Method for determining reliability of full-size strength test system based on multi-source data fusion

Abstract

The invention discloses a method for determining the reliability of a full-size strength test system based on multi-source data fusion, and the method comprises the steps: obtaining different types of reliability data from different ways for elements which need to collect the reliability data in the full-size strength test system; according to the collected reliability data of the elements of the full-size strength test system, combining the data types of the reliability data, and constructing a fusion strategy of the reliability data of the elements of the same type and different types; constructing a structural schematic diagram of the full-size strength test system; constructing a GO graph model of the full-size strength test system based on the schematic diagram; and obtaining the reliability of the original by using the fusion strategy, and obtaining the reliability of the full-size strength test system based on the GO graph model and a GO method. According to the method, the problems that the system is lack of reliability data of a unified type and the reliability result does not conform to reality are effectively solved, and the accuracy of reliability evaluation of the full-size strength test system is improved.

Description

Technical Field

[0001] The invention relates to the field of system reliability engineering, and in particular to a reliability determination method for a full-scale strength test system by multi-source data fusion. Background Art

[0002] The full-scale strength test system is used to test and evaluate the strength and durability of aircraft structures under real working conditions. It is a key equipment in the aviation field to ensure the safety and reliability of aircraft structures. By simulating various load conditions in the actual flight environment, the full-scale strength test can identify and verify potential problems in the aircraft structure design. The full-scale strength test system is usually composed of multiple complex mechanical, electrical and hydraulic components, and the reliability of these components directly affects the reliability of the entire system.

[0003] However, due to the complexity and diversity of components in full-scale strength test systems, there is often a lack of comprehensive and detailed component reliability data. There are many types of reliability data generated during the operation of full-scale strength test systems, including stress-strain data, vibration data, temperature data, historical failure data, maintenance records, operation logs, etc. The sources of these data include test data, data provided by manufacturers, historical accumulated data, reliability manual data, etc.

[0004] Due to the diverse types and sources of reliability data, traditional reliability assessment methods usually rely on a single data source and data type, and cannot fully utilize multi-source data during system operation, resulting in limited accuracy and comprehensiveness of assessment results. Summary of the invention

[0005] The purpose of the present invention is to provide a reliability determination method for a full-scale strength test system based on multi-source data fusion, which is based on the fusion of reliability data from different sources and different types to comprehensively and accurately evaluate the reliability of the full-scale strength test system.

[0006] In order to achieve the above tasks, the present invention adopts the following technical solutions:

[0007] A reliability determination method for a full-scale strength test system based on multi-source data fusion, comprising:

[0008] S1, for the components in the full-scale strength test system that need to collect reliability data, obtain different types of reliability data from different channels;

[0009] S2, based on the collected reliability data of the components of the full-scale strength test system and combined with the data type of the reliability data, a fusion strategy for the reliability data of each component of the same type and different types is constructed;

[0010] S3, construct the structural principle diagram of the full-scale strength test system;

[0011] S4, constructing a GO graph model of a full-scale strength test system based on the schematic diagram;

[0012] S5, based on the data type of the collected component reliability data, the reliability of the component is obtained by using the fusion strategy, and based on the GO graph model and the GO method, the reliability of the full-scale strength test system is obtained.

[0013] Furthermore, the specific process of S1 is as follows:

[0014] S1-1, determine the components in the full-scale strength test system for which reliability data collection is required;

[0015] S1-2, determine the type of reliability data that needs to be collected, including performance degradation data, failure rate data, and life data;

[0016] S1-3, determine the ways to obtain the reliability data that needs to be collected, including reliability manuals, historical records, experimental and test data, data provided by suppliers, and expert experience evaluation data;

[0017] S1-4, check the accuracy, consistency and completeness of the collected reliability data to ensure that the data is authentic and reliable.

[0018] Furthermore, the fusion strategy of reliability data of components of the same type and different types described in S2 includes:

[0019] If the reliability data types collected for the components are of the same type, the same type of reliability data fusion is performed:

[0020] Fusion between performance degradation data:

[0021] Record the performance degradation data of n groups of components collected, and fit n groups of performance degradation data to obtain n performance degradation trajectory curves; combined with the failure threshold, obtain n pseudo life data T i , by determining the distribution type of pseudo-lifetime data, the reliability function R1(t) based on performance degradation data is obtained;

[0022] Fusion between life span data:

[0023] Record the life data T of m components collected j , by determining the life distribution type, the reliability function R2(t) based on the life data is obtained;

[0024] Fusion between failure rate data:

[0025] Using the s failure rate data λ obtained from different channels k(k=1,2,..,s), and assign corresponding weights p through analytic hierarchy process k , and the comprehensive failure rate is obtained Constructing the reliability function R3(t)=e based on failure rate data -λt ; Where e is the base of natural logarithm, t is the time parameter;

[0026] After the reliability function is obtained, the reliability of the component is calculated using the reliability function.

[0027] Furthermore, the fusion strategy of reliability data of components of the same type and different types described in S2 also includes:

[0028] If the reliability data types collected for the components are of different types, different types of reliability data fusion are performed:

[0029] Fusion between performance degradation data and life data:

[0030] After obtaining the reliability function R1(t) based on the performance degradation data and the reliability function R2(t) based on the life data, the reliability function R1(t) based on the performance degradation data and the reliability function R2(t) based on the life data are assigned corresponding weights q1 and q2 through the hierarchical analysis method, and the reliability function R4(t) based on the performance degradation data and the life data is obtained as q1R1(t)+q2R2(t);

[0031] Fusion between performance degradation data and failure rate data:

[0032] The reliability function R1(t) based on the performance degradation data and the reliability function R3(t) based on the failure rate data are obtained as follows: -λt Then, the reliability function R1(t) based on the performance degradation data and the reliability function R3(t) based on the failure rate data are assigned corresponding weights h1 and h2 through the hierarchical analysis method, and the reliability function R5(t) based on the performance degradation data and the failure rate data is obtained as follows: h1R1(t)+h2R3(t);

[0033] Fusion between life data and failure data:

[0034] The reliability function R2(t) based on the life data and the reliability function R3(t) based on the failure rate data are obtained as follows: -λt Then, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights g1 and g2 through the hierarchical analysis method, and the reliability function R6(t) based on life data and failure rate data is obtained as follows: g1R2(t)+g2R3(t);

[0035] Fusion between performance degradation data, life data and failure rate data:

[0036] The reliability function R1(t) based on performance degradation data, the reliability function R2(t) based on life data, and the reliability function R3(t) based on failure rate data are obtained = e -λt Then, the reliability function R1(t) based on performance degradation data, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights ω1, ω2 and ω3 through the hierarchical analysis method, and the reliability function R7(t) based on life data and failure rate data is obtained: R7(t) = ω1R1(t) + ω2R2(t) + ω3R3(t).

[0037] Furthermore, the target layer of the hierarchical analysis method is the data type, including performance degradation data, life data, and failure data; the criterion layer is the factors affecting component reliability, including the reliability of the data source, the data sample size, and the data processing plan; the solution layer is the data source, including reliability manuals, historical records, experiments and tests, suppliers, and expert experience evaluations; the expert scoring method is used to construct a discriminant matrix to obtain the weights of each data source in the final solution layer relative to the reliability of the target layer components.

[0038] Furthermore, the component blocks in the schematic diagram correspond to the actual component elements of the system, and the signal flow in the schematic diagram is consistent with the information flow when the system is actually working.

[0039] Furthermore, the GO graph model of the full-scale strength test system constructed based on the schematic diagram in S4 includes:

[0040] S4-1, Determine the operator type of the full-scale strength test system

[0041] Operator types are divided into functional operators and logical operators. According to the structural schematic diagram of the full-scale strength test system constructed by S3, the two-state action element in the system is represented as a type 1 operator, the input signal action element in the system is represented as a type 5 operator, the multi-input and multi-output element in the system is represented as a type 13 operator, the AND logical relationship in the system is represented as a type 10 operator, and the multi-input and multi-closed-loop link in the system is represented as a type 24B operator.

[0042] S4-2, Establish the reliability GO diagram model of the full-scale strength test system

[0043] Based on the determined operator type and the structural schematic diagram of the full-scale strength test system, the first-level GO diagram model and the second-level GO diagram model of the full-scale strength test system are established with reference to "Principles and Applications of GO Method".

[0044] Further, S5, based on the data type of the collected component reliability data, the reliability of the component is obtained by using the fusion strategy, and based on the GO graph model and the GO method, the reliability of the full-scale strength test system is obtained, including:

[0045] For the components corresponding to the operators in the GO graph model, the type of the reliability data of the collected components is determined. If the reliability data of the components has only one type, the reliability data fusion method of the same type in the strategy is used to fuse them and calculate the reliability of the components; if the components have more than one type, the reliability data fusion method of different types in the strategy is used to fuse them and calculate the reliability of the components;

[0046] After determining the reliability of all components in the full-scale strength test system, the reliability of the output signal flow of the full-scale strength test system is determined according to the GO method. This reliability is the reliability quantification value of the full-scale strength test system.

[0047] Compared with the prior art, the present invention has the following technical features:

[0048] The method for determining the reliability of a full-scale strength test system with multi-source data fusion provided by the present invention overcomes the limitation of a single data source by integrating reliability data from multiple sources and types, such as test data, expert experience data, historical accumulated data, and reliability manual data, in the absence of reliability data, thereby improving the accuracy of reliability assessment. Accurate reliability assessment results can help formulate more reasonable maintenance plans, optimize maintenance strategies, avoid excessive or insufficient maintenance, thereby reducing maintenance costs, extending the service life of the system, and providing strong support for the safe operation and scientific management of the full-scale strength test system. In addition, the method proposed by the present invention is not only applicable to the reliability assessment of the full-scale strength test system, but can also be promoted and applied to the reliability analysis of other complex systems, promoting technological progress and application development in related fields, and has broad application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 It is a schematic diagram of the process of the present invention;

[0050] Figure 2 A hierarchical structure model for the data fusion of the actuator failure rate of the full-scale strength test system in the present invention;

[0051] Figure 3 It is a structural principle diagram of a full-scale strength test system in an embodiment of the present invention;

[0052] Figure 4 It is a GO graph model of the equivalent output structure of the multi-input and multi-feedback links of the full-scale strength test system in the embodiment of the present invention;

[0053] Figure 5 A GO graph model of an equivalent feedback structure of a multi-input and multi-feedback link of a full-scale strength test system in an embodiment of the present invention;

[0054] Figure 6 It is a GO graph model of multiple input and multiple feedback links of a full-scale strength test system in an embodiment of the present invention;

[0055] Figure 7 It is the GO diagram model of the full-scale strength test system in the embodiment of the present invention;

[0056] Figure 8 It is the reliability change curve of the full-scale strength test system in the embodiment of the present invention.

[0057] Explanation of the accompanying numbers: 1-strain gauge, 2-position sensor, 3-ST-16 / ST18 data acquisition system, 4-data acquisition computer, 5-test control system, 6-workstation, 7-oil station, 8-hydraulic pump, 9-oil distributor, 10-servo valve A, 11-actuator A, 12-lever assembly, 13-force sensor A, 14-servo valve B, 15-actuator B, 16-force sensor B, 17-AeroST loading control system. DETAILED DESCRIPTION

[0058] Referring to the accompanying drawings, a reliability determination method for a full-scale strength test system with multi-source data fusion provided by the present invention comprises the following steps:

[0059] S1, for the components in the full-scale strength test system that need to collect reliability data, different types of reliability data are obtained from different channels.

[0060] S1-1, determine the components in the full-scale strength test system for which reliability data collection is required;

[0061] S1-2, determine the type of reliability data that needs to be collected, including performance degradation data, failure rate data, and life data;

[0062] S1-3, determine the ways to obtain the reliability data that needs to be collected, including reliability manuals, historical records, experimental and test data, data provided by suppliers, and expert experience evaluation data;

[0063] S1-4, check the accuracy, consistency and completeness of the collected reliability data to ensure that the data is authentic and reliable.

[0064] Through the above specific steps, the reliability data of the full-scale strength test system components can be systematically collected to provide basic data support for subsequent system reliability analysis.

[0065] S2, based on the reliability data of the components of the full-scale strength test system collected as described in S1 and combined with the data type of the reliability data, a fusion strategy for the reliability data of components of the same type and different types is constructed.

[0066] (1) If the reliability data types collected for the component are of the same type, the same type of reliability data fusion is carried out, including the fusion of performance degradation data, the fusion of life data, and the fusion of failure data, to obtain the reliability data fusion result of the component.

[0067] Fusion between performance degradation data:

[0068] n sets of performance degradation data are fitted to obtain n performance degradation trajectory curves, and combined with the failure threshold, n pseudo life data T are obtained. i , by determining the life distribution type, the reliability function R1(t) based on the performance degradation data is obtained.

[0069] Fusion between life span data:

[0070] According to m life data T j , by determining the life distribution type, the reliability function R2(t) based on the life data is obtained.

[0071] Fusion between failure rate data:

[0072] Using the s failure rate data λ obtained from different channels k (k=1,2,..,s), and assign corresponding weights p through analytic hierarchy process k , we get the comprehensive failure rate λ, Constructing the reliability function R3(t)=e based on failure rate data -λt ; where e is the base of the natural logarithm and t is the time parameter.

[0073] (2) If the types of reliability data collected for a component are of different types, different types of reliability data fusion are performed, including the fusion between performance degradation data and life data, the fusion between performance degradation data and failure rate data, the fusion between life data and failure data, and the fusion between performance degradation data, life data and failure rate data, to obtain the reliability data fusion result of the component.

[0074] Fusion between performance degradation data and life data:

[0075] n sets of performance degradation data are fitted to obtain n performance degradation trajectory curves, and combined with the failure threshold, n pseudo life data T are obtained. i ; By determining the life distribution type, the reliability function R1(t) based on the performance degradation data is obtained; according to m life data T jBy determining the life distribution type, the reliability function R2(t) based on the life data is obtained; then the reliability function R1(t) based on the performance degradation data and the reliability function R2(t) based on the life data are assigned corresponding weights q1 and q2 through the hierarchical analysis method, and the reliability function R4(t) based on the performance degradation data and the life data is obtained as R4(t)=q1R1(t)+q2R2(t).

[0076] Fusion between performance degradation data and failure rate data:

[0077] n sets of performance degradation data are fitted to obtain n performance degradation trajectory curves, and combined with the failure threshold, n pseudo life data T are obtained. i , by determining the life distribution type, the reliability function R1(t) based on the performance degradation data is obtained; the s failure rate data λ obtained from different channels k (k=1,2,..,s), and assign corresponding weights p through analytic hierarchy process k , we get the comprehensive failure rate λ, The reliability function R3(t)=e based on the failure rate data is obtained -λt ; Then, the reliability function R1(t) based on the performance degradation data and the reliability function R3(t) based on the failure rate data are assigned corresponding weights h1 and h2 through the hierarchical analysis method, and the reliability function R5(t) based on the performance degradation data and the failure rate data is obtained as R5(t)=h1R1(t)+h2R3(t).

[0078] Fusion between life data and failure data:

[0079] According to m life data T j , by determining the life distribution type, the reliability function R2(t) based on the life data is obtained; the s failure rate data λ obtained from different channels k , assign corresponding weights through the hierarchical analysis method, and obtain the comprehensive failure rate λ, The reliability function R3(t)=e based on the failure rate data is obtained -λt ; Then, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights g1 and g2 through the hierarchical analysis method, and the reliability function R6(t) based on life data and failure rate data is obtained as R6(t)=g1R2(t)+g2R3(t).

[0080] Fusion between performance degradation data, life data and failure rate data:

[0081] n sets of performance degradation data are fitted to obtain n performance degradation trajectory curves, and combined with the failure threshold, n pseudo life data T are obtained. i, by determining the life distribution type, the reliability function R1(t) based on the performance degradation data is obtained; according to m life data T j , by determining the life distribution type, the reliability function R2(t) based on the life data is obtained; the s failure rate data λ obtained from different channels k , assign corresponding weights through the hierarchical analysis method, and obtain the comprehensive failure rate λ, The reliability function R3(t)=e based on the failure rate data is obtained -λt ; Then, the reliability function R1(t) based on performance degradation data, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights ω1, ω2 and ω3 through the hierarchical analysis method, and the reliability function R7(t) based on life data and failure rate data is obtained as R7(t)=ω1R1(t)+ω2R2(t)+ω3R3(t).

[0082] Among them, the target layer of the hierarchical analysis method is the data type, such as performance degradation data, life data, and failure data; the criterion layer is the factors affecting component reliability, such as the reliability of the data source, the data sample size, the data processing plan, etc.; the solution layer is the data source, including reliability manuals, historical records, experiments and tests, suppliers, and expert experience evaluations; the expert scoring method is used to construct a discriminant matrix to obtain the weight of each data source in the final solution layer relative to the reliability of the target layer components.

[0083] S3, constructing a structural schematic diagram of a full-scale strength test system; wherein the component boxes in the schematic diagram correspond to the actual component elements of the system, and the signal flow in the schematic diagram is consistent with the information flow when the system is actually working.

[0084] S4, constructing a GO graph model of a full-scale strength test system based on the schematic diagram.

[0085] S4-1, Determine the full-scale strength test system operator type.

[0086] Operator types are divided into functional operators and logical operators. According to the structural schematic diagram of the full-scale strength test system described in S3, referring to "Principles and Applications of GO Method" (edited by Shen Zupei and Huang Xiangrui, 2004.8) and the basic operator selection principle in the patent application with publication number CN114117811A, the two-state action element in the system itself is represented as a type 1 operator, the input signal action element in the system itself is represented as a type 5 operator, the multi-input and multi-output element in the system is represented as a type 13 operator, the AND logical relationship in the system is represented as a type 10 operator, and the multi-input and multi-closed-loop links in the system are represented as type 24B operators.

[0087] S4-2, establish the reliability GO diagram model of the full-scale strength test system.

[0088] According to the structural principle diagram of the full-scale strength test system described in S3, referring to "Principles and Applications of GO Method" (edited by Shen Zupei and Huang Xiangrui, 2004.8) and the GO diagram modeling principle in CN114117811A, the first-level GO diagram model and the second-level GO diagram model of the full-scale strength test system were established respectively.

[0089] S5, based on the data type of the collected component reliability data, the reliability of the component is obtained by using the fusion strategy, and based on the GO graph model and the GO method, the reliability of the full-scale strength test system is obtained.

[0090] S5-1, determine the reliability data of the components corresponding to the operators in the GO diagram model of the full-scale strength test system.

[0091] For the components corresponding to the operators in the GO graph model, the type of reliability data of the components collected in S1 is determined. If the reliability data of the component has only one type, the corresponding data type in (1) of S2 is used for fusion to calculate the reliability of the component. If the component has more than one type, the different types of data types corresponding to (2) of S2 are used for fusion to calculate the reliability of the component.

[0092] For example, for a certain component, when only a plurality of life data are collected during data collection, the "fusion of life data" method in (1) of S2 is used to take the fusion result of the component life data as the reliability of the component.

[0093] If a component collects both life data and performance degradation data during data collection, the method of "fusion of performance degradation data and life data" in (2) of S2 is used to take the fusion result as the reliability of the component.

[0094] S5-2, determine the reliability of the full-scale strength test system.

[0095] Determining the reliability of the full-scale strength test system means determining the success probability of the full-scale strength test system, that is, the reliability of the full-scale strength test system. After determining the reliability of all components in the full-scale strength test system, the reliability of the output signal flow of the full-scale strength test system is determined according to the quantitative calculation method of the GO method. The reliability of the output signal flow of the full-scale strength test system is the quantitative value of the reliability of the full-scale strength test system.

[0096] Example:

[0097] This embodiment provides a method for determining the reliability of a full-scale strength test system with multi-source data fusion, comprising the following steps:

[0098] S1, collect reliability data of components that make up the full-scale strength test system.

[0099] Table 1 is a list of components of a full-scale strength test system and the failure rate data of the full-scale strength test system components collected from the manufacturer, reliability manual and reliability test, where “-” indicates that there is no data for this item.

[0100] Table 1 Failure rate data of components of a full-scale strength test system from different sources

[0101]

[0102] S2, determine the reliability data of full-scale strength test system components based on data fusion

[0103] Based on the failure rate data of full-scale strength test system components collected from different sources and the fusion method of the same type of reliability data, taking the actuator as an example, the failure rate data fusion of full-scale strength test system components is carried out to determine the component reliability.

[0104] First construct Figure 2 The hierarchical model shown in FIG. 1 is divided into three levels, wherein the first level is the target level A, which represents the failure rate data of the actuator; the second level is the criterion level B, which represents the various factors affecting the failure rate weights of the actuators from different sources, including the reliability of the data source, the data sample size, and the data processing method; the third level is the object level F, which represents the source of the reliability data to be weighted, including data provided by the manufacturer, reliability manuals, and reliability experiments.

[0105] Secondly, the expert scoring scale of 1 to 9 is used to construct the judgment matrix of target layer A relative to criterion layer B and the judgment matrix of criterion layer B relative to object layer F respectively.

[0106] (1) Judgment matrix Λ of target layer A relative to criterion layer B A-B ;

[0107]

[0108] (2) Judgment matrix Λ of criterion layer B relative to object layer F B1-F ,Λ B1-F ,Λ B1-F .

[0109]

[0110] Finally, the weight vector ω of the object layers F1, F2, and F3 relative to the target layer A is calculated to be (0.2974, 0.2310, 0.4716). The actuator failure rate λ = 5.4734 / 10 can be obtained by integrating the data provided by the manufacturer, the reliability manual data and the reliability test data. -6 h -1 , the actuator reliability function based on failure rate data fusion is:

[0111] According to the above cases, the reliability data of the components of the full-scale strength test system based on data fusion can be determined, as shown in Table 2.

[0112] Table 2 Reliability data of components of a full-scale strength test system based on data fusion

[0113]

[0114] S3. Draw the structural schematic diagram of the full-scale strength test system whose reliability needs to be determined.

[0115] Without affecting the understanding of the working principle of the full-scale strength test system, the full-scale strength test system is reasonably simplified. The structural principle diagram of the full-scale strength test system is shown in the figure. Figure 3 As shown, the full-scale strength test system includes strain gauge 1, position sensor 2, ST-16 / ST18 data acquisition system 3, data acquisition computer 4, test control system 5, workstation 6, oil station 7, hydraulic pump 8, oil circuit distributor 9, servo valve A10, actuator A11, lever assembly 12, force sensor A13, servo valve B14, actuator B15, force sensor B16, and AeroST loading control system 17.

[0116] S4, draw the GO diagram model of the full-scale strength test system whose reliability needs to be determined.

[0117] The first step is to determine the operator type of the full-scale strength test system.

[0118] like Figure 3As shown in the figure, the structural schematic diagram of the full-scale strength test system, the strain gauge 1, the position sensor 2, the test control system 5, and the oil station 7 are the input elements in the full-scale strength test system, which are represented by the type 5 operator; the ST-16 / ST18 data acquisition system 3, the data acquisition computer 4, the oil circuit distributor 9, the actuator A11, the lever assembly 12, the force sensor A13, the actuator B15, and the force sensor B16 are the two-state elements in the full-scale strength test system, which are represented by the type 1 operator; the hydraulic pump 8 must be acted on by the input signal of the test control system 5 before the output signal acts on the oil circuit distributor 9, and the servo valve A10 and the servo valve B14 are acted on by the input load feedback signal of the AeroST loading control system 17 before the output The flow regulation signal acts on the actuator A11 and the actuator B15, so the hydraulic pump 8, the servo valve A10 and the servo valve B14 are the signal-conducting elements in the test system, and are represented by the type 6 operator; the AeroST loading control system 17 has two input signals, the workstation input and the force sensor, and one output signal acts on the servo valve A10 and the servo valve B14, so the AeroST loading control system 17 is a multi-input and multi-output element in the full-scale strength test system, so the type 13 operator is selected to represent it; the AND logical relationship in the full-scale strength test system is represented by the type 10 operator; the full-scale strength test system contains multi-input and multi-feedback links, so the type 24B operator is selected to describe the multi-input and multi-feedback links. Table 3 shows the operator types and operator numbers corresponding to the various elements and logical relationships of the full-scale strength test system. In order to facilitate the understanding of the GO graph model, a virtual input unit is introduced, and its success probability is 1, that is, the reliability is 1.

[0119] Table 3 Full-scale strength test system operator

[0120]

[0121]

[0122] The second step is to establish a reliability model of the full-scale strength test system.

[0123] (1) Establish a first-level GO graph model, that is, establish a GO graph model with multiple inputs and multiple feedback links. Figure 4 As shown in FIG, the equivalent output structure GO graph model of the full-scale strength test system contains multiple input and multiple feedback links, where operator 23 and operator 24 are virtual input operators with a success probability of 1 for the output signal; Figure 5 As shown in FIG. 1 , the equivalent feedback structure GO graph model of the full-scale strength test system contains multiple input and multiple feedback links, where operator 25, operator 26 and operator 27 are virtual input operators with an output signal success probability of 1; Figure 6As shown, the GO graph model containing multiple input and multiple feedback links in the full-scale strength test system.

[0124] (2) Establish a secondary GO graph model, that is, establish a complete GO graph model of the full-scale strength test system. Figure 7 As shown, the GO diagram model of the full-scale strength test system, the first digit of the operator represents the operator type, and the second digit represents the operator number; the number on the signal flow represents the signal flow number; the signal flow 22 is the output of the full-scale strength test system.

[0125] S5, determine the reliability of the full-scale strength test system based on the GO method.

[0126] According to the reliability data of various sources of the full-scale strength test system components in S1 and S2 and the component reliability determination method based on multi-source data fusion, the component reliability function corresponding to the operator of the full-scale strength test system is determined. Combined with the GO graph model of the full-scale strength test system constructed in S4, the reliability of the system output signal flow is determined according to the GO method quantitative calculation method in "Principles and Applications of GO Method" (edited by Shen Zupei and Huang Xiangrui, 2004.8). The reliability function of the full-scale strength test system is as follows Figure 8 shown.

[0127] The above results show that the method for determining the reliability of a full-scale strength test system using multi-source data fusion provided by the present invention can make full use of various types of reliability data from various sources in the absence of reliability data, and perform reliability evaluation on the full-scale strength test system through reliability data fusion, thereby overcoming the limitation of a single data source, thereby improving the accuracy of reliability evaluation. Accurate reliability evaluation results can help formulate more reasonable maintenance plans, optimize maintenance strategies, avoid excessive or insufficient maintenance, thereby reducing maintenance costs, extending the service life of the system, and providing strong support for the safe operation and scientific management of the full-scale strength test system.

[0128] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A reliability determination method for a full-scale strength test system based on multi-source data fusion, characterized in that: include: S1, for the components in the full-scale strength test system that need to collect reliability data, obtain different types of reliability data from different channels; S2, based on the collected reliability data of the components of the full-scale strength test system and combined with the data type of the reliability data, a fusion strategy for the reliability data of each component of the same type and different types is constructed; S3, construct the structural principle diagram of the full-scale strength test system; S4, constructing a GO graph model of a full-scale strength test system based on the schematic diagram; S5, based on the data type of the collected component reliability data, the reliability of the original component is obtained by using the fusion strategy, and based on the GO graph model and the GO method, the reliability of the full-scale strength test system is obtained.

2. The reliability determination method of a full-scale strength test system based on multi-source data fusion according to claim 1 is characterized in that: The specific process of S1 is as follows: S1-1, determine the components in the full-scale strength test system for which reliability data collection is required; S1-2, determine the type of reliability data that needs to be collected, including performance degradation data, failure rate data, and life data; S1-3, determine the ways to obtain the reliability data that needs to be collected, including reliability manuals, historical records, experimental and test data, data provided by suppliers, and expert experience evaluation data; S1-4, check the accuracy, consistency and completeness of the collected reliability data to ensure that the data is authentic and reliable.

3. The reliability determination method of a full-scale strength test system based on multi-source data fusion according to claim 1 is characterized in that: The fusion strategy of reliability data of components of the same type and different types described in S2 includes: If the reliability data types collected for the components are of the same type, the same type of reliability data fusion is performed: Fusion between performance degradation data: Record the performance degradation data of n groups of components collected, and fit n groups of performance degradation data to obtain n performance degradation trajectory curves; combined with the failure threshold, obtain n pseudo life data T i , by determining the distribution type of pseudo-lifetime data, the reliability function R1(t) based on performance degradation data is obtained; Fusion between life span data: Record the life data T of m components collected j , by determining the life distribution type, the reliability function R2(t) based on the life data is obtained; Fusion between failure rate data: Using the s failure rate data λ obtained from different channels k (k=1,2,..,s), and assign corresponding weights p through analytic hierarchy process k , and the comprehensive failure rate is obtained Constructing the reliability function R3(t)=e based on failure rate data -λt ; Where e is the base of natural logarithm, t is the time parameter; After the reliability function is obtained, the reliability of the component is calculated using the reliability function.

4. The reliability determination method of a full-scale strength test system based on multi-source data fusion according to claim 1 is characterized in that: The fusion strategy of reliability data of components of the same type and different types described in S2 also includes: If the reliability data types collected for the components are of different types, different types of reliability data fusion are performed: Fusion between performance degradation data and life data: After obtaining the reliability function R1(t) based on the performance degradation data and the reliability function R2(t) based on the life data, the reliability function R1(t) based on the performance degradation data and the reliability function R2(t) based on the life data are assigned corresponding weights q1 and q2 through the hierarchical analysis method, and the reliability function R4(t) based on the performance degradation data and the life data is obtained as q1R1(t)+q2R2(t); Fusion between performance degradation data and failure rate data: The reliability function R1(t) based on the performance degradation data and the reliability function R3(t) based on the failure rate data are obtained as follows: -λt Then, the reliability function R1(t) based on the performance degradation data and the reliability function R3(t) based on the failure rate data are assigned corresponding weights h1 and h2 through the hierarchical analysis method, and the reliability function R5(t) based on the performance degradation data and the failure rate data is obtained as follows: h1R1(t)+h2R3(t); Fusion between life data and failure data: The reliability function R2(t) based on the life data and the reliability function R3(t) based on the failure rate data are obtained as follows: -λt Then, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights g1 and g2 through the hierarchical analysis method, and the reliability function R6(t) based on life data and failure rate data is obtained as follows: g1R2(t)+g2R3(t); Fusion of performance degradation data, life data and failure rate data: The reliability function R1(t) based on performance degradation data, the reliability function R2(t) based on life data, and the reliability function R3(t) based on failure rate data are obtained = e -λt Then, the reliability function R1(t) based on performance degradation data, the reliability function R2(t) based on life data and the reliability function R3(t) based on failure rate data are assigned corresponding weights ω1, ω2 and ω3 through the hierarchical analysis method, and the reliability function R7(t) based on life data and failure rate data is obtained: R7(t) = ω1R1(t) + ω2R2(t) + ω3R3(t).

5. The reliability determination method of a full-scale strength test system with multi-source data fusion according to claim 1 is characterized in that: The target layer of the hierarchical analysis method is the data type, including performance degradation data, life data, and failure data; the criterion layer is the factors affecting component reliability, including the reliability of the data source, the data sample size, and the data processing plan; the solution layer is the data source, including reliability manuals, historical records, experiments and tests, suppliers, and expert experience evaluations; the expert scoring method is used to construct a discriminant matrix to obtain the weights of each data source in the final solution layer relative to the reliability of the target layer components.

6. The reliability determination method of a full-scale strength test system based on multi-source data fusion according to claim 1 is characterized in that: The component blocks in the schematic diagram correspond to the actual component elements of the system, and the signal flow in the schematic diagram is consistent with the information flow when the system is actually working.

7. The reliability determination method of a full-scale strength test system with multi-source data fusion according to claim 1 is characterized in that: The GO graph model of the full-scale strength test system constructed based on the schematic diagram described in S4 includes: S4-1, Determine the operator type of the full-scale strength test system Operator types are divided into functional operators and logical operators. According to the structural schematic diagram of the full-scale strength test system constructed by S3, the two-state action element in the system is represented as a type 1 operator, the input signal action element in the system is represented as a type 5 operator, the multi-input and multi-output element in the system is represented as a type 13 operator, the AND logical relationship in the system is represented as a type 10 operator, and the multi-input and multi-closed-loop link in the system is represented as a type 24B operator. S4-2, Establish the reliability GO diagram model of the full-scale strength test system Based on the determined operator type and the structural schematic diagram of the full-scale strength test system, the first-level GO diagram model and the second-level GO diagram model of the full-scale strength test system are established with reference to "Principles and Applications of GO Method".

8. The reliability determination method of a full-scale strength test system with multi-source data fusion according to claim 1 is characterized in that: S5: based on the data type of the collected component reliability data, the reliability of the original component is obtained by using the fusion strategy, and based on the GO graph model and the GO method, the reliability of the full-scale strength test system is obtained, including: For the components corresponding to the operators in the GO graph model, the type of the reliability data of the collected components is determined. If the reliability data of the components has only one type, the reliability data fusion method of the same type in the strategy is used to fuse them and calculate the reliability of the components; if the components have more than one type, the reliability data fusion method of different types in the strategy is used to fuse them and calculate the reliability of the components; After determining the reliability of all components in the full-scale strength test system, the reliability of the output signal flow of the full-scale strength test system is determined according to the GO method. This reliability is the reliability quantification value of the full-scale strength test system.

Images