Fault diagnosis method and device, equipment, storage medium and program product

Abstract

The invention discloses a fault diagnosis method and device, equipment, a storage medium and a program product, and belongs to the field of fault diagnosis. The fault diagnosis method comprises the following steps: acquiring fault feature data of industrial equipment, wherein the fault feature data comprises at least one type of feature data; inputting the fault feature data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment; calculating the membership degree of each type of feature data under each feature type by using the membership function under each feature type in a plurality of feature types corresponding to each type of feature data, and performing fuzzy reasoning according to the membership degree of each type of feature data under each feature type and a pre-constructed fuzzy rule base, obtaining a second fault diagnosis result of the industrial equipment; and determining a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result. Therefore, the fault diagnosis efficiency can be improved.

Description

Technical Field

[0001] The present application belongs to the field of fault diagnosis, and specifically relates to a fault diagnosis method, device, equipment, storage medium and program product. Background Art

[0002] In the field of fault diagnosis of industrial equipment, the data sources of quality inspection reports and fault reports are wide-ranging, including both traditional paper reports and electronic reports. The formats and contents of these reports are heterogeneous, and there are problems such as colloquial terms and serious non-standardization, making it difficult to achieve unified management and utilization of data. Current fault diagnosis methods mostly rely on manual diagnosis and evaluation, which has the problem of low efficiency. Summary of the invention

[0003] The purpose of the embodiments of the present application is to provide a fault diagnosis method, apparatus, device, storage medium and program product to solve the problem of low efficiency of current fault diagnosis methods.

[0004] In order to solve the above technical problems, this application is implemented as follows:

[0005] In a first aspect, a fault diagnosis method is provided, comprising:

[0006] Acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data;

[0007] Inputting the fault characteristic data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment;

[0008] Using the membership function of each feature type in the plurality of feature types corresponding to each type of the feature data, the degree of membership of each type of the feature data is calculated, and fuzzy reasoning is performed according to the degree of membership of each type of the feature data and a pre-constructed fuzzy rule base to obtain a second fault diagnosis result of the industrial equipment; wherein each fuzzy rule in the fuzzy rule base is used to characterize the association between the feature type to which the fault feature data belongs and the fault category;

[0009] A final fault diagnosis result of the industrial equipment is determined according to the first fault diagnosis result and the second fault diagnosis result.

[0010] In a second aspect, a fault diagnosis device is provided, comprising:

[0011] An acquisition module, used to acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data;

[0012] An analysis module, used for inputting the fault characteristic data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment;

[0013] An evaluation module, used to calculate the degree of membership of each type of the feature data in each of the plurality of feature types corresponding to each type of the feature data by using the membership function under each feature type, and to perform fuzzy reasoning according to the degree of membership of each type of the feature data in each feature type and a pre-built fuzzy rule base, so as to obtain a second fault diagnosis result of the industrial equipment; wherein each fuzzy rule in the fuzzy rule base is used to characterize the association between the feature type to which the fault feature data belongs and the fault category;

[0014] A determination module is used to determine a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result.

[0015] In a third aspect, an electronic device is provided, comprising a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method described in the first aspect.

[0016] In a fourth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0017] In a fifth aspect, a computer program product is provided, comprising computer instructions, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0018] Through the scheme of the embodiment of the present application, the final fault diagnosis result of the industrial equipment can be obtained based on the fault diagnosis model and fuzzy reasoning, which can improve the automation and efficiency of fault diagnosis compared with the current manual diagnosis method. In addition, the accuracy and reliability of fault diagnosis can be significantly improved by combining the fault diagnosis model and fuzzy reasoning to perform fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flow chart of a fault diagnosis method provided by an embodiment of the present application;

[0020] Figure 2 is a flowchart of a fault diagnosis process in an embodiment of the present application;

[0021] Figure 3 is a structural schematic diagram of a fault diagnosis device provided in an embodiment of the present application;

[0022] Figure 4It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0025] The fault diagnosis method, apparatus, device, storage medium and program product provided in the embodiments of the present application are described in detail below with reference to the accompanying drawings through specific embodiments and their application scenarios.

[0026] See also Figure 1 , Figure 1 is a flow chart of a fault diagnosis method provided in an embodiment of the present application, the method is applied to electronic equipment, such as Figure 1 As shown, the method comprises the following steps:

[0027] Step 11: Acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data.

[0028] In the embodiment of the present application, the industrial equipment may be, but is not limited to, industrial production equipment, industrial processing equipment, etc. The fault characteristic data may be understood as fault data to be processed / analyzed / evaluated, and is used to determine the fault category of the corresponding industrial equipment.

[0029] Optionally, the characteristic data may be temperature, oil leakage, pressure, deformation, etc., depending on the specific industrial equipment and is not limited thereto.

[0030] Step 12: Input the fault feature data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment.

[0031] In the embodiment of the present application, the fault diagnosis model is obtained by pre-training, and can be selected as a large language model (LLM), etc. The specific training method can adopt an existing training method, which is not limited.

[0032] Step 13: Utilize the membership function under each feature type among the multiple feature types corresponding to each type of feature data to calculate the membership degree of each type of feature data under each feature type, and perform fuzzy reasoning based on the membership degree of each type of feature data under each feature type and a pre-built fuzzy rule base to obtain a second fault diagnosis result of the industrial equipment; each fuzzy rule in the fuzzy rule base is used to characterize the association between the feature type to which the fault feature data belongs and the fault category.

[0033] In the embodiment of the present application, the number of feature types corresponding to each type of feature data may be determined based on actual conditions. For example, taking the feature data as temperature as an example, the corresponding feature types may include low temperature and high temperature, or low temperature, medium temperature and high temperature. For another example, taking the feature data as the amount of oil leakage as an example, the corresponding feature types may include a small amount and a large amount, or a small amount, a medium amount and a large amount, or a small amount, a medium amount, a large amount and an extremely large amount.

[0034] The membership function is used to describe a fuzzy set, and the corresponding value range is 0 to 1, indicating the degree to which an element belongs to the fuzzy set, that is, the closer to 0, the lower the degree to which an element belongs to the fuzzy set, and the closer to 1, the higher the degree to which an element belongs to the fuzzy set. The membership function can be selected from triangular membership, trapezoidal membership or Gaussian membership functions, etc.

[0035] The fuzzy rule can be expressed in the form of "if...then...", for example: if x1 is Ai1, x2 is Ai2, ..., xn is Ai n , then y is Bi; where x is the input variable, y is the output variable, A is the fuzzy set of the input variable, and B is the fuzzy set of the output variable. The corresponding fuzzy rule base can be understood as an expert system.

[0036] Step 14: Determine a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result.

[0037] Optionally, after obtaining the first fault diagnosis result and the second fault diagnosis result, the final fault diagnosis result of the industrial equipment can be determined based on a comparison of the first fault diagnosis result and the second fault diagnosis result, such as comparing the probability values ​​in the first fault diagnosis result and the second fault diagnosis result. The final fault diagnosis result of the industrial equipment can also be obtained by fusing the first fault diagnosis result and the second fault diagnosis result.

[0038] Through the scheme of the embodiment of the present application, the final fault diagnosis result of the industrial equipment can be obtained based on the fault diagnosis model and fuzzy reasoning, which can improve the automation and efficiency of fault diagnosis compared with the current manual diagnosis method. In addition, combining the fault diagnosis model and fuzzy reasoning to perform fault diagnosis can also significantly improve the accuracy and reliability of fault diagnosis, which is conducive to dealing with complex and changeable fault conditions.

[0039] Optionally, performing fuzzy reasoning according to the degree of membership of each type of feature data under each feature type and a pre-built fuzzy rule base to obtain the second fault diagnosis result of the industrial equipment may include:

[0040] According to the degree of membership of each type of feature data under each feature type, the degree of membership of the at least one type of feature data under each fuzzy rule is calculated by taking the minimum value operation;

[0041] A second fault diagnosis result of the industrial equipment is determined according to the degree of membership of the at least one type of characteristic data under each fuzzy rule.

[0042] In this way, by taking the minimum value operation, the accuracy of the diagnosis result can be improved.

[0043] Optionally, determining the second fault diagnosis result of the industrial equipment according to the degree of membership of the at least one type of feature data under each fuzzy rule may include:

[0044] According to the membership degree of the at least one type of characteristic data under each fuzzy rule and the precise value corresponding to the fault category under each fuzzy rule, the fuzzified precise value corresponding to the at least one type of characteristic data is calculated by using the centroid method;

[0045] According to the precise value corresponding to the fault category under each fuzzy rule and the fuzzified precise value, the second fault diagnosis result of the industrial equipment is determined. For example, if the fuzzified precise value is equal to the precise value corresponding to a certain fault category, the fault category can be directly used as the second fault diagnosis result of the industrial equipment; or, if the fuzzified precise value is between the precise values ​​corresponding to two fault categories, the two fault categories can be merged to obtain the second fault diagnosis result of the industrial equipment.

[0046] For example, taking the above-mentioned fuzzy rule expression form as an example, for the fault feature data, the minimum value operation can be adopted to calculate the membership degree μ of the input part of each fuzzy rule Ri. Ri :

[0047]

[0048] The membership degree μ of the output part Bi of the fuzzy rule RiBi (y) can be expressed as:

[0049] μ Bi (y) = min(μ Ri ,μ Bi _ initial (y))

[0050] Aggregate the outputs of all fuzzy rules and use the maximum operation to obtain the membership at y, that is, the membership of each fuzzy rule:

[0051] μB(y)=max(min(μ R1 ,μ B1 (y)),min(μ R2 ,μ B2 (y)),...,min(μ Rm ,μ Bm (y))) After that, the centroid method can be used to obtain the fuzzy precise value k * :

[0052]

[0053] Among them, k represents the exact value of each fuzzy rule, which can be set based on actual needs.

[0054] The diagnostic process in this application is described below with specific examples.

[0055] In the embodiment of the present application, taking the characteristic data as temperature and oil leakage as an example, the temperature and oil leakage can be understood as input variables and can be defined as follows:

[0056] -Temperature value: 70, 90, 110, 130 (unit: degree);

[0057] -Oil leakage amount: 5, 10, 15, 20 (unit: ml).

[0058] The output variable is the fault category and can be defined as follows:

[0059] -Fault categories: minor oil leakage, major oil leakage, minor over-temperature, moderate over-temperature, major over-temperature.

[0060] Taking the triangular membership function as an example, we define membership functions for temperature and oil leakage respectively; for example, the temperature membership function can be as follows:

[0061] -Low Temperature:

[0062] -Medium temperature:

[0063] -high temperature:

[0064] The membership function of oil leakage can be shown as follows:

[0065] - Small amount:

[0066] -Medium volume:

[0067] - Large quantity:

[0068] Assuming that the fault characteristic data include: temperature T = 100 degrees, oil leakage volume V = 15ml; based on the above temperature membership function and oil leakage membership function, the following temperature membership and oil leakage membership can be obtained:

[0069] -Temperature membership:

[0070] -Low temperature: μ 低温 (100) = 0;

[0071] -Medium temperature:

[0072] -high temperature:

[0073] -Oil leakage membership:

[0074] - Small amount: μ 少量 (15) = 0;

[0075] -Medium:

[0076] -Large quantities:

[0077] Assume that the fuzzy rule base includes the following fuzzy rules:

[0078] 1) If the temperature is low and the oil leakage is small, the fault category is slight oil leakage;

[0079] 2) If the temperature is low and the oil leakage is large, the fault category is severe oil leakage;

[0080] 3) If the temperature is medium and the oil leakage is medium, the fault category is medium overtemperature;

[0081] 4) If the temperature is high and the oil leakage is moderate, the fault category is severe overtemperature.

[0082] According to the temperature membership and oil leakage membership calculated above and the above fuzzy rules, the following fuzzy reasoning is performed:

[0083] -Rule 1 (low temperature and small amount): membership μ 轻微渗漏油 =min(0,0)=0;

[0084] -Rule 2 (low temperature and large amount): membership μ 严重渗漏油 =min(0,1)=0;

[0085] -Rule 3 (medium temperature and medium quantity): membership μ 中度超温 =min(0.5,1)=0.5;

[0086] -Rule 4 (high temperature and medium quantity): membership μ 严重超温 =min(0.5, 1)=0.5.

[0087] Further assume that the exact values ​​corresponding to different fault categories are:

[0088] -Minor oil leakage: 0.2:

[0089] -Severe oil leakage: 0.6:

[0090] - Moderate overheating: 0.5:

[0091] - Severe overtemperature: 0.8.

[0092] Then the center of gravity method can be used to calculate the fuzzy exact value k * :

[0093]

[0094] In getting k * Afterwards, it can be compared with the precise values ​​corresponding to each fault category. Since 0.6<0.65<0.8, it can be determined that the diagnosis result / fault category is "moderate overtemperature tends to severe overtemperature", which means that the industrial equipment has a moderate overtemperature problem and the situation is likely to gradually worsen.

[0095] In the embodiment of the present application, the fuzzy rule base and / or the fault diagnosis model are obtained in advance. The fault diagnosis method in this embodiment may also include:

[0096] The first large language model is used to extract information from the fault diagnosis report of the industrial equipment to obtain multiple data pairs; each data pair includes fault description information and fault category; the fault diagnosis report is a historical report, which may include a traditional paper report and / or an electronic report, and the format and content of these reports may be heterogeneous; the first large language model is used to extract information from the fault diagnosis report; thus, with the help of the information extraction of the large language model, the historical data can be used in a unified manner, thereby improving the utilization and management efficiency of the data;

[0097] The second largest language model is used to extract fuzzy rules from the plurality of data pairs; the second largest language model is different from the first largest language model, and the second largest language model is used to extract fuzzy rules; thus, the fuzzy rules can be automatically extracted, the complexity and error of manual operation can be reduced, and complex fault diagnosis scenarios can be better handled.

[0098] The extracted fuzzy rules are reviewed and those fuzzy rules that do not conform to the rules are eliminated. For example, during the review, the extracted fuzzy rules can be cleaned and deduplicated, and the LLM model can be used in combination with the relevant national standards of the equipment to review the rules and eliminate the fuzzy rules that are contrary to the standards. In addition, manual review can be performed to further verify the extracted fuzzy rules and improve the accuracy.

[0099] The fuzzy rule base is constructed using the audited fuzzy rules; and / or the fault diagnosis model is trained based on the audited fuzzy rules and the multiple data pairs.

[0100] Optionally, when constructing a fuzzy rule base using the reviewed fuzzy rules, all reviewed fuzzy rules can be directly added to the fuzzy rule base, or based on actual needs, some of the reviewed fuzzy rules can be selected to be added to the fuzzy rule base, and there is no limitation on this.

[0101] Optionally, the process of training the fault diagnosis model according to the audited fuzzy rules and the multiple data pairs may include: first, constructing multiple training data pairs according to the audited fuzzy rules and the multiple data pairs; each training data pair contains fault feature data and corresponding marked fault categories, for example, the multiple data pairs can be verified using the audited fuzzy rules to obtain training data pairs that meet the rules; then, the multiple training data pairs are used for model training to obtain the fault diagnosis model. The specific training method can adopt the existing training method, which is not limited to this.

[0102] Optionally, in order to improve the extraction effect of fuzzy rules, when extracting fuzzy rules using the second largest language model, an instruction set for extracting fuzzy rules can be constructed for supervised fine-tuning (SFT).

[0103] Optionally, the first fault diagnosis result includes a first diagnosis probability corresponding to each of the multiple fault categories obtained based on the fault diagnosis model, and the second fault diagnosis result includes a second diagnosis probability corresponding to each of the multiple fault categories obtained by the fuzzy reasoning; the final fault diagnosis result of the industrial equipment determined based on the first fault diagnosis result and the second fault diagnosis result may include

[0104] Using a Bayesian reasoning method, the first diagnosis probability corresponding to each fault category and the second diagnosis probability corresponding to each fault category are integrated to obtain a final diagnosis probability corresponding to each fault category;

[0105] The fault category with the highest final diagnosis probability among the multiple fault categories is determined as the final fault diagnosis result of the industrial equipment.

[0106] In this way, with the help of the Bayesian reasoning process, the accuracy and reliability of fault diagnosis can be improved.

[0107] For example, the formula for Bayesian inference can be as follows:

[0108]

[0109] Among them, D i represents the i-th fault category, n represents the number of fault categories; E actual Indicates the diagnostic result of fuzzy reasoning; M actual represents the diagnosis result of the fault diagnosis model; P(D i ) represents the prior probability of the occurrence of the i-th fault category, which can be obtained by statistics; P(E actual |D i ) indicates that when the i-th fault category occurs, the diagnosis result of fuzzy reasoning is D i The conditional probability of actual |D i ) indicates that when the i-th type of fault occurs, the diagnosis result of the fault diagnosis model is D i The conditional probability of i |E actual ,M actual ) represents the final diagnosis probability after fusion.

[0110] For example, suppose an industrial equipment has three fault categories D1, D2, and D3, and the corresponding prior probabilities are: P(D1) = 0.3, P(D2) = 0.5, P(D3) = 0.2;

[0111] The diagnostic probabilities obtained by fuzzy reasoning under D1, D2 and D3 with the help of the inference rule base are: P(E actual |D1)=0.8,P(E actual |D2)=0.6,P(E actual |D3)=0.4;

[0112] With the help of the fault diagnosis model, the diagnosis probabilities obtained under D1, D2 and D3 are: P(M actual |D1)=0.9,P(M actual |D2)=0.7,P(M actual|D1)=0.5;

[0113] Then: According to the formula of Bayesian inference, we can first calculate the denominator, that is: ∑ i n ＝1 P(E actual |D i )·P(M actual |D i )·P(D i ); the specific calculation is:

[0114] P(E actual |D1)·P(M actual |D1)·P(D1)=0.8·0.9·0.3=0.216;

[0115] P(E actual |D2)·P(M actual |D2)·P(D2)=0.6·0.7·0.5=0.21;

[0116] P(E actual |D3)·P(M actual |D3)·P(D3)=0.4·0.5·0.2=0.04;

[0117] Next, the final diagnosis probability is calculated for each fault category:

[0118] (1) For D1,

[0119] (2) For D2,

[0120] (3) For D3,

[0121] Since 0.463 corresponding to D1 is the highest in the final diagnosis probability, it can be determined that the corresponding industrial equipment has fault category D1.

[0122] In an optional implementation, when diagnosing a fault on an industrial device, Figure 2 As shown, the specific fault diagnosis process may include:

[0123] S1: Perform optical character recognition (OCR) on historical paper reports of industrial equipment, and directly read electronic reports; the format and content of these reports can be heterogeneous;

[0124] S2: Use a large language model (such as LLM1 for extracting information) to extract information from the information obtained in S1 to obtain data pairs, each of which contains fault description information and fault category;

[0125] S3: extract fuzzy rules from the data pairs extracted from S2 using a large language model (such as LLM2 for extracting fuzzy rules); wherein, in order to improve the extraction effect, an instruction set for extracting fuzzy rules can be constructed (for example, manually constructed) to perform targeted fine-tuning on LLM2;

[0126] S4: For the extracted fuzzy rules, duplicate cleaning is performed, and the rules are audited by using a large language model (such as LLM3 for auditing fuzzy rules) combined with the relevant national standards of the equipment, and the fuzzy rules that are contrary to the standards are eliminated, and manual review is performed to further verify the extracted fuzzy rules and improve the accuracy; in order to

[0127] S5: using the fuzzy rules reviewed in S4 and the data pairs extracted in S2, construct multiple training data pairs, each of which contains fault feature data and a corresponding marked fault category, and use the training data pairs to perform model training to obtain a fault diagnosis model (such as LLM4); in order to improve the accuracy of the model, the fault diagnosis model LLM4 can also be fine-tuned with targeted instructions, such as SFT fine-tuning;

[0128] S6: Using the fuzzy rules reviewed in S4, a fuzzy rule base for fuzzy reasoning is constructed, which can be understood as an expert system;

[0129] S7: In practical applications, when the fault feature data of the industrial equipment is obtained, the fault feature data can be analyzed using the fault diagnosis model LLM4 to obtain a fault diagnosis result M, and fuzzy reasoning can be performed with the help of the constructed fuzzy rule base to obtain a fault diagnosis result E; the specific fuzzy reasoning process can be referred to the above embodiment, which will not be repeated here;

[0130] S8: The fault diagnosis result M and the fault diagnosis result E obtained in S7 are fused using the Bayesian reasoning method to obtain a final fault diagnosis result; the specific fusion process can be found in the above embodiment and will not be described in detail here.

[0131] It should be noted that the fault diagnosis method provided in the embodiment of the present application can be executed by a fault diagnosis device or a control module in the fault diagnosis device for executing the fault diagnosis method. In the embodiment of the present application, the fault diagnosis device provided in the embodiment of the present application is described by taking the fault diagnosis method executed by the fault diagnosis device as an example.

[0132] See also Figure 3 , Figure 3is a schematic diagram of a fault diagnosis device provided in an embodiment of the present application, and the device is applied to electronic equipment, such as Figure 3 As shown, the fault diagnosis device 30 includes:

[0133] An acquisition module 31 is used to acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data;

[0134] An analysis module 32, configured to input the fault characteristic data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment;

[0135] The evaluation module 33 is used to calculate the membership degree of each type of the feature data under each feature type by using the membership function under each of the multiple feature types corresponding to each type of the feature data, and perform fuzzy reasoning according to the membership degree of each type of the feature data under each feature type and a pre-constructed fuzzy rule base to obtain a second fault diagnosis result of the industrial equipment; wherein each fuzzy rule in the fuzzy rule base is used to characterize the association relationship between the feature type to which the fault feature data belongs and the fault category;

[0136] The determination module 34 is used to determine a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result.

[0137] Optionally, the evaluation module 33 includes:

[0138] A calculation unit, configured to calculate the degree of membership of the at least one type of feature data under each fuzzy rule by taking a minimum value according to the degree of membership of each type of feature data under each feature type;

[0139] A determination unit is used to determine a second fault diagnosis result of the industrial equipment according to the membership degree of the at least one type of feature data under each fuzzy rule.

[0140] Optionally, the determination unit is specifically used to: calculate the fuzzy precise value corresponding to the at least one type of feature data using the centroid method according to the membership of the at least one type of feature data under each fuzzy rule and the precise value corresponding to the fault category under each fuzzy rule; determine the second fault diagnosis result of the industrial equipment according to the precise value corresponding to the fault category under each fuzzy rule and the fuzzy precise value.

[0141] Optionally, the fault diagnosis device 30 further includes:

[0142] An extraction unit, configured to extract information from the fault diagnosis report of the industrial equipment using the first language model to obtain a plurality of data pairs; each of the data pairs includes fault description information and a fault category;

[0143] an extraction unit, configured to extract fuzzy rules from the plurality of data pairs using a second language model;

[0144] An audit unit, used for auditing the extracted fuzzy rules and eliminating the fuzzy rules that do not conform to the rules;

[0145] The processing unit is used to construct the fuzzy rule base using the audited fuzzy rules; and / or to train the fault diagnosis model according to the audited fuzzy rules and the multiple data pairs.

[0146] Optionally, the processing unit is further used for:

[0147] According to the audited fuzzy rules and the multiple data pairs, multiple training data pairs are constructed; each of the training data pairs contains fault feature data and a corresponding marked fault category; and the multiple training data pairs are used to perform model training to obtain the fault diagnosis model.

[0148] Optionally, the first fault diagnosis result includes a first diagnosis probability corresponding to each of the multiple fault categories obtained based on the fault diagnosis model, and the second fault diagnosis result includes a second diagnosis probability corresponding to each of the multiple fault categories obtained by the fuzzy reasoning;

[0149] The determination module 34 is specifically used to: use the Bayesian reasoning method to fuse the first diagnosis probability corresponding to each fault category and the second diagnosis probability corresponding to each fault category to obtain the final diagnosis probability corresponding to each fault category; and determine the fault category with the highest final diagnosis probability among the multiple fault categories as the final fault diagnosis result of the industrial equipment.

[0150] The fault diagnosis device 30 of the embodiment of the present application can realize the above Figure 1 The various processes of the method embodiment shown can achieve the same technical effect, and will not be described again here to avoid repetition.

[0151] Optional, such as Figure 4 As shown, an embodiment of the present application also provides an electronic device 40, including a processor 41, a memory 42, and a program or instruction stored in the memory 42 and executable on the processor 41. When the program or instruction is executed by the processor 41, each process of the above-mentioned fault diagnosis method embodiment is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0152] The embodiment of the present application also provides a computer program product, including computer instructions. When the computer instructions are executed by a processor, the various processes of the above-mentioned fault diagnosis method embodiment can be implemented and the same technical effect can be achieved. To avoid repetition, they will not be repeated here.

[0153] An embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the various processes of the above-mentioned fault diagnosis method embodiment can be implemented and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0154] Computer readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0155] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0156] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0157] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a service classification device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0158] The above is only a preferred implementation of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A fault diagnosis method, characterized in that: include: Acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data; Inputting the fault characteristic data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment; Using the membership function of each feature type in the plurality of feature types corresponding to each type of the feature data, the degree of membership of each type of the feature data is calculated, and fuzzy reasoning is performed according to the degree of membership of each type of the feature data and a pre-constructed fuzzy rule base to obtain a second fault diagnosis result of the industrial equipment; wherein each fuzzy rule in the fuzzy rule base is used to characterize the association between the feature type to which the fault feature data belongs and the fault category; A final fault diagnosis result of the industrial equipment is determined according to the first fault diagnosis result and the second fault diagnosis result.

2. The method according to claim 1, characterized in that The step of performing fuzzy reasoning according to the degree of membership of each type of feature data under each feature type and a pre-built fuzzy rule base to obtain a second fault diagnosis result of the industrial equipment includes: According to the degree of membership of each type of feature data under each type of feature, the degree of membership of at least one type of feature data under each fuzzy rule is calculated by taking the minimum value; A second fault diagnosis result of the industrial equipment is determined according to the degree of membership of the at least one type of feature data under each fuzzy rule.

3. The method according to claim 2, characterized in that Determining the second fault diagnosis result of the industrial equipment according to the degree of membership of the at least one type of feature data under each fuzzy rule includes: Calculate the fuzzified precise value corresponding to the at least one type of characteristic data by using the centroid method according to the degree of membership of the at least one type of characteristic data under each fuzzy rule and the precise value corresponding to the fault category under each fuzzy rule; A second fault diagnosis result of the industrial equipment is determined according to the precise value corresponding to the fault category under each fuzzy rule and the fuzzified precise value.

4. The method according to claim 1, characterized in that: The method further comprises: Using the first language model to extract information from the fault diagnosis report of the industrial equipment to obtain a plurality of data pairs; each of the data pairs includes fault description information and a fault category; extracting fuzzy rules from the plurality of data pairs using a second language model; Reviewing the extracted fuzzy rules, and eliminating the fuzzy rules that do not conform to the rules; The fuzzy rule base is constructed using the audited fuzzy rules; and / or the fault diagnosis model is trained based on the audited fuzzy rules and the multiple data pairs.

5. The method according to claim 4, characterized in that The method of training a fault diagnosis model based on the reviewed fuzzy rules and the multiple data pairs includes: Constructing a plurality of training data pairs according to the reviewed fuzzy rules and the plurality of data pairs; each of the training data pairs comprises fault feature data and a corresponding marked fault category; Model training is performed using the multiple training data pairs to obtain the fault diagnosis model.

6. The method according to claim 1, characterized in that The first fault diagnosis result includes a first diagnosis probability corresponding to each of the multiple fault categories obtained based on the fault diagnosis model, and the second fault diagnosis result includes a second diagnosis probability corresponding to each of the multiple fault categories obtained by the fuzzy reasoning; The step of determining a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result includes: Using a Bayesian reasoning method, the first diagnosis probability corresponding to each fault category and the second diagnosis probability corresponding to each fault category are integrated to obtain a final diagnosis probability corresponding to each fault category; The fault category with the highest final diagnosis probability among the multiple fault categories is determined as the final fault diagnosis result of the industrial equipment.

7. A fault diagnosis device, characterized in that: include: An acquisition module, used to acquire fault characteristic data of industrial equipment, wherein the fault characteristic data includes at least one type of characteristic data; An analysis module, used for inputting the fault characteristic data into a pre-trained fault diagnosis model for analysis to obtain a first fault diagnosis result of the industrial equipment; An evaluation module, used to calculate the degree of membership of each type of the feature data in each of the plurality of feature types corresponding to each type of the feature data by using the membership function under each feature type, and to perform fuzzy reasoning according to the degree of membership of each type of the feature data in each feature type and a pre-built fuzzy rule base, so as to obtain a second fault diagnosis result of the industrial equipment; wherein each fuzzy rule in the fuzzy rule base is used to characterize the association between the feature type to which the fault feature data belongs and the fault category; A determination module is used to determine a final fault diagnosis result of the industrial equipment according to the first fault diagnosis result and the second fault diagnosis result.

8. An electronic device, characterized in that: The method comprises a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method according to any one of claims 1 to 6.

9. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the steps of the method according to any one of claims 1 to 6.

Images