Fuzzy test method and system based on program control flow

Abstract

The invention relates to the technical field of industrial automation control system testing, and particularly discloses a fuzzy testing method and system based on program control flow, and the method comprises the steps: obtaining parameter data; analyzing the process parameter data and the environment disturbance data to obtain a mutation point; eliminating the process parameter data mutated due to interference of the environmental disturbance data according to the mutational points to obtain target process parameter data; extracting environment disturbance data, execution path data and production quality data in the event window; generating triple data; performing clustering analysis on the triple data in the same event window, and determining an abnormal cluster and a corresponding abnormal type; and according to the abnormal triple data, constructing a fuzzy test case. According to the method, parameter fluctuation under a normal working condition is prevented from being misjudged as system abnormity, internal defects and external interference factors of the system are effectively distinguished, and accurate filtering of external environment interference is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial automation control system testing, and in particular to a fuzzy testing method and system based on program control flow. Background Art

[0002] In industrial automation control systems, the reliability of software control logic directly impacts the stability of the production process and product quality. Fuzz testing, as a dynamic testing technique, triggers abnormal program behavior by inputting unexpected random or semi-random data into the target program, thereby discovering potential software defects.

[0003] However, existing fuzz testing methods have significant limitations in their application in industrial control systems, especially the insufficient ability to detect anomalies in control flow paths. They are unable to accurately identify abnormal execution paths caused by sudden changes in process parameters and environmental disturbances. Moreover, without associating them with production quality data, it is impossible to determine whether abnormal execution paths lead to product quality problems. The generated fuzzy test cases are not targeted enough, and some key defects are difficult to detect during the testing phase, which may lead to production accidents or quality fluctuations. Summary of the Invention

[0004] In response to the above problems, the purpose of the present invention is to provide a fuzzy testing method and system based on program control flow, which avoids misjudging parameter fluctuations under normal working conditions as system abnormalities, effectively distinguishes internal system defects from external interference factors, and realizes accurate filtering of external environmental interference.

[0005] The present invention provides a fuzzy testing method based on program control flow, comprising: Acquiring parameter data; the parameter data includes process parameter data, environmental disturbance data, execution path data and production quality data; Analyzing the process parameter data and the environmental disturbance data by a sliding window technique to obtain a mutation point; Eliminate process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point to obtain target process parameter data; An event window is set with the mutation point of the target process parameter data as the center, and environmental disturbance data, execution path data and production quality data within the event window are extracted; Decompose the execution path data within the same event window into multiple branch decision points, and generate triplet data of parameter-branch decision point-production quality based on the process parameter data, environmental disturbance data, and production quality data corresponding to each branch decision point; Perform cluster analysis on the triplet data within the same event window to determine the abnormal clusters and corresponding abnormal types; The occurrence probability of each abnormal cluster is calculated based on the preset historical data; When the probability of occurrence of an abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster; When the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data; A fuzzy test case is constructed based on the abnormal triple data.

[0006] In a possible implementation, obtaining parameter data includes: Get original parameter data; Normalizing the original parameter data by timestamp to obtain first parameter data in the same time zone; Identify outliers and missing values ​​in the first parameter data, remove the outliers and fill the missing values ​​by linear interpolation to obtain parameter data.

[0007] In a possible implementation, analyzing the process parameter data and the environmental disturbance data using a sliding window technique to obtain a mutation point includes: Using a sliding window of fixed length to segment the process parameter data and environmental disturbance data; Calculate the mean of process parameter data and environmental disturbance data in each sliding window respectively; Calculating a change in the mean value of the process parameter data based on the mean value of the process parameter data in the current sliding window and the mean value of the process parameter data in the sliding window before the current sliding window; When the change value of the mean value of the process parameter data is greater than a preset mutation threshold, determining the starting point of the current sliding window as the mutation point; When the change values ​​of a plurality of consecutive sliding windows are all greater than a preset mutation threshold, the window center moment of the first sliding window among the plurality of consecutive sliding windows is determined as the mutation point.

[0008] In a possible implementation, eliminating process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point to obtain target process parameter data includes: Calculate the change value of the mean value of the environmental disturbance data in the sliding window where the mutation point is located; Calculate the correlation coefficient between the change in the mean value of the process parameter data and the change in the mean value of the environmental disturbance data within the same sliding window; When the correlation coefficient is greater than a preset coefficient, the process parameter data in the sliding window is eliminated.

[0009] In one possible implementation, the correlation coefficient is calculated according to the following formula: : ; Where, represents the correlation coefficient, Indicates the change in the mean value of process parameter data, represents the change in the mean value of the environmental disturbance data, represents the covariance calculation, Indicates the standard deviation of the change in the mean of process parameter data, Indicates the standard deviation of the change in the mean of the environmental disturbance data.

[0010] In a possible implementation, performing cluster analysis on triplet data within the same event window to determine abnormal clusters and corresponding abnormality types includes: Perform cluster analysis on the feature vectors of the triple data within the same event window and divide the triple data into multiple clusters; Calculating a local anomaly factor of triplet data within each cluster, and determining an anomaly score of the cluster based on the local anomaly factor; When the anomaly score is greater than a preset anomaly threshold, determining the cluster as an abnormal cluster; The feature vector of the abnormal cluster is matched with a preset abnormal type judgment statistical table to determine the abnormal type of the abnormal cluster.

[0011] In a possible implementation, when the probability of occurrence of the abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster includes: Count the number of occurrences of anomaly clusters of each anomaly type in all event windows; Calculating the probability of occurrence of abnormal clusters of each abnormal type based on the number of occurrences; When the probability of occurrence of the abnormal cluster is greater than the first preset threshold, a significance test is performed on the probability of occurrence of the abnormal cluster of the abnormal type, and a significance test result is output; When the occurrence probability of an abnormal cluster is less than a significance threshold, the abnormal cluster is determined to be a true abnormal cluster.

[0012] In a possible implementation, when the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data includes: Obtaining triplet data within all the true abnormal clusters; Extracting key fields of triplet data within the true anomaly cluster; When the data corresponding to the key field is greater than a second preset threshold, the triple data is determined to be abnormal triple data.

[0013] In a possible implementation, constructing a fuzzy test case according to the abnormal triple data includes: Extracting numerical distribution characteristics of process parameter data and numerical distribution characteristics of environmental disturbance data in the abnormal triple data, and generating fuzzy test cases according to the numerical distribution characteristics of the process parameter data and the numerical distribution characteristics of the environmental disturbance data; Obtaining execution path data of the fuzzy test case; When the deviation between the execution path data of the fuzzy test case and the expected abnormal path is greater than the preset fault tolerance threshold, the fuzzy test is re-executed to generate a new fuzzy test case; When the deviation value between the execution path data of the fuzzy test case and the expected abnormal path is greater than the preset fault tolerance threshold, the fuzzy test case will be generated as the final fuzzy test case.

[0014] The present invention also provides a fuzzy testing system based on program control flow, which is used to execute any of the above-mentioned fuzzy testing methods, including: An acquisition module, configured to acquire parameter data; the parameter data includes process parameter data, environmental disturbance data, execution path data, and production quality data; An analysis module, configured to analyze the process parameter data and the environmental disturbance data using a sliding window technique to obtain a mutation point; a removal module, configured to remove process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point, to obtain target process parameter data; An extraction module, configured to set an event window centered on a mutation point of the target process parameter data, and extract environmental disturbance data, execution path data, and production quality data within the event window; The triplet module is used to decompose the execution path data within the same event window into multiple branch decision points, and generate triplet data of parameter-branch decision point-production quality based on the process parameter data, environmental disturbance data and production quality data corresponding to each branch decision point; Clustering module, used to perform cluster analysis on triplet data within the same event window to determine abnormal clusters and corresponding abnormal types; A calculation module is used to calculate the occurrence probability of each abnormal cluster based on preset historical data; When the probability of occurrence of an abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster; When the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data; A construction module is used to construct a fuzzy test case based on the abnormal triple data.

[0015] The program control flow-based fuzzy testing method and system provided by the present invention can effectively identify pseudo-abnormal data caused by external interference by introducing bivariate correlation analysis, achieve accurate filtering of external environmental interference, avoid misjudging parameter fluctuations under normal operating conditions as system abnormalities, and effectively distinguish internal system defects from external interference factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A flowchart of a fuzzy testing method based on program control flow provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The following detailed description of the embodiments of the present invention is provided in conjunction with the accompanying drawings and examples. The following detailed description of the embodiments and the accompanying drawings are intended to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention. That is, the present invention is not limited to the preferred embodiments described, and the scope of the present invention is defined by the claims.

[0018] In the description of the present invention, it should be noted that, unless otherwise specified, “plurality” means two or more; the terms “first”, “second”, etc. are used for descriptive purposes only and cannot be understood as indicating or implying relative importance; for ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances.

[0019] Existing fuzzy testing methods for industrial automation control systems suffer from insufficient detection of control flow path anomalies. This is particularly problematic, as anomalous execution paths caused by the combined effects of process parameter mutations and environmental disturbances are difficult to effectively identify. Existing techniques fail to incorporate production quality data into their analysis frameworks, resulting in test cases failing to cover the anomalous paths that truly impact product quality. For example, in an injection molding production line, voltage fluctuations can trigger temperature parameter mutations, causing the control system execution path to deviate from expectations. However, existing methods are unable to distinguish whether parameter mutations are caused by environmental disturbances or equipment failures, resulting in a lack of targeted test cases.

[0020] In order to solve the above problems, the inventors observed that the generation of abnormal paths in industrial control systems is often accompanied by the coupling effect of process parameter mutations and environmental disturbances, while the existing methods only focus on a single data source, resulting in a high misjudgment rate; by analyzing historical data, it was found that there is a strong correlation between production quality indicators and abnormal execution paths, but the existing test case generation mechanism does not establish this correlation; based on this, the inventors proposed to perform spatiotemporal alignment analysis of process parameters, environmental disturbances, execution paths and production quality data, and construct fuzzy test cases by dynamically screening abnormal triplet data.

[0021] Figure 1A flowchart of a fuzzy testing method based on program control flow provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the present invention provides a fuzzy testing method based on program control flow, comprising: Step S1, obtaining parameter data; In one possible implementation, original parameter data is obtained; the original parameter data is timestamp-normalized to obtain first parameter data unified into the same time zone; outliers and missing values ​​in the first parameter data are identified, outliers are eliminated, and missing values ​​are filled by linear interpolation to obtain parameter data.

[0022] Parameter data includes process parameter data, environmental disturbance data, execution path data, and production quality data. Process parameter data includes timing parameters such as temperature, pressure, and flow. Environmental disturbance data includes voltage, current, temperature and humidity, electromagnetic interference, etc.; execution path data includes control process logs, state transition records, branch jumps, and other event sequences. Production quality data includes pass rate, number of defective products, defect type, detection indicators, etc. Process parameter data refers to the core physical quantity that reflects the operating status of the industrial automation control system. Specifically, it can be collected in real time using industrial instruments such as temperature sensors, pressure transmitters, and flow meters to characterize the steady-state characteristics of the production process.

[0023] Environmental disturbance data refers to external interference factors that may affect the stability of the control system. Specifically, it can be collected using equipment such as power monitoring devices, temperature and humidity sensors, and electromagnetic field detectors to identify the coupling effect of uncontrolled variables on process parameters.

[0024] Execution path data refers to the logical branch records generated during the operation of the control program. It can be extracted through the event log of the PLC controller or the debugging interface of the embedded system to track the dynamic changes of the control flow.

[0025] Production quality data refers to quantitative indicators that reflect the results of product manufacturing. It can be obtained through visual inspection systems, quality analyzers or manual sampling records, and is used to establish the correlation between control flow anomalies and quality defects.

[0026] Timestamp standardization refers to unifying the time base of multi-source heterogeneous data to the same time zone. It can be implemented using a network time protocol or a hardware clock synchronization module to eliminate timing deviations between data acquisition devices.

[0027] Outlier removal refers to identifying and removing outlier data points that are outside a reasonable range. This can be achieved using a box plot method or a standard deviation threshold method to prevent sensor failure or transient interference from interfering with subsequent analysis.

[0028] Missing value filling refers to the numerical restoration of gaps caused by interruptions in data collection. It can be implemented by using a linear interpolation algorithm of adjacent data points to ensure the continuity of the data sequence.

[0029] Specifically, in industrial automation control systems, process parameters such as temperature and pressure are collected in real time through distributed sensor networks, forming time series data at the second or millisecond level; environmental disturbance data such as voltage fluctuations and electromagnetic interference are collected through independent monitoring devices, forming a parallel data stream with the operating data of the control system; the control process log records branch jump events during the execution of the PLC program, and the state transition record saves the switching time point of the equipment operation mode; production quality data is periodically generated by online detection equipment, reflecting the distribution characteristics of the qualified rate of batch products; after collection, multi-source data is first timestamp aligned, for example, the device clocks in different time zones are uniformly converted to Coordinated Universal Time to ensure the temporal consistency of the event sequence; then the data is detected for anomalies. For example, when the pressure sensor records a sudden increase in value that exceeds the equipment safety threshold, the data point is marked as an anomaly and eliminated; for data missing due to network interruption, the average value of the two valid data points before and after is used for interpolation and filling to form a complete time series data set.

[0030] Traditional methods usually only collect process parameters and execution path data, ignoring the potential impact of environmental disturbances on the control system, resulting in a lack of correlation basis for abnormal path analysis with external interference factors; existing data cleaning technologies mostly use a single threshold filtering method, which cannot distinguish between abnormal values ​​caused by sensor failures and environmental disturbances, and easily leads to the accidental deletion of valid information; the present invention establishes a complete causal relationship chain between control flow anomalies and external interference and quality defects by constructing a multi-dimensional data system that includes environmental disturbances and production quality; uses timestamp standardization processing to solve the problem of time sequence misalignment caused by differences in acquisition frequency of multi-source data, such as matching the millisecond-level log of the PLC controller with the minute-level quality inspection data on the time axis; the combined strategy of outlier removal and missing value filling eliminates invalid noise and maintains data continuity. For example, in a voltage sag event, the process parameter fluctuations caused by real disturbances are retained, and only outliers caused by sensor anomalies are eliminated.

[0031] The present invention effectively solves the problems of fragmentation and time sequence asynchrony in multi-source data collection in industrial scenarios, providing a high-integrity data foundation for subsequent abnormal path identification; timestamp standardization processing ensures the precise correspondence between control flow events and process parameter changes, such as establishing a millisecond-level correlation between branch jump times and ambient temperature and humidity fluctuations; the outlier cleaning mechanism eliminates the interference of equipment fault data, so that process parameter mutations caused by real environmental disturbances can be accurately identified; missing value filling maintains the continuity of the data sequence, avoiding the failure of event window analysis due to collection interruption; the construction of a multi-dimensional data system enables control flow anomalies to not only be associated with execution path characteristics, but also to synchronously analyze external interference factors and quality indicator changes, providing more comprehensive input conditions for fuzzy test case generation.

[0032] Step S2, analyzing the process parameter data and the environmental disturbance data by a sliding window technique to obtain a mutation point; In one possible implementation, a sliding window of fixed length is used to segment process parameter data and environmental disturbance data; the mean of the process parameter data and the mean of the environmental disturbance data in each sliding window are calculated respectively; the change value of the mean of the process parameter data is calculated based on the mean of the process parameter data in the current sliding window and the mean of the process parameter data in the sliding window before the current sliding window; when the change value of the mean of the process parameter data is greater than a preset mutation threshold, the starting point of the current sliding window is determined as the mutation point; when the change values ​​of multiple consecutive sliding windows are all greater than the preset mutation threshold, the window center moment of the first sliding window in the multiple consecutive sliding windows is determined as the mutation point.

[0033] Among them, sliding window technology refers to the technology of using a fixed-length time window to segment time series data. Specifically, it can be implemented by sliding a window with a length of 5 seconds in a step size of 1 second to capture the mutation characteristics of process parameters and environmental disturbances.

[0034] Among them, the fixed-length sliding window refers to dividing the time series data into multiple continuous data segments of equal length. Specifically, it can be implemented by using a time series segmentation algorithm to dynamically analyze the correlation changes between process parameters and environmental disturbances.

[0035] The mean of process parameter data and environmental disturbance data refers to the arithmetic mean of the data in each sliding window. It can be calculated by accumulating the data in the window and dividing it by the window length. It is used to eliminate the interference of random fluctuations on mutation point detection.

[0036] The mutation threshold refers to the critical value for determining whether a process parameter has changed significantly. It can be set based on historical data statistics or expert experience to distinguish normal fluctuations from abnormal mutations.

[0037] Multiple consecutive sliding windows refer to multiple consecutive data segments that meet the mutation condition. Specifically, a counter can be used to record the number of windows that meet the condition continuously to eliminate misjudgments caused by occasional environmental disturbances.

[0038] Specifically, the process parameter data and environmental disturbance data are divided into sliding windows of fixed length, and the mean in each window is calculated and compared with the mean of its previous window; when the change value of the process parameter mean exceeds the preset mutation threshold, the starting point of the window is marked as a candidate mutation point. If the change values ​​of the process parameter mean in subsequent consecutive windows all exceed the threshold, the center moment of the first window is determined to be the true mutation point; through the judgment conditions of continuous windows, single mutations caused by occasional environmental disturbances are filtered out, and only data segments with continuous abnormal changes are retained, thereby accurately identifying true anomalies caused by the coupling of environmental disturbances and process parameters.

[0039] Traditional methods usually use single-point mutation detection or static threshold judgment, which cannot distinguish between instantaneous interference caused by environmental disturbances and real process anomalies. The present invention uses dynamic sliding window analysis combined with continuous window mutation consistency verification to effectively eliminate occasional noise interference and improve the accuracy of mutation point detection. For example, the existing technology may mistakenly judge temperature parameter jumps caused by instantaneous voltage fluctuations as process anomalies, while the present invention can identify such instantaneous interference through a continuous window verification mechanism to avoid misscreening.

[0040] The present invention can accurately identify the real mutation points in process parameter data caused by environmental disturbances, avoiding misjudging data fluctuations caused by occasional disturbances as process anomalies; the screened mutation points provide a reliable time benchmark for subsequent analysis, ensuring that the correlation analysis between abnormal execution paths and production quality data is based on real abnormal data, thereby improving the pertinence of fuzzy test cases.

[0041] Step S3, eliminating process parameter data that has been mutated due to interference from environmental disturbance data from the process parameter data according to the mutation points, to obtain target process parameter data; In one possible implementation, the change value of the mean value of the environmental disturbance data in the sliding window where the mutation point is located is calculated; the correlation coefficient between the change value of the mean value of the process parameter data and the change value of the mean value of the environmental disturbance data in the same sliding window is calculated; when the correlation coefficient is greater than a preset coefficient, the process parameter data in the sliding window is eliminated.

[0042] In one possible implementation, the correlation coefficient is calculated according to the following formula: : ; Where, represents the correlation coefficient, Indicates the change in the mean value of process parameter data, represents the change in the mean value of the environmental disturbance data, represents the covariance calculation, Indicates the standard deviation of the change in the mean of process parameter data, Indicates the standard deviation of the change in the mean of the environmental disturbance data.

[0043] The change in the mean value of the environmental disturbance data in the sliding window where the mutation point is located refers to the fluctuation amplitude of the environmental parameters obtained by the sliding window analysis method during the mutation period. It can be calculated by the adjacent window mean difference method and is used to characterize the potential impact of environmental factors on process parameters. The correlation coefficient refers to the correlation index of the change trend between process parameters and environmental disturbances in the same time window. It can be calculated using the Pearson correlation coefficient algorithm to quantify the synchronization of the changes between the two. The preset coefficient refers to a pre-configured screening threshold, which can be dynamically adjusted according to the tolerance of environmental interference in industrial scenarios, and is used to distinguish the boundary conditions between internal anomalies of the system and external interference.

[0044] Specifically, after detecting the process parameter mutation point, the mean change of the environmental disturbance data corresponding to the time window is extracted synchronously; by calculating the correlation coefficient between the process parameters and the environmental disturbance changes, it can be determined whether the two show a significant positive or negative correlation. When the correlation coefficient exceeds the set threshold, it indicates that the process parameter mutation is mainly caused by environmental factors. At this time, the process parameter data in the window is marked as invalid data and a elimination operation is performed. For example, in the temperature parameter mutation window, if the correlation coefficient between the voltage fluctuation and the temperature change is detected to reach 0.85, and the set threshold is 0.8, it is determined that the temperature change is a normal fluctuation caused by external interference, thereby eliminating the interference of the data on subsequent analysis.

[0045] Traditional methods only judge anomalies based on the mutation amplitude of the process parameters themselves, without considering the coupling relationship between environmental disturbances and process parameters. The present invention can effectively identify pseudo-abnormal data caused by external interference by introducing bivariate correlation analysis. For example, the existing technology may misjudge temperature fluctuations caused by voltage drops as temperature control system failures, while the present invention can accurately identify such interference events by calculating the correlation coefficient between the two.

[0046] Through the above technical solution, the present invention achieves accurate filtering of external environmental interference, avoids misjudging parameter fluctuations under normal operating conditions as system abnormalities, effectively distinguishes internal system defects from external interference factors, and provides a high-quality data foundation for subsequent abnormality detection.

[0047] Step S4, setting an event window with the mutation point of the target process parameter data as the center, and extracting environmental disturbance data, execution path data, and production quality data within the event window; The event window refers to a time interval extending from the parameter mutation point as the center, for example, 3 seconds before and after the mutation point, which is used to capture the control flow branch decisions associated with the parameter mutation.

[0048] In step S5, the execution path data within the same event window is decomposed into multiple branch decision points, which are numbered. Based on the process parameter data, environmental disturbance data, and production quality data corresponding to each branch decision point, a triplet of parameter-branch decision point-production quality data is generated. For example: process parameter + environmental disturbance, branch decision point + number, production quality.

[0049] Among them, triplet data refers to binding the branch decision point with the process parameters, environmental disturbances and quality indicators at the corresponding moment. This can be achieved through timestamp alignment to form a structured data record.

[0050] The triple data of all branch decision points are summarized to form a structured table of triple data, as shown in Table 1 below.

[0051] Table 1

[0052] Among them, the branch decision point refers to the logical judgment node in the control process that triggers the program execution path switch; Triple data refers to a data unit consisting of process parameters and environmental disturbance combination parameters, branch decision point numbers, and production quality indicators.

[0053] A structured table refers to a data storage format that is arranged in the order of branch decision point numbers. It can be implemented through a database table or a two-dimensional array structure, and is used to standardize the storage of multi-dimensional related data for batch processing.

[0054] Specifically, during the operation of the industrial automation control system, the execution path data is decomposed into N branch decision points and assigned unique numbers; for each branch decision point, the process parameters and environmental disturbance data within the set time range before and after the trigger moment are extracted, and the production quality indicators corresponding to the time period are associated, such as obtaining the defect rate data of the batch of products; by merging the process parameters and environmental disturbance data into combined parameters, and binding them with the branch decision point number and quality indicators, triple data with time and space correlation is formed, and the data of all branch decision points are summarized in a structured table.

[0055] Traditional methods only record execution path switching events without establishing a dynamic association between branch decision points and process parameters; the present invention achieves dynamic matching of control flow branches with process parameters and environmental disturbances by decomposing branch decision points and constructing a triplet data structure; in the prior art, production quality data is usually stored independently in a statistical system, while the present invention directly associates it with branch decision points, thereby revealing the causal relationship between abnormal paths and quality problems.

[0056] Through the above technical solution, the present invention solves the problem of difficulty in association analysis caused by the separate storage of control flow branches, process parameters and quality data, and realizes the spatiotemporal alignment and structured integration of multi-source heterogeneous data; this data association mechanism provides a directly processable data basis for subsequent analysis of the impact of abnormal execution paths on product quality.

[0057] Step S6: performing cluster analysis on the triplet data within the same event window to determine the abnormal cluster and the corresponding abnormal type; In one possible implementation, cluster analysis is performed on the feature vectors of triple data within the same event window to divide the triple data into multiple clusters; the local anomaly factor of the triple data within each cluster is calculated, and the anomaly score of the cluster is determined based on the local anomaly factor; when the anomaly score is greater than a preset anomaly threshold, the cluster is determined to be an abnormal cluster; the feature vector of the abnormal cluster is matched with a preset anomaly type criterion statistical table to determine the anomaly type of the abnormal cluster.

[0058] Abnormal cluster analysis involves classifying multidimensional feature vectors using a density clustering algorithm, such as the DBSCAN algorithm, to identify outlier clusters. Feature vectors include core features such as cluster center parameters, branches, and production quality. Table 2 shows a statistical table of abnormality classification criteria.

[0059] Table 2

[0060] Among them, feature vector processing refers to converting three types of heterogeneous data, namely process parameters, branch decision points, and production quality, into numerical vectors of unified dimensions. Specifically, it can be achieved by using normalization processing and principal component analysis methods to eliminate data dimensional differences and extract key features.

[0061] The local anomaly factor is an indicator that measures the difference in density between a data point and its neighboring data. It can be achieved by calculating the ratio of the average reachable distance between each data point and its K nearest neighbors, and is used to quantify the degree of anomaly within the cluster. The abnormality type judgment statistical table refers to a pre-established mapping relationship table between abnormal features and quality defects. Specifically, it can be constructed using a historical fault case library and is used to quickly match abnormality types based on clustering core features.

[0062] Specifically, the process parameters, environmental disturbances, branch decision points and production quality data are first converted into standardized feature vectors to eliminate the impact of parameter unit differences on the analysis; then a clustering algorithm is used to divide the triplet data in the same event window into several clusters, and the abnormal clusters formed by isolated points are identified through density distribution; for each abnormal cluster, its local anomaly factor is calculated and converted into an anomaly score. When the anomaly score exceeds the preset threshold, it is determined to be an abnormal cluster; finally, the central characteristic parameters of the abnormal cluster are extracted and matched with the parameter fluctuation range, branch jump pattern and quality defect type recorded in the preset criterion table to determine the anomaly type as parameter out-of-limit anomaly or control logic error anomaly.

[0063] Existing methods only judge parameter anomalies through a single threshold, without considering the correlation between control flow branches and quality data, resulting in a high misjudgment rate; the present invention fuses multidimensional data through feature vectors, combines cluster analysis with local anomaly factor calculation, and can accurately identify clustered anomaly clusters caused by parameter mutations and control flow anomalies; at the same time, through matching anomaly types through a preset criterion table, data anomalies are directly associated with quality defects, solving the problem that existing technologies cannot determine whether abnormal paths affect product quality.

[0064] The present invention can accurately identify abnormal execution paths caused by sudden changes in process parameters and environmental disturbances, and clarify the types of abnormalities based on production quality data; by associating abnormal clustering with quality defects, it can effectively screen out abnormal paths that truly affect product quality, providing data support for generating targeted fuzzy test cases, thereby improving the accuracy and effectiveness of industrial control system testing.

[0065] Step S7, calculating the occurrence probability of each abnormal cluster based on preset historical data; In one possible implementation, the occurrence probability refers to the frequency of occurrence of a specific abnormal pattern based on historical data statistics, for example, by using a Bayesian probability model to calculate the conditional probability.

[0066] The probability of occurrence is calculated according to the following formula: ; Where, Indicates the exception type The probability of occurrence of abnormal clustering, Indicates the exception type The number of occurrences of abnormal clusters in all event windows, Indicates the total number of event windows.

[0067] Step S8: When the probability of occurrence of the abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster; In one possible implementation, the number of occurrences of abnormal clusters of each abnormal type in all event windows is counted; the probability of occurrence of abnormal clusters of each abnormal type is calculated based on the number of occurrences; when the probability of occurrence of an abnormal cluster is greater than a first preset threshold, a significance test is performed on the probability of occurrence of abnormal clusters of the abnormal type, and the significance test result is output; when the probability of occurrence of an abnormal cluster is less than the significance threshold, the abnormal cluster is determined to be a true abnormal cluster.

[0068] Counting the number of occurrences of abnormal clusters of each abnormal type in all event windows refers to recording the abnormal cluster types and their occurrence frequencies by traversing all event windows. This can be achieved by using a hash table or a counter to quantify the frequency characteristics of abnormal occurrences. Calculating the probability of occurrence based on the number of occurrences means dividing the number of occurrences of a certain type of abnormal cluster by the total number of event windows. This can be achieved using frequency statistics to filter out high-frequency abnormal types. Conducting a significance test refers to verifying the statistical hypothesis of the probability of occurrence of high-frequency anomaly types. Specifically, it can be implemented using a chi-square test or a t-test to verify whether the anomaly pattern is statistically significant. Determining whether the significance test result is less than the significance threshold means comparing the p-value obtained by the test with the preset significance level, which can be specifically implemented by using a numerical comparator to determine the statistical credibility of the abnormal clustering.

[0069] Specifically, anomaly clustering first uses frequency statistics to filter out high-frequency anomaly types whose probability of occurrence exceeds a preset threshold. For example, when a certain type of anomaly occurs 20 times in a 100-event window, its probability of occurrence is 20%. Then, a significance test is performed on the high-frequency anomalies that exceed the threshold, such as using a chi-square test to verify whether the probability of the anomaly occurring is significantly higher than the normal fluctuation level of the system; If the p-value obtained by the test is less than the significance threshold of 0.05, it is determined to be a true abnormal cluster. This double verification mechanism reduces the interference of accidental anomalies through frequency screening, and then eliminates the influence of random fluctuations through statistical testing, and finally identifies true abnormal patterns with statistical significance.

[0070] Existing methods usually rely only on a single frequency threshold to judge anomalies. For example, if the number of occurrences exceeds 5, it is directly determined to be a real anomaly. This method cannot distinguish between high-frequency accidental fluctuations and systematic anomalies, and it is easy to misjudge anomalies caused by temporary environmental interference as real defects. The present invention introduces a significance test and adds statistical inference verification on the basis of frequency screening. For example, when the probability of an anomaly occurring is high but the statistical test shows that it has no significant difference from the random distribution, it will still be excluded. This dual verification mechanism significantly improves the accuracy of anomaly discrimination.

[0071] The present invention effectively solves the problem of misjudgment of anomalies caused by environmental disturbances in industrial control systems. It verifies the significance of high-frequency anomaly patterns through statistical tests, avoiding incorporating accidental anomalies caused by temporary interference into test case generation. The significance test can identify anomaly patterns with stable statistical characteristics. For example, a certain type of anomaly exhibits a significantly high incidence feature in different time periods of morning, noon and evening, thereby ensuring that the constructed fuzzy test cases can cover real control flow defects.

[0072] Step S9, when the triple data in the true abnormal cluster is greater than the second preset threshold, determining the triple data as abnormal triple data; In one possible implementation, triple data within all true abnormal clusters are obtained; key fields of the triple data within the true abnormal clusters are extracted; and when the data corresponding to the key field is greater than a second preset threshold, the triple data is determined to be abnormal triple data.

[0073] The second preset threshold refers to the critical value used to quantitatively judge production quality anomalies. It can be implemented by using a non-conforming rate threshold, a defect type ratio threshold, or a parameter extreme value threshold. For example, when the ratio of a certain type of defect exceeds the preset 5%, the judgment is triggered. The threshold is set through statistical analysis of historical quality data and is used to screen abnormal data that directly affects product quality.

[0074] Key field extraction refers to separating process parameters, control flow branch identifiers, and production quality indicators from triples. Specifically, structured query statements can be used to implement field screening and establish an association model between control flow paths and quality defects.

[0075] Specifically, after obtaining a true anomaly cluster that has been probabilistically verified, all triplet data sets within the cluster are first loaded. The second preset threshold is dynamically configured through the quality inspection standard database, for example, the threshold for the number of defective products is set to 2% of the batch output. The data parsing module extracts key fields from the triplet, such as the process parameter fluctuation range, the number of control flow branch jumps, and the distribution of defect types. The failure rate indicator of each triplet is compared item by item with the judgment threshold. When the parameter fluctuation range of a triplet is detected to exceed the safety threshold and is accompanied by a specific defect type, it is determined to be an abnormal triplet. This process ensures that the screened abnormal data meets the dual conditions of control flow anomalies and quality defects through a quality association verification mechanism.

[0076] Traditional fuzz testing methods only generate test cases based on program execution path anomalies, and do not establish a correlation analysis between control flow anomalies and production quality data; the present invention introduces a multi-dimensional judgment system and uses quality defect indicators as necessary conditions for anomaly judgment, such as requiring that abnormal paths lead to an increase in the failure rate or the generation of specific defects; test cases in the existing technology often cover invalid abnormal paths, while the present invention uses a double verification mechanism to only retain abnormal data that actually affects product quality as the basis for test case generation, significantly improving test effectiveness.

[0077] Through the above technical solution, the present invention solves the problem that fuzzy test cases cannot effectively cover actual quality defects, and realizes the precise association between control flow anomalies and production quality defects; by setting quantitative judgment thresholds, non-quality-related program anomalies can be filtered out; through key field extraction and threshold comparison, abnormal paths that directly affect product qualification rates can be identified; through the joint analysis of quality indicators and control flow data, it is ensured that the generated test cases are targeted at actual production defects, thereby improving the targeting and reliability of industrial control system testing.

[0078] Step S10: construct a fuzzy test case based on the abnormal triple data.

[0079] In one possible implementation, the numerical distribution characteristics of the process parameter data and the numerical distribution characteristics of the environmental disturbance data in the abnormal triple data are extracted, and a fuzzy test case is generated based on the numerical distribution characteristics of the process parameter data and the numerical distribution characteristics of the environmental disturbance data; the execution path data of the fuzzy test case is obtained; when the deviation value of the execution path data of the fuzzy test case and the expected abnormal path is greater than a preset fault tolerance threshold, the fuzzy test is re-executed to generate a new fuzzy test case; until the deviation value of the execution path data of the fuzzy test case and the expected abnormal path is greater than the preset fault tolerance threshold, the fuzzy test case is generated as the final fuzzy test case.

[0080] Among them, the numerical distribution characteristics refer to the statistical laws exhibited by process parameters and environmental disturbances in abnormal triples. Specifically, they can be modeled using probability density functions or Gaussian mixture models to describe the correlation between the parameter fluctuation range and abnormal conditions.

[0081] Fuzzy test cases refer to a set of test inputs generated based on abnormal data features. They can be generated through Monte Carlo sampling or genetic algorithms to simulate the combined conditions of process parameter mutations and environmental disturbances in actual production environments.

[0082] The deviation value between the execution path data and the expected exception path refers to the degree of difference between the actual control flow branch and the preset exception path. Specifically, it can be calculated using Hamming distance or cosine similarity to quantify the effectiveness of the test case triggering the exception path.

[0083] The fault tolerance threshold refers to the maximum range of execution path deviation allowed. It can be set based on the statistical variance of the historical normal path and is used to determine whether the test case requires iterative optimization.

[0084] The fuzz test is re-executed through the test case iteration signal. The test case iteration signal refers to the instruction that triggers the regeneration of the test case. It can be specifically achieved by setting a threshold of the number of times the deviation value exceeds the limit, which is used to establish a dynamic optimization mechanism.

[0085] Specifically, after the numerical distribution characteristics of process parameters and environmental disturbances in the abnormal triplet data are extracted, fuzzy test cases with similar statistical characteristics are generated through a probabilistic model; after these fuzzy test cases are injected into the industrial automation control system, the system execution path data and production quality data are collected in real time; by comparing the deviation value between the actual execution path and the expected abnormal path, it can be verified whether the test case effectively triggers the target abnormal path; when the deviation value exceeds the fault tolerance threshold, the system automatically triggers the test case iteration process and regenerates the optimized fuzzy test case based on the latest collected data.

[0086] In some specific embodiments, the numerical distribution characteristics of process parameters and environmental disturbances can be modeled by a Gaussian mixture model, for example, decomposing the joint distribution of temperature parameters and voltage disturbances into multiple Gaussian components; the fault tolerance threshold can be set to 3 times the standard deviation of the Hamming distance of the historical normal path. For example, when the actual deviation exceeds the threshold, the system automatically marks the test case as an inefficient case and starts iteration.

[0087] Traditional fuzz testing methods rely on random input generation, cannot associate the synergistic effects of process parameter mutations and environmental disturbances, and lack a dynamic optimization mechanism based on execution path deviations; the present invention uses abnormal triplet data feature extraction and closed-loop verification feedback to enable test cases to accurately reproduce abnormal conditions in real production environments, and continuously improve test coverage through iterative optimization.

[0088] Through the above technical solution, the present invention solves the problem that traditional fuzzy test cases cannot cover the combined effects of process parameter mutations and environmental disturbances, ensuring that the test input can effectively trigger the path branch that causes production quality abnormalities; through deviation value calculation and fault tolerance threshold judgment, quantitative evaluation and dynamic optimization of test case effectiveness are realized, avoiding the repeated execution of invalid test cases.

[0089] In summary, after the process parameter data and environmental disturbance data are processed through a sliding window, parameter mutation points are identified through mean change detection. For example, in the injection molding machine temperature control scenario, when the temperature mean in a certain window changes by more than 10°C compared to the previous window, it is marked as a mutation point. The correlation coefficient between the environmental disturbance parameter and the process parameter in this window is then calculated. When the correlation coefficient is greater than 0.7, it is determined to be a pseudo-mutation caused by environmental interference and is eliminated. For true mutation points, a 30-second event window is set with this point as the center, and the environmental disturbance data, execution path data, and production quality data within this period are simultaneously extracted. Each branch decision point is decomposed into independent decision units, such as conditional judgment nodes in the PLC control process, and triple data is constructed with the corresponding environmental disturbance data, execution path data, and production quality data. In the cluster analysis stage, the temperature fluctuation amplitude, branch jump frequency, product defect rate and other features in the triplet data are standardized, and abnormal clusters are identified through the clustering algorithm. For example, if a cluster of data is detected to have the combined characteristics of high temperature deviation, frequent branch jump and high defective rate, it is determined to be an overheating abnormality type; by calculating the probability of occurrence of this type of abnormality in historical events, when the probability exceeds the preset threshold, it is determined to be a real abnormal cluster that needs to be tested, and then the abnormal triples that meet the judgment threshold are extracted to generate test cases.

[0090] Traditional methods usually analyze environmental disturbance data or process parameters separately without establishing multi-dimensional data association. For example, in the test of a certain injection molding machine control system, existing tools detected temperature parameter anomalies but did not associate them with voltage fluctuation data, resulting in the generated test cases failing to reproduce the actual fault scenario. However, this method synchronously analyzes environmental disturbances and process parameter changes through event windows. At the same time, existing test case generators do not consider production quality data and may perform redundant tests on path anomalies that do not affect product quality. However, the present invention generates test cases only for abnormal execution paths that cause product defects through criterion threshold filtering.

[0091] The present invention can effectively distinguish between pseudo-anomalies caused by environmental disturbances and real equipment failures, and accurately locate abnormal execution paths that lead to product quality degradation; by establishing a dynamic correlation model between control flow branches and process parameters, environmental disturbances and quality indicators, it significantly improves the pertinence and effectiveness of test cases.

[0092] The present invention also provides a fuzzy testing system based on program control flow, which is used to execute any of the above-mentioned fuzzy testing methods, including: An acquisition module is used to acquire parameter data; the parameter data includes process parameter data, environmental disturbance data, execution path data and production quality data; An analysis module is used to analyze process parameter data and environmental disturbance data using a sliding window technique to obtain mutation points; A elimination module is used to eliminate process parameter data that has been mutated due to interference from environmental disturbance data according to mutation points, and obtain target process parameter data; An extraction module is used to set an event window centered on the mutation point of the target process parameter data and extract environmental disturbance data, execution path data, and production quality data within the event window; The triplet module is used to decompose the execution path data within the same event window into multiple branch decision points, and generate triplet data of parameter-branch decision point-production quality based on the process parameter data, environmental disturbance data and production quality data corresponding to each branch decision point; Clustering module, used to perform cluster analysis on triplet data within the same event window to determine abnormal clusters and corresponding abnormal types; A calculation module is used to calculate the occurrence probability of each abnormal cluster based on preset historical data; When the probability of occurrence of the abnormal cluster is greater than a first preset threshold, determining the abnormal cluster as a true abnormal cluster; When the triple data in the true abnormal cluster is greater than a second preset threshold, determining the triple data as abnormal triple data; The construction module is used to construct fuzzy test cases based on abnormal triple data.

[0093] Compared with the prior art, the program control flow-based fuzz testing method and system provided by the present invention have the following beneficial effects: 1) By introducing bivariate correlation analysis, the present invention can effectively identify pseudo-abnormal data caused by external interference, achieve accurate filtering of external environmental interference, avoid misjudging parameter fluctuations under normal operating conditions as system anomalies, and effectively distinguish internal system defects from external interference factors.

[0094] 2) This invention fuses multidimensional data through feature vectors, combines cluster analysis with local anomaly factor calculation, and can accurately identify clustered anomaly clusters caused by parameter mutations and control flow anomalies. At the same time, it matches anomaly types through a preset criterion table and directly associates data anomalies with quality defects. This solves the problem that existing technologies are unable to determine whether abnormal paths affect product quality, and provides data support for generating targeted fuzzy test cases.

[0095] 3) Through the extraction of abnormal triplet data features and closed-loop verification feedback, the present invention enables test cases to accurately reproduce abnormal conditions in real production environments, and continuously improves test coverage through iterative optimization. This solves the problem that traditional fuzzy test cases cannot cover the combined effects of process parameter mutations and environmental disturbances, ensuring that the test input can effectively trigger the path branch that leads to production quality anomalies.

[0096] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A fuzz testing method based on program control flow, characterized in that: include: Acquiring parameter data; the parameter data includes process parameter data, environmental disturbance data, execution path data and production quality data; Analyzing the process parameter data and the environmental disturbance data by a sliding window technique to obtain a mutation point; Eliminate process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point to obtain target process parameter data; An event window is set with the mutation point of the target process parameter data as the center, and environmental disturbance data, execution path data and production quality data within the event window are extracted; Decompose the execution path data within the same event window into multiple branch decision points, and generate triplet data of parameter-branch decision point-production quality based on the process parameter data, environmental disturbance data, and production quality data corresponding to each branch decision point; Perform cluster analysis on the triplet data within the same event window to determine the abnormal clusters and corresponding abnormal types; The occurrence probability of each abnormal cluster is calculated based on the preset historical data; When the probability of occurrence of an abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster; When the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data; A fuzzy test case is constructed based on the abnormal triple data.

2. The fuzz testing method according to claim 1, wherein: The acquiring parameter data includes: Get original parameter data; Normalizing the original parameter data by timestamp to obtain first parameter data in the same time zone; Identify outliers and missing values ​​in the first parameter data, remove the outliers and fill the missing values ​​by linear interpolation to obtain parameter data.

3. The fuzz testing method according to claim 1, wherein: The analyzing the process parameter data and the environmental disturbance data by the sliding window technology to obtain the mutation point includes: Using a sliding window of fixed length to segment the process parameter data and environmental disturbance data; Calculate the mean of process parameter data and environmental disturbance data in each sliding window respectively; Calculating a change in the mean value of the process parameter data based on the mean value of the process parameter data in the current sliding window and the mean value of the process parameter data in the sliding window before the current sliding window; When the change value of the mean value of the process parameter data is greater than a preset mutation threshold, determining the starting point of the current sliding window as the mutation point; When the change values ​​of a plurality of consecutive sliding windows are all greater than a preset mutation threshold, the window center moment of the first sliding window among the plurality of consecutive sliding windows is determined as the mutation point.

4. The fuzzy testing method according to claim 3, wherein: Eliminating process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point to obtain target process parameter data includes: Calculate the change value of the mean value of the environmental disturbance data in the sliding window where the mutation point is located; Calculate the correlation coefficient between the change in the mean value of the process parameter data and the change in the mean value of the environmental disturbance data within the same sliding window; When the correlation coefficient is greater than a preset coefficient, the process parameter data in the sliding window is eliminated.

5. The fuzz testing method according to claim 4, wherein: Also includes: The correlation coefficient is calculated according to the following formula : ； Where, represents the correlation coefficient, Indicates the change in the mean value of process parameter data, represents the change in the mean value of the environmental disturbance data, represents the covariance calculation, Indicates the standard deviation of the change in the mean of process parameter data, Indicates the standard deviation of the change in the mean of the environmental disturbance data.

6. The fuzz testing method according to claim 1, wherein: The cluster analysis of the triplet data within the same event window to determine the abnormal cluster and the corresponding abnormal type includes: Perform cluster analysis on the feature vectors of the triple data within the same event window and divide the triple data into multiple clusters; Calculating a local anomaly factor of triplet data within each cluster, and determining an anomaly score of the cluster based on the local anomaly factor; When the anomaly score is greater than a preset anomaly threshold, determining the cluster as an abnormal cluster; The feature vector of the abnormal cluster is matched with a preset abnormal type judgment statistical table to determine the abnormal type of the abnormal cluster.

7. The fuzz testing method according to claim 1, wherein: When the probability of occurrence of the abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster includes: Count the number of occurrences of anomaly clusters of each anomaly type in all event windows; Calculating the probability of occurrence of abnormal clusters of each abnormal type based on the number of occurrences; When the probability of occurrence of the abnormal cluster is greater than the first preset threshold, a significance test is performed on the probability of occurrence of the abnormal cluster of the abnormal type, and a significance test result is output; When the occurrence probability of an abnormal cluster is less than a significance threshold, the abnormal cluster is determined to be a true abnormal cluster.

8. The fuzz testing method according to claim 1, wherein: When the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data includes: Obtaining triplet data within all the true abnormal clusters; Extracting key fields of triplet data within the true anomaly cluster; When the data corresponding to the key field is greater than a second preset threshold, the triple data is determined to be abnormal triple data.

9. The fuzz testing method according to claim 1, wherein: The constructing of a fuzzy test case according to the abnormal triple data includes: Extracting numerical distribution characteristics of process parameter data and numerical distribution characteristics of environmental disturbance data in the abnormal triple data, and generating fuzzy test cases according to the numerical distribution characteristics of the process parameter data and the numerical distribution characteristics of the environmental disturbance data; Obtaining execution path data of the fuzzy test case; When the deviation between the execution path data of the fuzzy test case and the expected abnormal path is greater than the preset fault tolerance threshold, the fuzzy test is re-executed to generate a new fuzzy test case; When the deviation value between the execution path data of the fuzzy test case and the expected abnormal path is greater than the preset fault tolerance threshold, the fuzzy test case will be generated as the final fuzzy test case.

10. A fuzzy testing system based on program control flow, used to execute the fuzzy testing method according to any one of claims 1 to 9, characterized in that: include: Acquisition module, used to obtain parameter data; The parameter data includes process parameter data, environmental disturbance data, execution path data and production quality data; An analysis module, configured to analyze the process parameter data and the environmental disturbance data using a sliding window technique to obtain a mutation point; a removal module, configured to remove process parameter data that is mutated due to interference from environmental disturbance data from the process parameter data according to the mutation point, to obtain target process parameter data; An extraction module, configured to set an event window centered on a mutation point of the target process parameter data, and extract environmental disturbance data, execution path data, and production quality data within the event window; The triplet module is used to decompose the execution path data within the same event window into multiple branch decision points, and generate triplet data of parameter-branch decision point-production quality based on the process parameter data, environmental disturbance data and production quality data corresponding to each branch decision point; Clustering module, used to perform cluster analysis on triplet data within the same event window to determine abnormal clusters and corresponding abnormal types; A calculation module is used to calculate the occurrence probability of each abnormal cluster based on preset historical data; When the probability of occurrence of an abnormal cluster is greater than a first preset threshold, determining that the abnormal cluster is a true abnormal cluster; When the triple data in the true abnormal cluster is greater than a second preset threshold, determining that the triple data is abnormal triple data; A construction module is used to construct a fuzzy test case based on the abnormal triple data.

Images