Fault Diagnosis Method for the Control System of Electronic Equipment Based on Fault Tree

Abstract

The present invention relates to a fault diagnosis method for an electronic device control system based on a fault tree. By introducing the fault tree method and utilizing the probabilities of occurrence of various events and various importance degrees in the fault tree model, fault diagnosis of the electronic device control system is realized, which can effectively improve the current situation where there is no mature fault diagnosis method for the electronic device control system; by introducing the method of improved linear weighted summation, the probabilities of occurrence of various faults are effectively and accurately sorted, and according to the sorting results and in combination with the set fault isolation threshold, possible faults are isolated, which can effectively solve the problem of misoperation of fault isolation when the fault diagnosis result is inaccurate.

Description

Technical Field

[0001] The present invention relates to a device fault diagnosis technology, and particularly to a fault diagnosis method for an electronic device control system based on a fault tree. Background Art

[0002] The control system is the core component of an electronic device. Once a fault occurs in the control system, it often leads to the shutdown of the entire electronic device. Studying its fault diagnosis and fault isolation methods is of great significance for improving the reliability of the electronic device. At present, the research in the field of fault diagnosis has been relatively in-depth, and diagnostic algorithms based on signal processing, machine learning, deep learning, etc. are constantly developing, such as wavelet transform, support vector machine, neural network and other methods, which are involved in the fields of machinery, electronic and electrical, chemistry, communication network, etc. However, most of the current research in the field of equipment health management focuses on the fault diagnosis of mechanical equipment, and there is less research on the fault diagnosis of electronic devices. Moreover, since the common device state data in electronic devices are voltage, current, temperature, pulse signals, while the common device states in mechanical equipment are vibration, acceleration and other data, the device fault diagnosis models established in the field of mechanical equipment cannot be directly and effectively applied to the field of electronic device fault diagnosis. Therefore, establishing a fault diagnosis model for electronic devices is a current research hotspot. For example, in Reference [1], the analytic hierarchy process is used to establish a functional and structural hierarchy model of the CPU board in an electronic equipment, and combined with neural network technology, automatic fault reasoning and positioning are realized.

[0003] However, after conducting an extensive literature review, it was found that although there is already a certain research foundation for the fault diagnosis of electronic device control systems, the current research is still very lacking. There is a lack of typical fault diagnosis methods for electronic device control systems. Therefore, it is very necessary to research and propose targeted and effective fault diagnosis algorithms and frameworks for electronic device control systems. Looking at the current fields of system reliability assessment and fault diagnosis, the model structure of the fault tree analysis method is easy to understand and has been widely applied. In reference [2], the reliability of the office building fire point detection system was studied through fault tree analysis. In reference [3], aiming at the fault diagnosis problem of electric vehicle charging devices, a fault diagnosis framework was constructed. By establishing a fault tree model, the faults of the charging devices were accurately located and quickly solved. In reference [4], aiming at the reliability assessment problem of network systems, a reliability model for sequential fault handling was established and studied based on the fault tree analysis method. In practical applications, since it is difficult to obtain sufficient probability data of the bottom events of the fault tree, the probabilities of the bottom events in the fault tree are usually estimated according to the opinions of experts or engineers. In reference [5], similarity aggregation and fuzzy set theory were combined to comprehensively process the opinions of different experts, so as to obtain the occurrence probabilities of the bottom events in the fault tree and complete the safety assessment of natural gas storage tanks. The fault tree has been developed relatively early in fault diagnosis and is becoming mature. Cooperating with other algorithms, it can be effectively applied to the problem of equipment fault diagnosis. Considering the current situation that there is no mature fault diagnosis method for electronic device control systems, researching the fault diagnosis of electronic device control systems based on the fault tree method has certain research significance.

[0004] In addition, simply conducting fault diagnosis cannot solve the fundamental problem. It is also necessary to isolate the fault source and recover the fault after a fault occurs in order to achieve the purpose of maintaining the system. Fault isolation means that when designing the system, various fault situations should be considered as much as possible. Taking fault isolation measures can control the fault range locally, prevent the expansion of the fault range, thus increasing the impact on the availability of the upper-level system. And when a fault occurs, the fault source can be quickly located, providing necessary conditions for subsequent fault recovery. For circuit faults, components such as circuit breakers and relay protectors can be set to isolate the fault point from the system [6][7], or switch to redundant devices to continue working, and then reconnect to the system after the fault is recovered. When a fault occurs in an electronic device, the importance of each fault is different and there is often a certain correlation between various faults. If a possible fault cause judged by simple fault diagnosis is used for fault isolation, the effect of fault isolation is often not achieved when the fault diagnosis result is inaccurate. Therefore, when a fault occurs, how to effectively and accurately rank the occurrence probabilities of each fault and isolate the possible faults according to the ranking results in combination with the set fault isolation threshold has certain research significance.

[0005] Literature [1] Duan Xiusheng, Wang Zhiqiang, Cheng Yuanzeng. Research on Fault Diagnosis Technology of CPU Board Based on AHP [A]. Professional Committee of Intelligent Automation of Chinese Association of Automation, Institute of Automation, Chinese Academy of Sciences. Proceedings of the 2005 Chinese Conference on Intelligent Automation.

[0006] Literature [2] Macleod J, Tan S, Moinuddin K. Reliability of fire (point) detection system in office buildings in Australia – A fault tree analysis [J]. Fire Safety Journal, 2020, 115: 103150.

[0007] Literature [3] Gao D X, Hou J J, Liang K, et al. Fault Diagnosis System for Electric Vehicle Charging Devices Based on Fault Tree Analysis [C] / / 2018 37th Chinese Control Conference (CCC). IEEE, 2018.

[0008] Literature [4] Sun X, Liu Y, Deng L. Reliability assessment of cyber-physical distribution network based on the fault tree [J]. Renewable Energy, 2020, 155: 1411 - 1424.

[0009] Literature [5] Hailong Y, Changhua L, Wei W, et al. Safety assessment of natural gas storage tank using similarity aggregation method based fuzzy fault tree analysis (SAM-FFTA) approach [J]. Journal of Loss Prevention in the Process Industries, 2020, 66: 104159.

[0010] Reference [6] Mengfei Z, Jinghua W, Haitai Z. Cooperative Fault Isolation Technology for Relay Protection and Distribution Automation [C] / / 2018 China International Conference on Electricity Distribution (CICED), Tianjin, 2018: 1390-1394.

[0011] Reference [7] Li G, Wu H, Wang F. Bayesian network approach based on fault isolation for power system fault diagnosis [C] / / 2014 International Conference on Power System Technology, Chengdu, 2014: 601-606. Summary of the Invention

[0012] Aiming at the problem of effective fault judgment and fault isolation in the control system of electronic equipment, a fault diagnosis method for the control system of electronic equipment based on a fault tree is proposed, which can accurately realize the fault diagnosis of the control system of electronic equipment and provide a solid theoretical basis for subsequent health management and maintenance work of electronic equipment.

[0013] The technical solution of the present invention is: A fault diagnosis method for the control system of electronic equipment based on a fault tree, including the following steps:

[0014] 1) Establish a fault tree model according to the research object and fault causes: According to the structural composition and operation process of the control system of electronic equipment, clarify the fault objects and corresponding fault causes, determine the top event, intermediate events and bottom events of the fault tree model. The top event is the fault object, and the bottom event is the most basic or fundamental fault cause that causes the top event to occur. The intermediate events connect the top event and the bottom event. The intermediate events or bottom events belonging to the same top event are associated using logic gates, and are decomposed layer by layer to establish and simplify the fault tree model;

[0015] 2) Conduct qualitative analysis on the fault tree model established in step 1), and find the minimum cut sets of the top event and all intermediate events according to the logical relationship between events;

[0016] 3) Quantitative analysis is carried out according to the fault tree model, corresponding minimal cut sets and relevant fault data to obtain the occurrence probabilities of each event: Statistical analysis of the occurrence probabilities of each basic event is carried out based on historical data or expert evaluation, and the occurrence probabilities of the top event and all intermediate event minimal cut sets are obtained according to the fault tree model and corresponding minimal cut sets;

[0017] 4) Calculate various relevance importance degrees of the top event and basic events and minimal cut set importance degrees according to the fault tree model;

[0018] 5) Determine the most likely fault causes according to the results of the previous step, and use linear weighted summation to obtain the ranking of the final fault probabilities;

[0019] 6) Finally, locate and isolate the fault points: According to the fault ranking results, combined with the set fault isolation threshold and the redundancy design of the research object, use the dynamic switching method for fault isolation.

[0020] Furthermore, the occurrence probability calculation method in step 3) is as follows:

[0021] Suppose the fault tree model has m basic events, n minimal cut sets, T represents the top event, C k represents the k-th minimal cut set, P(·) represents the occurrence probability, and the occurrence probability g of the top event is:

[0022]

[0023] Or the occurrence probability of the top event can also be directly solved according to each basic event. For the occurrence probability of the top event connected by an "AND" gate:

[0024]

[0025] The occurrence probability of the top event connected by an "OR" gate is:

[0026]

[0027] Among them, q j represents the occurrence probability of the j-th intermediate event in the next layer of the top event, and S represents the number of intermediate events in the next layer of the top event.

[0028] Furthermore, various relevance importance degrees of the top event and basic events in step 4) include structural importance degree, probability importance degree, and critical importance degree;

[0029] The structural importance degree: Without considering the occurrence probability of the basic event itself, analyze the influence degree of each basic event on the occurrence of the top event only from the structure. The specific implementation method: List the truth table and count the number of state switches of the top event from non-occurrence to occurrence when the basic event changes from non-occurrence to occurrence to obtain the structural importance degrees of all basic events i represents the i-th basic event, and st represents the structural importance;

[0030] The probability importance: It represents the degree of change in the probability of the top event caused by the change in the probability of the basic event. The probability importance of the i-th basic event is The critical importance: It represents the ratio of the relative change rate of the probability of the top event to the relative change rate of the probability of the basic event. The critical importance of the i-th basic event is The importance of the minimal cut set: It represents the contribution to the occurrence of the top event when all events in the minimal cut set occur. The importance of the k-th minimal cut set

[0031] Furthermore, the specific implementation method of step 5): The fault tree analysis method gives priority to the importance of the minimal cut set, followed by the structural importance. Set the weight matrix of the structural importance, probability importance, critical importance, and minimal cut set importance as W. Considering that the more basic events a minimal cut set contains, the smaller the probability of the simultaneous occurrence of the basic events, a coefficient is multiplied during linear weighting to balance the relationship between the two, that is, the coefficient where N k is the total number of basic events included in the k-th minimal cut set; the final weighted importance of the i-th basic event The calculation formula is as follows:

[0032]

[0033] where means that when the i-th basic event belongs to the k-th minimal cut set C k at this time, calculate once, and sum after traversing the minimal cut sets k = 1, 2,..., n.

[0034] The beneficial effects of the present invention are as follows: The fault diagnosis method for the electronic device control system established based on the fault tree in the present invention realizes the fault diagnosis of the electronic device control system by introducing the fault tree method and using the probabilities and various importances of the events occurring in the fault tree model, which can effectively improve the current situation where there is no mature fault diagnosis method for the electronic device control system; by introducing the method of improved linear weighted summation, the probabilities of various faults are effectively and accurately sorted, and according to the sorting results, combined with the set fault isolation threshold, the possible faults are isolated, which can effectively solve the problem of misoperation of fault isolation when the fault diagnosis result is inaccurate. Description of the Drawings

[0035] Figure 1 is the flowchart of the fault diagnosis method for the electronic device control system established based on the fault tree in the present invention;

[0036] Figure 2Schematic diagram of the fault tree of the electronic device control system according to an embodiment of the present invention;

[0037] Figure 3 Corresponding to the present invention Figure 2 Visualization diagram of the fault tree model. Detailed implementation manners

[0038] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. This embodiment is implemented on the premise of the technical solution of the present invention, and gives detailed implementation manners and specific operation processes, but the protection scope of the present invention is not limited to the following embodiments.

[0039] As Figure 1 shown in the flowchart of the fault diagnosis method for the electronic device control system established based on the fault tree, it mainly includes the following parts:

[0040] S1. According to the research object and fault causes, establish a fault tree model. The specific implementation steps include:

[0041] In the fault tree model, the top event corresponds to the most concerned fault event in fault diagnosis, the bottom event is the most basic or fundamental fault cause that causes the top event to occur. When there is more than one level in case of a fault, the intermediate event is what connects the top event and the bottom event. Among them, the top event and the bottom event are necessary. The intermediate events or bottom events belonging to the same top event need to be associated using logic gates. The commonly used logic gates mainly include the logical "AND" gate and the logical "OR" gate.

[0042] Taking the CPU module of the electronic controller of the gas turbine control system as an example in this embodiment, by consulting relevant materials, understand the possible fault events and corresponding fault causes of the CPU module. The focus of the fault diagnosis of the CPU module is whether the CPU is faulty. Therefore, the top event corresponding to the fault tree is the CPU module fault, and the bottom events are all the underlying fault events that cause the CPU fault, mainly including unreasonable wiring system, excessive power consumption load, component quality problems, crashing, network interface faults. The intermediate events are the fault events of the intermediate modules connecting the CPU and the underlying layer, mainly including hardware faults, software faults, CPU overheating, grounding faults, motherboard overheating. The established fault tree model is as Figure 2 shown, and the event names represented by each symbol are also listed one by one. T represents the top event of CPU module fault, x i (i = 1, 2, 3, 4, 5) represents the bottom event, M j (j = 1, 2, 3, 4, 5) represents the intermediate event. The top event and the bottom event are connected by logic gates. Among them, according to the event logical relationship, the logical gates used are all logical "OR" gates.

[0043] S2. Conduct qualitative analysis on the fault tree, and find the minimal cut sets of the top event and intermediate events according to the logical relationships between events. The specific implementation steps include:

[0044] After establishing the CPU fault tree, qualitative analysis is needed to locate the possible fault causes that may lead to the occurrence of the top event, and find the minimal cut sets that cause the occurrence of the top event. According to the definition of cut sets, the basic events of M3 are connected by logical "OR" gates, so its minimal cut sets are {x1}, {x2}. Similarly, the minimal cut sets of M4 are {x1}, {x3}, and the minimal cut sets of M5 are {x2}, {x3}. The basic events of M2 are connected by logical "OR" gates, so its minimal cut sets are {x4}, {x5}. And M1 = M3 ∪ M4 ∪ M5, so the minimal cut sets of M1 are {x1}, {x2}, {x3}. T = M1 ∪ M2, and its minimal cut sets are {x1}, {x2}, {x3}, {x4}, {x5}. Among them, {x1} means that when x1 occurs simultaneously, T will definitely occur, and so on.

[0045] S3. Conduct quantitative analysis based on the fault tree model, corresponding minimal cut sets, and relevant fault data to find the occurrence probabilities of each event. The specific implementation steps include:

[0046] After finding all the minimal cut sets of the top event, quantitative analysis can be carried out to find the occurrence probability of the top event described above, and determine the weak links of the system and the components with the highest fault probability. According to historical data or expert evaluation, statistical analysis of the occurrence probabilities of each basic event shows that: q1 = 0.001, q2 = 0.002, q3 = 0.004, q4 = 0.0009, q5 = 0.0011, where q i represents the occurrence probability of the basic event x i . Let C k (k = 1, 2, 3, 4, 5) represent the kth minimal cut set. According to logical relationships and probability theory, the occurrence probabilities of each minimal cut set can be obtained:

[0047] P(C1) = q1 = 0.001,

[0048] P(C2) = q2 = 0.002,

[0049] P(C3) = q3 = 0.004,

[0050] P(C4) = q4 = 0.0009,

[0051] P(C5) = q5 = 0.0011,

[0052] The occurrence probability of the top event is obtained from the calculation formula:

[0053]

[0054] S4. Calculate a series of importance degrees of each event and the importance degree of the minimal cut sets according to the fault tree model. The specific implementation steps include:

[0055] Step401: Calculate the structural importance degree.

[0056] Table 1

[0057]

[0058] First, according to the structural importance degree calculation formula, it is necessary to list the truth table to count the number of state switches of the top event from not occurring to occurring when a specific bottom event changes from not occurring to occurring. The truth table shows the Boolean logic relationship of the events, where 0 represents that the event does not occur and 1 represents that the event occurs. It is obtained from the structural importance degree calculation formula.

[0059] Structural importance degree: Without considering the probability of the bottom event itself occurring, analyze the influence degree of each bottom event on the occurrence of the top event only from the structure. It represents the importance of the component corresponding to the bottom event in the system structure and completely depends on the position of the component in the system. The calculation formula is as follows:

[0060]

[0061] Where In addition, when solving [Φ(1, x i ) - Φ(0, x i )], it is necessary to list the truth table to count the number of state switches of the top event from not occurring to occurring when a certain bottom event changes from not occurring to occurring.

[0062] m represents the number of bottom events, and x in ∑[Φ(1, x i ) - Φ(0, x i )] i represents the i-th bottom event. This formula is just a statistical symbol, indicating that when the i-th bottom event x i changes from not occurring to occurring, that is, from 0 → 1, ∑[Φ(1, x i ) - Φ(0, x i )] counts the number of state switches of the top event from not occurring to occurring. In it, i represents the i-th bottom event, and st represents the structural importance degree.

[0063] For event x1, when x1 changes from 0 → 1 and T changes from 0 → 1, there is 1 case in total, and the structural importance degree

[0064] For event x2, when x2 changes from 0 → 1 and T changes from 0 → 1, there is 1 case in total, and the structural importance degree

[0065] For event x3, when x3 changes from 0→1 and T changes from 0→1, there is 1 case in total, and the structural importance

[0066] For event x4, when x4 changes from 0→1 and T changes from 0→1, there is 1 case in total, and the structural importance

[0067] For event x5, when x5 changes from 0→1 and T changes from 0→1, there is 1 case in total, and the structural importance

[0068] Step402: Calculate the probability importance.

[0069] Assume that the fault tree model has m basic events, n minimal cut sets, T represents the top event, and C k represents the kth minimal cut set, and P(·) represents the occurrence probability. The occurrence probability of the top event is obtained according to the following formula:

[0070]

[0071] Alternatively, the occurrence probability of the top event can also be directly solved according to each basic event. For the occurrence probability of the top event connected by an "AND" gate:

[0072]

[0073] For the occurrence probability of the top event connected by an "OR" gate:

[0074]

[0075] Among them, q j represents the occurrence probability of the jth intermediate event in the layer below the top event, and S represents the number of intermediate events in the layer below the top event.

[0076] The probability importance represents the degree of change in the occurrence probability of the top event caused by the change in the occurrence probability of the basic event. The general formula for calculating the probability importance of the ith basic event is shown as follows:

[0077]

[0078] Among them, g is the occurrence probability of the top event, q i represents the occurrence probability of the basic event x i i = 1, 2,..., m, and pr represents the probability importance.

[0079] In addition, regarding the problem that the same letter cannot represent different meanings, it has been distinguished. Among them, the number of basic events is m, and the order is represented by i; the number of minimal cut sets is n, and the order is represented by k; the number of intermediate events is S, and the order is represented by j;

[0080] Since the top event in this embodiment is connected by an "OR" gate, according to the occurrence probability formula of the top event connected by an "OR" gate, there are 2 intermediate events M1 and M2 in the next layer of the top event in this embodiment. Therefore, the occurrence probability g of the top event is g = 1 - (1 - P(M1))(1 - P(M2)), where

[0081] P(M1) = P(M3 ∪ M4 ∪ M5) = P((x1 ∪ x2) ∪ (x1 ∪ x3) ∪ (x2 ∪ x3))

[0082] = P(x1 ∪ x2 ∪ x3) = 1 - (1 - q1)(1 - q2)(1 - q3)

[0083] = 6.986008×10 -3

[0084] P(M2) = P(M4 ∪ M5) = 1 - (1 - q4)(1 - q5) = 0.00199901

[0085] From the calculation formula of probability importance degree,

[0086]

[0087] Therefore, the probability importance degrees of x1, x2, x3, x4, and x5 are respectively:

[0088]

[0089]

[0090]

[0091]

[0092]

[0093] Step403: Calculate the critical importance degree.

[0094] The critical importance degree represents the ratio of the relative change rate of the occurrence probability of the top event to the relative change rate of the occurrence probability of the bottom event. The general calculation formula for the critical importance degree of the i-th bottom event is shown as follows:

[0095]

[0096] Among them, g is the occurrence probability of the top event, q i represents the occurrence probability of the bottom event x i , i = 1, 2,..., m, represents the probability importance degree of the bottom event x i , and Cr represents the critical importance degree.

[0097] According to the key importance calculation formula, the key importance degrees of x1, x2, x3, x4, and x5 are respectively:

[0098]

[0099]

[0100]

[0101]

[0102]

[0103] Step404: Calculate the importance degree of the minimum cut set.

[0104] The importance degree of the minimum cut set represents the contribution to the occurrence of the top event when all events in the minimum cut set occur. The general formula for calculating the importance degree of the k-th minimum cut set is shown as follows:

[0105]

[0106] Among them, g is the occurrence probability of the top event, Q k represents the occurrence probability of the k-th minimum cut set, k = 1, 2,..., n, and FV represents the importance degree of the minimum cut set.

[0107] According to the minimum cut set importance calculation formula, the importance degrees of the minimum cut sets of C1, C2, C3, and C4 are respectively:

[0108]

[0109]

[0110]

[0111]

[0112]

[0113] This embodiment conducts simulation based on the pycharm platform. The program input is the fault tree model, including the top event, intermediate events, bottom events, logic gates, and the occurrence probabilities of each bottom event. For example, Figure 3 is to visualize the input fault tree model. The output is the occurrence probability of the top event and a series of importance degrees.

[0114] S5. Determine the most likely fault cause, and use linear weighted summation to obtain the ranking of the final fault probability. The specific implementation steps include:

[0115] The comprehensive and organized calculation results of all the above steps are shown in Tables 2 and 3. As can be seen from Table 2, the structural importance degrees of x1, x2, x3, x4, and x5 are all 0.0625. Therefore, the health degrees of the components corresponding to all basic events need to be equally emphasized. The probability importance degree of x3 is 0.995009, which is the highest among the five basic events, indicating that the change in the probability of x3 has the greatest impact on the occurrence probability of the top event. Thus, improving the reliability of component x3 is the most helpful for enhancing the overall reliability. Similarly, the probability importance degree of x4 is 0.991922, which is the lowest among the five basic events, indicating that the change in the probability of x4 has the least impact on the occurrence probability of the top event, and improving the reliability of component x4 is also the least helpful for enhancing the overall reliability.

[0116] The probability importance degree only analyzes from the aspect of sensitivity and does not consider the occurrence probability of the basic events themselves. The critical importance degree is judged from both the event occurrence probability and sensitivity. In Table 2, the critical importance degree of x3 is 0.443653, which is the highest among all basic events, and the critical importance degree of x4 is 0.099512, which is the lowest among all basic events. Besides indicating that the system is most sensitive to the change in the occurrence probability of x3, it also shows that due to the relatively high occurrence probability of x3, the reliability of its corresponding component is very important. For x4, on the contrary, both its probability importance degree and critical importance degree are the lowest, indicating that the reliability of the component corresponding to x4 has a relatively small impact on the system.

[0117] As can be seen from Table 3, among the five minimum cut sets, the importance degree of {x3} is the highest, with a value of 0.445879, {x2} comes second, with a value of 0.222939, and the importance degree of {x4} is the lowest, with a value of 0.100323. Therefore, when the top event T occurs, {x3} makes the greatest contribution to the occurrence of T, {x2} comes second, and {x4} makes the smallest contribution. Since the sum of the importance degrees of the two cut sets {x2} and {x3} reaches 0.668818, exceeding half, when the top event T occurs, the basic events x2 and x3 should be focused on to check whether they occur.

[0118] Table 2

[0119]

[0120] Table 3

[0121]

[0122] Taking all importance factors into comprehensive consideration, the linear weighted summation method is adopted to obtain the ranking of the final failure probability. According to the above analysis, the fault tree analysis method usually gives priority to the importance of the minimal cut sets, and the structural importance of components in the system is secondary. In this embodiment, the weights of the structural importance, probability importance, critical importance, and minimal cut set importance are set as W = [0.25, 0.10, 0.15, 0.50]. Considering that the more basic events a minimal cut set contains, the smaller the probability of the simultaneous occurrence of the basic events, only considering the importance of the minimal cut sets cannot conduct a comprehensive evaluation. Therefore, in this embodiment, a coefficient needs to be multiplied during linear weighting to balance the relationship between the two, that is, the coefficient where N k is the total number of basic events included in the k-th minimal cut set, k = 1, 2,..., n. Therefore, the final weighted importance of the i-th basic event is calculated as follows, where i = 1, 2,..., m, represents the structural importance of the i-th basic event, represents the probability importance of the i-th basic event, represents the critical importance of the i-th basic event, represents the importance of the k-th minimal cut set, represents that when the i-th basic event belongs to the k-th minimal cut set C k , calculate the value of once, and sum it up after traversing the minimal cut sets k = 1, 2,..., n. The calculation results according to this formula are shown in Table 4. It can be seen from Table 4 that when the CPU control module fails, the ranking of the failure probability possibilities is x3, x2, x5, x1, x4.

[0123]

[0124] Table 4

[0125]

[0126] S6. Finally, locate and isolate the fault point, and the specific implementation steps include:

[0127] The control system is the core brain of the electronic device, controlling the safe and reliable operation of the entire electronic device. Once a fault occurs, in addition to quickly determining the location and type of the fault, it is also necessary to quickly isolate the fault point to reduce further damage. According to the above steps, the minimal cut sets most likely to cause the top event T to occur are {x2} and {x3}. According to the fault sorting results, combined with the redundancy design of the research object and the set fault isolation threshold (set to 0.2 here), a dynamic switching method is adopted. Through disconnecting switches such as circuit breakers, the relevant components corresponding to x2 and x3 are separated from the system respectively and switched to the corresponding secondary control modules to achieve fault isolation. After the fault is restored, they are reconnected, effectively realizing the fault diagnosis and fault isolation of the electronic device control system.

[0128] The preferred specific embodiments of the present application have been described in detail above. It should be understood that those of ordinary skill in the art can make many modifications and variations based on the concept of the present application without creative work. Therefore, all technical solutions that can be obtained by those skilled in the art in the technical field of the present application through logical analysis, reasoning, or limited experiments based on the concept of the present application on the basis of the prior art should be within the protection scope determined by the present application.

[0129] The above embodiments only represent several implementation manners of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several deformations and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A fault diagnosis method for an electronic device control system based on a fault tree, characterized in that Including the following steps: 1) Establish a fault tree model according to the research object and fault cause: Based on the structural composition and operation process of the electronic device control system, clarify the fault object and the corresponding fault cause, determine the top event, intermediate events, and bottom events of the fault tree model. The top event is the fault object, and the bottom events are the basic fault causes that lead to the occurrence of the top event. The intermediate events connect the top event and the bottom events. For the top event and intermediate events, and intermediate events and bottom events belonging to the same top event, logical gates are used for association, and layer-by-layer decomposition is carried out to establish the fault tree model; 2) Conduct qualitative analysis on the fault tree model established in step 1), and obtain the minimal cut sets of the top event and the minimal cut sets of all intermediate events according to the logical relationship between events; 3) Conduct quantitative analysis according to the fault tree model, the corresponding minimal cut sets, and relevant fault data. Conduct statistical analysis on the occurrence probabilities of each bottom event based on historical data or expert evaluation, and obtain the occurrence probabilities of the minimal cut sets of the top event and the minimal cut sets of all intermediate events; 4) Obtain various correlation importance degrees between the top event and the bottom events according to the fault tree model, and obtain the minimal cut set importance degree according to the occurrence probabilities of the minimal cut sets of the top event and the minimal cut sets of all intermediate events. The minimal cut set importance degree represents the contribution to the occurrence of the top event when all events in the minimal cut set occur; 5) Set the weights of various correlation importance degrees and minimal cut set importance degrees according to the results of various correlation importance degrees and minimal cut set importance degrees between the top event and the bottom events, and use linear weighted summation based on the results and weights to obtain the ranking of the final fault probabilities; 6) Finally, locate and isolate the fault point: According to the ranking result of the fault probabilities, combined with the set fault isolation threshold and the redundancy of the research object, adopt the dynamic switching method for fault isolation.

2. The fault diagnosis method for the electronic device control system established based on the fault tree according to claim 1, wherein The calculation methods for the occurrence probabilities of the minimal cut sets of the top event and the minimal cut sets of all intermediate events in step 3) are as follows: Suppose the fault tree model has a total of m basic events, n minimal cut sets, T represents the top event, and C k represents the k-th minimal cut set, P(·) represents the occurrence probability of each minimal cut set, and the occurrence probability g of the top event is: Alternatively, the occurrence probability of the top event can also be directly solved according to each bottom event. For the occurrence probability of the top event connected by an "AND" gate: The occurrence probability of the top event connected by an "OR" gate is: Among them, q j represents the occurrence probability of the j-th intermediate event in the next layer under the top event, and S represents the number of intermediate events in the next layer under the top event.

3. The fault diagnosis method for the electronic device control system established based on the fault tree according to claim 2, wherein, The various correlation importance degrees between the top event and the bottom events in step 4) include structural importance degree, probability importance degree, and critical importance degree; The structural importance: Without considering the probability of the occurrence of the basic event itself, only analyze the influence degree of each basic event on the occurrence of the top event from the structure. The specific implementation method: List the truth table and count the number of state switches of the top event from non-occurrence to occurrence when the basic event changes from non-occurrence to occurrence, and obtain the structural importance of all basic events i represents the i-th basic event, and st represents the structural importance; The probability importance: It represents the degree of change in the probability of the top event caused by the change in the probability of the bottom event. The probability importance of the i-th bottom event is pr represents the probability importance; The key importance: It represents the ratio of the relative change rate of the top event occurrence probability to the relative change rate of the bottom event occurrence probability. The key importance of the i-th bottom event is Cr represents the key importance; The importance of the minimal cut set: It represents the contribution to the occurrence of the top event when all events in the minimal cut set occur. The importance of the k-th minimal cut set FV represents the importance of the minimal cut set.

4. The fault diagnosis method for the electronic device control system established based on the fault tree according to claim 3, characterized in that, The specific implementation method of step 5) is as follows: The fault tree analysis method gives priority to the importance of the minimum cut sets, and the structural importance is secondary. Set the weight matrices of the structural importance, probability importance, critical importance, and minimum cut set importance as W. Considering that the more basic events included in the minimum cut set, the smaller the probability of simultaneous occurrence of the basic events, a coefficient is multiplied during linear weighting to balance the relationship between the two, that is, the coefficient where N k is the total number of basic events included in the k-th minimum cut set; the final weighted importance of the i-th basic event is calculated as follows: where means that when the \(i\)-th basic event belongs to the \(k\)-th minimal cut set \(C\) k calculate the value of once, and sum up after traversing all the minimal cut sets \(k = 1, 2, \cdots, n\).

Images