A method for determining the state of an airborne triple-redundant inertial navigation system

Abstract

The present invention discloses a state determination method for an airborne triple-redundant inertial navigation system, comprising the following steps: Step 1: constructing a fault detection quantity based on differential information between redundant signals and prior statistical characteristics; Step 2: calculating the membership of each evidence set value based on the fault detection quantity and a detection threshold; Step 3: calculating the membership of the redundant signal state set value using the evidence set value membership based on Dependent-Sensitive Reasoning (D-S) evidence reasoning; Step 4: making a state determination decision for each redundant signal based on the magnitude relationship of the membership of each state attribute in the state set value; Step 5: making a decision on the state attribute of the redundant inertial navigation system as a whole based on the fault determination decision and the state membership value of each redundant signal, and calculating the corresponding confidence value; Step 6: making a decision on the working state of the redundant inertial navigation system based on the confidence value of each state attribute of the redundant inertial navigation system. The present invention can provide a more reasonable decision for the possible "single redundant signal no fault" state.

Description

Technical Field

[0001] The invention belongs to the technical field of flight control, and in particular relates to a state judgment method for an airborne triple-redundant configuration inertial navigation system. Background Art

[0002] Airborne redundant inertial navigation systems often use system-level hardware redundancy to ensure mission reliability and safety. Taking both mission reliability and hardware reliability requirements into account, airborne redundant inertial navigation systems often adopt a triple-redundancy configuration.

[0003] The inertial navigation system with triple redundancy configuration usually realizes the functions of fault detection isolation, fault reconstruction, voting output, etc. based on the mutual detection between redundant signals. The basic structure is as follows: Figure 1 As shown. Honeywell proposed a fault detection method based on the mutual inspection of three redundant signals. By constructing differential components between the three redundant signals and comparing the differential components with the expected distribution, the redundant signals with possible faults can be identified. For the possible dual-system failure, Boeing proposed a fault detection and identification method for three redundant signals based on cyclic check values. The cyclic check values are used to record the historical fault information of each redundant signal. Each time the redundant signals are mutually inspected, in addition to considering the consistency between the differential signal and the expected distribution, the fault history of each redundant signal must also be considered. Regarding the fault detection and judgment logic of multiple redundant signals, Cai Yanan and others from the University of Electronic Science and Technology of China proposed a decision table for fault detection quantity and redundant signal fault decision. Jing Yiming and others from Nanjing University of Aeronautics and Astronautics used fuzzy sets to implement quality evaluation and fault judgment of each navigation system.

[0004] Current research on fault detection methods for redundant inertial navigation systems lacks analysis and research on the logical relationship between fault detection quantities and fault judgments. Fault detection logic lacks effective theoretical support, and it is impossible to take precautions in advance against possible decision-making risks. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for determining the state of an airborne triple-redundant inertial navigation system.

[0006] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0007] A method for determining the state of an airborne triple-redundant inertial navigation system comprises the following steps:

[0008] Step 1: Construct fault detection quantity based on differential information between redundancy signals and prior statistical characteristics;

[0009] Step 2: Calculate the membership of each evidence set value based on the fault inspection amount and detection threshold;

[0010] Step 3: Calculate the redundancy signal state set-valued membership using the evidence set-valued membership based on DS evidence reasoning;

[0011] Step 4: Each redundancy signal makes a state judgment decision based on the size relationship of the membership degree of each state attribute in the state set value;

[0012] Step 5: Make a decision on the overall state attribute of the redundant inertial navigation system based on the fault judgment decision and state membership value of each redundancy signal, and calculate the corresponding confidence value;

[0013] Step 6: Make a decision on the working state of the redundant inertial navigation system based on the confidence value of each state attribute of the redundant inertial navigation system.

[0014] Furthermore, in step 1, the first redundancy signal Δx whose output noise obeys Gaussian distribution is i,t and the second redundancy signal Δx j,t , using the chi-square fault detection quantity λ ij,t Construct differential information, as shown in formula (1)

[0015] λ ij,t =Δx ij,t T A ij -1 Δx ij,t (1)

[0016] in

[0017] Δx ij,t : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Difference vector, the vector dimension is 6;

[0018] A ij :Δx ij,t The variance prior distribution diagonal matrix of the difference vector is denoted as

[0019]

[0020] Where σ ij 2 is the prior variance of the difference vector elements.

[0021] Furthermore, in step 2, a fault decision logic based on a set-valued decision information system I = (U, A, F) is established. The set-valued decision information system consists of a decision object set U, an attribute set A, and a mapping set F. The attribute set A consists of an evidence set E and a state set S, which is denoted as

[0022]

[0023] Among them, the evidence set E is denoted as

[0024]

[0025] The definitions of each evidence element in the formula are as follows:

[0026] H ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t consistent;

[0027] The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Inconsistency;

[0028] h ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Uncertain consistency,

[0029] The state set S is denoted as

[0030]

[0031] The definitions of each state element in the formula are as follows:

[0032] H i : The first redundancy signal Δx i,t No faults;

[0033] The first redundancy signal Δx i,t Fault;

[0034] h i : The first redundancy signal Δx i,t Status uncertain,

[0035] The mapping set F includes:

[0036] Evidence element group (e ij ,e ik ) to the state set value s.

[0037]

[0038] Mapping f2 from evidence set value e to state set value

[0039] f2(e)=f1(e ij ,e ik )∩f1(e ij ,e jk )∩f1(e jk ,e ik ) (7)

[0040] Among them, in the mapping f2 there is

[0041]

[0042] Furthermore, in step 2, the membership function of the evidence element is:

[0043]

[0044]

[0045]

[0046] In the formula

[0047] P(H ij ):Evidence element H ij The degree of membership;

[0048] Evidence Elements The degree of membership;

[0049] P(h ij ):Evidence element h ij The degree of membership;

[0050] λ ij,t : Chi-square fault detection quantity;

[0051] α: significance level of chi-square hypothesis test, set to 1×10 -4 ;

[0052] T D,α : The fault detection threshold corresponding to the significance level α of the chi-square hypothesis test is 27.8563.

[0053] Furthermore, in step 3, the mapping relationship between the evidence set value e and the state set value s is shown in Table 1:

[0054] Table 1 Mapping table of evidence set values and state set values

[0055]

[0056]

[0057] Table 1e ij is the evidence element, and the value range set is s i is the state element of each redundancy signal, and the value range set is There are 26 different combinations of evidence elements, and the corresponding evidence set value is denoted as n = [0,26].

[0058] Furthermore, the formula for calculating the membership of the evidence set value based on the membership value of each element in the evidence set value in step 3 is:

[0059] m(e(n))=P(e 12 )P(e 13 )P(e 23 ) (12)

[0060] Where P(e ij ) is the evidentiary element e ij The membership value of the evidence set is used to calculate the state element s of each redundancy signal. i The formula for membership is:

[0061]

[0062] Where N is the normalization coefficient, denoted as

[0063]

[0064] Furthermore, in step 4, according to the membership degree m(s) of each state attribute of the redundancy signal, i ) The decision strategy for making state judgment is to select the state attribute with the largest membership value as the state attribute of the redundancy signal, which is recorded as:

[0065] d i =s i |m(s i )=max(m(H i ),m(H i ),m(h i )) (15).

[0066] Furthermore, in step 5, the state attributes of the redundant inertial navigation system include:

[0067] SF: Single redundancy signal failure;

[0068] AF: triple redundancy signal failure;

[0069] NF: triple redundant signal has no fault;

[0070] SN: Single redundancy signal without fault,

[0071] The confidence calculation formula for each state attribute is:

[0072]

[0073] Furthermore, in step 6, the decision on the working state of the redundant inertial navigation system is made based on the state with the highest confidence level of the state attribute, which is expressed as:

[0074] A S=a|B(a)=max(B(SF),B(SN),B(AF),B(NF)) (17)

[0075] Where A S is the redundant inertial navigation system working state decision, a is the redundant inertial navigation system working state attribute variable, and the value range set is {SF, SN, AF, NF}.

[0076] Compared with the prior art, the present invention has the following significant advantages:

[0077] 1. The state judgment strategy of the airborne triple-redundant inertial navigation system proposed in this paper adopts a set-valued decision information system to construct a decision logic framework, clarifies the logical relationship between fault detection quantity and state judgment, solves the problem that traditional fault judgment strategies have no theoretical support, and provides an effective method for the analysis and evaluation of redundant systems.

[0078] 2. By introducing the uncertainty term h in the evidence set and state set ij and h i , which solves the decision risk of incomplete judgment of the redundancy signal state due to the "second type error" of the chi-square hypothesis test in the traditional redundancy signal state judgment.

[0079] 3. Simulation experiments show that compared with traditional fault judgment strategies, the decision-making strategy proposed in this invention can identify slowly varying faults in fault margin signals in advance, and can make more reasonable decisions for the possible "single redundancy signal no fault" state. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 Structural diagram of the inertial navigation system configured for the existing airborne triple redundancy.

[0081] Figure 2 Schematic diagram of sudden fault.

[0082] Figure 3 This is a schematic diagram of a slow-changing fault.

[0083] Figure 4 Schematic diagram of the change of fault redundancy signal membership under sudden fault.

[0084] Figure 5 A comparison diagram of state judgment under sudden fault.

[0085] Figure 6 Schematic diagram of the change of fault margin signal membership under slow-varying faults.

[0086] Figure 7 A comparison diagram of state judgment under slowly varying faults.

[0087] Figure 8Schematic diagram of the change of chi-square detection quantity under the abnormal mode of "single redundant signal without fault".

[0088] Figure 9 For decision system D F Status judgment result diagram.

[0089] Figure 10 This is the D state judgment result diagram of the decision system.

[0090] Figure 11 This is the status judgment diagram of the redundant inertial navigation system in the "single redundancy signal no fault" abnormal mode. DETAILED DESCRIPTION

[0091] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0092] A method for determining the state of an airborne triple-redundant inertial navigation system comprises the following steps:

[0093] Step 1: Construct fault detection quantity based on differential information between redundancy signals and prior statistical characteristics;

[0094] Step 2: Calculate the membership of each evidence set value based on the fault inspection amount and detection threshold;

[0095] Step 3: Calculate the redundancy signal state set-valued membership using the evidence set-valued membership based on DS evidence reasoning;

[0096] Step 4: Each redundancy signal makes a state judgment decision based on the size relationship of the membership degree of each state attribute in the state set value;

[0097] Step 5: Make a decision on the overall state attribute of the redundant inertial navigation system based on the fault judgment decision and state membership value of each redundancy signal, and calculate the corresponding confidence value;

[0098] Step 6: Make a decision on the working state of the redundant inertial navigation system based on the confidence value of each state attribute of the redundant inertial navigation system.

[0099] Furthermore, in step 1, the first redundancy signal Δx whose output noise obeys Gaussian distribution is i,t and the second redundancy signal Δx j,t , using the chi-square fault detection quantity λ ij,t Construct differential information, as shown in formula (1)

[0100] λ ij,t =Δx ij,t T A ij-1 Δx ij,t (1)

[0101] in

[0102] Δx ij,t : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Difference vector, the vector dimension is 6;

[0103] A ij :Δx ij,t The variance prior distribution diagonal matrix of the difference vector is denoted as

[0104]

[0105] Where σ ij 2 is the prior variance of the difference vector elements.

[0106] Furthermore, in step 2, a fault decision logic based on a set-valued decision information system I = (U, A, F) is established. The set-valued decision information system consists of a decision object set U, an attribute set A, and a mapping set F. The attribute set A consists of an evidence set E and a state set S, which is denoted as

[0107]

[0108] Among them, the evidence set E is denoted as

[0109]

[0110] The definitions of each evidence element in the formula are as follows:

[0111] H ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t consistent;

[0112] The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Inconsistency;

[0113] h ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Uncertain consistency,

[0114] The state set S is denoted as

[0115]

[0116] The definitions of each state element in the formula are as follows:

[0117] H i : The first redundancy signal Δx i,t No faults;

[0118] The first redundancy signal Δx i,t Fault;

[0119] h i : The first redundancy signal Δx i,t Status uncertain,

[0120] The mapping set F includes:

[0121] Evidence element group (e ij ,e ik ) to the state set value s.

[0122]

[0123] Mapping f2 from evidence set value e to state set value

[0124] f2(e)=f1(e ij ,e ik )∩f1(e ij ,e jk )∩f1(e jk ,e ik ) (7)

[0125] Among them, in the mapping f2 there is

[0126]

[0127] Furthermore, in step 2, the membership function of the evidence element is:

[0128]

[0129]

[0130]

[0131] In the formula

[0132] P(H ij ):Evidence element H ij The degree of membership;

[0133] Evidence Elements The degree of membership;

[0134] P(h ij ):Evidence element h ij The degree of membership;

[0135] λij,t : Chi-square fault detection quantity;

[0136] α: significance level of chi-square hypothesis test, set to 1×10 -4 ;

[0137] T D,α : The fault detection threshold corresponding to the significance level α of the chi-square hypothesis test is 27.8563.

[0138] Furthermore, in step 3, the mapping relationship between the evidence set value e and the state set value s is shown in Table 1:

[0139] Table 1 Mapping table of evidence set values and state set values

[0140]

[0141]

[0142]

[0143] Table 1e ij is the evidence element, and the value range set is s i is the state element of each redundancy signal, and the value range set is There are 26 different combinations of evidence elements, and the corresponding evidence set value is denoted as n = [0,26].

[0144] Furthermore, the formula for calculating the membership of the evidence set value based on the membership value of each element in the evidence set value in step 3 is:

[0145] m(e(n))=P(e 12 )P(e 13 )P(e 23 ) (12)

[0146] Where P(e ij ) is the evidentiary element e ij The membership value of the evidence set is used to calculate the state element s of each redundancy signal. i The formula for membership is:

[0147]

[0148] Where N is the normalization coefficient, denoted as

[0149]

[0150] Furthermore, in step 4, according to the membership degree m(s) of each state attribute of the redundancy signal, i) The decision strategy for making state judgment is to select the state attribute with the largest membership value as the state attribute of the redundancy signal, which is recorded as:

[0151] d i =s i |m(s i )=max(m(H i ),m(H i ),m(h i )) (15).

[0152] Furthermore, in step 5, the state attributes of the redundant inertial navigation system include:

[0153] SF: Single redundancy signal failure;

[0154] AF: triple redundancy signal failure;

[0155] NF: triple redundant signal has no fault;

[0156] SN: Single redundancy signal without fault,

[0157] The confidence calculation formula for each state attribute is:

[0158]

[0159] Furthermore, in step 6, the decision on the working state of the redundant inertial navigation system is made based on the state with the highest confidence level of the state attribute, which is expressed as:

[0160] A S =a|B(a)=max(B(SF),B(SN),B(AF),B(NF)) (17)

[0161] Where A S is the redundant inertial navigation system working state decision, a is the redundant inertial navigation system working state attribute variable, and the value range set is {SF, SN, AF, NF}.

[0162] In order to verify the effectiveness and advancement of the fault judgment method proposed in this paper, three abnormal modes, namely, sudden fault, slow fault and "single redundancy signal no fault", are set in the simulation data of the triple redundant inertial navigation system. The fault judgment method proposed in this paper and the traditional fault judgment method are used to judge the fault redundancy signal and redundant system status under the three abnormal modes, and the judgment effects of the two methods are compared. Among them, the fault judgment method proposed in this paper is recorded as "decision system D F The traditional fault judgment method based on differential information is recorded as "decision system D". The comparison results are shown in the attached table and figure.

[0163] Table 2 Simulation noise bias stability parameters

[0164]

[0165] Table 2 shows the simulation error parameter settings of the redundant inertial reference system. The error parameters refer to Honeywell's Laser V micro inertial reference system, with the gyroscope model being GG1320AN and the accelerometer model being QA2000-030.

[0166] Table 3 Fault simulation setting table

[0167]

[0168] Table 3 shows the fault amplitude settings for the two fault types. It is generally believed that when the noise amplitude exceeds three times the standard deviation of the prior distribution, the redundancy signal is in a fault state.

[0169] Figure 2 and Figure 3 It is used to describe the relationship between the measured value of the fault margin signal IRS1, the measured value of the no-fault margin signal IRS2, and the simulation true value.

[0170] Figure 4 This paper describes the changes in the membership of various state attributes of the fault redundancy signal under sudden fault mode. When a fault occurs, the "no fault state" membership value immediately drops to near 0, the "fault state" membership value is distributed in the interval (0.5, 1.0), and the "abnormal state" membership value is distributed in the interval (0, 0.5). At this time, the redundancy signal is judged to be in a fault state.

[0171] Figure 5 The proposed fault decision logic and traditional fault decision logic are described to identify sudden faults. The sudden faults have large values, and both fault decision logics can effectively identify them.

[0172] Figure 6 The evolution of the membership of each state attribute of the fault margin signal under a slowly varying fault mode is described. From 1500s to 1520s, the membership value of the "non-fault state" gradually decreases to near 0 as the fault signal increases, while the membership value of the "abnormal state" gradually increases within this interval and gradually exceeds the membership value of the "normal state." From 1520s to 1532.3s, the membership value of the "fault state" gradually increases, while the membership value of the "abnormal state" decreases. After 1532.3s, the membership value of the "fault state" exceeds the membership values of the other two state attributes. The state attribute judgment result of the fault margin signal is "non-fault state - abnormal state - fault state."

[0173] Figure 7This paper describes the effectiveness of the proposed fault decision logic compared to traditional fault decision logic in identifying slowly varying faults. Because the proposed fault decision logic takes into account the fault information implicit in the increased amplitude of the fault detection variable and places higher demands on the "no fault state" decision condition, it can detect the occurrence of slowly varying faults in advance, with the first fault detection occurring 1 second earlier than traditional fault detection logic.

[0174] Figure 8 The fault detection value λ is described when the IRS2 measurement value decreases in accuracy due to a slow-changing fault in the IRS1 measurement value, resulting in a "single redundancy signal no fault" state in the redundant inertial navigation system. 12,t and λ 13,t With the detection threshold T D,α The size relationship of the chi-square test λ 13,t At 1519.1s, the detection threshold T was exceeded. D,α , and the chi-square test λ 12,t It is not until 1530.3s that it exceeds the detection threshold T D,α , and the chi-square test λ 12,t The changing trend is to first become smaller and then become larger.

[0175] Figure 9-10 Describes the decision system D under the "single redundant signal fault-free" mode F And the status judgment results of each redundancy signal by decision system D. Among them, decision system D can only obtain the status judgment of IRS2 measurement information in this abnormal mode. In engineering practice, to ensure the overall safety of the system, the redundancy signals of the other two status positions are usually regarded as fault redundancy signals. F The status judgment of the three redundancy signals can be given in this abnormal mode, and the status judgment of each redundancy signal can be reasonably explained under the fault judgment logic.

[0176] Figure 11 This paper describes how the proposed fault judgment strategy changes over time in the "single redundant signal intact" anomaly mode. As the slowly varying fault in IRS1 increases, the redundant inertial reference system's operating state gradually switches from "three redundant signals intact" to "single redundant signal normal" and finally to "single redundant signal faulty."

[0177] Table 4 Comparison of IRS1 measurement information status judgment effects

[0178]

[0179] Table 4 shows the comparison of the judgment results of the two decision strategies for fault margin signals under sudden faults and slow faults. The judgment results of the two decision strategies under sudden fault mode are basically the same; under slow fault mode, the fault judgment method D proposed in this paper is F Compared with traditional fault judgment methods, D has advantages in fault coverage and time to first fault detection.

[0180] Table 5 Comparison of the number of false detection points in IRS2 and IRS3 status judgments

[0181]

[0182] Table 5 shows the comparison of the judgment results of the two decision strategies for the non-fault margin signal under sudden fault and slow fault. The two decision strategies have no errors in the judgment results of the non-fault margin signal under sudden fault mode; under slow fault mode, the fault judgment method D proposed in this paper is F No "fault state" judgment will be made for a signal without fault margin, but the traditional fault judgment method D will make an erroneous "fault state" judgment for a signal without fault margin.

[0183] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the foregoing embodiments. The foregoing embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for determining the state of an airborne triple-redundant inertial navigation system, characterized in that: The following steps are involved: Step 1: Construct fault detection quantity based on differential information between redundancy signals and prior statistical characteristics; Step 2: Calculate the membership of each evidence set value based on the fault inspection amount and detection threshold; Step 3: Calculate the redundancy signal state set-valued membership using the evidence set-valued membership based on DS evidence reasoning; Step 4: Each redundancy signal makes a state judgment decision based on the size relationship of the membership degree of each state attribute in the state set value; Step 5: Make a decision on the overall state attribute of the redundant inertial navigation system based on the fault judgment decision and state membership value of each redundancy signal, and calculate the corresponding confidence value; Step 6: Make a decision on the working state of the redundant inertial navigation system based on the confidence value of each state attribute of the redundant inertial navigation system.

2. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 1, wherein: In step 1, the output noise obeys the first redundancy signal Δx i,t and the second redundancy signal Δx j,t , using the chi-square fault detection quantity λ ij,t Construct differential information, as shown in formula (1) l ij,t =Δx ij,t T A ij -1 Δx ij,t (1) in Δx ij,t : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Difference vector, the vector dimension is 6; A ij :Δx ij,t The variance prior distribution diagonal matrix of the difference vector is denoted as Where σ ij 2 is the prior variance of the difference vector elements.

3. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 2, wherein: In step 2, a fault decision logic based on a set-valued decision information system I = (U, A, F) is established. The set-valued decision information system consists of a decision object set U, an attribute set A, and a mapping set F, wherein the attribute set A consists of an evidence set E and a state set S, denoted as Among them, the evidence set E is denoted as The definitions of each evidence element in the formula are as follows: H ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t consistent; The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Inconsistency; h ij : The first redundancy signal Δx i,t and the second redundancy signal Δx j,t Uncertain consistency. The state set S is denoted as The definitions of each state element in the formula are as follows: H i : The first redundancy signal Δx i,t No faults; The first redundancy signal Δx i,t Fault; h i : The first redundancy signal Δx i,t Status uncertain, The mapping set F includes: Evidence element group (e ij ,e ik ) to the state set value s. Mapping f2 from evidence set value e to state set value f2(e)=f1(e ij ,And ik )∩f1(e ij ,And jk )∩f1(e jk ,And ik ) (7) Among them, in the mapping f2 there is 4. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 3, wherein: In step 2, the membership function of the evidence element is: In the formula P(H ij ):Evidence element H ij The degree of membership; Evidence Elements The degree of membership; P(h ij ):Evidence element h ij The degree of membership; λ ij,t : Chi-square fault detection quantity; α: significance level of chi-square hypothesis test, set to 1×10 -4 ; T D,α : The fault detection threshold corresponding to the significance level α of the chi-square hypothesis test is 27.8563.

5. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 4, wherein: In step 3, the mapping relationship between the evidence set value e and the state set value s is shown in Table 1: Table 1 Mapping table of evidence set values and state set values Table 1e ij is the evidence element, and the value range set is s i is the state element of each redundancy signal, and the value range set is There are 26 different combinations of evidence elements, and the corresponding evidence set value is denoted as n = [0,26].

6. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 5, wherein: The formula for calculating the membership of the evidence set value based on the membership value of each element in the evidence set value in step 3 is: m(e(n))=P(e 12 )P(ie 13 )P(ie 23 ) (12) Where P(e ij ) is the evidentiary element e ij The membership value of the evidence set is used to calculate the state element s of each redundancy signal. i The formula for membership is: Where N is the normalization coefficient, denoted as 7. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 6, wherein: In step 4, the membership degree m(s) of each state attribute of the redundancy signal is calculated. i ) The decision strategy for making state judgment is to select the state attribute with the largest membership value as the state attribute of the redundancy signal, which is recorded as: d i =s i |m(s i )=max(m(H i ),m(H i ),m(h i )) (15)。 8. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 7, wherein: In step 5, the state attributes of the redundant inertial navigation system include: SF: Single redundancy signal failure; AF: triple redundancy signal failure; NF: triple redundant signal has no fault; SN: Single redundancy signal without fault, The confidence calculation formula for each state attribute is:

9. The method for determining the state of an airborne triple-redundant inertial navigation system according to claim 8, wherein: In step 6, the decision on the working state of the redundant inertial navigation system is the state with the highest confidence level of the state attribute, which is recorded as: A S =a|B(a)=max(B(SF),B(SN),B(AF),B(NF)) (17) Where A S is the redundant inertial navigation system working state decision, a is the redundant inertial navigation system working state attribute variable, and the value range set is {SF, SN, AF, NF}.

Images