Redundant strapdown inertial measurement unit fault diagnosis and isolation method and system

Abstract

The invention relates to the technical field of fault detection, and discloses a redundant strapdown inertial measurement unit fault diagnosis and isolation method and system, so as to improve the overall performance of fault processing. The method comprises the following steps: determining target objects of fault diagnosis, wherein the target objects are grouped gyroscopes or add tables in the same redundant strapdown inertial measurement unit; the method comprises the following steps: calculating a variation trend and a constant of any gyroscope or accelerometer in a target object based on previous continuous frames based on a least square method, then calculating a predicted value of a current frame according to the variation trend and the constant, then calculating whether a difference value between the predicted value of the current frame and a measured value is greater than a set prediction threshold value, and if so, judging whether the difference value is greater than the set prediction threshold value. And judging that the corresponding gyroscope or accelerometer has a fault, and performing isolation processing on the gyroscope or accelerometer having the fault.

Description

Technical Field

[0001] The present invention relates to the technical field of fault detection, and in particular to a redundant strapdown inertial group fault diagnosis and isolation method and system. Background Art

[0002] The inertial group is one of the core units of current high-value spacecraft such as launch vehicles, manned spaceflight, and ballistic missiles. It provides full-parameter navigation information for space vehicles, and its reliability is an important factor in determining whether the space mission can achieve its purpose.

[0003] In order to improve the reliability of the inertial group, a multi-table redundant strapdown inertial group is proposed, which can continue to provide effective navigation information for the spacecraft after some components of the inertial group fail. Fault diagnosis technology is one of the core technologies of the redundant strapdown inertial group system. Through fault diagnosis, the fault and the source of the fault can be effectively identified, and then the source of the fault can be processed and isolated, thereby realizing a complete set of fault monitoring and processing mechanisms. Therefore, a fault diagnosis and isolation technology is designed for the redundant strapdown inertial group, which can promote the rapid development and application of the core technology of the redundant strapdown inertial group, and has high application value in the aerospace and military fields. Summary of the invention

[0004] The present invention aims to disclose a redundant strapdown inertial unit fault diagnosis and isolation method and system to improve the overall performance of fault handling.

[0005] To achieve the above-mentioned purpose, the redundant strapdown inertial unit fault diagnosis and isolation method disclosed in the present invention comprises:

[0006] Step S1, determining a target object for fault diagnosis, wherein the target object is a group of gyroscopes or an additional meter in the same redundant strapdown inertial unit;

[0007] Step S2, calculate any gyroscope or add-on meter in the target object based on the change trend and constant between previous consecutive frames based on the least squares method, then calculate the predicted value of the current frame based on the change trend and the constant, and then calculate whether the difference between the predicted value and the measured value of the current frame is greater than the set prediction threshold. If so, determine that the corresponding gyroscope or add-on meter is faulty, and perform isolation processing on the faulty gyroscope or add-on meter.

[0008] Preferably, the least squares processing for a single gyroscope includes:

[0009] Step S21: Use the five frames of data measured by the gyroscope at the first five moments Construct the output model of the gyroscope:

[0010]

[0011] Among them, i represents the order of the gyroscope in the group, T represents the measurement time, t represents the frame sequence, and dgi , D i are the changing trend and constant of the gyroscope respectively; and:

[0012]

[0013] Step S22: Predict the gyro output data of the current frame The calculated value of is:

[0014]

[0015] Step S23: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. g If yes, the corresponding gyro or meter is judged to be faulty, and the calculation formula is:

[0016] Preferably, the least squares processing for a single addition table includes:

[0017] Step S26: Use the five frames of data measured at the first five moments of the addition table Construct the output model for adding tables:

[0018]

[0019] Among them, j represents the order of the addition table in the group, T represents the measurement time, t represents the frame sequence, and da j , K j They are the changing trends and constants of the added table respectively; and:

[0020]

[0021] Step S27: Predict the output data of the table for the current frame The calculated value of is:

[0022]

[0023] Step S28: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. a If yes, the corresponding addition meter or addition meter is judged to be faulty, and the calculation formula is:

[0024] Preferably: calculate based on the null space solution whether the overall residual of all gyroscopes or adding tables participating in the calculation in the target object group is less than the set redundant detection threshold; if not, calculate based on the null space solution whether the local residuals of the remaining members participating in the calculation after removing one member are less than the set redundant detection threshold, and then combine the local residual results of each remaining combination to obtain the member that has caused the fault; wherein, the number of gyroscopes or adding tables based on redundant settings in the group is at least 4.

[0025] Preferably, the redundant strapdown inertial group is a ten-meter redundant strapdown inertial group consisting of five add-on meters and five gyroscopes, and the measurement model of the ten-meter redundant strapdown inertial group gyroscope and the add-on meter is:

[0026] m g =H g ω;

[0027] m a =H a g;

[0028] Among them, m g 、m a are the measurement information of the redundant inertial group gyro and the added meter, H g , H a is the spatial configuration matrix of the redundant inertial group gyro and the added table of the ten meters, ω is the orthogonal three-axis oxyz angular velocity information, and g is the orthogonal three-axis oxyz acceleration information;

[0029] Assume Nu g 、Nu a H g '、H a 'The null space solution satisfies:

[0030] H g 'Nu g =0;

[0031] H a 'Nu a =0;

[0032] get:

[0033] Nu g 'H g ω=Nu g 'm g =0;

[0034] Nu a 'H a g=Nu a 'm a =0;

[0035] Among them, H g 'For H g The transpose of H a 'For H a The transpose of

[0036] The calculated results of the measured values ​​of the fault detection combination are the residual error e of the gyro or meter fault detection combination. g and e a , the calculation formulas are:

[0037] eg =Nu g 'm g ;

[0038] e a =Nu a 'm a .

[0039] Preferably, the five gyroscopes and the five adding meters are installed on the platform in an inverted cone distribution, and the adding meters and the gyroscopes are arranged in pairs so that the spatial arrangement matrix is ​​consistent.

[0040] Preferably, the present invention also determines whether there is a hardware fault by combining a single measurement value of the target object, whether the change between two consecutive data exceeds the measurement range, and whether the consecutive frames are fixed values.

[0041] Preferably, the present invention also uses a median filtering algorithm to calculate whether the median of the current frame and the corresponding number of previous continuous frames exceeds a range or a set threshold to determine whether a fault exists.

[0042] Preferably, the present invention introduces a weighted least squares algorithm to calculate the orthogonal three-axis navigation data, and the calculation formulas are:

[0043]

[0044] Among them, W g , W a is a 5×5 weighted matrix. When there is no fault, W g , W a It is a unit array. When the ith gyro or the jth adding table fails, the row corresponding to the ith gyro or the jth adding table in the unit array is modified to 0, and the data of other rows and columns remain unchanged.

[0045] To achieve the above object, the present invention also discloses a redundant strapdown inertial group fault diagnosis and isolation system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above method when executing the computer program.

[0046] In summary, the present invention has the following beneficial effects:

[0047] The least squares method is innovatively applied to the fault diagnosis and isolation of redundant strapdown inertial system, which has good reliability in practical applications. It can also be combined with other diagnostic technologies to produce complementary effects. Taking the redundant strapdown inertial system of ten meters as an example, based on five acceleration and five angular velocity information, the five fault diagnoses of hardware abnormality judgment, basic information detection, median filter detection, least squares algorithm prediction comparison, and redundancy detection can be performed on the measurement information. Once any fault detection identifies the abnormality of the device of the redundant strapdown inertial system of ten meters, it is considered that the device has a fault. Finally, the weighted least squares algorithm is used to isolate the fault, and finally a set of multi-level fault identification, detection and isolation processing of the redundant strapdown inertial system of ten meters is completed. Therefore, it can handle and deal with most faults. Based on the multiple detection technologies adopted, the effective identification capability of the inertial system fault can be guaranteed.

[0048] The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0050] Figure 1 The present invention is a schematic diagram of the deployment structure of the additional meter and the gyroscope in the ten-meter redundant strapdown inertial unit disclosed in the embodiment of the present invention.

[0051] Figure 2 The present invention discloses a method for diagnosing and isolating a redundant strapdown inertial unit fault according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The embodiments of the present invention are described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.

[0053] Example 1

[0054] This embodiment discloses a fault diagnosis and isolation method for a ten-meter redundant strapdown inertial group. The ten-meter redundant strapdown inertial group is composed of five gyroscopes and five accelerometers, and the specific configuration structure is shown in Figure 1. Oxyz is the strapdown inertial group shell coordinate system, which satisfies the right-hand rule. The five gyroscopes and five accelerometers are installed on the platform in an inverted cone distribution. The gyroscopes and accelerometers with the same serial number theoretically coincide. The installation angles of the five axes can all be determined by two angles to determine their spatial positions, which are the angles α between the axis and the ox axis. mn (m=g,a;n=1,2,3,4,5) and the angle β between the projection on the oyz plane and the oz axis mn (m=g,a;n=1,2,3,4,5). Figure 1 As shown in1 With table J 1 coaxial, and α mn (m=g,a;n=1,2,3,4,5) and β mn (m=g,a;n=1,2,3,4,5) The specific value is determined by the inertial group structure design, and different angles can be designed according to different needs.

[0055] That is, the measurement model of the ten-meter redundant strapdown inertial unit gyro and the added meter:

[0056] m g =H g ω(1)

[0057] m a =H a g(2)

[0058] Among them, m g 、m a is the measurement information of the redundant inertial group gyro and the additional meter, H g , H a is the spatial configuration matrix of the redundant inertial group gyro and the additional table, that is, the measurement matrix, ω is the orthogonal three-axis oxyz angular velocity information, and g is the orthogonal three-axis oxyz acceleration information. It is specifically expressed as:

[0059]

[0060]

[0061] Reference Figure 2 The fault diagnosis and isolation method of redundant strapdown inertial unit group in this embodiment 10 comprises the following steps:

[0062] Step 1: First, test the hardware part of the ten-meter redundant strapdown inertial unit, mainly including:

[0063] (1) Sampling signal sending abnormality detection.

[0064] (2) Gyro plus meter output data reception abnormality detection.

[0065] (3) Gyro table data and frame number anomaly detection.

[0066] (4) Data processing chip anomaly detection.

[0067] Then collect the original measurement information of the ten redundant strapdown inertial units. The measurement information of the five gyroscopes is g 1 , g 2 , g 3 , g 4 , g 5 , the measurement information of the five additional meters is a 1 、a 2、a 3 、a 4 、a 5 , determine whether the measured values ​​of the gyro and the meter exceed the range:

[0068]

[0069] Where g = 9.80665 m / s 2 , is the acceleration due to gravity. Continue to add the change in two consecutive data to the gyro, Δg i (i=1...5), Δa j (j=1...5) to make a judgment:

[0070]

[0071] At the same time, the gyro plus table data is tested to see if it is a fixed value for multiple consecutive frames. Although the output is theoretically different when the inertial group is stationary, there are errors and noise in the measurement, so it is impossible to output the same data continuously. When the gyro plus table outputs a continuous fixed number, the navigation information is no longer available:

[0072]

[0073] That is, when the difference between two consecutive frames of the gyro meter is 0, N g With N a Start counting when N g With N a If the number of times exceeds the set fixed number, it is considered that the gyro or the meter is faulty. If the difference is not 0, N g With N a Cleared to 0.

[0074] Step 2: Use the median filter algorithm to process the data of the current three frames and compare them to get the median Subtract it from the measured value of the next frame. If the difference is greater than the threshold, it is considered that the device has a fault:

[0075]

[0076] Step 3: Use the gyroscope and the five frames of data measured recently by the meter and Construct the output model of the gyroscope and the addition table:

[0077]

[0078] Where T represents the measurement time. The short-term change trend dg of the gyro and the addition table can be calculated by the least squares algorithm through equations (13) and (14): i 、 j and the constant D i, K i :

[0079]

[0080] in

[0081]

[0082] The gyro table output data of the next frame can be predicted from the first five frames of data:

[0083] g' i =dg i +D i (i=1...5)(22)

[0084] a' j =da j +K j (j=1...5)(23)

[0085]

[0086] where Y g , Y a It is the gyro plus meter prediction threshold, which can be set according to the actual measurement results of different ten-meter redundant strapdown inertial group.

[0087] Step 4: Redundancy detection is the core of the ten-meter redundant inertial group fault detection. Its essence is to compare the outputs of multiple meters and adopt the multi-meter "voting" method to achieve the goal of fault judgment. When a fault occurs, it is impossible to tell "who is right and who is wrong" by comparing two gyroscopes or two gyroscopes alone. It is necessary to add one party as a true and false reference value to form a many-to-one fault identification mode.

[0088] Assume Nu g 、Nu a H g '、H a The null space solution of ' can be calculated by mathematical tools and satisfies:

[0089] H g 'Nu g =0(27)

[0090] H a 'Nu a =0(28)

[0091] Combining equation (1) and equation (2), we can get:

[0092] Nu g 'H g ω=Nu g 'm g =0(29)

[0093] Nu a 'H a g=Nu a 'm a =0(30)

[0094] H g 'For H g The transpose of H a 'For H a The transpose of . The fault detection algorithm can be constructed by equations (29) and (30), that is, by selecting at least 4 gyroscopes or 4 adding meters to form a fault detection combination, substituting the measured value of the fault detection combination into equations (29) and (30), a calculation result can be obtained. This calculation result is called the residual of the gyroscope and adding meter fault detection combination:

[0095] e g =Nu g 'm g (31)

[0096] e a =Nu a 'm a (32)

[0097] According to equations (29) and (30), under ideal measurement conditions, that is, m g 、m a is ideal, at this time e g 、e a The value of e is 0. g 、e a If the value is not 0, then one or more gyros and meters in the corresponding fault detection combination are faulty, and the residual of the fault detection combination without faulty gyros and meters must be 0. Finally, by comparing the residual values ​​of different combinations, the fault source can be specifically located. Redundant fault diagnosis is divided into two steps:

[0098] (1) Calculate all gyroscopes and table information:

[0099] e g12345 =Nu g ' 12345 m g12345 (33)

[0100] e a12345 =Nu a ' 12345 m a12345 (34)

[0101] where e g12345 、e a12345 is the residual of the five gyroscopes and five addition tables, Nu'g12345 、Nu' a12345 For the complete five-table H g With H a Transpose of the null space solution, m g12345 、m a12345 For the data of five gyroscopes and the addition table, if ||e g12345 ‖ and ‖e a12345 ‖ is less than the set redundant detection threshold, it is determined that no fault has occurred and no subsequent detection is required; if ||e g12345 ||with||e a12345 When || is greater than the redundant detection threshold, it is considered that a fault has occurred in the five gyros or the addition table, and the specific fault location calculation is continued, that is, the residuals of the four gyros or the addition table are calculated;

[0102] (2) Calculate the fault information of the combination of four gyroscopes and the addition table

[0103] Gyroscopes No. 2345, No. 1345, No. 1245, No. 1235, and No. 1234 are used as a group with the addition table for fault diagnosis. The residual calculations of these five combinations are as follows:

[0104]

[0105] Among them, e g2345 、e a2345 They are the residuals of the combination of gyro No. 2345 and the addition table, Nu' g2345 、Nu' a2345 They are the spatial relationship H formed by the 2345 gyroscope and the addition table. g2345 With H a2345 The null space solution of g2345 、m a2345 They are the data of gyro No. 2345 and addition meter respectively. Similarly, the variables in the other four formulas are the residuals, null space matrix and output data of the subsequent four gyro and addition meter combinations. When the residual is greater than the set threshold, it is determined that the inertial device in the combination is faulty. The threshold is determined by the specific characteristics of the gyro and addition meter and the measured value. If the residual is less than the set threshold, it is considered that the gyro or addition meter in the combination is not faulty.

[0106] For example, if the first gyro or adder fails, then only e2345 among the residuals calculated by equation (35) is less than the threshold, and the remaining residuals exceed the threshold. It can be identified that gyro 1 or adder 1 fails, that is, the fault identification and location are completed.

[0107] Step 5:

[0108] The ultimate need for navigation output is to obtain orthogonal three-axis navigation information, which is derived from the data of the cone five table by the least squares algorithm, that is:

[0109]

[0110] in They are the orthogonal three-axis navigation information deduced from the five gyroscopes and the five adding meters' measurements, and the above formula is the navigation output when there is no fault.

[0111] From step 1 to step 4, four kinds of detection can be obtained. These four kinds of detection results are independent of each other and do not affect each other's detection results. As long as any one of the detection results detects a fault, it is determined that the gyro or the add-on meter has a fault. These four detection results jointly determine the fault status of the gyro and the add-on meter. After obtaining these four detection results, for the redundant inertial group with faults, the weighted least squares algorithm is introduced to calculate the orthogonal three-axis navigation data:

[0112]

[0113] Where W g , W a is a 5×5 weighted matrix. When there is no fault, W g , W a is a unit array. When the ith gyro or the addition table fails, the row corresponding to the ith gyro or the jth addition table in the unit array is modified to 0, and the data of other rows and columns remain unchanged. Assuming that the 1st gyro fails, W g for:

[0114]

[0115] When two gyro or two gyro failures occur, similar sampling processing is sufficient, so that the navigation information output after isolating the failure can be obtained through weighted least squares.

[0116] Example 2

[0117] This embodiment discloses a redundant strapdown inertial group fault diagnosis and isolation system, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method corresponding to the above embodiment when executing the computer program. The method at least includes:

[0118] Step S1, determining a target object for fault diagnosis, wherein the target object is a group of gyroscopes or an additional meter in the same redundant strapdown inertial unit.

[0119] Step S2, calculate any gyroscope or add-on meter in the target object based on the change trend and constant between previous consecutive frames based on the least squares method, then calculate the predicted value of the current frame based on the change trend and the constant, and then calculate whether the difference between the predicted value and the measured value of the current frame is greater than the set prediction threshold. If so, determine that the corresponding gyroscope or add-on meter is faulty, and perform isolation processing on the faulty gyroscope or add-on meter.

[0120] Preferably, the least squares processing for a single gyroscope includes:

[0121] Step S21: Use the five frames of data measured by the gyroscope at the first five moments Construct the output model of the gyroscope:

[0122]

[0123] Among them, i represents the order of the gyroscope in the group, T represents the measurement time, t represents the frame sequence, and dg i , D i are the changing trend and constant of the gyroscope respectively; and:

[0124]

[0125] Step S22: Predict the gyro output data of the current frame The calculated value of is:

[0126]

[0127] Step S23: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. g If yes, the corresponding gyro or meter is judged to be faulty, and the calculation formula is:

[0128] Preferably, the least squares processing for a single addition table includes:

[0129] Step S26: Use the five frames of data measured at the first five moments of the addition table Construct the output model for adding tables:

[0130]

[0131] Among them, j represents the order of the addition table in the group, T represents the measurement time, t represents the frame sequence, and da j , K j They are the changing trends and constants of the added table respectively; and:

[0132]

[0133] Step S27: Predict the output data of the table for the current frame The calculated value of is:

[0134]

[0135] Step S28: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. a If yes, the corresponding addition meter or addition meter is judged to be faulty, and the calculation formula is:

[0136] Preferably: calculate based on the null space solution whether the overall residual of all gyroscopes or adding tables participating in the calculation in the target object group is less than the set redundant detection threshold; if not, calculate based on the null space solution whether the local residuals of the remaining members participating in the calculation after removing one member are less than the set redundant detection threshold, and then combine the local residual results of each remaining combination to obtain the member that has caused the fault; wherein, the number of gyroscopes or adding tables based on redundant settings in the group is at least 4.

[0137] Preferably, the redundant strapdown inertial group is a ten-meter redundant strapdown inertial group consisting of five add-on meters and five gyroscopes, and the measurement model of the ten-meter redundant strapdown inertial group gyroscope and the add-on meter is:

[0138] m g =H g ω;

[0139] m a =H a g;

[0140] Among them, m g 、m a are the measurement information of the redundant inertial group gyro and the added meter, H g , H a is the spatial configuration matrix of the redundant inertial group gyro and the added table of the ten meters, ω is the orthogonal three-axis oxyz angular velocity information, and g is the orthogonal three-axis oxyz acceleration information;

[0141] Assume Nu g 、Nu a H g '、H a 'The null space solution satisfies:

[0142] H g 'Nu g =0;

[0143] H a 'Nu a =0;

[0144] get:

[0145] Nu g 'H gω=Nu g 'm g =0;

[0146] Nu a 'H a g=Nu a 'm a =0;

[0147] Among them, H g 'For H g The transpose of H a 'For H a The transpose of

[0148] The calculated results of the measured values ​​of the fault detection combination are the residual error e of the gyro or meter fault detection combination. g and e a , the calculation formulas are:

[0149] e g =Nu g 'm g ;

[0150] e a =Nu a 'm a .

[0151] Preferably, the five gyroscopes and the five adding meters are installed on the platform in an inverted cone distribution, and the adding meters and the gyroscopes are arranged in pairs so that the spatial arrangement matrix is ​​consistent.

[0152] Preferably, the present invention also determines whether there is a hardware fault by combining a single measurement value of the target object, whether the change between two consecutive data exceeds the measurement range, and whether the consecutive frames are fixed values.

[0153] Preferably, the present invention also uses a median filtering algorithm to calculate whether the median of the current frame and the corresponding number of previous continuous frames exceeds a range or a set threshold to determine whether a fault exists.

[0154] Preferably, the present invention introduces a weighted least squares algorithm to calculate the orthogonal three-axis navigation data, and the calculation formulas are:

[0155]

[0156] Among them, W g , W a is a 5×5 weighted matrix. When there is no fault, W g , W a It is a unit array. When the ith gyro or the jth adding table fails, the row corresponding to the ith gyro or the jth adding table in the unit array is modified to 0, and the data of other rows and columns remain unchanged.

[0157] In summary, the method and system disclosed in the embodiments of the present invention have at least the following beneficial effects:

[0158] The least squares method is innovatively applied to the fault diagnosis and isolation of redundant strapdown inertial system, which has good reliability in practical applications. It can also be combined with other diagnostic technologies to produce complementary effects. Taking the redundant strapdown inertial system of ten meters as an example, based on five acceleration and five angular velocity information, the five fault diagnoses of hardware abnormality judgment, basic information detection, median filter detection, least squares algorithm prediction comparison, and redundancy detection can be performed on the measurement information. Once any fault detection identifies the abnormality of the device of the redundant strapdown inertial system of ten meters, it is considered that the device has a fault. Finally, the weighted least squares algorithm is used to isolate the fault, and finally a set of multi-level fault identification, detection and isolation processing of the redundant strapdown inertial system of ten meters is completed. Therefore, it can handle and deal with most faults. Based on the multiple detection technologies adopted, the effective identification capability of the inertial system fault can be guaranteed.

[0159] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A redundant strapdown inertial unit fault diagnosis and isolation method, characterized in that: include: Step S1, determining a target object for fault diagnosis, wherein the target object is a group of gyroscopes or an additional meter in the same redundant strapdown inertial unit; Step S2, calculate any gyroscope or add-on meter in the target object based on the change trend and constant between previous consecutive frames based on the least squares method, then calculate the predicted value of the current frame based on the change trend and the constant, and then calculate whether the difference between the predicted value and the measured value of the current frame is greater than the set prediction threshold. If so, determine that the corresponding gyroscope or add-on meter is faulty, and perform isolation processing on the faulty gyroscope or add-on meter.

2. The method according to claim 1, characterized in that The least squares processing for a single gyroscope includes: Step S21: Use the five frames of data measured by the gyroscope at the first five moments Construct the output model of the gyroscope: Among them, i represents the order of the gyroscope in the group, T represents the measurement time, t represents the frame sequence, and dg i , D i are the changing trend and constant of the gyroscope respectively; and: Step S22: Predict the gyro output data of the current frame The calculated value of is: Step S23: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. g If yes, the corresponding gyro or meter is judged to be faulty, and the calculation formula is:

3. The method according to claim 1, characterized in that The least squares processing for a single addition table includes: Step S26: Use the five frames of data measured at the first five moments of the addition table Construct the output model for adding tables: Among them, j represents the order of the addition table in the group, T represents the measurement time, t represents the frame sequence, and da j , K j They are the changing trends and constants of the added table respectively; and: Step S27: Predict the output data of the table for the current frame The calculated value of is: Step S28: Calculate whether the difference between the current frame prediction value and the measured value is greater than the set prediction threshold Y. a If yes, the corresponding addition meter or addition meter is judged to be faulty, and the calculation formula is:

4. The method according to claim 2 or 3, characterized in that: Also includes: Based on the null space solution, calculate whether the overall residual of all gyroscopes or addition meters participating in the calculation in the target object group is less than the set redundant detection threshold. If not, calculate based on the null space solution whether the local residuals of the remaining members participating in the calculation after removing one member are less than the set redundant detection threshold, and then combine the local residual results of each remaining combination to obtain the member that has caused the fault; wherein, the number of gyroscopes or addition meters based on redundant settings in the group is at least 4.

5. The method according to claim 4, characterized in that The redundant strapdown inertial group is a ten-meter redundant strapdown inertial group consisting of 5 additional meters and 5 gyroscopes. The measurement model of the gyroscope and the additional meter of the ten-meter redundant strapdown inertial group is: m g =H g oh; m a =H a g; Among them, m g 、m a are the measurement information of the redundant inertial group gyro and the added meter, H g , H a is the spatial configuration matrix of the redundant inertial group gyro and the added table of the ten meters, ω is the orthogonal three-axis oxyz angular velocity information, and g is the orthogonal three-axis oxyz acceleration information; Assume Nu g 、Nu a H g '、H a 'The null space solution satisfies: H g 'Not g =0; H a 'Not a =0; get: Not g 'H g ω=No g "M g =0; Not a 'H a g=No a "M a =0; Among them, H g ' is H g The transpose of H a ' is H a The transpose of The calculated results of the measured values ​​of the fault detection combination are the residual error e of the gyro or meter fault detection combination. g and e a , the calculation formulas are: it is g =No g "M g ; it is a =No a "M a 。 6. The method according to claim 5, characterized in that Five gyroscopes and five adding meters are installed on the platform in an inverted cone distribution, and the adding meters and the gyroscopes are arranged in pairs to make the spatial arrangement matrix consistent.

7. The method according to claim 6, characterized in that Also includes: Whether there is a hardware fault is determined by checking the single measurement value of the target object, whether the change between two consecutive data exceeds the measurement range, and whether the consecutive frames are fixed values.

8. The method according to claim 7, characterized in that Also includes: The median filtering algorithm is used to calculate whether the median of the current frame and the corresponding number of previous continuous frames exceeds the range or the set threshold to determine whether there is a fault.

9. The method according to claim 8, characterized in that The weighted least squares algorithm is introduced to calculate the orthogonal three-axis navigation data. The calculation formulas are: Among them, W g , W a is a 5×5 weighted matrix. When there is no fault, W g , W a It is a unit array. When the ith gyro or the jth adding table fails, the row corresponding to the ith gyro or the jth adding table in the unit array is modified to 0, and the data of other rows and columns remain unchanged.

10. A redundant strapdown inertial unit fault diagnosis and isolation system, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 9 is implemented.

Images