Methods, devices, storage media and electronic equipment for identifying abnormal equipment conditions

Abstract

This application discloses a method, apparatus, storage medium, and electronic device for identifying abnormal equipment states. The method extracts multi-dimensional features from massive test logs, including power supply current characteristics, ADC sampling characteristics, attenuation value sequences, and electronic lock characteristics. It comprehensively utilizes outlier detection based on statistical distribution, dual detection of median deviation and internal fluctuations in the attenuation value sequence, batch offset analysis, and multi-parameter correlation modeling. This enables the identification of hidden defects that traditional methods cannot detect, such as parameters within the specification threshold range but with significant offsets, overall batch drift, abnormal coupling between parameters, and abnormal sequence fluctuations. Furthermore, by calculating a comprehensive defect score for each device, devices that would otherwise be mistakenly judged as acceptable based solely on a single threshold are accurately marked as potentially defective, significantly reducing the probability of potentially defective devices being mistakenly judged as acceptable.

Description

Technical Field

[0001] This application relates to the field of electric vehicle charging equipment production and testing technology, specifically to a method, device, storage medium, and electronic device for identifying abnormal equipment states. Background Technology

[0002] With the rapid development of electric vehicle charging infrastructure, the Supply Equipment Communication Controller (SECC) and Electric Vehicle Communication Controller (EVCC), as the core execution units of the charging communication protocol, directly affect charging compatibility and safety. During the production testing phase, extensive functional and electrical parameter tests are required on the SECC and EVCC devices.

[0003] Currently, in the field of SECC and EVCC product testing, the identification of abnormal equipment conditions generally adopts the specification threshold judgment method, that is, the qualified threshold range of each test item is preset in the test script. During the test execution, the equipment collects various parameters in real time and compares them with the preset thresholds. If a parameter exceeds the threshold range, the equipment is judged as a defective product and is intercepted; if all parameters are within the threshold range, it is judged as a qualified product and enters the next process.

[0004] However, current equipment anomaly identification methods can only identify explicit anomalies where test parameters exceed specification thresholds, but cannot identify implicit anomalies such as parameter shifts, batch drifts, multi-parameter coupling anomalies, and sequence fluctuation anomalies that are within the threshold range. This leads to equipment with quality risks being misjudged as qualified. Summary of the Invention

[0005] This application provides a method, apparatus, storage medium, and electronic device for identifying abnormal equipment states, which can reduce the probability that equipment with quality risks is mistakenly judged as qualified.

[0006] In a first aspect, embodiments of this application provide a method for identifying abnormal device states, including: Extract a set of characteristic parameters for each device from massive test logs. The set of characteristic parameters includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each of the aforementioned feature parameter sets to obtain the outlier detection results for each device. Based on the attenuation value sequence, the intermediate value deviation of each device is detected to obtain the degree of offset anomaly of each device; Based on the attenuation value sequence, internal fluctuation anomaly detection is performed on each device to obtain the degree of fluctuation anomaly of each device. The equipment is grouped according to production batches, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of characteristic parameters. Construct a correlation model between the characteristic parameters of each device to calculate the correlation coefficient between the characteristic parameters; Based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, a comprehensive defect score for each device is calculated, and potentially defective devices are marked based on the comprehensive defect score.

[0007] In the device abnormal state identification method provided in this application embodiment, the step of detecting outliers in the feature parameters of each feature parameter set based on the statistical distribution of historical normal devices to obtain the outlier detection results for each device includes: For each feature parameter in each set of the aforementioned feature parameters, calculate the first historical mean and the first historical standard deviation of historically normal devices; When the value of the characteristic parameter of the corresponding device is less than the difference between the first historical mean and three times the first historical standard deviation, or greater than the sum of the first historical mean and three times the first historical standard deviation, the corresponding device is marked as an outlier device, thereby obtaining the outlier detection results for each device.

[0008] In the device abnormal state identification method provided in this application embodiment, the step of performing intermediate value deviation detection on each device based on the attenuation value sequence to obtain the degree of offset abnormality of each device includes: Calculate the batch mean and batch standard deviation of the median attenuation value of all devices within the same batch; The median attenuation value of each device is obtained based on the attenuation value sequence; When the absolute difference between the median value of the attenuation value and the average value of the batch is greater than a preset multiple multiplied by the standard deviation of the batch, it is determined that the corresponding device has an abnormal median value deviation. The degree of offset anomaly of each device is determined based on the magnitude of the absolute difference.

[0009] In the device abnormal state identification method provided in this application embodiment, the step of detecting intra-sequence fluctuation anomalies of each device based on the attenuation value sequence to obtain the fluctuation anomaly degree of each device includes: The second historical mean and second historical standard deviation of the standard deviation of attenuation values ​​of normally functioning equipment in history; The standard deviation of the attenuation value of each device is obtained based on the attenuation value sequence. When the standard deviation of the attenuation value is greater than the sum of the second historical mean and three times the second historical standard deviation, it is determined that the corresponding device has a sequence fluctuation abnormality, and the degree of fluctuation abnormality is determined according to the multiple by which the standard deviation of the attenuation value exceeds the second historical mean.

[0010] In the device abnormal state identification method provided in this application embodiment, the step of constructing a correlation model between the feature parameters of each device to calculate the correlation coefficient between the feature parameters includes: A first correlation model is constructed between the power supply current characteristics and the ADC sampling characteristics to calculate the correlation coefficient between the power supply current of the corresponding device and the ADC sampling value; A second correlation model is constructed between the median attenuation value and the positional distance in the electronic lock feature, so as to calculate the correlation coefficient between the median attenuation value and the positional distance; A third correlation model is constructed between the power supply current characteristics and the offset error in the ADC sampling characteristics to calculate the correlation coefficient between the power supply current and the offset error.

[0011] In the equipment abnormality status identification method provided in this application embodiment, the step of calculating the comprehensive defect score of each device based on the outlier detection result, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient includes: Assign corresponding weights to the outlier detection result, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, respectively. The outlier detection results of each device are quantified into outlier scores, the degree of deviation anomaly is quantified into deviation scores, the degree of fluctuation anomaly is quantified into fluctuation scores, the degree of batch deviation is used as batch deviation scores, and the correlation coefficient is mapped into association anomaly scores. The outlier score, the offset score, the fluctuation score, the batch offset score, and the associated anomaly score are weighted and summed to obtain the comprehensive defect score for each device.

[0012] In the equipment abnormality identification method provided in this application embodiment, the step of grouping each piece of equipment by production batch and calculating the batch offset of the parameter mean of each batch relative to the global mean based on the feature parameter set includes: Group all equipment by production batch, calculate the mean of parameters for each batch based on the set of characteristic parameters, and calculate the global mean and global standard deviation for all equipment; For each batch, calculate the absolute difference between the mean of the parameters in that batch and the global mean, and then divide it by the global standard deviation to obtain the batch offset of that batch.

[0013] Secondly, embodiments of this application provide a device for identifying abnormal equipment states, comprising: The parameter extraction unit is used to extract the feature parameter set of each device from massive test logs. The feature parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. The outlier detection unit is used to detect outliers in the feature parameters of each set of feature parameters based on the statistical distribution of historical normal devices, and to obtain the outlier detection results for each device. The deviation detection unit is used to perform intermediate value deviation detection on each device according to the attenuation value sequence to obtain the degree of deviation anomaly of each device. The fluctuation detection unit is used to detect internal fluctuation anomalies of each device based on the attenuation value sequence, and to obtain the degree of fluctuation anomaly of each device. The offset calculation unit is used to group each device according to the production batch and calculate the batch offset of the parameter mean of each batch relative to the global mean based on the set of feature parameters. A coefficient calculation unit is used to construct a correlation model between the characteristic parameters of each device, so as to calculate the correlation coefficient between the characteristic parameters. The scoring calculation unit is used to calculate the comprehensive defect score of each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, and to mark potentially defective devices based on the comprehensive defect score.

[0014] Thirdly, this application provides a storage medium storing a plurality of instructions that are adapted for loading by a processor to execute the device abnormal state identification method described in any of the preceding claims.

[0015] Fourthly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the device abnormal state identification method described in any of the above claims.

[0016] In summary, the device abnormality identification method provided in this application includes extracting a set of feature parameters for each device from massive test logs. The set of feature parameters includes power supply current characteristics, ADC sampling characteristics, attenuation value sequences, and electronic lock characteristics. Based on the statistical distribution of historically normal devices, outlier detection is performed on the feature parameters in each set of feature parameters to obtain outlier detection results for each device. Median deviation detection is performed on each device based on the attenuation value sequences to obtain the degree of deviation anomaly for each device. Internal fluctuation anomaly detection is performed on each device based on the attenuation value sequences to obtain the degree of fluctuation anomaly for each device. Devices are grouped by production batch, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of feature parameters. A correlation model is constructed between the feature parameters of each device to calculate the correlation coefficient between the feature parameters. A comprehensive defect score is calculated for each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the batch deviation, and the correlation coefficient. Potentially defective devices are then marked based on the comprehensive defect score. This application's embodiments extract multi-dimensional features such as power supply current characteristics, ADC sampling characteristics, attenuation value sequences, and electronic lock characteristics from massive test logs. It also comprehensively utilizes outlier detection based on statistical distribution, dual detection of median deviation and internal fluctuations in the attenuation value sequence, batch offset analysis, and multi-parameter correlation modeling. This enables the identification of hidden defects that traditional methods cannot detect, such as parameters within the specification threshold range but with significant offsets, overall batch drift, abnormal parameter coupling, and abnormal sequence fluctuations. Furthermore, by calculating a comprehensive defect score for each device, devices that would otherwise be mistakenly judged as qualified based solely on a single threshold are accurately marked as potentially defective devices, significantly reducing the probability of risky devices being mistakenly judged as qualified. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a schematic diagram illustrating an application scenario of the device abnormal state identification method provided in this application embodiment.

[0019] Figure 2 This is a flowchart illustrating the device abnormal state identification method provided in the embodiments of this application.

[0020] Figure 3 This is a schematic diagram of the device abnormal status identification device provided in the embodiments of this application.

[0021] Figure 4 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0022] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0023] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, components, features, and elements with the same names in different embodiments of this application may have the same meaning or different meanings, the specific meaning of which must be determined by its interpretation in that specific embodiment or further in conjunction with the context of that specific embodiment.

[0024] It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit this application.

[0025] In the following description, the use of suffixes such as "module," "part," or "unit" to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, "module," "part," or "unit" may be used interchangeably.

[0026] In the description of this application, it should be noted that the terms "upper," "lower," "left," "right," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application. In addition, terms such as "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0027] Currently, in the field of SECC and EVCC product testing, the identification of abnormal equipment conditions generally adopts the specification threshold judgment method, that is, the qualified threshold range of each test item is preset in the test script. For example, the supply current does not exceed 200mA, the SLAC attenuation value is between 10dB and 35dB, the CP voltage is between 8.5V and 9.3V, the PP voltage is between 1.2V and 1.8V, and the electronic lock locking distance is between 5mm and 15mm. During the test execution, the equipment collects the above parameters in real time and compares them with the preset thresholds. If any parameter exceeds the threshold range, the equipment is judged as unqualified and is intercepted; if all parameters are within the threshold range, it is judged as qualified and enters the next process.

[0028] However, current equipment anomaly identification methods can only identify explicit anomalies where test parameters exceed specification thresholds, but cannot identify implicit anomalies such as parameter shifts, batch drifts, multi-parameter coupling anomalies, and sequence fluctuation anomalies that are within the threshold range. This leads to equipment with quality risks being misjudged as qualified.

[0029] Based on this, embodiments of this application provide a method, apparatus, storage medium, and electronic device for identifying abnormal device states. Specifically, the device for identifying abnormal device states can be integrated into an electronic device, which can be a server or a terminal, etc. The terminal can include mobile phones, wearable smart devices, tablet computers, laptops, and personal computers (PCs), etc., as well as other computers and auxiliary devices. The server can be a single server or a server cluster composed of multiple servers, and can be a physical server or a virtual server.

[0030] For example, such as Figure 1 As shown, electronic devices can extract feature parameter sets for each device from massive test logs. These feature parameter sets include power supply current characteristics, ADC sampling characteristics, attenuation value sequences, and electronic lock characteristics. Based on the statistical distribution of historically normal devices, outlier detection is performed on the feature parameters in each feature parameter set to obtain outlier detection results for each device. Median deviation detection is performed on each device based on the attenuation value sequence to obtain the degree of deviation anomaly for each device. Internal fluctuation anomaly detection is performed on each device based on the attenuation value sequence to obtain the degree of fluctuation anomaly for each device. Devices are grouped by production batch, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the feature parameter set. A correlation model is constructed between the feature parameters of each device to calculate the correlation coefficient between the feature parameters. Based on the outlier detection results, deviation anomaly degree, fluctuation anomaly degree, batch deviation degree, and correlation coefficient, a comprehensive defect score for each device is calculated, and potentially defective devices are marked based on the comprehensive defect score.

[0031] The technical solutions shown in this application will be described in detail below through specific embodiments. It should be noted that the order of description of the following embodiments is not intended to limit the priority of the embodiments.

[0032] Please see Figure 2 , Figure 2 This is a flowchart illustrating the equipment abnormal state identification method provided in this application embodiment. This method can be applied to production testing scenarios for electric vehicle charging equipment, specifically for identifying abnormal states of SECC and EVCC. The specific flow of this method is as follows: 101. Extract the characteristic parameter set of each device from massive test logs. The characteristic parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics.

[0033] Specifically, the first step is to extract multi-dimensional feature parameters for each device from the massive amount of test logs generated during the production testing phase, forming a feature parameter set for that device. This feature parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequences, and electronic lock characteristics.

[0034] The power supply current characteristics include the average power supply current and standard deviation of the device in idle mode, as well as the peak surge current when the device switches from sleep mode to operating mode. These power supply current characteristics are key indicators for assessing the leakage current status of power management chips and printed circuit boards (PCBs). By extracting the average and standard deviation of the current in idle mode, the stability of the device's static power consumption can be evaluated; by extracting the peak surge current, the health status of the power supply startup circuit and decoupling capacitors can be determined.

[0035] The ADC sampling characteristics include sampled values ​​of the Control Pilot (CP) voltage and Proximity Pilot (PP) voltage, and the offset error of each channel is calculated. The offset error is calculated as the absolute value of the difference between the ADC acquired value and the measured value from an external high-precision source. This offset error reflects the accuracy of the ADC reference source and sampling circuit, such as ADC reference source drift and voltage divider network deviation.

[0036] This attenuation value sequence is a Signal Level Attenuation Characterization (SLAC) attenuation value sequence. During SLAC testing, the device generates multiple sets of attenuation values, typically 50 to 60 sets, forming an attenuation value sequence. This step extracts the raw data from this sequence for subsequent statistical feature analysis.

[0037] The electronic lock features include the position distance value when the electronic lock is locked and a lock-in status indicator. Abnormal electronic lock distance may reflect mechanical assembly problems (such as lock hook misalignment, guide rail jamming) or sensor calibration deviations.

[0038] 102. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each feature parameter set to obtain the outlier detection results for each device.

[0039] Specifically, for each feature parameter in each feature parameter set, the first historical mean (denoted as μ) and the first historical standard deviation (denoted as σ) of historically normal devices can be calculated. When the value of the feature parameter of the corresponding device is less than the difference between the first historical mean and three times the first historical standard deviation (i.e., <μ - 3σ), or greater than the sum of the first historical mean and three times the first historical standard deviation (i.e., >μ + 3σ), the corresponding device is marked as an outlier device, thus obtaining the outlier detection results for each device.

[0040] Taking the average power supply current as an example, assuming the historical average idle current of normal equipment is 85mA and the first historical standard deviation is 8mA, then the lower limit for outlier detection is 85 - 3 × 8 = 61mA, and the upper limit is 85 + 3 × 8 = 109mA. If the average idle current of a certain equipment is 118mA, which is greater than 109mA, then it is marked as an outlier equipment.

[0041] Furthermore, for the SLAC attenuation value sequence, this embodiment not only detects outliers in the median attenuation value (A_median), but also outliers in the standard deviation of the attenuation value (A_std). That is, when the standard deviation of the attenuation value A_std of a certain device is significantly greater than the mean of the standard deviation of the attenuation value of historical normal devices (denoted as μ_std) plus three times the standard deviation (σ_std), i.e., A_std>μ_std + 3σ_std, even if its median value is normal, it is marked as a device with abnormal fluctuations, indicating that there may be a problem with PLC communication instability.

[0042] This application employs the three-standard-deviation criterion from statistics, which can effectively identify devices whose characteristic parameters significantly deviate from the normal distribution. It also specifically targets outlier detection for the standard deviation of SLAC attenuation values, identifying abnormal devices with normal median values ​​but excessive fluctuations. Compared to traditional specification threshold methods, this application can detect hidden risks that, even if not exceeding the specification range, have shown significant deviations or abnormal fluctuations. For example, if the current gradually increases from 85mA to 100mA (still within the specification limit of 200mA), but has already been identified as an outlier, it can provide early warning of potential quality risks.

[0043] For example, in testing a batch of 1000 SECC devices, device SN:ABC123 was found to have an average idle current of 118mA, exceeding the historical average plus three standard deviations (109mA), and was therefore marked as an outlier. Subsequent re-inspection revealed a PCB leakage problem. Another device, SN:ABC456, had a median SLAC attenuation value of 31dB, while the historical normal average was 22dB. The upper limit of three standard deviations was 31dB, which was just on the borderline, and it was also marked as an outlier. Furthermore, the standard deviation of this device's attenuation value was 5.2dB, while the historical normal attenuation value had a mean standard deviation of 2.5dB and a standard deviation of 0.8dB. The upper limit of three standard deviations was 2.5 + 2.4 = 4.9dB. Since 5.2dB > 4.9dB, this device was also marked as having abnormal fluctuations.

[0044] 103. Based on the attenuation value sequence, perform intermediate value deviation detection on each device to obtain the degree of offset anomaly of each device.

[0045] This application embodiment specifically targets the SLAC attenuation value sequence to perform intermediate value deviation detection in order to assess whether the typical level of signal attenuation of the device deviates from other devices in the same batch.

[0046] Specifically, the batch mean (denoted as μ_median) and batch standard deviation (denoted as σ_median) of the median attenuation values ​​of all devices within the same batch can be calculated. The median attenuation value (A_median) of each device is obtained from the attenuation value sequence. When the absolute difference between the median attenuation value and the batch mean is greater than a preset multiple multiplied by the batch standard deviation (i.e., |A_median - μ_median|>δ × σ_median), the corresponding device is determined to have an abnormal median value deviation. Then, the degree of deviation abnormality for each device is determined based on the magnitude of the absolute difference. The preset multiple δ is typically set to 2.0.

[0047] For example, if the median attenuation value of a batch of equipment has a batch average of 22dB and a batch standard deviation of 3dB, and the preset multiplier is 2.0, then the deviation threshold is 2×3=6dB. If the median attenuation value of a certain equipment is 29dB, and the absolute difference is 7dB > 6dB, then it is determined that there is an abnormal median deviation, and the degree of deviation is relatively high.

[0048] This application's embodiments, through batch-based relative comparison, can identify devices within the same production batch whose attenuation levels significantly deviate from the mainstream trend. This deviation may be caused by individual differences in PLC communication modules, capacitance drift in coupling capacitors, or accumulated assembly tolerances. Even if the attenuation value of the device is still within the specified range (e.g., 10dB-35dB), it can be accurately identified, preventing it from flowing into the next process.

[0049] For example, in a real-world application scenario, in a batch of SECC equipment, the median SLAC attenuation value for most devices was between 20dB and 24dB. However, device SN:ABC789 had a median attenuation value of 18dB, differing from the batch average of 22dB by 4dB. The preset threshold for a multiplier of 2.0 is 6dB, and the 4dB difference was within the threshold, therefore it was not considered an abnormal deviation. Another device, SN:ABC890, had a median attenuation value of 29dB, differing by 7dB, exceeding the threshold and being marked as having an abnormal median deviation. Analysis revealed that a faulty solder joint in the PLC communication module of this device caused the higher attenuation value.

[0050] 104. Based on the attenuation value sequence, perform internal fluctuation anomaly detection on each device to obtain the degree of fluctuation anomaly of each device.

[0051] It should be noted that this embodiment complements step 103, focusing on the fluctuation of the attenuation value sequence within a single device, rather than the shift of the intermediate value.

[0052] Specifically, the second historical mean (denoted as μ_std) and the second historical standard deviation (denoted as σ_std) of the attenuation value standard deviation of historically normal devices can be calculated. The attenuation value standard deviation (A_std) of each device is obtained based on the attenuation value sequence. When the attenuation value standard deviation is greater than the sum of the second historical mean and three times the second historical standard deviation (i.e., A_std > μ_std + 3σ_std), the corresponding device is determined to have abnormal sequence fluctuations, and the degree of fluctuation abnormality is determined based on the multiple by which the attenuation value standard deviation exceeds the second historical mean.

[0053] For example, if the mean standard deviation of attenuation values ​​for historically normal equipment is 2.5 dB, and the second historical standard deviation is 0.8 dB, then the threshold for abnormal fluctuations is 2.5 + 3 × 0.8 = 4.9 dB. If the standard deviation of attenuation values ​​for a certain equipment is 5.2 dB, which is greater than 4.9 dB, it is judged as having abnormal series fluctuations, indicating a high degree of abnormal fluctuation. Even if the intermediate value is normal, it is still considered a potentially defective piece of equipment.

[0054] This embodiment can identify devices with unstable and excessively fluctuating attenuation value sequences. Such devices may exhibit intermittent communication failures, with intermediate values ​​possibly being normal, but prone to intermittent connection failures in actual use. Traditional thresholding methods are completely incapable of detecting such anomalies, while this embodiment effectively compensates for this deficiency by analyzing the statistical characteristics (standard deviation) within the sequence.

[0055] 105. Group each piece of equipment according to production batch, and calculate the batch deviation of the parameter mean of each batch relative to the global mean based on the set of characteristic parameters.

[0056] It is understood that this embodiment analyzes systemic production deviations from a batch perspective, rather than focusing on a single device.

[0057] Specifically, all devices can be grouped by production batch, and the mean parameter value for each batch (e.g., the mean idle current of all devices in that batch, denoted as μ_batch) can be calculated based on the set of characteristic parameters. The global mean (denoted as μ_global) and global standard deviation (denoted as σ_global) for all devices (across batches) can also be calculated. For each batch, the absolute difference between the batch's mean parameter value and the global mean is calculated, and then divided by the global standard deviation to obtain the batch shift: Batch_Shift = |μ_batch - μ_global| / σ_global.

[0058] Specifically, when Batch_Shift exceeds a preset threshold (usually 2.0), even if the parameters of all devices within the batch are within the specification threshold range, the system will still mark the batch as a process deviation risk batch and apply a defect score weighting to all devices within the batch. This batch deviation is usually caused by process fluctuations in production processes such as SMT placement and reflow soldering.

[0059] For example, the global average idle current is 85mA, and the global standard deviation is 8mA. If the average idle current of a batch of equipment is 97mA, then the batch offset is |97-85| / 8 = 1.5. If the preset batch offset threshold is 2.0, then no alarm is triggered for that batch; if the average idle current of a batch is 101mA, and the offset is 2.0, then a process offset risk alarm is triggered.

[0060] This embodiment identifies systemic deviations in an entire batch caused by process fluctuations such as SMT placement accuracy drift, reflow soldering temperature variations, and batch differences in incoming materials. Even if the individual parameters of all equipment within the batch are within the specified threshold range, this method can still issue an early warning, prompting quality engineers to check the production process, thereby avoiding the generation of a large number of potentially defective products.

[0061] 106. Construct a correlation model between the characteristic parameters of each device to calculate the correlation coefficient between the characteristic parameters.

[0062] It should be noted that this embodiment can construct a physical correlation model between feature parameters and calculate the correlation coefficient to discover coupling anomalies that cannot be identified by a single parameter.

[0063] Specifically, this embodiment includes the construction of the following three association models: First, a primary correlation model is constructed between the power supply current characteristics and the ADC sampling characteristics to calculate the correlation coefficient (denoted as ρ(I, V_cp)) between the power supply current and the ADC sampling value of the corresponding device. The ADC sampling value mainly refers to the CP voltage. In the theoretical model, changes in the CP voltage affect the load of the power management chip, thereby altering the power supply current. When the absolute value of the correlation coefficient between the power supply current and the CP voltage is lower than a preset threshold (this threshold can be determined based on historical normal device data, for example, 0.5) and the power supply current is abnormal, it is determined that the power management chip is malfunctioning.

[0064] Second, a second correlation model is constructed between the median attenuation value and the positional distance in the electronic lock features to calculate the correlation coefficient (denoted as ρ(A_median, D_lock)) between the median attenuation value and the positional distance. Since deviations in the electronic lock's locking distance can cause changes in the signal transmission path, resulting in abnormal attenuation, if the two show a strong correlation (e.g., the absolute value of the correlation coefficient is greater than 0.6) and the attenuation value is abnormal, it is determined to be an assembly position deviation of the electronic lock.

[0065] Third, a third correlation model is constructed between the supply current characteristics and the offset error in the ADC sampling characteristics to calculate the correlation coefficient between the supply current and the offset error. If a strong correlation exists, it is determined to be a power supply problem in the sampling circuit, such as the ADC reference voltage being affected by power supply ripple interference.

[0066] Understandably, this embodiment overcomes the limitations of independent parameter judgment in traditional methods. By utilizing the physical coupling relationship between parameters, it can accurately locate the cause of the fault. For example, the supply current may be too high and the CP voltage may be too low, but both are within the threshold when viewed individually. Through correlation analysis, it is found that they are highly correlated, thus inferring that the power management chip is abnormal, rather than a simple random fluctuation.

[0067] For example, in a practical application scenario, the idle current of device SN:ABC321 was 98mA (slightly high, but still within the specification of 200mA), and the CP voltage was 8.2V (the specification lower limit is 8.5V, and the actual voltage was below the lower limit, but it is assumed that it might pass under another lenient specification). The correlation coefficient between the two was calculated to be 0.72, with an absolute value greater than 0.5, and the power supply current was abnormal, indicating a faulty power management chip. After replacing the power chip, the current and CP voltage returned to normal. Another device, SN:ABC654, had a median SLAC attenuation value of 29dB and an electronic lock locking distance of 12mm. The correlation coefficient between the two was 0.65, with an absolute value greater than 0.6, indicating that the abnormal signal attenuation was caused by a misalignment in the electronic lock assembly. Disassembly revealed that the electronic lock mounting bracket was offset by approximately 2mm.

[0068] 107. Calculate the comprehensive defect score for each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, and mark potentially defective devices based on the comprehensive defect score.

[0069] This embodiment integrates all the aforementioned test results and uses a weighted scoring model to perform a comprehensive quantitative evaluation of each device.

[0070] Specifically, we can first configure corresponding weights for outlier detection results, deviation anomaly degree, fluctuation anomaly degree, batch deviation degree, and correlation coefficient. The specific values ​​of the weights can be set based on historical data statistical experience, or optimized and adjusted through data-driven methods (such as logistic regression), with the sum of all weights being 1. For example, we can set the weights w_I for power supply current outlier score and surge current anomaly score, w_ADC for ADC deviation error score, w_SLAC for SLAC attenuation value median deviation score and fluctuation anomaly score, w_Lock for electronic lock distance anomaly score, w_Batch for batch deviation score, and w_Corr for multi-parameter correlation anomaly score.

[0071] Then, the outlier detection results of each device are quantified into outlier scores (e.g., linearly mapped to 0-1 based on the multiple of the deviation from the standard deviation), the degree of deviation anomaly is quantified into a deviation score, the degree of fluctuation anomaly is quantified into a fluctuation score, the batch deviation degree is directly used as the batch deviation score, and the correlation coefficient is mapped into an association anomaly score (e.g., a higher score is taken when the correlation coefficient is abnormal). For power supply current anomalies, this application further performs graded processing: based on the degree of current deviation, current anomalies are divided into three levels—Level 1 anomaly (slight deviation): current value between μ+2σ and μ+3σ, marked as a device of concern; Level 2 anomaly (moderate deviation): current value between μ+3σ and μ+4σ, marked as a suspected defective device; Level 3 anomaly (severe deviation): current value greater than μ+4σ, marked as a high-risk defective device. For devices with anomalies at each level, the system automatically retains complete current timing data, corresponding test stage information, associated CP voltage and ADC sampling data, for quality analysts to perform root cause tracing.

[0072] Finally, the outlier score, offset score, fluctuation score, batch offset score, and correlation anomaly score are weighted and summed to obtain the comprehensive defect score for each device: Defect_Score = w_I × Score_I + w_ADC × Score_ADC + w_SLAC × Score_SLAC + w_Lock × Score_Lock + w_Batch × Batch_Shift + w_Corr × Correlation_Score. When the comprehensive defect score exceeds a preset threshold (e.g., 0.6), the corresponding device is marked as a potentially defective device. It is worth noting that even if the device shows all tests as passed in the traditional specification threshold judgment, it will still be marked as a potentially defective device if the comprehensive defect score exceeds the threshold, triggering a re-inspection or interception process.

[0073] This application's embodiments employ multi-dimensional weighted fusion and introduce a current anomaly classification and retention mechanism to avoid misjudgment or omission based on a single indicator, achieving quantitative assessment, automatic screening, and root cause tracing of equipment anomalies. Compared to the traditional binary judgment of "either good or bad," this application provides a more refined risk classification, effectively identifying equipment that "passes testing but has hidden risks," thereby significantly reducing the probability of equipment with quality risks being misjudged as qualified and improving the overall reliability of the product.

[0074] For example, in a batch of 1000 SECC devices, all devices passed the traditional specification threshold test (i.e., all parameters were within the specification range). After calculating the comprehensive defect score using the method of this application, a total of 67 potentially defective devices were identified (defect rate 6.7%). Among them, 28 devices had abnormal power supply current (including some of the 47 devices with first-level abnormalities, 12 devices with second-level abnormalities, and 3 devices with third-level abnormalities), 15 devices had abnormal deviations or fluctuations in the median value of SLAC attenuation, 12 devices had ADC sampling deviations, 8 devices had abnormal electronic lock distances, and 4 devices had multi-parameter correlation anomalies. The system automatically retained the complete current timing data of all devices with abnormal current for subsequent analysis. After manual re-inspection, 59 of the 67 devices were found to have varying degrees of latent defects (such as capacitor solder joints, PCB leakage, ADC reference source temperature drift, electronic lock assembly deviations, and PLC communication module instability). One device had a comprehensive defect score of 0.68 (threshold 0.6), passed all traditional tests, but re-inspection revealed that its CP voltage sampling circuit had a resistor solder joint, resulting in a large temperature drift. The method described in this application successfully intercepted the device, preventing potential charging interruption failures after it entered the market.

[0075] The device abnormality identification method provided in this application includes extracting a set of feature parameters for each device from massive test logs. The set of feature parameters includes power supply current features, ADC sampling features, attenuation value sequences, and electronic lock features. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each feature parameter set to obtain outlier detection results for each device. Median deviation detection is performed on each device based on the attenuation value sequence to obtain the degree of deviation abnormality for each device. Internal fluctuation anomaly detection is performed on each device based on the attenuation value sequence to obtain the degree of fluctuation anomaly for each device. Each device is grouped by production batch, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of feature parameters. A correlation model between the feature parameters of each device is constructed to calculate the correlation coefficient between the feature parameters. A comprehensive defect score for each device is calculated based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the batch deviation, and the correlation coefficient, and potential defective devices are marked based on the comprehensive defect score. This application extracts multi-dimensional features such as power supply current, ADC sampling, SLAC attenuation value sequence, and electronic lock from massive test logs. It comprehensively utilizes statistical distribution-based outlier detection (identifying parameter deviations from the normal distribution but not exceeding thresholds), dual detection of median deviation and internal fluctuations in the SLAC attenuation value sequence (identifying relative shifts and abnormal sequence fluctuations within a batch), batch shift analysis (identifying systematic batch drift), and multi-parameter correlation modeling (identifying coupling anomalies such as power supply current and CP voltage, attenuation value and electronic lock distance). This allows for the comprehensive capture of various latent anomalies that traditional threshold methods cannot detect. Furthermore, by performing a comprehensive defect scoring on the equipment, equipment that was mistakenly judged as qualified due to all parameters being within specifications and thus posing a quality risk can be accurately marked as potentially defective. This effectively solves the technical problem of existing technologies that can only identify explicit anomalies while missing latent anomalies, leading to misjudgments of qualification. In other words, this application can reduce the probability of equipment with quality risks being misjudged as qualified.

[0076] To facilitate better implementation of the equipment abnormality status identification method provided in this application, this application also provides an equipment abnormality status identification device. The meanings of the terms used are the same as in the above-described equipment abnormality status identification method, and specific implementation details can be found in the descriptions within the method embodiments.

[0077] Please see Figure 3 , Figure 3 This is a schematic diagram of the equipment abnormality status identification device provided in an embodiment of this application. The device may include a parameter extraction unit 201, an outlier detection unit 202, a deviation detection unit 203, a fluctuation detection unit 204, an offset calculation unit 205, a coefficient calculation unit 206, and a scoring calculation unit 207. The parameter extraction unit 201 is used to extract the feature parameter set of each device from massive test logs. The feature parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. Outlier detection unit 202 is used to detect outliers in the feature parameters of each feature parameter set based on the statistical distribution of historical normal devices, and to obtain the outlier detection results of each device. The deviation detection unit 203 is used to perform intermediate value deviation detection on each device according to the attenuation value sequence to obtain the degree of deviation abnormality of each device. The fluctuation detection unit 204 is used to detect internal fluctuation anomalies of each device based on the attenuation value sequence, and to obtain the degree of fluctuation anomalies of each device. The offset calculation unit 205 is used to group each device according to the production batch and calculate the batch offset of the parameter mean of each batch relative to the global mean based on the feature parameter set. The coefficient calculation unit 206 is used to construct a correlation model between the characteristic parameters of each device in order to calculate the correlation coefficient between the characteristic parameters. The scoring calculation unit 207 is used to calculate the comprehensive defect score of each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, and to mark potentially defective devices based on the comprehensive defect score.

[0078] For specific implementation methods of each of the above units, please refer to the embodiments of the above-described equipment abnormality status identification method, which will not be repeated here.

[0079] In summary, the device abnormal state identification device provided in this application embodiment can extract the feature parameter set of each device from massive test logs through the parameter extraction unit 201. The feature parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence, and electronic lock characteristics. The outlier detection unit 202 performs outlier detection on the feature parameters in each feature parameter set based on the statistical distribution of historical normal devices to obtain the outlier detection results for each device. The deviation detection unit 203 performs median deviation detection on each device according to the attenuation value sequence to obtain the degree of deviation abnormality of each device. The fluctuation detection unit 204 performs attenuation detection based on the attenuation value sequence. The depreciation sequence performs internal fluctuation anomaly detection on each device to obtain the fluctuation anomaly degree of each device; the offset calculation unit 205 groups each device according to production batch and calculates the batch offset degree of the parameter mean of each batch relative to the global mean based on the feature parameter set; the coefficient calculation unit 206 constructs the correlation model between the feature parameters of each device to calculate the correlation coefficient between each feature parameter; the scoring calculation unit 207 calculates the comprehensive defect score of each device based on the outlier detection results, offset anomaly degree, fluctuation anomaly degree, batch offset degree and correlation coefficient, and marks potentially defective devices based on the comprehensive defect score. This application extracts multi-dimensional features such as power supply current, ADC sampling, SLAC attenuation value sequence, and electronic lock from massive test logs. It comprehensively utilizes statistical distribution-based outlier detection (identifying parameter deviations from the normal distribution but not exceeding thresholds), dual detection of median deviation and internal fluctuations in the SLAC attenuation value sequence (identifying relative shifts and abnormal sequence fluctuations within a batch), batch shift analysis (identifying systematic batch drift), and multi-parameter correlation modeling (identifying coupling anomalies such as power supply current and CP voltage, attenuation value and electronic lock distance). This allows for the comprehensive capture of various latent anomalies that traditional threshold methods cannot detect. Furthermore, by performing a comprehensive defect scoring on the equipment, equipment that was mistakenly judged as qualified due to all parameters being within specifications and thus posing a quality risk can be accurately marked as potentially defective. This effectively solves the technical problem of existing technologies that can only identify explicit anomalies while missing latent anomalies, leading to misjudgments of qualification. In other words, this application can reduce the probability of equipment with quality risks being misjudged as qualified.

[0080] This application also provides an electronic device that may integrate the device abnormal state identification device of this application embodiment, such as... Figure 4 As shown, it illustrates a structural schematic diagram of the electronic device involved in the embodiments of this application, specifically: The electronic device may include components such as a processor 301 with one or more processing cores and a memory 302 with one or more computer-readable storage media. Those skilled in the art will understand that... Figure 4The electronic device structure shown does not constitute a limitation on the electronic device and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein: The processor 301 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs stored in the memory 302 and / or the methods provided in this application, and by calling data stored in the memory 302, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. Optionally, the processor 301 may include one or more processing cores; preferably, the processor 301 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operation of the storage medium, user interface, and application programs, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 301.

[0081] The memory 302 can be used to store software programs and the methods provided in this application. The processor 301 executes various functional applications and data processing by running the software programs stored in the memory 302 and the methods provided in this application. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store applications required for operating the storage medium and at least one function; the data storage area may store data created based on the use of the electronic device. In addition, the memory 302 may include high-speed random access memory and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 302 may also include a memory controller to provide the processor 301 with access to the memory 302.

[0082] Although not shown, the electronic device may also include a display unit, an input unit, and a power supply, etc., which will not be described in detail here. Specifically, in this embodiment, the processor 301 in the electronic device loads the executable files corresponding to the processes of one or more application programs into the memory 302 according to the following instructions, and the processor 301 runs the application programs stored in the memory 302 to realize various functions, as follows: Extract the set of characteristic parameters for each device from massive test logs. The set of characteristic parameters includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each feature parameter set to obtain the outlier detection results for each device. The intermediate value deviation of each device is detected based on the attenuation value sequence to obtain the degree of offset anomaly of each device. Based on the attenuation value sequence, internal fluctuation anomaly detection is performed on each device to obtain the degree of fluctuation anomaly of each device. The equipment is grouped according to production batches, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of characteristic parameters. Construct a correlation model between the characteristic parameters of each device to calculate the correlation coefficient between the characteristic parameters; The comprehensive defect score for each device is calculated based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient. Potential defective devices are then marked based on the comprehensive defect score.

[0083] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0084] Therefore, embodiments of this application provide a storage medium storing a plurality of instructions that can be loaded by a processor to execute steps in any of the methods provided in embodiments of this application. For example, the instructions can execute the following steps: Extract the set of characteristic parameters for each device from massive test logs. The set of characteristic parameters includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each feature parameter set to obtain the outlier detection results for each device. The intermediate value deviation of each device is detected based on the attenuation value sequence to obtain the degree of offset anomaly of each device. Based on the attenuation value sequence, internal fluctuation anomaly detection is performed on each device to obtain the degree of fluctuation anomaly of each device. The equipment is grouped according to production batches, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of characteristic parameters. Construct a correlation model between the characteristic parameters of each device to calculate the correlation coefficient between the characteristic parameters; The comprehensive defect score for each device is calculated based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient. Potential defective devices are then marked based on the comprehensive defect score.

[0085] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0086] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0087] Since the instructions stored in the storage medium can execute the steps of any method provided in the embodiments of this application, the beneficial effects that any method provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0088] The above provides a detailed description of the device abnormality identification method, apparatus, storage medium, and electronic device provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for identifying abnormal equipment status, characterized in that, include: Extract a set of characteristic parameters for each device from massive test logs. The set of characteristic parameters includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each of the aforementioned feature parameter sets to obtain the outlier detection results for each device. Based on the attenuation value sequence, the intermediate value deviation of each device is detected to obtain the degree of offset anomaly of each device; Based on the attenuation value sequence, internal fluctuation anomaly detection is performed on each device to obtain the degree of fluctuation anomaly of each device. The equipment is grouped according to production batches, and the batch deviation of the parameter mean of each batch relative to the global mean is calculated based on the set of characteristic parameters. Construct a correlation model between the characteristic parameters of each device to calculate the correlation coefficient between the characteristic parameters; Based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, a comprehensive defect score for each device is calculated, and potentially defective devices are marked based on the comprehensive defect score.

2. The equipment abnormality identification method as described in claim 1, characterized in that, Based on the statistical distribution of historical normal devices, outlier detection is performed on the feature parameters in each of the aforementioned feature parameter sets to obtain outlier detection results for each device, including: For each feature parameter in each set of the aforementioned feature parameters, calculate the first historical mean and the first historical standard deviation of historically normal devices; When the value of the characteristic parameter of the corresponding device is less than the difference between the first historical mean and three times the first historical standard deviation, or greater than the sum of the first historical mean and three times the first historical standard deviation, the corresponding device is marked as an outlier device, thereby obtaining the outlier detection results for each device.

3. The equipment abnormality identification method as described in claim 2, characterized in that, The step of detecting intermediate value deviations for each device based on the attenuation value sequence to obtain the degree of offset anomaly for each device includes: Calculate the batch mean and batch standard deviation of the median attenuation value of all devices within the same batch; The median attenuation value of each device is obtained based on the attenuation value sequence; When the absolute difference between the median value of the attenuation value and the average value of the batch is greater than a preset multiple multiplied by the standard deviation of the batch, it is determined that the corresponding device has an abnormal median value deviation. The degree of offset anomaly of each device is determined based on the magnitude of the absolute difference.

4. The equipment abnormality identification method as described in claim 3, characterized in that, The step of detecting intra-sequence fluctuation anomalies in each device based on the attenuation value sequence to obtain the degree of fluctuation anomalies in each device includes: The second historical mean and second historical standard deviation of the standard deviation of attenuation values ​​of normally functioning equipment in history; The standard deviation of the attenuation value of each device is obtained based on the attenuation value sequence. When the standard deviation of the attenuation value is greater than the sum of the second historical mean and three times the second historical standard deviation, it is determined that the corresponding device has a sequence fluctuation abnormality, and the degree of fluctuation abnormality is determined according to the multiple by which the standard deviation of the attenuation value exceeds the second historical mean.

5. The equipment abnormality identification method as described in claim 3, characterized in that, The construction of the correlation model between the feature parameters of each device, and the calculation of the correlation coefficient between the feature parameters, includes: A first correlation model is constructed between the power supply current characteristics and the ADC sampling characteristics to calculate the correlation coefficient between the power supply current of the corresponding device and the ADC sampling value; A second correlation model is constructed between the median attenuation value and the positional distance in the electronic lock feature, so as to calculate the correlation coefficient between the median attenuation value and the positional distance; A third correlation model is constructed between the power supply current characteristics and the offset error in the ADC sampling characteristics to calculate the correlation coefficient between the power supply current and the offset error.

6. The equipment abnormality identification method as described in claim 1, characterized in that, The step of calculating the comprehensive defect score for each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient includes: Assign corresponding weights to the outlier detection result, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, respectively. The outlier detection results of each device are quantified into outlier scores, the degree of deviation anomaly is quantified into deviation scores, the degree of fluctuation anomaly is quantified into fluctuation scores, the degree of batch deviation is used as batch deviation scores, and the correlation coefficient is mapped into association anomaly scores. The outlier score, the offset score, the fluctuation score, the batch offset score, and the associated anomaly score are weighted and summed to obtain the comprehensive defect score for each device.

7. The equipment abnormality identification method as described in claim 1, characterized in that, The process of grouping equipment by production batch and calculating the batch offset of the parameter mean of each batch relative to the global mean based on the set of feature parameters includes: Group all equipment by production batch, calculate the mean of parameters for each batch based on the set of characteristic parameters, and calculate the global mean and global standard deviation for all equipment; For each batch, calculate the absolute difference between the mean of the parameters in that batch and the global mean, and then divide it by the global standard deviation to obtain the batch offset of that batch.

8. A device for identifying abnormal equipment status, characterized in that, include: The parameter extraction unit is used to extract the feature parameter set of each device from massive test logs. The feature parameter set includes power supply current characteristics, ADC sampling characteristics, attenuation value sequence and electronic lock characteristics. The outlier detection unit is used to detect outliers in the feature parameters of each set of feature parameters based on the statistical distribution of historical normal devices, and to obtain the outlier detection results for each device. The deviation detection unit is used to perform intermediate value deviation detection on each device according to the attenuation value sequence to obtain the degree of deviation anomaly of each device. The fluctuation detection unit is used to detect internal fluctuation anomalies of each device based on the attenuation value sequence, and to obtain the degree of fluctuation anomaly of each device. The offset calculation unit is used to group each device according to the production batch and calculate the batch offset of the parameter mean of each batch relative to the global mean based on the set of feature parameters. A coefficient calculation unit is used to construct a correlation model between the characteristic parameters of each device, so as to calculate the correlation coefficient between the characteristic parameters. The scoring calculation unit is used to calculate the comprehensive defect score of each device based on the outlier detection results, the degree of deviation anomaly, the degree of fluctuation anomaly, the degree of batch deviation, and the correlation coefficient, and to mark potentially defective devices based on the comprehensive defect score.

9. A storage medium, characterized in that, The storage medium stores multiple instructions, which are applicable to a processor for loading to execute the device abnormal state identification method according to any one of claims 1-7.

10. An electronic device, characterized in that, The device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the device abnormal state identification method as described in any one of claims 1-7.

Images