Sample generation method for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields

Abstract

The embodiment of the invention provides a sample generation method for general verification of fault detection and diagnosis of multi-field machinery and electronic products, and the method comprises the steps: according to the project demands of a target project, carrying out the fault detection and diagnosis of the multi-field machinery and electronic products; initial samples stored in a database are screened, and a plurality of intermediate samples related to the target item are obtained; determining a fault mode corresponding to each intermediate sample, and extracting the fault modes meeting a consistency index in the fault modes to obtain a to-be-verified fault mode set; and determining a weight corresponding to each fault mode in the to-be-verified fault mode set, and determining a target sample related to the target item based on the weight.

Description

Technical Field

[0001] The embodiments in this specification relate to the field of computer technology, and in particular to a sample generation method for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields. Background Technology

[0002] Fault detection and diagnosis models can provide advance estimates of equipment health and defects based on various information data, enabling timely measures to be taken before catastrophic damage or failure occurs. In addition, fault detection and diagnosis models can also guide the formulation of equipment maintenance plans, thereby extending the lifespan of equipment.

[0003] When using fault detection and diagnosis models to design or apply fault diagnosis and prediction technologies for typical fault modes in a system, the challenge lies in addressing a large number of diverse system fault diagnosis and prediction needs. Different systems typically exhibit varying diagnostic and prediction capabilities, and these capabilities also differ significantly under different engineering conditions, including varying operating conditions, fault modes, and data conditions. Therefore, quickly determining suitable fault diagnosis and prediction samples based on existing engineering information and user requirements is extremely difficult. In other words, how to select and obtain suitable verification samples to ensure the actual performance of the fault detection and diagnosis model becomes a pressing issue. Summary of the Invention

[0004] In view of this, embodiments of this specification provide a sample generation method for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields. One or more embodiments of this specification also relate to a sample generation device, a computing device, a computer-readable storage medium, and a computer program for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, to address the technical deficiencies existing in the prior art.

[0005] According to a first aspect of the embodiments of this specification, a sample generation method for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields is provided, comprising: Based on the project requirements of the target project, the initial samples stored in the database are filtered to obtain multiple intermediate samples related to the target project; Determine the failure mode corresponding to each intermediate sample, and extract the failure modes that meet the consistency index from each failure mode to obtain the set of failure modes to be verified. Determine the weight corresponding to each fault mode in the set of fault modes to be verified, and determine the target samples related to the target project based on the weights.

[0006] Optionally, the step of extracting the failure modes that meet the consistency index from each failure mode to obtain a set of failure modes to be verified includes: Divide each fault mode into sets to generate at least two initial fault mode sets; Determine the hierarchical fault data corresponding to each fault mode in the target initial fault mode set, wherein each hierarchical fault data consists of fault modes at different levels, and the target initial fault mode set is each of the at least two initial fault mode sets; A judgment matrix is ​​constructed based on the hierarchical fault data, and the fault modes that meet the consistency index in each fault mode contained in the hierarchical fault data are extracted according to the judgment matrix to obtain a set of fault modes to be verified.

[0007] Optionally, the step of extracting fault modes that meet the consistency index from each fault mode included in the hierarchical fault data according to the judgment matrix to obtain a set of fault modes to be verified includes: Solve the characteristic equation constructed based on the judgment matrix and the identity matrix to obtain the maximum eigenvalue of the judgment matrix; The consistency index value is determined based on the maximum eigenvalue, and the consistency ratio is calculated based on the index value. If the consistency ratio is determined to meet the condition, a set of failure modes to be verified is constructed based on the failure modes associated with the judgment matrix.

[0008] Optionally, determining the target samples related to the target project based on the weights includes: The fault modes in the set of fault modes to be verified are sorted according to the weights, and the target fault modes are filtered according to the sorting results. Identify the samples to be verified corresponding to the target failure mode, and perform a sufficiency test on the samples to be verified; If the verification passes, the sample to be verified is identified as a target sample related to the target project.

[0009] Optionally, the sufficiency test on the sample to be verified includes: The sample to be verified is split to generate at least two sample sets; Obtain the mapping relationship between the failure modes related to the target project and the project parameter values ​​of the target project, and determine the equivalent set of failure modes corresponding to each sample set based on the mapping relationship; The sufficiency measure of the corresponding sample set is determined based on the fault mode equivalence set, and the sufficiency test of the sample to be verified is performed based on the sufficiency measure.

[0010] Optionally, after obtaining multiple intermediate samples related to the target project, the method further includes: Determine the target sample size for the target samples related to the target project based on the project requirements; If the sample size of the plurality of intermediate samples meets the requirements based on the target sample size, the step of determining the fault mode corresponding to each intermediate sample is performed.

[0011] Optionally, the sample generation method for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields further includes: If it is determined that the sample size of the multiple intermediate samples is insufficient, fault data generated by performing forward diffusion noise injection on the initial samples is obtained, and the target condition variable corresponding to the fault data is determined. The fault data and the target condition variable are input into the trained sample generation model for processing to generate intermediate samples to be screened. The sample generation model includes a generator network, a discriminator network, and a degradation feature reconstruction network. Random sampling is performed on the plurality of intermediate samples and the intermediate samples to be screened according to the target sample size to generate target samples related to the target project.

[0012] According to a second aspect of the embodiments of this specification, a sample generation device for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields is provided, comprising: The filtering module is configured to filter the initial samples stored in the database according to the project requirements of the target project to obtain multiple intermediate samples related to the target project. The extraction module is configured to determine the fault mode corresponding to each intermediate sample and extract the fault modes that meet the consistency index from each fault mode to obtain the set of fault modes to be verified. The determination module is configured to determine the weight corresponding to each fault mode in the set of fault modes to be verified, and to determine the target sample related to the target project based on the weight.

[0013] According to a third aspect of the embodiments of this specification, a computing device is provided, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement any of the steps of the sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0014] According to a fourth aspect of the embodiments of this specification, a computer-readable storage medium is provided that stores computer-executable instructions, which, when executed by a processor, implement the steps of any one of the sample generation methods for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0015] According to a fifth aspect of the embodiments of this specification, a computer program is provided, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the sample generation method described above for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0016] This embodiment of the specification filters initial samples stored in a database according to the project requirements of the target project to obtain multiple intermediate samples related to the target project; determines the fault mode corresponding to each intermediate sample, and extracts the fault modes that meet the consistency index from each fault mode to obtain a set of fault modes to be verified; determines the weight corresponding to each fault mode in the set of fault modes to be verified, and determines the target samples related to the target project based on the weights. By filtering samples according to the project requirements of different target projects to obtain target samples that meet the conditions, it is beneficial to ensure the accuracy and applicability of the generated samples, and also to ensure the accuracy of the detection results obtained by fault detection and diagnosis through target samples. Attached Figure Description

[0017] Figure 1 This is a flowchart of a sample generation method for a general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, provided in one embodiment of this specification; Figure 2 This is a schematic diagram of the structure of a sample generation device for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, provided in one embodiment of this specification. Figure 3 This is a structural block diagram of a computing device provided in one embodiment of this specification. Detailed Implementation

[0018] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0019] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0020] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0021] This specification provides a sample generation method for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields. This specification also relates to a sample generation device for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, a computing device, a computer-readable storage medium, and a computer program, which will be described in detail in the following embodiments.

[0022] Figure 1 A flowchart is shown of a sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to an embodiment of this specification, which specifically includes the following steps.

[0023] Step 102: According to the project requirements of the target project, filter the initial samples stored in the database to obtain multiple intermediate samples related to the target project.

[0024] Specifically, the target project can be a verification system for mechanical and electronic products in multiple fields. The project requirements can be the number and type of target samples needed to detect and diagnose typical failure modes of the target project.

[0025] Based on this, in the process of screening and obtaining the required target samples, the initial samples stored in the database can be screened according to the project requirements of the target project to obtain intermediate samples related to the target project.

[0026] Specifically, data discrimination criteria are a preliminary screening step for samples in the database to determine which samples can be used for fault detection and diagnosis.

[0027] Data discrimination criteria are used to perform multi-dimensional, joint screening of initial samples in the database to identify high-confidence samples that can be used for fault detection and diagnostic modeling. The embodiments in this specification construct a data discrimination mechanism that integrates rule constraints, statistical discrimination, and adaptive threshold updates to comprehensively judge samples from multiple levels, including data quality, information validity, and engineering consistency.

[0028] Specific discriminant dimensions include: 1) Data integrity assessment: The system detects whether samples have missing key sensor channels, data interruptions, or structural damage from both the time and feature dimensions. Samples that do not meet the minimum information completeness requirements are directly deemed invalid and removed.

[0029] 2) Data correlation judgment: Combine the statistical characteristics of historical failure samples to evaluate the degree of correlation between data features and failure modes, and retain only the feature data that contributes to the discrimination of the target failure mode, thereby avoiding irrelevant information from interfering with subsequent model training.

[0030] 3) Data timeliness assessment: Introduce system configuration and operating condition tags to perform timeliness matching analysis on historical data, ensuring that the selected samples are consistent with the current system structure, software and hardware version and operating conditions, and avoiding data mismatch problems caused by system evolution.

[0031] 4) Data Anomaly Detection: By jointly analyzing statistical distribution characteristics and local mutation features, abnormal fluctuations and abrupt changes in the data are identified. Anomalies may originate from sensor malfunctions or reflect potential system failure symptoms. Therefore, they are marked rather than simply deleted during the detection process for use in subsequent analysis.

[0032] 5) Data consistency judgment: For multi-source, multi-channel data, analyze the measurement consistency of the same physical quantity by different sensors, identify data conflicts caused by sensor drift, inaccuracy or communication abnormalities, and adjust the sample credibility accordingly.

[0033] Through the aforementioned multidimensional joint discrimination mechanism, the system can assign different quality and confidence levels to the original samples, providing a high-quality input data foundation for subsequent connotation analysis, sample size determination, and sample generation.

[0034] Furthermore, intrinsic analysis is used to conduct in-depth analysis of the screened data to extract intrinsic information that characterizes the stability, reliability, and representativeness of the samples. This specification provides an intrinsic analysis method for samples oriented towards noise sensitivity. By controlling the adjustment of noise type, noise amplitude, and signal-to-noise ratio, the method analyzes the performance changes of generated samples under different noise conditions, thereby extracting the intrinsic characteristics of sample stability in uncertain environments.

[0035] In practical applications, the initial samples stored in the database can be subjected to connotation analysis, i.e., in-depth analysis of the initial samples to extract useful information and features. Specifically, when determining the samples required for detecting and diagnosing typical failure modes in verification systems for multi-domain mechanical and electronic products, it is first necessary to clarify the basis for sample selection recommendations. Based on a summary of the basic process of verification system failure detection and diagnosis and the theoretical methods and experimental operations involved in each step, a connotation analysis of the verification system typical failure mode detection and diagnosis sample library was conducted. A set of sample attribute information was selected to adequately characterize the applicability of comprehensive test samples for various verification systems, serving as the basis for determining the target samples.

[0036] This specification provides a comprehensive noise sensitivity evaluation tool for analyzing and assessing changes in the stability of generated samples under different noise conditions. By precisely controlling the type, amplitude, and signal-to-noise ratio of noise, it helps users understand the robustness and reliability of generated samples in various noise environments.

[0037] Specific sample content analysis includes: (1) Noise type selection.

[0038] Specifically, you can select from a list of predefined noise types, which mainly include Gaussian noise, uniform noise, salt and pepper noise, Poisson noise, and mixed noise.

[0039] Gaussian noise refers to a type of noise whose probability density function follows a Gaussian distribution (i.e., a normal distribution). If a noise's amplitude distribution follows a Gaussian distribution, and its power spectral density is uniformly distributed, it is called Gaussian white noise. The second moments of Gaussian white noise are uncorrelated, and the first moments are constants, representing the temporal correlation between successive signals. Gaussian white noise includes thermal noise and shot noise. In communication channel testing and modeling, Gaussian noise is used as additive white noise to generate additive white Gaussian noise.

[0040] Uniform noise is a type of random noise that is evenly distributed within a specified range. Its energy is uniformly distributed across all frequencies and is unaffected by frequency; it is also known as white noise. Uniform noise is commonly used in image and audio processing to simulate noise in real-world environments.

[0041] Salt-and-pepper noise, also known as impulse noise, is a common type of noise in images. It appears as randomly occurring white or black dots, where black pixels appear in bright areas or white pixels appear in dark areas (or both). Salt-and-pepper noise can be caused by sudden, strong interference with the image signal, analog-to-digital converter errors, or bit transmission errors. For example, a malfunctioning sensor can cause a pixel value to be at its minimum, while a saturated sensor can cause a pixel value to be at its maximum.

[0042] Poisson noise is a type of statistical noise, typically used to describe the distribution of the frequency of random events occurring within a given time or space. In image processing, Poisson noise can be used to simulate certain natural phenomena in images, such as grainy textures. A key characteristic of Poisson noise is that its intensity is related to the local brightness of the image; the brighter the area, the greater the noise intensity. This noise distribution can be described by the Poisson distribution, a discrete probability distribution that describes the frequency of random events occurring within a fixed time interval or space.

[0043] Mixed noise refers to the presence of multiple types of noise in a signal simultaneously. These noises may include, but are not limited to, impulse noise, Gaussian noise, Poisson noise, etc.

[0044] Different noise types are used to simulate signal disturbances under different interference mechanisms. By comparing and analyzing the performance changes of samples under different noise types, we can identify the differences in the sensitivity of samples to different noise mechanisms, and avoid samples from being stable only under a single noise assumption, thereby improving the comprehensiveness and reliability of the intrinsic analysis of samples.

[0045] In the implementation, noise can be selectively injected according to signal channel, characteristic frequency band or time segment to specifically evaluate the robustness of the sample in key feature regions.

[0046] (2) Noise amplitude control.

[0047] After selecting the noise type, the noise amplitude is controlled. The system provides a noise amplitude adjustment method, allowing users to input a specific noise amplitude value or select different noise levels such as low, medium, and high.

[0048] The noise amplitude is adjusted progressively to create a continuous performance response sequence for the sample as it transitions from weak to strong perturbations. In this implementation, the progressive noise amplitude can employ an adaptive search strategy, dynamically adjusting the noise step size based on changes in sample performance to accelerate the location of noise intervals where sample performance significantly degrades, thereby reducing computational costs.

[0049] (3) Signal-to-noise ratio adjustment.

[0050] Users can set different signal-to-noise ratio levels to simulate changes from low-noise to high-noise environments.

[0051] The system automatically calculates and applies the corresponding noise amplitude based on the set signal-to-noise ratio, so that the samples are tested under a uniform noise intensity scale. This allows for accurate analysis of the performance trend of the samples as noise increases, and avoids analytical bias caused by differences in noise parameter settings.

[0052] (4) Operating condition disturbance simulation.

[0053] The simulation examines the performance changes of generated samples under different operating conditions, including factors such as temperature, humidity, and pressure.

[0054] By repeatedly performing noise sensitivity analysis under different operating conditions, the stability changes of the sample under the combined effects of noise and operating conditions are evaluated. In the implementation method, the combination of noise level and operating condition parameters is generated using a preset coverage matrix to ensure consistent coverage of the target operating condition space and the reproducibility of the analysis results, thereby making the intrinsic analysis results of the sample closer to the actual operating environment.

[0055] (5) Signal noise simulation and sample stability connotation extraction During the sample generation or processing, real-time or phased noise perturbation is applied to the signal to obtain perturbation samples under different noise conditions, and the perturbation samples are tested and verified.

[0056] By comparing the changes in key characteristics or performance indicators of samples under different noise levels, a response curve of sample performance as noise increases is constructed, and its degradation behavior characteristics are analyzed. When sample performance shows a significant decrease, abrupt change, or accelerated degradation, the corresponding noise level or change behavior is determined as the noise sensitivity characteristic of the sample.

[0057] The noise sensitivity characteristics include at least one of the following or a combination thereof: The degradation slope characteristics of sample performance as a function of noise; The instability noise level characteristic of the sample performance falling below a preset threshold for the first time; The consistency characteristics of the performance response of the samples under different noise types or operating conditions.

[0058] Under the premise of meeting the preset performance threshold conditions, the maximum noise level at which the sample can maintain effective performance is further determined as a quantitative upper limit indicator of the sample stability.

[0059] (6) Sample content classification Based on the noise sensitivity characteristics, the stability connotation of the samples is graded and a corresponding sample connotation level label is generated. The level label is used to characterize the reliability of the samples under complex noise and operating condition disturbance environments.

[0060] In this implementation, the sample content level label is written back to the database for subsequent determination of whether the sample size meets the requirements.

[0061] After constructing the sample attribute information set of the fault detection and diagnosis samples of the verification system, the initial samples stored in the database can be screened based on this information set to obtain samples that meet the conditions. These samples can be used as intermediate samples. Alternatively, to ensure the accuracy of the screening results, the samples can be further screened to obtain intermediate samples.

[0062] Specifically, the aforementioned samples will be further screened, which can be done based on factors such as data completeness, data relevance, data timeliness, data anomalies, and data consistency.

[0063] Data integrity involves checking whether the data is complete and free of missing or corrupted values. For example, if a fault sample lacks crucial sensor readings, it may be unsuitable for fault diagnosis. Data relevance assesses the correlation between the data and the type of fault. For example, if a database contains temperature readings of a machine during normal operation, this data may be useful for diagnosing overheating faults. Data timeliness determines whether the data is up-to-date and relevant to current system configuration and operating conditions. For example, older fault data may not be applicable to newer versions of software or hardware. Data anomaly identification identifies outliers in the data that may indicate a fault. For example, if a sensor typically returns values ​​between 0 and 100 but suddenly shows readings of -10 or 110, this could be a sign of a fault. Data consistency checks whether the data is consistent with other data sources. For example, if two different sensors measure the same physical quantity but their readings differ significantly, this indicates a problem with one of the sensors.

[0064] In one optional implementation, after obtaining multiple intermediate samples related to the target project, the method further includes: Determine the target sample size for the target samples related to the target project based on the project requirements; If the sample size of the plurality of intermediate samples meets the requirements based on the target sample size, the step of determining the fault mode corresponding to each intermediate sample is performed.

[0065] Specifically, after obtaining intermediate samples through the aforementioned screening method, it is necessary to determine whether the number of intermediate samples meets the requirements. If not, a certain number of samples need to be generated so that the sum of the number of generated samples and the number of intermediate samples meets the requirements.

[0066] In the embodiments of this specification, based on the intensional analysis, a sample size determination algorithm is used to evaluate whether the sample size of the intermediate samples meets the requirements. The algorithm considers the following factors: Fault type diversity: Ensure sufficient samples for each fault type; Feature coverage: Ensure that the selected samples can cover all changes in key features; Model complexity: More complex models may require more data to avoid overfitting.

[0067] Based on this, the embodiments in this specification employ a fault detection and diagnosis sample size determination algorithm based on Bayesian posterior risk to evaluate whether the sample size of the intermediate samples meets the requirements. The specific steps are as follows: According to Bayesian statistics, the sample The generation of this involves two steps. First, imagine randomly generating an observation value y from π (y). i Then, the observations of sample X are generated from the conditional distribution L(x|y). The joint conditional density function of sample x is expressed as follows:

[0068] Wherein, the joint density function It is called the likelihood function.

[0069] Because y i It is based on the prior distribution Since it is randomly generated, in order to integrate all prior information, we cannot only consider y. i However, it is necessary to calculate all possible values ​​of y. Using a distribution... If we consider the combined sample, then the joint probability distribution of sample x and parameter y is:

[0070] The above formula incorporates prior information. Population information and sample information L(x|y). When statistical inference is needed about parameter y, it is combined with sample observations. Then, h(x|γ) should be used to infer γ based on h(x|y). To do this, h(x|y) needs to be decomposed as follows:

[0071] Where m(x) is the marginal probability density function of x.

[0072]

[0073] It can be seen that the marginal probability density function m(x) is independent of y and contains no information about y. Therefore, π(y|x) can be used to estimate the parameter y, and its calculation method is as follows:

[0074] In the above formula, the conditional probability distribution π(y|x) is called the posterior distribution of the distribution parameter y. It integrates all information from the prior, the population, and the sample. Statistical inference of γ through the posterior distribution π(y|x) is the most effective method.

[0075] Based on the aforementioned algorithm, determine whether the number of intermediate samples meets the requirements. If it does, proceed to step 104.

[0076] Furthermore, this specification proposes a method for determining the sample size for fault detection and diagnosis based on hierarchical information fusion and posterior risk criteria. This method, while maintaining the interpretability and engineering feasibility of the posterior risk criteria, can fully integrate prior information at the unit / subsystem / system level with success-failure test data, and solve for the minimum test sample size under the risk constraints of both. The specific implementation process is as follows: (1) Parameter and data definition Let the testability metric be the fault detection rate. During fault injection verification, each injection corresponds to a success-or-failure observation: success indicates "fault detected," and failure indicates "fault not detected." System-level observation data is denoted as... This includes system-level or equivalent success / failure data obtained through structural fusion (such as the number of trials, the number of failures, etc.).

[0077] (2) Hierarchical prior fusion A hierarchical Bayesian network (HBN) is established based on the hierarchical structure of the equipment system. The fault detection rate of each level node is used as the network transmission parameter to achieve layer-by-layer information integration from the bottom unit to the system layer. HBN can characterize the dependence of the system's fault detection capability on the detection capability of the lower-level unit and support the modeling of the influence relationship between nodes at the same level, thereby improving the structural consistency and interpretability of prior fusion.

[0078] For bottom-level units without parent nodes, their detection priors can be represented using a Beta distribution:

[0079] And combined with the success or failure test data of this unit Obtain the posterior:

[0080] The above Beta-binary conjugate update form is used to ensure that the inference process is stable and reproducible.

[0081] For upper-level nodes with parent nodes, their "inherited prior" (derived from lower-level information) is obtained through the conditional probability table (CPT) of the HBN and uncertainty inference. Since inherited priors usually lack an analytical form, a Monte Carlo sampling approximation method can be used. A distribution fitting strategy is then employed to obtain the prior representation of the upper-level nodes, which is then fused with the node's self-prior and success / failure data to finally obtain the posterior distribution of the system-level nodes. .

[0082] (4) Posterior risk criterion constraints Let the sample size be Under the given test protocol, the maximum number of failures allowed is .make This represents the number of observed failures. Define the sampling feature function (pass / fail probability):

[0083] This function describes "when the true detection rate is..." At that time, the number of failures does not exceed The probability of "".

[0084] Under the risk constraints of both parties, a posterior risk inequality is constructed. This differs from traditional methods that directly use... This method uses system posterior. Entry constraints are used to reflect the degree of support from the fused objective evidence. This can be written in the following form:

[0085] in, The threshold values ​​agreed upon by both parties. These represent the risk ceilings for the manufacturer and the purchaser / user, respectively. This risk modeling, which "replaces prior with system post-hoc," more realistically reflects the existing evidence. The acceptance of the acceptance conclusion is supported by both parties, thereby reducing the sample size while ensuring credibility.

[0086] (5) Sample size calculation algorithm because Often arising from hierarchical reasoning and fusion, it is difficult to directly analyze integrals, but can be derived from... The posterior sample set is obtained by random sampling. The risk term is approximated using Monte Carlo integration:

[0087] By traversing the candidates Combination, calculation And check if the conditions are met simultaneously. Constraints: Select the minimum sample size that satisfies the constraints. and corresponding As an output of the experimental design.

[0088] (6) Output results The output includes: minimum experimental sample size Maximum number of failures allowed and the risk assessment value under this plan. .

[0089] Step 104: Determine the fault mode corresponding to each intermediate sample, and extract the fault modes that meet the consistency index from each fault mode to obtain the set of fault modes to be verified.

[0090] In one optional implementation, the step of extracting fault modes that meet the consistency index from each fault mode to obtain a set of fault modes to be verified includes: Divide each fault mode into sets to generate at least two initial fault mode sets; Determine the hierarchical fault data corresponding to each fault mode in the target initial fault mode set, wherein each hierarchical fault data consists of fault modes at different levels, and the target initial fault mode set is each of the at least two initial fault mode sets; A judgment matrix is ​​constructed based on the hierarchical fault data, and the fault modes that meet the consistency index in each fault mode contained in the hierarchical fault data are extracted according to the judgment matrix to obtain a set of fault modes to be verified.

[0091] Furthermore, the step of extracting fault modes that meet the consistency index from each fault mode included in the hierarchical fault data according to the judgment matrix to obtain a set of fault modes to be verified includes: Solve the characteristic equation constructed based on the judgment matrix and the identity matrix to obtain the maximum eigenvalue of the judgment matrix; The consistency index value is determined based on the maximum eigenvalue, and the consistency ratio is calculated based on the index value. If the consistency ratio is determined to meet the condition, a set of failure modes to be verified is constructed based on the failure modes associated with the judgment matrix.

[0092] Specifically, due to the complex and uncertain relationships among the various characteristics of the evaluation indicators, these indicators may be incomparable, and each indicator may have a different degree of influence on the evaluation results. Therefore, the Analytic Hierarchy Process (AHP) is introduced to analyze the fault sample set and fault mode set, thereby obtaining the target sample.

[0093] In the embodiments of this specification, each fault mode is first divided into sets to generate at least two initial fault mode sets, that is, at least two replaceable units are generated.

[0094] The target is the replaceable units of the fault mode set, where K is the number of replaceable units, and each replaceable unit U i By N i It consists of several failure modes.

[0095] Assuming replaceable unit U i The failure mode set is , For the fault mode set F i The corresponding hierarchical fault data. Hierarchical fault data refers to data that organizes and classifies fault modes according to a certain hierarchical structure. For example, fault modes can be divided into broad categories such as "functional faults," "hardware faults," and "software faults," and then further subdivided within each category. Specifically, all fault modes that may affect the unit can be identified through historical data, expert knowledge, fault reports, and system analysis, thus forming the hierarchical fault data corresponding to the unit's fault mode set.

[0096] Furthermore, after determining the hierarchical fault data corresponding to the fault mode set, the specific implementation process of the fault mode random extraction algorithm based on hierarchical analysis is as follows: Step 202: Construct the judgment matrix.

[0097] Statistical sets The number of fault modes k i and this k i Each failure mode is extracted and combined into a set. and order Construct a judgment matrix.

[0098] Where P is the judgment matrix. These are the eigenvalues ​​of the matrix.

[0099] The judgment matrix P is composed as follows: The goal is to identify failure modes that meet consistency metrics within a set of failure modes. , ( ) represents each failure mode in the failure mode set. express for The relative importance values. Then the judgment matrix P can be written as:

[0100] In the established indicator system, indicators (failure modes) at the same level are comparable. However, within the same level, each indicator has a different weight. Considering attributes such as experimental risk, accuracy, and confidence level, the 1-9 scaling method shown in Table 1 is used to measure the relative merits of the judgment objects when constructing the judgment matrix. This scaling method is simple, clear, and easy to use.

[0101] Table 1 Scale Description

[0102] Step 204: Solve the characteristic equation.

[0103] In the Analytic Hierarchy Process (AHP), the decision matrix P is used to represent the relative importance of each factor (failure mode). The eigenvector corresponding to the largest eigenvalue (eigenvalue) of the decision matrix P is obtained by solving the characteristic equation det(P−λI)=0, where I is the identity matrix.

[0104] Step 206: Perform a consistency check.

[0105] Specifically, the largest eigenvalue (eigenvalue) of the judgment matrix P is usually used for consistency testing. The consistency index can be determined based on the largest eigenvalue, and the consistency ratio can be calculated based on the index value. If the largest eigenvalue is close to the order n of the matrix (i.e., λmax≈n), the consistency of the judgment matrix is ​​considered to be good.

[0106] The test formula is as follows:

[0107]

[0108] In the formula, CR is the consistency ratio of the judgment matrix; CI is the general consistency index of the judgment matrix; and RI is the average random consistency index of the judgment matrix. The RI values ​​of judgment matrices of order 1 to 9 are shown in the table.

[0109] Table 2. Values ​​of the average random consistency index RI

[0110] When the CR of the judgment matrix P is less than 0.1 or When CI=0, P is considered to have satisfactory consistency; otherwise, the elements in P need to be adjusted to ensure satisfactory consistency.

[0111] Given that the consistency ratio meets the condition, a set of failure modes to be verified is constructed based on the failure modes associated with the judgment matrix. Specifically, failure modes that meet the consistency index are extracted from the failure mode set to construct the set of failure modes to be verified, F. Fi .

[0112] Step 106: Determine the weight corresponding to each fault mode in the set of fault modes to be verified, and determine the target samples related to the target project based on the weights.

[0113] In one optional implementation, determining the target samples related to the target project based on the weights includes: The fault modes in the set of fault modes to be verified are sorted according to the weights, and the target fault modes are filtered according to the sorting results. Identify the samples to be verified corresponding to the target failure mode, and perform a sufficiency test on the samples to be verified; If the verification passes, the sample to be verified is identified as a target sample related to the target project.

[0114] Specifically, after obtaining the set of failure modes to be verified, the weight corresponding to each failure mode in the set of failure modes to be verified can be determined, and the target samples related to the target project can be determined based on the weights.

[0115] In the Analytic Hierarchy Process (AHP), the relative weight of each factor (failure mode) can be calculated using the judgment matrix, thereby ranking their importance. Specifically, the eigenvector corresponding to the largest eigenvalue can be normalized, and the result can be used as the relative weight of each factor to rank their importance, i.e., weight allocation.

[0116] In the embodiments of this specification, indicators (failure modes) at the same level are comparable, but each indicator at the same level has a different weight. Furthermore, these indicators include both positive ("high-potential") and negative ("low-potential") indicators. Indicators with high potential include: minimum sample size requirement, critical function failure coverage, functional failure coverage, test activation rate, testability rate, test utilization rate, injection cost consumption rate (by default, costs are incurred whenever a fault injection test is conducted), fault injection rate, time-domain clustering Gini coefficient, frequency-domain uniform KL divergence, and statistical domain feature density coefficient of variation. Indicators with low potential include: fault undiagnostic rate, severe fault coverage, and injection hazard rate. Moreover, considering the test costs and the number of fault modes allocated, the unit fault allocation rate and complex unit allocation rate are mutually restrictive with the injected fault consumption rate. Therefore, as long as the required number of fault samples is met, these are also set as positive indicators in the embodiments of this specification. To make the analytic hierarchy process (AHP) better applicable to sample sampling, the inverse indicators in this specification are all reversed in the embodiments. That is, a reference value is set for each inverse indicator, which is equal to 1 minus the actual true value.

[0117] The sample imbalance index for multi-domain fault detection and diagnosis is defined as follows: 1) In the time domain, extract features including but not limited to the following: mean, variance, RMS, peak value, peak-to-peak value, skewness, kurtosis, waveform factor, impulse factor, margin factor, zero-crossing rate, and rise / fall time; based on the extracted time domain features, use the K-means algorithm to perform clustering and calculate the sample proportion of each cluster. ; Calculate the time-domain clustering Gini coefficient:

[0118] 2) In the frequency domain, for the Fourier-Bessel domain spectrum of the signal The following method is used to determine the local maxima in the spectrum. Frequency band boundary Assume the number of local maxima found is ,but: At that time, keep the previous one The maximum points are used as the frequency band boundaries; When resetting parameters ,make

[0119] For each sample, calculate the energy proportion of each frequency band, and take the dominant frequency band (the band with the highest energy) as its frequency domain category. Statistically calculate the proportion of samples in each category. .

[0120] Calculate the uniform KL divergence in the frequency domain:

[0121] 3) Within the statistical domain, extract statistical features including, but not limited to, the following: higher-order moment statistics such as 3rd-order skewness and 4th-order kurtosis, autocorrelation function decay rate, and information entropy. Calculate the local kernel density of the sample points within the statistical domain formed by these statistical features:

[0122] in, For bandwidth parameters, The number of statistical features. The total number of samples, For the first A vector formed by the statistical characteristics of each sample.

[0123] Calculate the coefficient of variation of the characteristic density in the statistical domain:

[0124] Once a unified set of judgment indicators for decision-making is established, the weights of the analytic hierarchy process (AHP) can be determined. Methods for calculating the AHP weight vector W include the geometric mean, arithmetic mean, eigenvector method, and least squares method, which will be introduced one by one below.

[0125] 1) Geometric mean method.

[0126]

[0127] The relevant calculation steps are as follows: a) Multiply the elements of the judgment matrix P by row to obtain a new vector; b) Calculate the nth root of each component of the new vector; c) Normalize the resulting vector into a weight vector.

[0128] 2) Arithmetic mean method.

[0129] Since each column in the judgment matrix P approximately reflects the distribution of weights, the weight vector is estimated by using the arithmetic mean of all column vectors, resulting in:

[0130] The relevant calculation steps are as follows: a) Obtain the results by normalizing the elements of the judgment matrix P column by column. ; b) Add the normalized columns together; c) Divide the summed vector by n to obtain the weight vector.

[0131] 3) Eigenvector method Multiply the judgment matrix P by the weight vector W on the right, as shown in the following formula:

[0132] In the formula, To determine the largest eigenvalue of matrix P, it must exist and be unique. All components of W are positive components.

[0133] Then, the obtained weight vector is normalized to obtain the desired result.

[0134] 4) Least squares method The weight vector is determined using a fitting method to minimize the sum of squared residuals for each term. The solution is as follows:

[0135]

[0136] Once the consistency check and weight determination of the fault mode set are completed, a better fault sample set can be determined based on the weight distribution of the hierarchical analysis.

[0137] Specifically, after sorting each fault mode in the set of fault modes to be verified in descending order of weight, a certain number of fault modes at the top of the sorting can be identified as target fault modes, and the corresponding samples to be verified can be identified. Then, a sufficiency test is performed on the samples to be verified. If the test passes, the samples to be verified are identified as target samples related to the target project.

[0138] In addition, after completing the consistency check of the judgment matrix and obtaining the weight vectors corresponding to each fault mode... Subsequently, to improve the randomness and representativeness of fault sample sampling, this embodiment of the specification introduces a weight-guided random sampling strategy while maintaining the weight calculation and consistency check process of the analytic hierarchy process (AHP). Specifically, the weight vector obtained by the AHP is... Normalization is performed, and the result is used as the probability distribution for the extraction of fault modes:

[0139] After determining the replaceable unit The number of fault modes to be extracted is Under the premise of probability distribution Using random sampling without replacement, from the failure mode set Randomly selected from Each failure mode constitutes a failure sample set for this unit.

[0140] This method ensures that high-weighted fault modes have a high sampling probability while allowing medium- and low-weighted fault modes that meet consistency requirements to still be selected in multiple sampling processes, thereby improving the statistical diversity and coverage integrity of the fault sample set. By repeatedly executing the weighted random sampling process, the fault sample set used in the test validation experiment is finally determined, denoted as […]. .

[0141] In one optional implementation, the sufficiency test of the sample to be verified includes: The sample to be verified is split to generate at least two sample sets; Obtain the mapping relationship between the failure modes related to the target project and the project parameter values ​​of the target project, and determine the equivalent set of failure modes corresponding to each sample set based on the mapping relationship; The sufficiency measure of the corresponding sample set is determined based on the fault mode equivalence set, and the sufficiency test of the sample to be verified is performed based on the sufficiency measure.

[0142] Specifically, after determining the sample to be verified corresponding to the target failure mode, a sufficiency test can be performed on the sample to be verified. If the test passes, the sample to be verified is determined as the target sample related to the target project. This target sample can be used to detect and diagnose the target project.

[0143] Sample sufficiency testing involves verifying the sufficiency of the generated samples to be validated, ensuring that they cover all possible failure scenarios. For test validation experiments, the sample size is fixed, and the sample set is generated through sampling. Therefore, it is necessary to measure the sufficiency of the generated sample set to be validated and select the optimal sample set for test experiments.

[0144] In one optional implementation, the sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products further includes: If it is determined that the sample size of the multiple intermediate samples is insufficient, fault data generated by performing forward diffusion noise injection on the initial samples is obtained, and the target condition variable corresponding to the fault data is determined. The fault data and the target condition variable are input into the trained sample generation model for processing to generate intermediate samples to be screened. The sample generation model includes a generator network, a discriminator network, and a degradation feature reconstruction network. Random sampling is performed on the plurality of intermediate samples and the intermediate samples to be screened according to the target sample size to generate target samples related to the target project.

[0145] Specifically, after generating multiple intermediate samples and determining that the sum of the sample sizes of these intermediate samples does not meet the requirements, fault data generated by forward diffusion noise injection on the initial samples can be obtained. The target condition variable corresponding to the fault data is then determined. The fault data and the target condition variable are input into a trained sample generation model for processing to generate intermediate samples to be screened. The sample generation model includes a generator network, a discriminator network, and a degradation feature reconstruction network. Random sampling is performed on the multiple intermediate samples and the intermediate samples to be screened according to the target sample size to generate target samples related to the target project. These target samples can be used for fault detection and diagnosis of the target project.

[0146] In the embodiments described in this specification, the sample generation model includes a generator network, a discriminator network, and a degenerate feature reconstruction network.

[0147] Specifically, in the data preparation stage, it is mainly used to extract conditional variables of known failure sample degradation trends and to construct a training sample set composed of failure samples and conditional variables.

[0148] In the degradation trend condition variable extraction stage, the fault samples with known fault levels, i.e., historical fault data, are sent to the degradation trend condition variable extraction module. From each fault sample, k degradation features that can characterize the degree of equipment fault are extracted, i.e., k data carrying fault information are extracted and spliced ​​together to form the degradation trend condition variable, i.e., the initial condition variable.

[0149] As mentioned earlier, the initial condition variable can be expressed as:

[0150] in, It can be composed of time-domain / frequency-domain / time-frequency-domain features (such as RMS, kurtosis, dominant frequency energy, envelope spectrum amplitude, etc.) and used to characterize the degree of degradation, the direction of degradation and the rate of change.

[0151] In practical applications, the extracted k features that characterize the degree of equipment failure can be the mean, variance, and standard deviation of the equipment failure degree. Taking k = 3, failure samples 1 and 2, and the extracted features being the mean, variance, and standard deviation as an example, the mean, variance, and standard deviation of the equipment failure degree are extracted from failure samples 1 and 2 respectively. The three features extracted from failure sample 1 are concatenated to generate the initial condition variable for failure sample 1; the three features extracted from failure sample 2 are also concatenated to generate the initial condition variable for failure sample 2.

[0152] After obtaining the initial condition variables, forward diffusion noise injection is performed on the target historical fault data according to the preset diffusion step number to generate target fault data: Assuming the target historical fault data (initial sample) is The forward diffusion noise is injected according to the preset diffusion step number to obtain... :

[0153] in, , These are the diffusion scheduling parameters (determined by a preset noise schedule).

[0154] Let the number of diffusion steps be ,but .

[0155] The target fault data, initial condition variables, and preset diffusion steps are then concatenated to generate training samples.

[0156] In practical applications, for a sample set with a known fault severity (historical fault data, also known as fault samples) After the degradation trend condition variable extraction module, feature data carrying fault information is calculated for each sample. Then, each historical fault data is compared with its respective... The fault sample pairs are combined to form the training dataset. ,in k were extracted by different methods The conditional variable composed of the values ​​(second degenerate characteristic). .

[0157] During the model training phase, forward propagation is first performed using fault samples generated in the data preparation phase to calculate the generated samples, interpolated samples, and the discrimination results of the discriminator network. The discriminator network's loss is calculated based on its discrimination results for fault samples and generated samples. The generator network's loss is calculated based on its discrimination results for generated samples. The conditional mutual information maximization loss between the generator network and the Q network is calculated based on the reconstruction result of the degradation trend conditional variable contained in the generated samples by the degradation feature reconstruction network Q (i.e., the first degradation feature). Based on these loss terms, the discriminator network, generator network, and Q network are sequentially optimized and updated until the loss converges, completing the model training and obtaining the trained sample generation model.

[0158] Specifically, after generating fault samples, the fault samples are input into the generator network for processing to obtain generated samples under the control of condition variable c. .

[0159] Let the generator (denoising network) be The input is The output is a prediction of the noise:

[0160] Depend on An estimate of the original sample can be obtained:

[0161] Then, linear interpolation is performed on the fault samples and the generated samples to obtain the interpolated samples. ,in, , is a random number that is uniformly distributed between 0 and 1.

[0162] Next, the fault samples, generated samples, and interpolated samples are input into the discriminator network for processing to generate corresponding discrimination results. Based on the first discrimination result corresponding to the fault sample, the second discrimination result corresponding to the generated sample, and the interpolated sample, the first loss value corresponding to the discriminator network is calculated; based on the second discrimination result corresponding to the generated sample, the second loss value of the generator network is calculated.

[0163] The first loss value of the discriminator network is calculated as follows:

[0164] The second loss value of the generator network is calculated as follows:

[0165] For fault sample generation tasks that include conditional information, the more information about the input conditional variables the generated fault samples contain, the better; that is, the degradation information contained in the input conditional variables should be preserved to the greatest extent possible in the generated samples. Mutual information can measure the degree of dependence between two random variables:

[0166] Therefore, consider using the mutual information between the input condition variables and the generated samples. This is used to measure the degree of preservation of degraded information and is defined as conditional mutual information. Therefore, in addition to the loss values ​​of the generator and discriminator networks, it is necessary to optimize this conditional mutual information to its maximum during training. However, directly... Performing the calculations is very difficult, so an auxiliary distribution is constructed. This provides a lower bound for conditional mutual information: At this point, the objective of maximizing conditional mutual information is to maximize... At this point, a network Q is constructed to predict the mean and standard deviation of the condition variable. Assuming the condition variable follows a normal distribution, for a mean of... Standard deviation is For a random variable, its probability density function is:

[0167] To calculate Part 1 It means to calculate ,Right now Assume the mean and standard deviation of the condition variable output by network Q are respectively... and The objective of maximizing conditional mutual information can be written as:

[0168] Since the distribution of the condition variable c is fixed, Equivalent to:

[0169] If not considered The impact of this further optimizes the target. Equivalent to:

[0170] Therefore, the output of network Q is a vector with the same dimension as the condition variable c.

[0171] In actual training, besides calculating the loss values ​​of the generator and discriminator networks in the Conditional Temporal Deep Generative Adversarial Network (CTGAN) module, the generated samples can also be input into the degradation feature reconstruction network. The degradation feature reconstruction network reconstructs the degradation features of the generated samples to generate the first degradation feature. Based on the first degradation feature, the conditional mutual information maximization loss value between the generator network and the degradation feature reconstruction network is calculated. The specific processing method is as follows: Introducing degradation trends to rebuild networks This allows the degenerate semantics of the generated samples to revert to the condition variable:

[0172] Additionally, diffusion uniformity can be calculated using standard noise to predict MSE:

[0173] After calculating the aforementioned loss value, we can obtain: Total loss of generator (denoising network):

[0174] Reconstructing network loss (can be shared with generator) ):

[0175] Discriminator loss is .

[0176] Based on this, the model parameters of the sample generation model to be trained are adjusted according to each loss value to obtain the trained sample generation model. Specifically, according to... , , The generator network, discriminator network, and degradation trend reconstruction network are trained separately to obtain the trained sample generation model G. .

[0177] The fault data and the target condition variable can then be input into the trained sample generation model, and the generator network can process the fault data and the target condition variable to generate corresponding intermediate samples to be screened.

[0178] Specifically, the sample generation model generates intermediate samples to be screened, corresponding to the sample generation stage under unknown fault levels. This stage first samples fault samples and then concatenates a degradation trend condition variable (target condition variable) under the target fault level onto these samples. The fault samples and the target condition variable are then input into the trained sample generation model, where the generator network processes them to obtain the generator network's output. This output represents the fault sample corresponding to the input degradation trend condition variable for the fault level. By interpolating the degradation trend condition variables extracted from existing fault samples, the degradation trend condition variable under the unknown fault level is calculated, thus completing the interpolation generation of missing fault samples under the unknown fault level.

[0179] In practical applications, for a known set of condition variables Interpolation is performed to obtain the conditional variable representing the unknown degree of the target fault. For example, under linear interpolation:

[0180] in, Let be the set of condition variables corresponding to the a-th sample out of n initial samples; Let b be the set of condition variables corresponding to the b-th sample out of n initial samples.

[0181] Then, the sample generation is gradually inverted from pure noise, that is, the sample generation stage starts with high noise initialization:

[0182] right Stepwise noise reduction:

[0183]

[0184] The final generated sample is:

[0185] In addition, by interpolating the second degradation feature, the degradation feature corresponding to the target fault level can be obtained, and the target fault level can be preset.

[0186] This embodiment of the specification filters initial samples stored in a database according to the project requirements of the target project to obtain multiple intermediate samples related to the target project; determines the fault mode corresponding to each intermediate sample, and extracts the fault modes that meet the consistency index from each fault mode to obtain a set of fault modes to be verified; determines the weight corresponding to each fault mode in the set of fault modes to be verified, and determines the target samples related to the target project based on the weights. By filtering samples according to the project requirements of different target projects to obtain target samples that meet the conditions, it is beneficial to ensure the accuracy and applicability of the generated samples, and also to ensure the accuracy of the detection results obtained by fault detection and diagnosis through target samples.

[0187] Corresponding to the above method embodiments, this specification also provides embodiments of a sample generation device for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields. Figure 2 This specification illustrates a schematic diagram of a sample generation device for universal verification of fault detection and diagnosis in multi-domain mechanical and electronic products, provided in one embodiment. Figure 2 As shown, the device includes: The filtering module 202 is configured to filter the initial samples stored in the database according to the project requirements of the target project to obtain multiple intermediate samples related to the target project. The extraction module 204 is configured to determine the fault mode corresponding to each intermediate sample and extract the fault modes that meet the consistency index from each fault mode to obtain a set of fault modes to be verified. The determination module 206 is configured to determine the weight corresponding to each fault mode in the set of fault modes to be verified, and to determine the target sample related to the target project based on the weight.

[0188] In an optional implementation, the extraction module 204 is further configured to: Divide each fault mode into sets to generate at least two initial fault mode sets; Determine the hierarchical fault data corresponding to each fault mode in the target initial fault mode set, wherein each hierarchical fault data consists of fault modes at different levels, and the target initial fault mode set is each of the at least two initial fault mode sets; A judgment matrix is ​​constructed based on the hierarchical fault data, and the fault modes that meet the consistency index in each fault mode contained in the hierarchical fault data are extracted according to the judgment matrix to obtain a set of fault modes to be verified.

[0189] In an optional implementation, the extraction module 204 is further configured to: Solve the characteristic equation constructed based on the judgment matrix and the identity matrix to obtain the maximum eigenvalue of the judgment matrix; The consistency index value is determined based on the maximum eigenvalue, and the consistency ratio is calculated based on the index value. If the consistency ratio is determined to meet the condition, a set of failure modes to be verified is constructed based on the failure modes associated with the judgment matrix.

[0190] In an optional implementation, the determining module 206 is further configured to: The fault modes in the set of fault modes to be verified are sorted according to the weights, and the target fault modes are filtered according to the sorting results. Identify the samples to be verified corresponding to the target failure mode, and perform a sufficiency test on the samples to be verified; If the verification passes, the sample to be verified is identified as a target sample related to the target project.

[0191] In an optional implementation, the determining module 206 is further configured to: The sample to be verified is split to generate at least two sample sets; Obtain the mapping relationship between the failure modes related to the target project and the project parameter values ​​of the target project, and determine the equivalent set of failure modes corresponding to each sample set based on the mapping relationship; The sufficiency measure of the corresponding sample set is determined based on the fault mode equivalence set, and the sufficiency test of the sample to be verified is performed based on the sufficiency measure.

[0192] In one optional implementation, the sample generation device for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products further includes a processing module configured to: Determine the target sample size for the target samples related to the target project based on the project requirements; If the sample size of the plurality of intermediate samples meets the requirements based on the target sample size, the step of determining the fault mode corresponding to each intermediate sample is performed.

[0193] In an optional implementation, the processing module is further configured to: If it is determined that the sample size of the multiple intermediate samples is insufficient, fault data generated by performing forward diffusion noise injection on the initial samples is obtained, and the target condition variable corresponding to the fault data is determined. The fault data and the target condition variable are input into the trained sample generation model for processing to generate intermediate samples to be screened. The sample generation model includes a generator network, a discriminator network, and a degradation feature reconstruction network. Random sampling is performed on the plurality of intermediate samples and the intermediate samples to be screened according to the target sample size to generate target samples related to the target project.

[0194] The above is a schematic scheme of a sample generation device for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to this embodiment. It should be noted that the technical solution of this sample generation device for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products belongs to the same concept as the technical solution of the sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products described above. Details not described in detail in the technical solution of the sample generation device for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products can be found in the description of the technical solution of the sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products described above.

[0195] Figure 3 A structural block diagram of a computing device 300 according to one embodiment of this specification is shown. The components of the computing device 300 include, but are not limited to, a memory 310 and a processor 320. The processor 320 is connected to the memory 310 via a bus 330, and a database 350 is used to store data.

[0196] The computing device 300 also includes an access device 340, which enables the computing device 300 to communicate via one or more networks 360. Examples of these networks include a Public Switched Telephone Network (PSTN), a Local Area Network (LAN), a Wide Area Network (WAN), a Personal Area Network (PAN), or a combination of communication networks such as the Internet. The access device 340 may include one or more of any type of wired or wireless network interface (e.g., a Network Interface Card (NIC)), such as an IEEE 802.11 Wireless Local Area Network (WLAN) interface, a Wi-MAX interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth interface, a Near Field Communication (NFC) interface, and so on.

[0197] In one embodiment of this specification, the aforementioned components of the computing device 300 and Figure 3 Other components, not shown, can also be connected to each other, for example, via a bus. It should be understood that... Figure 3 The block diagram of the computing device shown is for illustrative purposes only and is not intended to limit the scope of this specification. Those skilled in the art can add or replace other components as needed.

[0198] The computing device 300 can be any type of stationary or mobile computing device, including mobile computers or mobile computing devices (e.g., tablet computers, personal digital assistants, laptop computers, notebook computers, netbooks, etc.), mobile phones (e.g., smartphones), wearable computing devices (e.g., smartwatches, smart glasses, etc.) or other types of mobile devices, or stationary computing devices such as desktop computers or PCs. The computing device 300 can also be a mobile or stationary server.

[0199] The processor 320 is used to execute the following computer-executable instructions, which, when executed by the processor, implement the steps of the sample generation method for the general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields.

[0200] The above is a schematic representation of a computing device according to this embodiment. It should be noted that the technical solution of this computing device belongs to the same concept as the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products. Details not described in detail in the technical solution of the computing device can be found in the description of the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0201] An embodiment of this specification also provides a computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the sample generation method described above for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0202] The above is an illustrative scheme of a computer-readable storage medium according to this embodiment. It should be noted that the technical solution of this storage medium belongs to the same concept as the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products. Details not described in detail in the technical solution of the storage medium can be found in the description of the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0203] An embodiment of this specification also provides a computer program, wherein when the computer program is executed in a computer, it causes the computer to perform the steps of the sample generation method described above for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields.

[0204] The above is an illustrative example of a computer program in this embodiment. It should be noted that the technical solution of this computer program belongs to the same concept as the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products. Details not described in detail in the computer program's technical solution can be found in the description of the above-described sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products.

[0205] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0206] The computer instructions include computer program code, which may be in the form of source code, object code, executable file, or certain intermediate forms. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium may be appropriately added to or subtracted according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media may not include electrical carrier signals and telecommunication signals.

[0207] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the embodiments in this specification are not limited to the described order of actions, because according to the embodiments in this specification, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in this specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the embodiments in this specification.

[0208] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0209] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the embodiments described herein. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the embodiments, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A sample generation method for general verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, comprising: Based on the project requirements of the target project, the initial samples stored in the database are filtered to obtain multiple intermediate samples related to the target project; Determine the failure mode corresponding to each intermediate sample, and extract the failure modes that meet the consistency index from each failure mode to obtain the set of failure modes to be verified. Determine the weight corresponding to each fault mode in the set of fault modes to be verified, and determine the target samples related to the target project based on the weights.

2. The sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 1, wherein the step of extracting fault modes that meet the consistency index from each fault mode to obtain a set of fault modes to be verified includes: Divide each fault mode into sets to generate at least two initial fault mode sets; Determine the hierarchical fault data corresponding to each fault mode in the target initial fault mode set, wherein each hierarchical fault data consists of fault modes at different levels, and the target initial fault mode set is each of the at least two initial fault mode sets; A judgment matrix is ​​constructed based on the hierarchical fault data, and the fault modes that meet the consistency index in each fault mode contained in the hierarchical fault data are extracted according to the judgment matrix to obtain a set of fault modes to be verified.

3. The sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 2, wherein the step of extracting fault modes that meet the consistency index from each fault mode included in the hierarchical fault data according to the judgment matrix to obtain a set of fault modes to be verified includes: Solve the characteristic equation constructed based on the judgment matrix and the identity matrix to obtain the maximum eigenvalue of the judgment matrix; The consistency index value is determined based on the maximum eigenvalue, and the consistency ratio is calculated based on the index value. If the consistency ratio is determined to meet the condition, a set of failure modes to be verified is constructed based on the failure modes associated with the judgment matrix.

4. The sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 1, wherein determining the target sample related to the target project based on the weight includes: The fault modes in the set of fault modes to be verified are sorted according to the weights, and the target fault modes are filtered according to the sorting results. Identify the samples to be verified corresponding to the target failure mode, and perform a sufficiency test on the samples to be verified; If the verification passes, the sample to be verified is identified as a target sample related to the target project.

5. The sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 4, wherein the sufficiency test of the sample to be verified includes: The sample to be verified is split to generate at least two sample sets; Obtain the mapping relationship between the failure modes related to the target project and the project parameter values ​​of the target project, and determine the equivalent set of failure modes corresponding to each sample set based on the mapping relationship; The sufficiency measure of the corresponding sample set is determined based on the fault mode equivalence set, and the sufficiency test of the sample to be verified is performed based on the sufficiency measure.

6. The sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 1, further comprising, after obtaining multiple intermediate samples related to the target project: Determine the target sample size for the target samples related to the target project based on the project requirements; If the sample size of the plurality of intermediate samples meets the requirements based on the target sample size, the step of determining the fault mode corresponding to each intermediate sample is performed.

7. The sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products according to claim 1 or 6, further comprising: If it is determined that the sample size of the multiple intermediate samples is insufficient, fault data generated by performing forward diffusion noise injection on the initial samples is obtained, and the target condition variable corresponding to the fault data is determined. The fault data and the target condition variable are input into the trained sample generation model for processing to generate intermediate samples to be screened. The sample generation model includes a generator network, a discriminator network, and a degradation feature reconstruction network. Random sampling is performed on the plurality of intermediate samples and the intermediate samples to be screened according to the target sample size to generate target samples related to the target project.

8. A sample generation device for universal verification of fault detection and diagnosis of mechanical and electronic products in multiple fields, comprising: The filtering module is configured to filter the initial samples stored in the database according to the project requirements of the target project to obtain multiple intermediate samples related to the target project. The extraction module is configured to determine the fault mode corresponding to each intermediate sample and extract the fault modes that meet the consistency index from each fault mode to obtain the set of fault modes to be verified. The determination module is configured to determine the weight corresponding to each fault mode in the set of fault modes to be verified, and to determine the target sample related to the target project based on the weight.

9. A computing device, comprising: Memory and processor; The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions. When the computer-executable instructions are executed by the processor, they implement the steps of the sample generation method for general verification of fault detection and diagnosis of multi-domain mechanical and electronic products as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing computer-executable instructions that, when executed by a processor, implement the steps of the sample generation method for universal verification of fault detection and diagnosis of multi-domain mechanical and electronic products as described in any one of claims 1 to 7.

Images