gyroscopes in the embodiment is:

VG _i = [-0.52483389, -0.62547267, 0.57735027] ^T ;

_VGj = [0.80409216, -0.14178314, 0.57735027] ^T ;

VG _k = [-0.27925828, 0.76725581, 0.57735027] ^T ;

VG _l = [0.70710678, 0.40824829, 0.57735027] ^T ;

VG _m = [-0.70710678, 0.40824829, 0.57735027] ^T ;

Set gyro i as fault, the measurement output of each gyro is:

Δg _i =-0.008384377516774;

Δg _j =0.014048585562026;

Δg _k = -0.004859446568903;

Δg _l =0.012355871354728;

Δg _m = -0.012326811593622;

According to the normal measurement error index b _norm of the gyroscope is less than 0.00002, and the threshold selection method selects the bonus threshold δ _gelim1 and the deduction threshold δ _gelim2 as

δ _gelim1 = 0.00002, δ _gelim2 = 0.00004;

(2) Taking the gyro group {i, j, k, l} as an example, the algorithm for calculating ε _ijkl is as follows:

k _{ijkl_sum} = 2.5862568255;

ε _ijkl = k _ijkl × [Δg _i Δg _j Δg _k Δg _l ] ^T = 2.23121e-004;

k _{ijkl_sum} × δ _gelim2 = 0.00008776;

Since ε _ijkl is greater than k _{ijkl_sum} × δ _gelim2 , the scores of gyro i, j, k, and l will be reduced by 1 point respectively.

The gyro groups {i, j, k, m}, {i, j, l, m}, {i, k, l, m}, {j, k, l, m} are all compared with the gyro groups {i, j, k, l} are treated similarly, and the scores obtained for each gyro group are shown in Table 1.

(3) Accumulate the scores of each gyro in different gyro groups. From the results, it can be seen that the score of -4 for gyro i is the lowest, so it can be determined that gyro i is faulty.

Table 1 The gyroscope scores judged by each group of balance equations

Gyro group GC _i GC _{j GC} GC _l GC _m {i,j,k,l} -1 -1 -1 -1 0 {i,j,k,m} -1 -1 -1 0 -1 {i,j,l,m} -1 -1 0 -1 -1 {i, k, l, m} -1 0 -1 -1 -1 {j,k,l,m} 0 +1 +1 +1 +1 Score accumulation -4 -2 -2 -2 -2

The content that is not described in detail in the description of the present invention belongs to the well-known technology of those skilled in the art.

Claims (4)

1. A gyroscope fault diagnosis method based on balance equation equivalence, it is characterized in that, comprising: S1, number the five tops as i, j, k, l, m, and divide them into five groups in the form of four tops: {i, j, k, l}, {i, j, k, m}, {i, j, l, m}, {i, k, l, m}, {j, k, l, m} carry out the fault detection of the balance equation respectively; S2, for the gyro group {i, j, k, l}, the installation unit vectors VG _i , VG _j , VG _k , VG _l of the gyro i, j, k, l measuring axes on the star body and their gyro corresponding measurement output Δg _i , Δg _j , Δg _k , Δg _l calculate the balance equation coefficient vector k _ijkl and the balance equation error ε _ijkl : <mrow><msub><mi>A</mi><mrow><mi>i</mi><mi>j</mi><mi>k</mi></mrow></msub><mo>=</mo><msup><mfenced open = "[" close = "]"><mtable><mtr><mtd><mrow><msubsup><mi>VG</mi><mi>i</mi><mi>T</mi></msubsup></mrow></mtd></mtr><mtr><mtd><mrow><msubsup><mi>VG</mi><mi>j</mi><mi>T</mi></msubsup></mrow></mtd></mtr><mtr><mtd><mrow><msubsup><mi>VG</mi><mi>k</mi><mi>T</mi></msubsup></mrow></mtd></mtr></mtable></mfenced><mrow><mo>-</mo><mn>1</mn></mrow></msup></mrow> <mrow><msub><mi>k</mi><mrow><mi>i</mi><mi>j</mi><mi>k</mi><mi>l</mi></mrow></msub><mo>=</mo><mfenced open = "[" close = "]"><mtable><mtr><mtd><mrow><msubsup><mi>VG</mi><mi>l</mi><mi>T</mi></msubsup><mo>&amp;CenterDot;</mo><msub><mi>A</mi><mrow><mi>i</mi><mi>j</mi><mi>k</mi></mrow></msub></mrow></mtd><mtd><mrow><mo>-</mo><mn>1</mn></mrow></mtd></mtr></mtable></mfenced></mrow> <mrow><msub><mi>&amp;epsiv;</mi><mrow><mi>i</mi><mi>j</mi><mi>k</mi><mi>l</mi>mi></mrow></msub><mo>=</mo><msub><mi>k</mi><mrow><mi>i</mi><mi>j</mi><mi>k</mi><mi>l</mi></mrow></msub><mo>&amp;times;</mo><mfenced open = "[" close = "]"><mtable><mtr><mtd><mrow><msub><mi>&amp;Delta;g</mi><mi>i</mi></msub></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>&amp;Delta;g</mi><mi>j</mi></msub></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>&amp;Delta;g</mi><mi>k</mi></msub></mrow></mtd></mtr><mtr><mtd><mrow><msub><mi>&amp;Delta;g</mi><mi>l</mi></msub></mrow></mtd></mtr></mtable></mfenced></mrow> According to the process of calculating k _ijkl and ε _ijkl for the above gyro group {i, j, k, l}, for the other four gyro groups {i, j, k, m}, {i, j, l, m}, {i , k, l, m}, {j, k, l, m} to obtain their balance equation coefficient vectors k _ijkm , k _ijlm , k _iklm , k _jklm and their balance equation errors ε _ijkm , ε _ijlm , ε _iklm , ε _jklm ; S3, for the gyro group {i, j, k, l}, according to the sum variable _{kijkl_sum of the absolute value of each element of the vector k ijkl and} the set gyro bonus point judgment threshold δ _gelim1 and deduction point judgment threshold δ _gelim2 to combine the gyro Each gyroscope in the center carries out the addition and subtraction processing of the same score: If ε _ijkl is less than k _{ijkl_sum} × δ _gelim1 , add points to the scores of gyroscopes i, j, k, and l; If ε _ijkl is greater than k _{ijkl_sum} × δ _gelim2 , the scores of gyroscopes i, j, k, and l are all subtracted; According to the same process of gyro group {i, j, k, l}, for the other four gyro groups {i, j, k, m}, {i, j, l, m}, {i, k, l, m }, {j, k, l, m}, respectively calculate the sum of the absolute values of the coefficient vector elements of each balance equation k _{ijkm_sum} , k _{ijlm_sum} , k _{iklm_sum} , k _{jklm_sum} , and combine the set thresholds δ _gelim1 , δ _gelim2 to The gyroscopes in each gyroscope group are added and subtracted; S4, faulty gyro judgment: After the processing of S2 and S3 above is completed and there has been a case where the score of the gyro is reduced, the gyro i, j, k, l, m with the most reduced score is determined to be a faulty gyro. 2 . The method according to claim 1 , wherein the thresholds δ _gelim1 and δ _gelim2 are set according to the configuration determined by the gyro installation unit vector, and satisfy δ _gelim2 >δ _gelim1 . 3 . The method according to claim 2 , wherein the selected threshold value δ _gelim1 is larger than the normal measurement error index range b _norm given by the gyro. 4 . 4. The method according to claim 2, characterized in that: Select δ _gelim2 to satisfy δ _gelim2 >2δ _gelim1 .