A method for failure analysis of an electronic system

Abstract

The application relates to the technical field of electronic system fault analysis, and discloses a fault analysis method of an electronic system, which comprises the following steps: creating an initial attribute model of the electronic system by using an attribute set family; constructing an initial calculation parameter set of the attribute model; recursively constructing a fault analysis parameter set based on the initial calculation parameter set; and finally, carrying out fault influence evaluation by using the fault analysis parameter set, so as to construct a final attribute model and complete the fault analysis of the electronic system. The application solves the problems of the prior art, such as large quantitative analysis deviation, low reliability and narrow application scene.

Description

Technical Field

[0001] This invention relates to the field of electronic system fault analysis technology, specifically a method for fault analysis of electronic systems. Background Technology

[0002] For large and complex electronic systems with highly integrated functions, flexible resource requirements, and diverse application environments, conducting analysis of all possible failure modes and their potential impacts is essential to identify design flaws and weaknesses, thereby guiding designers to take targeted improvement and compensation measures to enhance reliability. However, traditional failure analysis methods such as FMECA (Failure Mode and Effects Analysis), FTA (Fault Tree Analysis), and ETA (Event Tree Analysis) are far from perfect for failure analysis of electronic systems with multi-resource and flexible scheduling architectures. Specifically, this is manifested in the following ways: (1) FMECA analysis has a simple hierarchy and analysis factors. The traditional product-based hierarchical analysis process can only perform recursive analysis of single fault transmission relationships between upper and lower levels, and cannot analyze the network-like fault transmission situation in multi-product-layer nested electronic systems where multiple causes in the lower level lead to multiple faults in the upper level; (2) In quantitative analysis of FMECA... In the determination of the probability β of the failure mode, the method of calculation is usually set according to experience, and there is no effective calculation method. The deviation may be large. (3) The FMECA analysis of each level of product depends on the basic failure rate of the product at that level. It is not combined with the actual failure situation of the product. The failure rate generated by the failure mode is too conservative and the quantitative analysis deviation is large. (4) The basic data calculated in the quantitative analysis of FTA and ETA all need to rely on the analysis data of each level of FMECA. Due to the defects of the FMECA analysis process and the inaccuracy of the analysis data, the quantitative calculation results of FTA and ETA deviate greatly from the actual situation. The above defects make it very difficult to carry out quantitative analysis of failure in newly developed large and complex electronic systems. The analysis results usually cannot fully and accurately reflect the design reality, which has become a problem in the field of failure analysis. The failure analysis and impact assessment of large, complex, multi-functional, and multi-state electronic systems with multiple nested levels and multiple product forms is a complicated task. Due to the limitations of the above traditional failure analysis methods, the results of the work have not been integrated with the reliability prediction work, which weakens the application value of traditional failure analysis in system design. Summary of the Invention

[0003] To overcome the shortcomings of existing fault analysis techniques, this invention provides a fault analysis method for electronic systems, which solves the problems of large quantitative analysis bias, low reliability, and narrow application scenarios in existing technologies.

[0004] The technical solution adopted by the present invention to solve the above problems is:

[0005] A fault analysis method for electronic systems involves creating an initial attribute model of the electronic system using a family of attribute sets, then constructing an initial set of calculation parameters for the attribute model, recursively constructing a fault analysis parameter set based on the initial set of calculation parameters, and finally using the fault analysis parameter set to conduct a fault impact assessment, thereby constructing the final attribute model and completing the fault analysis of the electronic system.

[0006] As a preferred technical solution, the attribute model of the electronic system is represented as a family of attribute sets S, where S = {P, G, FM, U, T, C, V};

[0007] in,

[0008] P is a set of products, represented as: P = {(P...} ij X ij |1≤i≤H,1≤j≤X i}, where H is the total number of product levels, H≥2; P ij It is the product with product level i and node number j; X i X is the number of product nodes at level i; ij It is P ij Quantity;

[0009] G is the set of product functions, represented as: G = {G} ijg |1≤i≤H,1≤j≤X i , 1≤g≤Z ij}, G ijg It is P ij The g-th function, Z ij It is P ij The number of functions;

[0010] FM is a set of product failure modes, represented as: FM = {FM ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, FM ijk It is P ij The kth failure mode, Y ij It is P ij The number of failure modes;

[0011] U is the set of product failure analysis parameters, and U is represented as: U={(λ Pij , λ FMijk α ijk ,β ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, λ Pij It is P ij Failure rate; λ FMijk It's FMijk Failure rate; α ijk It's FM ijk The frequency ratio; β ijk It is by FM ijk The set of probabilities of generating higher-level effects;

[0012] T is the set of product working times, represented as: T = {t} ij |1≤i≤H,1≤j≤X i}, t ij It is P ij Product working hours;

[0013] C is the set of product hazard levels, represented as: C = {(C ij (q), C ijk (q))|1≤i≤H,1≤j≤X i , 1≤k≤Y ij ,q∈Q},C ij (q) represents P under the qth severity category. ij The degree of harm, C ijk (q) represents the severity category q under FM ijk The degree of harm, where Q is the set of severity categories defined by the system;

[0014] V is a set of product failure propagation relationships, represented as: V = {VC, VE}, where VC is the set of product failure causes, represented as VC = {VC...}. ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}; VE is the set of product failure impacts, VE is represented as VE = {VE} ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, VC ijk It's FM ijk The set of causes of failure, VE ijk It's FM ijk The set of fault impacts.

[0015] As a preferred technical solution, the steps include:

[0016] S1, Create the initial attribute model of the electronic system: Define the attribute model of the electronic system as a family of attribute sets, and determine the initial state set of the electronic system;

[0017] S2, Construct the initial set of computational parameters for the attribute model: Determine the set of parameters required for the initial computation of the multi-nested network fault propagation attribute model;

[0018] S3, Constructing the fault analysis parameter set of the attribute model: Perform recursive analysis on the fault analysis parameters of the all-electronic system hierarchical products to determine the influence probability of the multi-nested network fault propagation attribute model;

[0019] S4, Evaluate the impact of the fault: Based on steps S1 to S3, conduct a fault impact assessment, determine the hazard set, and thus construct the final attribute model S of the electronic system, completing the electronic system fault analysis.

[0020] As a preferred technical solution, step S1 includes the following steps:

[0021] S11 defines the attribute set family S of the electronic system;

[0022] S12 defines the initial state set S0 of the attribute model of the electronic system, expressed as S0 = {P, G, FM, V}. S0∪U∪T∪C=S;

[0023] S13, Construct the product set P of the attribute model of the electronic system: Define the elements of the product set P level by level from the top layer to the bottom layer of the electronic system (P... ij X ij );

[0024] S14, Construct the product function set G of the attribute model of the electronic system: Analyze the functional information of products at each level of the electronic system, and traverse to construct P. ij Functional subset G ij , Finally, a complete set of product features, G, is formed;

[0025] S15, Constructing the Product Failure Mode Set (FM) of the Attribute Model of the Electronic System: Combining the functional information of the electronic system products, analyze the failure modes of products at each level of the electronic system, and traverse to construct P. ij Failure Mode Subset FM ij , FM ij ={FM ijk |1≤k≤Y ij This ultimately forms a complete set of product failure modes (FM); among which, FM ij All Y ij Each failure mode element should be an independent event and cover a functional subset G. ij All Z ij Failure characteristics of each function;

[0026] S16, Construct the product fault propagation relationship set V of the attribute model of the electronic system: Based on the product set P obtained in step S13, the product function set G obtained in step S14, and the product fault mode set FM obtained in step S15, determine the propagation relationship of multi-nested network faults.

[0027] As a preferred technical solution, step S16 includes the following steps:

[0028] S161, Construct the product failure cause set VC: Traverse the product set P, where the product level is l (1≤l≤H) and the node number is m (P lm X lm ), constructing product P lm Fault Mode FM lmn (n is P) lm Fault mode number) Related fault cause subset VC lmn , VC lmn ={FM ijkc |1≤c≤M ijk_lmn}, FM ijkc For FM lmn The cth cause of failure, M ijk_lmn For FM lmn The number of fault causes; when l≤H-1, FM ijkc ∈FM i FM i ={FM ijk |i=l+1,1≤j≤X i , 1≤k≤Y ij}, FM i FM is a subset of FM at level i = l + 1; when l = H, it corresponds to the bottom level of the product hierarchy tree, so only FM is defined. lmn The cause of the malfunction;

[0029] S162, Construct the product failure impact set VE: Traverse the product set P, focusing on elements at product level i (1≤i≤H) and node number j (P... ij X ij ), constructing product P ij Fault Mode FM ijk The relevant fault-affected subset VE ijk , VE ijk ={FM lmnf |1≤f≤E ijk_lmn}, FM lmnf For FM ijk The f-th fault effect, E ijk_lmn For FM ijkThe number of faults affecting the number of cases; when i≥2, FM lmnf ∈FM l FM l ={FM lmn |l=i-1,1≤m≤X l , 1≤n≤Y lm}, FM l FM is a subset of FM at level l = i-1; when i = 1, this corresponds to the highest level of product composition, so only FM is defined. ijk The impact of the malfunction.

[0030] As a preferred technical solution, step S4 includes the following steps:

[0031] S21, Define the initial set of computational parameters SA: denoted as SA = {U H0 ,T};where, U H0 The initial parameter set for fault analysis is denoted as U. H0 ={(λ PHj , λ FMHjk α Hjk |1≤j≤X H , 1≤k≤Y Hj};

[0032] S22, Construct the initial parameter set U for fault analysis H0 By collecting product data, the lowest level of product P is formed. Hj All reliability statistics, determine U H0 The element (λ) PHj , λ FMHjk α Hjk The value of );

[0033] S23, Construct the product work time set T: By analyzing the product P at each level ij The working time parameter is used to determine the element t of the product working time set T. ij The value.

[0034] As a preferred technical solution, step S3 includes the following steps:

[0035] S31, Determine the recursive calculation level: Let the current level in the adjacent product level be the i-th level, and the next higher level be the l-th level; where l = i-1, and i = H during the initial recursive calculation;

[0036] S32, Construct a higher-level influence probability set β ijk According to FM ijk The relevant fault-affected subset VE ijk Constructing β ijk ={β ijk _lmnf |1≤f≤E ijk_lmn}, β ijk _ lmnf For FM ijk Faults affect FM lmnf The probability of a higher level of influence, E ijk_lmn For FM ijk The number of impacts caused by the faults;

[0037] S33, Calculate the product P of the next higher level in the adjacent hierarchy. lm Fault Mode FM lmn failure rate λ FMlmn :

[0038] According to FM lmn Related fault cause subset VC lmn When l ≤ H-1, the failure rate of all failure modes at level l is calculated using the following formula:

[0039]

[0040] in,

[0041] λ FMlmn FM lmn Failure rate;

[0042] X ijkc Indicates VC lmn FM element ijkc The corresponding number of identical products;

[0043] λ FMijkc FM ijkc Failure rate;

[0044] β ijkc_lmn FM ijkc Generate a higher-level fault mode (FM) lmn The probability of influence;

[0045] t ijkc FM ijkc The corresponding product working hours;

[0046] t lm Product P lm Working hours;

[0047] S34, Calculate the product P of the next higher level in the adjacent hierarchy. lm failure rate λ Plm :

[0048] After completing product P lm All Y lm Failure rate λ for each failure modeFMlmn After calculation, the failure rate of the product is calculated using the following formula:

[0049]

[0050] Where, λ Plm P represents lm Failure rate;

[0051] S35, Calculate the product P of the next higher level in the adjacent hierarchy. lm Failure mode frequency ratio α lmn :

[0052] Calculate the product P of the next higher level in adjacent levels using the following formula. lm Failure mode frequency ratio:

[0053]

[0054] Where, α lmn FM lmn The frequency ratio;

[0055] S36, traverse and calculate the fault analysis parameters of all products at the next higher level: Repeat steps S33 to S35 until all X values ​​at the l-th level have been traversed. l Fault analysis parameters for each product;

[0056] S37, Traverse and determine the fault analysis parameters of all product levels: Repeat steps S31 to S36 until the fault analysis parameters of all H product levels are calculated.

[0057] As a preferred technical solution, in step S3, when FM lmn ∈VE ijk When using β ijk_lmn Indicates fault mode FM ijk Generate a higher-level fault mode (FM) lmn The probability, β ijk_lmn The value range is [0, 1];

[0058] β ijk_lmn The determination method varies depending on the circumstances, as follows:

[0059] When a partial single-fault mode produces the effect of a single fault:

[0060] When FM ijk Number of failures affected by E ijk_lmn =1 and will necessarily produce FM lmn At that time, β ijk_lmn =1;

[0061] When FM ijk Number of failures affected by Eijk_lmn =1 and FM is generated occasionally. lmn At that time, β ijk_lmn The value is determined based on the actual engineering situation;

[0062] When a partial single-fault mode produces multiple fault effects:

[0063] When FM ijk Number of failures affected by E ijk_lmn When ≥2, β ijk_lmnf The value is determined in conjunction with the actual engineering situation and satisfies

[0064]

[0065] When a localized multi-fault mode produces the effect of a single fault:

[0066] When N (N≥2) failure modes work together to produce the same higher-level impact FM lmn At that time, there exists a VC lmn subset of VC lmn_KN ={FM ijkc |1≤c≤N}, and VC lmn_KN The elements in the formula satisfy the following condition: ① When the number of simultaneous occurrences is not less than K (2≤K≤N), FM is generated. lmn ② When the number of simultaneous occurrences is less than K, no FM is generated. lmn If VC lmn_KN FM element ijkc The failure rate is λ. ijk And it is a constant, with a working time of t. ijk The unit failure distribution follows an exponential distribution. The probability of failure mode impact is calculated using the following formula:

[0067]

[0068] in,

[0069] β ijkc_lmn Indicated by FM ijkc Generate FM lmn The probability of;

[0070] W c Indicates fault mode FM ijkc The corresponding product's efficiency coefficient, 0≤W c ≤1, and satisfy:

[0071]

[0072] As a preferred technical solution, in step S3, when a single fault effect occurs due to a local multi-fault mode,

[0073] If the unit is in hot standby mode, λijk For failure rate during operation, t ijk Corresponding product usage time;

[0074] If the unit is in cold standby mode, λ ijk The failure rate under non-working conditions, t ijk Corresponding product storage time.

[0075] As a preferred technical solution, step S4 includes the following steps:

[0076] S41, define the product failure modes (FM) at each level based on the system-defined severity category set Q. lmn Severity category q;

[0077] S42, confirm with VE ijk Zhongyuan FM lmnf Related FM ijk Harm level C ijk_lmnf The calculation method is as follows:

[0078] C ijk_lmnf =λ FMijk ·β ijk_lmnf ·t ijk ;

[0079] in,

[0080] β ijk_lmnf Indicates fault mode FM ijk To generate a higher level of influence FM lmn The probability of occurrence;

[0081] t ijk Indicates fault mode FM ijk The corresponding product working hours.

[0082] S43, Determine the severity C of each level of failure mode. ijk (q), calculated as follows:

[0083]

[0084] in,

[0085] C ijk (q) represents the FM under the qth severity category. ijk Harm level:

[0086] R represents VE ijk The number of elements with a medium severity category q, R≥1;

[0087] FM lmnr Indicates VE ijk Elements with a severity category of q;

[0088] C ijk_lmnr Indicates with VE ijk Zhongyuan FM lmnr Related FM ijk The degree of harm;

[0089] S44, Determine the hazard level C of each product level. ij (q), calculated as follows:

[0090]

[0091] in,

[0092] C ij (q) represents product P under the qth severity category. ij The degree of harm.

[0093] Compared with the prior art, the present invention has the following advantages:

[0094] This invention proposes a fault analysis method for electronic systems, improving the completeness and accuracy of FMECA quantitative analysis for such fault transmission systems. It applies FMECA quantitative analysis results and enumeration concepts to reliability prediction work, enriching the quantitative reliability prediction methods. It can be widely applied in the field of reliability engineering (such as FTA analysis, hazard analysis, etc.), improving the comprehensiveness, sufficiency, and effectiveness of reliability analysis work. It is suitable for large and complex electronic systems with highly integrated functions, flexible resource requirements, and diverse application environments. Attached Figure Description

[0095] Figure 1 A schematic diagram illustrating the attribute model and construction method of an electronic system;

[0096] Figure 2 Create a flowchart for the initial property model of the electronic system;

[0097] Figure 3 A schematic diagram of the product composition of an electronic system;

[0098] Figure 4 A schematic diagram illustrating the functional information of products at various levels of an electronic system;

[0099] Figure 5 A schematic diagram of failure modes for products at various levels of an electronic system;

[0100] Figure 6 A schematic diagram of the failure mode relationships under common-cause faults in electronic systems;

[0101] Figure 7 This is a schematic diagram of the initial property model of the electronic system;

[0102] Figure 8A flowchart for constructing the initial set of calculation parameters for the attribute model of the electronic system;

[0103] Figure 9 A flowchart for constructing the fault analysis parameter set for the attribute model of electronic systems;

[0104] Figure 10 A schematic diagram of the attribute transfer model for the effects of multiple faults;

[0105] Figure 11 A schematic diagram of the attribute propagation model for multi-unit voting faults;

[0106] Figure 12 Schematic diagram of the fault propagation attribute model for two parallel units

[0107] Figure 13 This is a schematic diagram of a multi-factor fault propagation attribute model. Detailed Implementation

[0108] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0109] Example 1

[0110] like Figures 1 to 13 As shown, this invention provides a fault analysis method for electronic systems. Taking the electronic system as the analysis object, it employs an attribute model for quantitative FMECA analysis, making it particularly suitable for large, complex, multi-functional, and multi-state electronic systems with multiple nested levels and various product forms. It is especially applicable to multi-nested network fault propagation modes. In this invention, the explanation of multi-nested network fault propagation modes is as follows: multi-nesting refers to fault modes with multiple levels; network refers to the intersecting propagation relationships of multiple fault causes and multiple fault effects; and fault propagation refers to the process of fault propagation from the bottom layer to the top layer.

[0111] This invention proposes a fault analysis method for electronic systems, which improves the completeness and accuracy of FMECA quantitative analysis of electronic systems. It applies FMECA quantitative analysis results and enumeration concepts to reliability prediction work, thus enriching the methods for quantitative reliability prediction.

[0112] This invention discloses a fault analysis method for electronic systems, applicable to large and complex electronic systems with highly integrated functions, flexible resource requirements, and diverse application environments.

[0113] Includes the following steps:

[0114] (I) Define the attribute model of the electronic system as a family of attribute sets. Based on the composition, function, fault information, and fault information transmission relationship of the analysis object, the function-fault enumeration method is used to determine the initial state set of the electronic system and create the initial attribute model of the electronic system. (II) Combining reliability engineering data, based on the failure rate of the lowest level product, the frequency ratio of failure modes, the failure rate of failure modes, and the working time of each level product, the initial calculation parameter set of the multi-nested network fault transmission attribute model is constructed. (III) Using the decomposition method and the recursive method, the fault transmission relationship attribute model is simplified, and the fault analysis parameters of all system-level products are recursively determined to construct the fault analysis parameter set of the attribute model. (IV) Complete the fault impact evaluation of each level product, and finally form the attribute model of the electronic system. This invention not only improves the completeness and accuracy of FMECA quantitative analysis of electronic systems, but also applies the FMECA quantitative analysis results and the function-fault enumeration concept to reliability prediction work, enriching the quantitative reliability prediction method and improving the application value of the work results.

[0115] To facilitate understanding, this embodiment presents a specific example of the implementation of the attribute model and construction method applicable to electronic systems, as follows:

[0116] 1) Attribute Models and Construction Methods of Electronic Systems

[0117] See Figure 1 . Figure 1 The paper presents a schematic diagram of the attribute model and construction method of an electronic system. The method includes four steps: ① creating an initial attribute model of the electronic system → ② constructing an initial set of calculation parameters for the attribute model → ③ constructing a set of fault analysis parameters for the attribute model → ④ evaluating the impact of faults to form the attribute model of the electronic system.

[0118] 2) Create the initial property model of the electronic system

[0119] See Figure 2 . Figure 2 The paper presents the process for creating the initial attribute model of an electronic system, which includes six steps: ① Define the attribute set family S of the electronic system → ② Define the initial state set family S0 of the attribute model → ③ Construct the product set P of the attribute model → ④ Construct the product function set G of the attribute model → ⑤ Construct the product failure mode set FM of the attribute model → ⑥ Construct the product failure propagation relationship set V of the attribute model.

[0120] (1) Define the family S of electronic systems

[0121] The electronic system is represented as a family of attribute sets S, S = {P, G, FM, U, T, C, V}.

[0122] in,

[0123] P is a set of products, represented as: P = {(P ij X ij |1≤i≤H,1≤j≤X i}, where H is the total number of product levels, H≥2; P ij It is the product with product level i and node number j; X i X is the number of product nodes at level i; ij It is P ij Quantity;

[0124] G is the set of product features, represented as: G = {G} ijg |1≤i≤H,1≤j≤X i , 1≤g≤Z ij}, where G ijg It is P ij The g-th function, Z ij It is P ij The number of functions;

[0125] FM is a set of product failure modes, represented as: FM = {FM ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, where FM ijk It is P ij The kth failure mode, Y ij It is P ij The number of failure modes;

[0126] U is the set of product failure analysis parameters, represented as: U={(λ Pij , λ FMijk α ijk ,β ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, where λ Pij It is P ij Failure rate; λ FMijk It's FM ijk Failure rate; α ijk It's FM ijk The frequency ratio; β ijk It is by FM ijk The set of probabilities of generating higher-level effects;

[0127] T is the set of product working times, represented as: T = {t} ij |1≤i≤H,1≤j≤X i}, where t ij It is P ij Working hours;

[0128] C is the set of product hazard levels, represented as: C = {(C ij (q), C ijk (q))|1≤i≤H,1≤j≤X i , 1≤k≤Y ij , q∈Q}, where C ij (q) represents P under the qth severity category. ij The degree of harm, C ijk (q) represents the severity category q under FM ijk The degree of harm, where Q is the set of severity categories defined by the system;

[0129] V is a family of product failure propagation relationships, represented as: V = {VC, VE}, where VC is the set of product failure causes, represented as VC = {VC... ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}; VE is the set of product failure impacts, represented as VE = {VE} ijk |1≤i≤H,1≤j≤X i , 1≤k≤Y ij}, VC ijk It's FM ijk The set of causes of failure, VE ijk It's FM ijk The set of fault impacts.

[0130] (2) Define the initial state set family S0 of the attribute model.

[0131] Based on the requirements of initial fault analysis, a state set family S0 is defined for the initial fault analysis, expressed as:

[0132] S0={P, G, FM, V}, S0∈S, S0∪U∪T∪C=S.

[0133] (3) Construct the product set P of the attribute model

[0134] From the top level of the electronic system being analyzed to its bottom level, the elements of the product set P are defined level by level (P...). ij X ij See also Figure 3 , Figure 3 The document outlines the product composition of an electronic system at level H (H≥2). Following the hierarchical relationship of the system, it describes each level from top to bottom, generally categorized into system level, subsystem level, subsystem level, equipment level, component level, module level, functional circuit level, and device level. Each product node should include at least the product name, quantity, and product level information. Figure 3In this context, level 1 is the top level (i.e., the highest product level), and level H is the bottom level (i.e., the lowest product level); (P lm X lm ) is an element in set P with product level l (1≤l<i=H) and node number m. Product P lm By X iA Product P iA ... X ij Product P ij ... X iB Product P iB Composition (1≤A≤j≤B).

[0135] (4) Construct the product function set G of the attribute model

[0136] See Figure 4 Based on the product composition of the electronic system, analyze the functional information of each level of the product, and traverse and construct P. ij Functional subset G ij G ij ∈G, G ij ={G ijg |1≤g≤Z ij Finally, a complete set of product functions G is formed.

[0137] (5) Construct the product failure mode set FM of the attribute model

[0138] See Figure 5 Based on the product function set G, analyze the failure modes of products at each level of the electronic system, and construct P by traversal. ij Failure Mode Subset FM ij FM ij ∈FM, FM ij ={FM ijk |1≤k≤Y ij This ultimately forms a complete set of product failure modes (FM); among which, FM ij All Y ij Each failure mode element should be an independent event and cover a functional subset G. ij All Z ij Failure characteristics of each function.

[0139] See Figure 6 For product P ij If fault mode FM exists ija Characterizing the product's function G ija Failure Mode FM ijb Characterizing the product's function G ijb Failure occurs when a common cause fault causes G to fail. ija and G ijbWhen both fail simultaneously, the fault mode FM is defined. ijc To characterize.

[0140] When the lowest level is components, GJB299 can be used to obtain the product's fault mode information.

[0141] (6) Construct the set of product fault propagation relationships V of the attribute model

[0142] See Figure 7 Based on the product set P, product function set G, and product failure mode set FM obtained from the aforementioned analysis, the transmission relationship of multi-nested network failures is determined.

[0143] First, analyze the causes of product failures and construct a set of product failure causes (VC). Then, iterate through and construct the set of product failure causes (P). lm Fault Mode FM (Product level is l, l = i-1, node number is m) lmn (n is P) lm Fault mode number) Related fault cause subset VC lmn VC lmn ∈VC,VC lmn ={FM ijkc |1≤c≤M ijk_lmn}, FM ijkc For FM lmn The cth cause of failure, M ijk_lmn For FM lmn The number of fault causes; when l≤H-1, FM ijkc ∈FM i FM i ={FM ijk |i=l+1,1≤j≤X i , 1≤k≤Y ij}, FM i FM is a subset of FM at level i = l + 1; when l = H, it corresponds to the lowest level of product composition, so only FM is defined. lmn The cause of the malfunction.

[0144] Then, analyze the impact of product failures and construct the product failure impact set VE. Iterate through and construct the set of failure impacts for product P. ij Fault Mode FM ijk The relevant fault-affected subset VE ijk VE ijk ∈VE,VE ijk ={FM lmnf |1≤f≤E ijk_lmn}, FM lmnf For FM ijk The f-th fault effect, E ijk_lmn For FM ijkThe number of faults affecting the number of cases; when i≥2, FM lmnf ∈FM l FM l ={FM lmn |l=i-1,1≤m≤X l , 1≤n≤Y lm}, FM l FM is a subset of FM at level l = i-1; when i = 1, this corresponds to the highest level of product composition, so only FM is defined. ijk The impact of the malfunction.

[0145] 3) Construct the initial set of calculation parameters for the attribute model

[0146] See Figure 8 The figure illustrates the process of constructing the initial calculation parameter set for the electronic system attribute model, including: defining the initial calculation parameter set SA and constructing the initial parameter set U for fault analysis. H0 Construct the product working time set T.

[0147] (1) Define the initial calculation parameter set SA: expressed as SA = {U H0 ,T}, where,U H0 The initial parameter set for fault analysis is denoted as U. H0 ={(λ PHj , λ FMHjk α Hjk |1≤j≤X H ,

[0148] 1≤k≤Y Hj};

[0149] (2) Determine the quantitative reliability parameters of the lowest-level products and construct the initial parameter set U for failure analysis. H0

[0150] By collecting product composition data at the lowest level (i.e., level H), product P... Hj Use all reliability statistics to determine the initial parameter set U for fault analysis. H 0 The element (λ) PHj , λ FMHjk α Hjk The values ​​include product failure rate, failure mode frequency ratio, and failure mode failure rate.

[0151] When the lowest level is components, GJB299 can be used to obtain the product's failure rate and failure mode frequency ratio. When the above parameters of the lowest level products cannot be obtained through GJB299, they can be determined by combining fault statistics data from engineering applications.

[0152] For a specific product P at the lowest level HjLet Y be the number of failure modes for this product. Hj Then the product failure rate λ PNj Fault mode frequency ratio α Hjk and failure mode failure rate λ FMHjk The three parameters satisfy the following relationship:

[0153] λ FMHjk =α Hjk ·λ FMHjk (1)

[0154]

[0155] (3) Construct the product working time set T

[0156] By analyzing product P at each level ij The working time parameter is used to determine the element t of the product working time set T. ij The value should be t, and it should have t. ij ≤t 11 (t 11 (This indicates the specified system operating time).

[0157] 4) Constructing the fault analysis parameter set for the attribute model

[0158] See Figure 9 From bottom to top, the fault analysis parameters of all system-wide products are recursively determined through the fault propagation relationship between adjacent product levels, as follows:

[0159] (1) Determine the recursive calculation level

[0160] Let the current level of the adjacent product be the i-th level, and the next level up be the l-th level; where l = i-1, and i = H during the initial recursive calculation.

[0161] (2) Construct a higher-level set of influence probability β ijk

[0162] According to FM ijk The relevant fault-affected subset VE ijk Constructing β ijk ={β ijk _ lmnf |1≤f≤E ijk_lmn}, β ijk _ lmnf For FM ijk Faults affect FM lmnf The probability of a higher level of influence, E ijk_lmn For FM ijk The number of impacts caused by the faults;

[0163] When i≥2, FM exists. lmn ∈VEijk Using β ijk_lmn Indicates fault mode FM ijk Generate a higher-level fault mode (FM) lmn The probability, β ijk_lmn The value range is [0, 1].

[0164] For a multi-nested network fault propagation attribute model, a decomposition method is used to simplify the complex fault propagation relationship into three cases based on local fault modes: local single fault mode produces single fault effect, local single fault mode produces multiple fault effect, and local multi fault mode produces single fault effect.

[0165] ① Local single-fault mode produces single-fault effects

[0166] When FM ijk Number of failures affected by E ijk_lmn =1 and will necessarily produce FM lmn At that time, β ijk_lmn =1;

[0167] When FM ijk Number of failures affected by E ijk_lmn =1 and FM is generated occasionally. lmn At that time, β ijk_lmn The value is determined based on the actual situation of the project.

[0168] ② Local single-fault mode generates multiple fault effects

[0169] See Figure 10 When a certain fault mode FM ijk When multiple faults are involved, record the fault mode FM. ijk The number of higher-level influences generated is E ijk_lmn ,β ijk_lmnf Let f be the probability of the f-th fault affecting FM. ijk Number of failures affected by E ijk_lmn When ≥2, β ijk_lmnf The value is determined in conjunction with the actual engineering situation and satisfies

[0170]

[0171] ③ Localized multi-fault modes produce the effects of single faults

[0172] See Figure 11 When N (N≥2) failure modes work together to produce the same higher-level effect FM lmn At that time, there exists a VC lmn subset of VC lmn_KN ={FM ijkc |1≤c≤N}, and VC lmn_KNThe elements in the formula satisfy the following condition: a) When the number of simultaneous occurrences is not less than K (2≤K≤N), an FM is generated. lmn b) When the number of simultaneous occurrences is less than K, no FM is generated. lmn If VC lmn_KN FM element ijkc The failure rate is λ. ijk And it is a constant, with a working time of t. ijk The unit failure distribution follows an exponential distribution. The probability of failure mode impact is calculated using the following formula:

[0173]

[0174] in,

[0175] β ijkc_lmn Indicated by FM ijkc Generate FM lmn The probability of;

[0176] W c Indicates fault mode FM ijkc The corresponding product's efficiency coefficient, 0≤W c ≤1, and satisfy:

[0177]

[0178] For a multi-unit voting attribute model and a fault propagation attribute model, when the unit's operating state is hot standby, λ ijk For failure rate during operation, t ijk Corresponding product usage time; when the unit is in cold standby mode, λ ijk The failure rate under non-working conditions, t ijk Corresponding product storage time.

[0179] Specifically, for the 2-unit parallel fault propagation attribute model, refer to... Figure 12 Two fault modes FM ijk1 and FM ijk2 Both must occur together for a higher-level failure mode (FM) to be generated. lmn It occurs, and assuming the failure rate is constant, let β ijk2_lmn =0, then the fault mode is FM ijk1 Its next-level impact factor β ijk1_lmn for

[0180]

[0181] (3) Calculate the product P of the next higher level in the adjacent level. lm Fault Mode FM lmn failure rate λ FMlmn

[0182] FMlmn In this regard, the failure rate should be determined by the failure rate of all its causes of failure and the corresponding probability contribution of the failure impact, combined with the product P. lm The relative working time of each component product is determined.

[0183] See Figure 13 According to FM lmn Related fault cause subset VC lmn When l ≤ H-1, the failure rate of all failure modes at level l is calculated using the following formula:

[0184]

[0185] in,

[0186] λ FMlmn FM lmn Failure rate;

[0187] X ijkc Indicates VC lmn FM element ijkc The corresponding number of identical products;

[0188] λ FMijkc FM ijkc Failure rate;

[0189] β ijkc_lmn FM ijkc Generate a higher-level fault mode (FM) lmn The probability of influence;

[0190] t ijkc FM ijkc The corresponding product working hours;

[0191] t lm Product P lm Working hours;

[0192] (4) Calculate the product P of the next higher level in the adjacent level. lm failure rate λ Plm .

[0193] After completing product P lm All Y lm Failure rate λ for each failure mode FMlmn After calculation, the failure rate of the product is calculated using the following formula:

[0194]

[0195] Where, λ Plm P represents lm Failure rate;

[0196] (5) Calculate the product P of the next higher level in the adjacent level. lm Failure mode frequency ratio α lmn :

[0197] Calculate the product P of the next higher level in adjacent levels using the following formula. lm Failure mode frequency ratio:

[0198]

[0199] Where, α lmn FM lmn The frequency ratio.

[0200] (6) Repeat (3) to (5) until all X in the l-th level have been traversed. l Fault analysis parameters for each product.

[0201] (7) Repeat (1) to (6) until the fault analysis parameters of all H-level products are calculated.

[0202] 5) Evaluate the impact of the fault and form an attribute model.

[0203] (1) In the fault propagation attribute model, the product fault modes FM at each level of the system are defined based on the product hazard set Q. lmn The severity category q.

[0204] (2) Determine the relationship with VE ijk Zhongyuan FM lmnf Related FM ijk Harm level C ijk_lmnf The calculation method is as follows:

[0205] C ijk_lmnf =λ FMijk ·β ijk_lmnf ·t ijk (9)

[0206] in,

[0207] β ijk_lmnf Indicates fault mode FM ijk To generate a higher level of influence FM lmn The probability of occurrence;

[0208] t ijk Indicates fault mode FM ijk The corresponding product working hours.

[0209] Determine the severity C of each level of failure mode ijk (q), calculated as follows:

[0210]

[0211] in,

[0212] C ijk (q) represents the FM under the qth severity category. ijk Harm level:

[0213] R represents VE ijk The number of elements with a medium severity category q, R≥1;

[0214] FM lmnr Indicates VE ijk Elements with a severity category of q;

[0215] C ijk_lmnr Indicates with VE ijk Zhongyuan FM lmnr Related FM ijk The degree of harm.

[0216] (3) Determine the hazard level C of each product level ij (q), calculated as follows:

[0217]

[0218] in,

[0219] C ij (q) represents product P under the qth severity category. ij The degree of harm.

[0220] 6) Application and expansion of attribute models and construction methods for electronic systems

[0221] The failure rate of each level of electronic system products calculated by equation (7) in the method provided in 4) can be regarded as the reliability prediction value of each level of products obtained by the function-fault enumeration method. It can be widely used in the field of reliability engineering (such as FTA analysis, ETA analysis, etc.), which improves the comprehensiveness, sufficiency and effectiveness of reliability analysis work.

[0222] The analytical process and method proposed in this invention not only provide product designers with a complete and accurate method for conducting quantitative FMECA analysis of large and complex electronic systems with highly integrated functions, flexible resource requirements, and diverse application environments, but also provide a solution for implementing quantitative FMECA analysis of large and complex electronic systems with highly integrated functions, flexible resource requirements, and diverse application environments using computer-readable program code.

[0223] As described above, the present invention can be implemented well.

[0224] All features disclosed in all embodiments of this specification, or steps in all methods or processes implied in the disclosure, may be combined and / or extended or replaced in any way, except for mutually exclusive features and / or steps.

[0225] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Based on the technical essence of the present invention, any simple modifications, equivalent substitutions, and improvements made to the above embodiments within the spirit and principles of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A fault analysis method for an electronic system, characterized in that, Includes the following steps: S1, Create the initial attribute model of the electronic system: Define the attribute model of the electronic system as a family of attribute sets, determine the initial state set of the electronic system, and determine the transmission relationship of multi-nested network faults; S2, Construct the initial set of calculation parameters for the attribute model: Determine the set of parameters required for the initial calculation of the electronic system attribute model; S3, Construct the fault analysis parameter set of the attribute model: Perform recursive analysis on the fault analysis parameters of the full electronic system level product to determine the influence probability of the electronic system attribute model; S4, Evaluate the impact of the fault: Based on steps S1 to S3, conduct a fault impact assessment, determine the hazard set, and thus construct the final attribute model of the electronic system. S Complete electronic system fault analysis; In step S1, the attribute model of the electronic system is represented as a family of attribute sets. S , S ={ P , G , FM , U , T , C , V },in: P It is a collection of products. P Represented as: P ={( P ij , X ij )|1≤ i ≤ H , 1≤ j ≤ X i }, H It is the total number of product levels. H ≥2; P ij The product level is i、 Node number is j Products ；X i It is the first i Number of product nodes in a hierarchical structure; X ij yes P ij Quantity; G It is a collection of product features. G Represented as: G ={ G ijg |1≤ i ≤ H ,1≤ j ≤ X i ,1≤ g ≤ Z ij }, G ijg yes P ij The g One function, Z ij yes P ij The number of functions; FM It is a set of product failure modes. FM Represented as: FM ={ FM ijk |1≤ i ≤ H ,1≤ j ≤ X i ,1≤ k ≤ Y ij }, FM ijk yes P ij The k One failure mode, Y ij yes P ij The number of failure modes; U It is a set of product failure analysis parameters. U Represented as: U ={( λ Pij , λ FMijk , α ijk , β ijk )| 1≤ i ≤ H , 1≤ j ≤ X i ,1≤ k ≤ Y ij }, λ Pij yes P ij Failure rate; λ FMijk yes FM ijk Failure rate; α ijk yes FM ijk The frequency ratio; β ijk It is by FM ijk The set of probabilities of generating higher-level effects; T It is the collection of product working hours. T Represented as: T ={ t ij |1≤ i ≤ H ,1≤ j ≤ X i }, t ij It is P ij Product working hours; C It is a collection of product hazard levels. C Represented as: C ={( C ij ( q ), C ijk ( q ))|1≤ i ≤ H ,1≤ j ≤ X i ,1≤ k ≤ Y ij , q ∈ Q }, C ij ( q ) is the first q Severity category P ij The degree of harm, C ijk ( q ) is the first q Severity category FM ijk The degree of harm, Q A set of severity categories defined for the system; V It is a set of product failure propagation relationships. V Represented as: V ={ VC , VE }, VC is A collection of reasons for product failure. VC Represented as VC ={ VC ijk | 1≤ i ≤ H ,1≤ j ≤ X i , 1≤ k ≤ Y ij }; VE It is the set of product failure impacts. VE Represented as VE ={ VE ijk | 1≤ i ≤ H ,1≤ j ≤ X i , 1≤ k ≤ Y ij }, VC ijk yes FM ijk The set of causes of failure, VE ijk yes FM ijk The set of failure impacts; The initial state set of the electronic system is represented as a family of initial state sets. S 0, S 0={ P , G , FM , V }, S 0 S , S 0∪ U ∪ T ∪ C = S ; Determining the propagation relationship of multi-nested mesh faults includes: based on the product set P Product Function Set G and product failure mode set FM Construct a family of product failure propagation relationships for the attribute model of an electronic system. V ; Step S2 includes the following steps: S21, Define the initial set of computational parameters SA : Represented as SA ={ U H0 , T };in, U H0 The initial parameter set for fault analysis is denoted as: U H0 ={( λ PHj , λ FMHjk , α Hjk )|1≤ j ≤ X H ,1≤ k ≤ Y Hj }; S22, Construct the initial parameter set for fault analysis U H0 The lowest level of products is composed of collected products. P Hj All reliability statistics, determine the elements λ PHj , λ FMHjk and α Hjk The value; S23, Construct a product work time set T By analyzing products at various levels P ij Working time parameters determine the product working time set. T elements t ij The value; Step S3 includes the following steps: S31, Determine the recursive calculation level: Let the current level in the adjacent level products be the [number]. i The next level is the first. l Hierarchy; among which, l = i -1, during the initial recursive calculation i=H ; S32, Construct a higher-level set of influence probabilities β ijk According to FM ijk Related fault impact subset VE ijk Build β ijk ={ β ijk _ lmnf |1≤ f ≤ E ijk_lmn }, β ijk _ lmnf for FM ijk Causes malfunction FM lmnf The probability of a higher level of influence. E ijk_lmn for FM ijk The number of impacts caused by the faults; S33, Calculate the product of the next higher level in the adjacent hierarchy. P lm Failure modes FM lmn failure rate λ FMlmn ; S34, Calculate the product of the next higher level in the adjacent hierarchy. P lm failure rate λ Plm : After completing the product P lm All Y lm Failure rate of each failure mode λ FMlmn After calculation, the failure rate of the product is calculated using the following formula: ； in, λ Plm express P lm Failure rate; S35, Calculate the product of the next higher level in the adjacent hierarchy. P lm Failure mode frequency ratio α lmn : The following formula is used to iterate and calculate the products of the next higher level in adjacent levels. P lm Failure mode frequency ratio: in, α lmn express FM lmn The frequency ratio; S36, traverse and calculate the fault analysis parameters of all products at the next higher level: repeat steps S33 to S35 until the traversal is complete. l All levels X l Fault analysis parameters for each product; S37, Traverse and determine the fault analysis parameters for all product levels: Repeat steps S31 to S36 until the traversal is complete and all calculations are finished. H Fault analysis parameters for products at each level; In step S3, when FM lmn ∈ VE ijk At that time, adopt β ijk_lmn Indicates the fault mode FM ijk Generate a higher-level failure mode FM lmn The probability, β ijk_lmn The value range is [0, 1]; in, β ijk_lmn The determination method varies depending on the circumstances, as follows: When a partial single-fault mode produces the effect of a single fault: when FM ijk Number of failures E ijk_lmn =1 and will necessarily produce FM lmn hour, β ijk_lmn =1; when FM ijk Number of failures E ijk_lmn =1 and occurs occasionally FM lmn hour, β ijk_lmn The value is determined based on the actual engineering situation; When a partial single-fault mode produces multiple fault effects: when FM ijk Number of failures E ijk_lmn When ≥2, β ijk_lmnf The value is determined in conjunction with the actual engineering situation and satisfies ； When a localized multi-fault mode produces the effect of a single fault: when N The combined effects of multiple failure modes produce the same higher-level impact. FM lmn At that time, there exists a VC lmn subset of VC lmn_KN ={ FM ijkc |1≤ c ≤ N },and VC lmn_KN The elements in the set satisfy: ① The number of simultaneous occurrences is not less than K At that time, it produces FM lmn , where 2≤ K ≤ N , N ≥2; ② The number of simultaneous occurrences is less than K At that time, no production FM lmn ;like VC lmn_KN elements in FM ijkc The failure rate is 100%. λ ijk And it is a constant, the working time is t ijk The unit failure distribution follows an exponential distribution. The probability of failure mode impact is calculated using the following formula: ； in, β ijkc_lmn Indicates by FM ijkc produce FM lmn The probability of; W c Indicates the fault mode FM ijkc The corresponding product's effect coefficient, 0≤ W c ≤1, and satisfy: 。 2. The fault analysis method for an electronic system according to claim 1, characterized in that, Step S1 includes the following steps: S11, Defines the attribute set family of the electronic system. S ; S12 defines the initial state set family of the attribute model of the electronic system. S 0; S13, Product set for constructing attribute models of electronic systems P Define the product set step by step, from the top level to the bottom level of the electronic system. P elements ( P ij , X ij ); S14, Product Function Set for Constructing Attribute Models of Electronic Systems G Analyze the functional information of products at each level of the electronic system, and traverse the structure. P ij Functional subset G ij , G ij G , G ij ={ G ijg |1≤ g ≤ Z ij Finally, a complete set of product features is formed. G ; S15, Constructing the set of product failure modes for the attribute model of the electronic system. FM By combining functional information of electronic system products, analyze the failure modes of products at each level of the electronic system, and traverse the structure. P ij Failure mode subset FM ij , FM ij FM , FM ij ={ FM ijk |1≤ k ≤ Y ij This ultimately forms a complete set of product failure modes. FM ;in, FM ij All Y ij Each failure mode element should be an independent event and cover a subset of functions. G ij all Z ij Failure characteristics of each function; S16, Constructing a family of product fault propagation relationships for the attribute model of the electronic system. V The product set obtained from step S13 P The product function set obtained in step S14 G The product failure mode set obtained in step S15 FM To determine the propagation relationship of multi-nested mesh faults.

3. The fault analysis method for an electronic system according to claim 2, characterized in that, Step S16 includes the following steps: S161, Construct a set of product failure causes VC Traverse the product collection P The product level is l The node number is m elements P lm and X lm Building and Products P lm Failure modes FM lmn Related subset of fault causes VC lmn , VC lmn VC , VC lmn ={ FM ijkc |1≤ c ≤ M ijk_lmn }, FM ijkc for FM lmn The c One cause of the malfunction, M ijk_lmn for FM lmn The number of causes of failure n for P lm Fault mode number, 1≤ l ≤ H ;when l ≤ H At -1, FM ijkc ∈ FM i , FM i ={ FM ijk | i = l +1, 1≤ j ≤ X i ,1≤ k ≤ Y ij }, FM i for FM In the i = l +1 level subset; when l = H At this point, which corresponds to the bottom layer of the product hierarchy tree, only the definition is needed. FM lmn The cause of the malfunction; S162, Construct the product failure impact set VE Traverse the product collection P The product level is i The node number is j elements P ij and X ij Building and Products P ij Failure modes FM ijk Related fault impact subset VE ijk , VE ijk VE , VE ijk ={ FM lmnf |1≤ f ≤ E ijk_lmn }, FM lmnf for FM ijk The first f The impact of the fault E ijk_lmn for FM ijk The number of faults affecting, 1≤ i ≤ H ;when i When ≥2, FM lmnf ∈ FM l , FM l ={ FM lmn | l = i -1, 1≤ m ≤ X l ,1≤ n ≤ Y lm }, FM l for FM In the l = i -1 level subset; when i When =1, this corresponds to the highest level of product composition, therefore only one level is defined. FM ijk The impact of the malfunction.

4. The fault analysis method for an electronic system according to claim 3, characterized in that, In step S33, calculate the product of the next higher level in the adjacent layers. P lm Failure modes FM lmn failure rate λ FMlmn ,include: According to FM lmn Related subset of fault causes VC lmn ,when l ≤ H When -1, use the following formula to iterate and calculate the th... l Failure rate of all failure modes at each level: ； in, λ FMlmn express FM lmn Failure rate; X ijkc express VC lmn elements in FM ijkc The corresponding number of identical products; λ FMijkc express FM ijkc Failure rate; β ijkc_lmn express FM ijkc Generate a higher-level failure mode FM lmn The probability of influence; t ijkc express FM ijkc The corresponding product working hours; t lm Indicates product P lm Working hours.

5. The fault analysis method for an electronic system according to claim 4, characterized in that, In step S3, when a local multi-fault mode produces the effect of a single fault, If the unit is in hot standby mode. λ ijk Failure rate during work operation t ijk Corresponding product usage time; If the unit is in cold standby mode. λ ijk Failure rate in non-working state t ijk Corresponding product storage time.

6. A fault analysis method for an electronic system according to any one of claims 1 to 5, characterized in that, Step S4 includes the following steps: S41, based on the system-defined set of severity categories Q Define product failure modes at each level FM lmn Severity category q ; S42, confirm with VE ijk medium elements FM lmnf Related FM ijk degree of harm C ijk_lmnf The calculation method is as follows: ； in, β ijk_lmnf Indicates the fault mode FM ijk To produce a higher level of influence FM lmn The probability of occurrence; t ijk Indicates the fault mode FM ijk The corresponding product working hours; S43, Determine the severity of each failure mode. C ijk ( q The calculation method is as follows: ； in, C ijk ( q ) indicates the first q Severity category FM ijk Harm level: R express VE ijk Medium severity category q The number of elements, R ≥1; FM lmnr express VE ijk Medium severity category q Element; C ijk_lmnr Indicates and VE ijk medium elements FM lmnr Related FM ijk The degree of harm; S44, Determine the hazard level of products at each level C ij ( q The calculation method is as follows: ； in, C ij ( q ) indicates the first q Products under the severity category P ij The degree of harm.

Images