Equipment quality evaluation method and device based on multi-source heterogeneous data, terminal equipment and storage medium

Abstract

The invention discloses an equipment quality assessment method and device based on multi-source heterogeneous data, terminal equipment and a storage medium, and belongs to the technical field of equipment quality assessment, and the method comprises the steps: obtaining multi-source heterogeneous time sequence data used for representing the manufacturing process of to-be-assessed equipment; dividing according to a time sequence change form, a first slice used for representing time sequence data of a slow change process and a second slice used for representing time sequence data of a short-time event, and calculating according to all the slices to obtain a plurality of statistical characteristic quantities of the to-be-evaluated equipment; and finally, determining a quality evaluation result of the to-be-evaluated equipment according to all the statistical characteristic quantities. By implementing the method and the device, the problem that in the prior art, a unified and fixed mode is used for carrying out sliding slicing on all types of data, and then quality evaluation is carried out based on the slices, so that the quality evaluation result is inaccurate can be solved.

Description

Technical Field

[0001] This invention relates to the field of equipment quality assessment technology, and in particular to a method, apparatus, terminal equipment and storage medium for equipment quality assessment based on multi-source heterogeneous data. Background Technology

[0002] As the power industry undergoes a profound transformation towards digitalization and intelligence, the production model and quality control system in the field of power grid main equipment manufacturing are experiencing systemic changes. Against this backdrop, the manufacturing process of core power grid equipment, represented by gas-insulated switchgear (GIS), covers seven key stages: design, raw materials, processing, assembly, commissioning, type testing, and delivery. To achieve precise control over the entire process, enterprises are gradually integrating multi-source heterogeneous data systems—such as Manufacturing Execution System (MES), Product Lifecycle Management System (PLM), and Enterprise Resource Planning System (ERP)—to conduct quality assessments of the equipment.

[0003] In existing technologies, a uniform and fixed method is usually used to perform sliding slicing on all types of data, and then quality assessment is performed based on the slices. However, in the actual process of equipment manufacturing, there are long-cycle smooth change data such as drying, aging, and pressure holding, as well as short-term event data such as opening and closing actions. Therefore, if a uniform sliding window is used for slicing, the resulting slices cannot capture key detailed features, which leads to inaccurate quality assessment results. Summary of the Invention

[0004] This invention provides a device quality assessment method, apparatus, terminal device, and storage medium based on multi-source heterogeneous data, which can solve the problem of inaccurate quality assessment results caused by using a uniform and fixed method to perform sliding slicing on all types of data and then performing quality assessment based on the slices in the prior art.

[0005] An embodiment of the present invention provides a device quality assessment method based on multi-source heterogeneous data, comprising: Acquire multi-source heterogeneous time-series data to represent the manufacturing process of the equipment to be evaluated; The above multi-source heterogeneous time series data are divided according to the temporal change pattern to obtain slowly changing process time series data and short-time event time series data; The time series data of the above-mentioned slowly changing process is divided according to the preset time window length to obtain several first slices; A time window is constructed based on the trigger time corresponding to the short-term event time series data mentioned above, and the corresponding short-term event time series data is segmented according to the time window to obtain several second slices. Based on all the first slices and all the second slices, several statistical characteristics of the device to be evaluated were calculated. Based on all statistical characteristics, the quality assessment results of the above-mentioned equipment to be evaluated are determined.

[0006] Furthermore, before dividing the aforementioned multi-source heterogeneous time-series data according to the temporal variation pattern to obtain slowly varying process time-series data and short-term event time-series data, the following steps are also included: Obtain the reference time and the time offset corresponding to the above multi-source heterogeneous time series data; Based on the aforementioned reference time and time offset, the time of the aforementioned multi-source heterogeneous time series data after the unified time base is calculated, and the corresponding multi-source heterogeneous time series data is time-aligned according to the time after the unified time base.

[0007] Furthermore, based on all the first slices and all the second slices, several statistical characteristics of the device to be evaluated are calculated, including: Obtain the preset standard value and allowable deviation range corresponding to each multi-source heterogeneous time series data; Based on the preset standard values ​​and allowable deviation ranges corresponding to all slices, the deviation characteristics of the above-mentioned equipment to be evaluated are calculated; wherein, the above-mentioned slices include: the first slice and the second slice; Based on the median of the current slice, normalize the current slice to obtain the data robustness strength value; Based on the robustness strength values ​​of the above data, the abnormal data in the corresponding slices are determined, and the abnormal data is removed to obtain the preprocessed slices. Least squares fitting is performed on each preprocessed slice to calculate the stationarity deviation of the current preprocessed slice. Extract the third slice corresponding to the pressure test time series data of the pressure holding section from all slices, and perform linear regression on the third slice to obtain the steady-state measure value. Extract the fourth slice from all slices to represent the mechanical property test results, and obtain the first speed threshold for determining the start of the action and the second speed threshold for determining the stability of the action from the fourth slice. Based on the aforementioned fourth slice, first speed threshold, and second speed threshold, the mechanical quality factor of the aforementioned equipment to be evaluated is determined. The aforementioned deviation characteristics, data robustness strength value, stationarity deviation, steady-state measurement value, and mechanical quality factor are used as the above statistical characteristic quantities.

[0008] Furthermore, the aforementioned quality assessment results include: compliance verification results and operational risk assessment results; Based on all statistical characteristics, the quality assessment results for the equipment to be evaluated are determined, including: Obtain the historical mean and historical standard deviation of the above-mentioned deviation characteristics, the historical mean and historical standard deviation of the above-mentioned data robustness strength values, and the historical mean and historical standard deviation of the above-mentioned stationarity deviation. Based on the corresponding historical mean and historical standard deviation, the deviation characteristics, data robustness strength value and stationarity deviation are standardized to obtain the corresponding standardized characteristic values. The standardized feature values ​​corresponding to the above deviation characteristics are compared with the preset deviation threshold; If the standardized feature value corresponding to the above-mentioned deviation characteristics exceeds the corresponding preset deviation threshold, the compliance verification result of the above-mentioned equipment to be evaluated is determined to be passed; otherwise, the compliance verification result of the above-mentioned equipment to be evaluated is determined to be failed. The robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the above data are compared with the corresponding preset safety threshold ranges. If the robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the above data all exceed the corresponding preset safety threshold range, the operational risk assessment result of the above-mentioned equipment to be evaluated is determined to be that there is an operational risk.

[0009] Furthermore, it also includes: Obtain the defect count and abnormal inspection report text of the above-mentioned equipment to be evaluated during the manufacturing process; The above-mentioned abnormal inspection report text is compared with the preset abnormal clauses to obtain the abnormal clauses triggered by the above-mentioned equipment to be evaluated. After obtaining the quality assessment results of the aforementioned equipment to be evaluated, feedback is provided based on the aforementioned defect count and the triggered anomaly clauses as evidence for the quality assessment.

[0010] Furthermore, before determining the quality assessment results of the aforementioned equipment based on all statistical characteristics, the following steps are also included: Based on the above reference time and multi-source heterogeneous time series data, the data source time deviation is calculated; Based on the time deviation of the above data sources, the time synchronization reliability is calculated; Missing and invalid values ​​are detected for data points within each slice to obtain the missing rate and coverage. The time synchronization reliability, missing rate, and coverage are compared with the corresponding preset normal threshold ranges. If the time synchronization reliability, missing rate, and coverage all meet the corresponding preset normal threshold ranges, the quality assessment result of the device to be evaluated is determined based on all statistical features. Otherwise, supplementary sampling or verification of the corresponding multi-source heterogeneous time series data should be performed.

[0011] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments; This invention provides a device for evaluating equipment quality based on multi-source heterogeneous data, comprising: The system includes a data acquisition module, a data type classification module, a time series data segmentation module for slowly changing processes, a time series data segmentation module for short-term events, a statistical feature calculation module, and a quality assessment module. The aforementioned data acquisition module is used to acquire multi-source heterogeneous time-series data representing the manufacturing process of the equipment to be evaluated; The aforementioned data type segmentation module is used to segment the multi-source heterogeneous time series data according to the temporal change pattern, thereby obtaining slowly changing process time series data and short-term event time series data. The aforementioned time series data segmentation module for the gradual change process is used to segment the aforementioned time series data for the gradual change process according to a preset time window length to obtain several first slices; The aforementioned short-term event time-series data segmentation module is used to construct a time window based on the trigger time corresponding to the aforementioned short-term event time-series data, and to segment the corresponding short-term event time-series data according to the aforementioned time window to obtain several second slices; The aforementioned statistical characteristic calculation module is used to calculate several statistical characteristic quantities of the device to be evaluated based on all first slices and all second slices. The aforementioned quality assessment module is used to determine the quality assessment result of the equipment to be assessed based on all statistical characteristics.

[0012] Furthermore, it also includes: a time alignment module; The aforementioned time alignment module is used to obtain a reference time and the corresponding time offset of the aforementioned multi-source heterogeneous time series data before dividing the aforementioned multi-source heterogeneous time series data according to the time series change pattern to obtain the slowly changing process time series data and the short-term event time series data; based on the aforementioned reference time and time offset, calculate the time of the aforementioned multi-source heterogeneous time series data after the unified time base, and perform time alignment of the corresponding multi-source heterogeneous time series data according to the time after the unified time base.

[0013] Based on the above method embodiments, the present invention provides a corresponding terminal device embodiment; The present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the device quality assessment method based on multi-source heterogeneous data described in any embodiment of the present invention.

[0014] Based on the above method embodiments, the present invention provides a corresponding storage medium embodiment; The present invention provides a storage medium including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the device quality assessment method based on multi-source heterogeneous data described in any embodiment of the present invention.

[0015] The embodiments of the present invention have the following beneficial effects: This invention provides a method, apparatus, terminal device, and storage medium for equipment quality assessment based on multi-source heterogeneous data. The method includes: acquiring multi-source heterogeneous time-series data representing the manufacturing process of the equipment to be assessed; then dividing the multi-source heterogeneous time-series data according to the time-series change pattern to obtain slowly changing process time-series data and short-term event time-series data; further dividing the slowly changing process time-series data according to a preset time window length to obtain several first slices; then constructing a time window based on the trigger time corresponding to the short-term event time-series data, and dividing the corresponding short-term event time-series data according to the time window to obtain several second slices; then calculating several statistical characteristic quantities of the equipment to be assessed based on all first slices and all second slices; and finally determining the quality assessment result of the equipment to be assessed based on all statistical characteristic quantities. Therefore, in this invention, the acquired multi-source heterogeneous time-series data are divided according to the temporal change pattern. Then, for the time-series data of the slow-change process, it is segmented according to the conventional slicing method, while for the time-series data of short-term events, it is sliced ​​according to the event trigger time. This allows the obtained slices to accurately cover the key time periods, and the subsequent quality assessment can be based on the data of these core time periods, thus improving the accuracy of the assessment results. Attached Figure Description

[0016] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating a device quality assessment method based on multi-source heterogeneous data, provided by an embodiment of the present invention.

[0018] Figure 2 This is a schematic diagram of the structure of a device quality assessment device based on multi-source heterogeneous data provided in an embodiment of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0021] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0022] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0023] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0024] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0025] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0026] See Figure 1 To address the problem of inaccurate quality assessment results caused by the existing technology's use of a uniform and fixed method for sliding slicing all types of data before quality assessment, an embodiment of the present invention provides a device quality assessment method based on multi-source heterogeneous data, comprising: Step S101: Acquire multi-source heterogeneous time-series data to represent the manufacturing process of the equipment to be evaluated; Specifically, the equipment to be evaluated refers to main power grid equipment (such as GIS, gas-insulated switchgear). Therefore, the aforementioned multi-source heterogeneous time-series data consists of process time-series data (including process and environmental data) and test time-series data collected during the entire manufacturing process of the main power grid equipment, including manufacturing and performance measurement. For illustrative purposes, the aforementioned process time-series data includes temperature, pressure, humidity, torque, etc., while the aforementioned test time-series data includes sealing pressure data, impact records, partial discharge counts, speed, displacement, etc.

[0027] As an illustration, a unified primary key is established using "Project-Equipment-Batch-Component-Workstation", and the set of manufacturing objects is denoted as... Therefore, the above-mentioned process timing data and experimental timing data can be expressed as follows: In the formula, This represents the time sequence data of the process corresponding to the i-th process parameter, where t represents time. This represents the data point at time t in the time series data corresponding to the i-th process parameter. This indicates the total number of process parameters. This represents the experimental time series data corresponding to the j-th experimental time series parameter. This represents the data point at time t in the experimental time series data corresponding to the j-th experimental time series parameter. This indicates the total number of timing parameters in the experiment. This represents the u-th manufacturing object, i.e., the device to be evaluated in this invention. This indicates the total number of devices to be evaluated. This represents the unified time index set corresponding to the u-th manufacturing object.

[0028] Step S102: Divide the above multi-source heterogeneous time series data according to the temporal change pattern to obtain the time series data of the slowly changing process and the time series data of the short-term event; Specifically, the aforementioned multi-source heterogeneous time-series data includes both long-cycle smooth change segments such as drying, aging, and pressure holding, the core value of which is trend change (e.g., pressure slowly decreasing, temperature gradually increasing). Simultaneously, the data also includes short-term events such as circuit breaker opening and closing actions, the start and end of withstand voltage and partial discharge tests, equipment start-up and shutdown, and alarms. The core value of these short-term events is instantaneous details (e.g., speed change at the moment of circuit breaker opening, partial discharge signal at the start of withstand voltage test). Therefore, it is necessary to classify these two types of time series according to their change patterns into "slow-change processes" and "event-type processes (i.e., the processes corresponding to short-term event time-series data)" before proceeding with subsequent operations.

[0029] In a preferred embodiment, before dividing the multi-source heterogeneous time-series data according to the time-series change pattern to obtain slowly changing process time-series data and short-term event time-series data, the method further includes: Obtain the reference time and the time offset corresponding to the above multi-source heterogeneous time series data; Based on the aforementioned reference time and time offset, the time of the aforementioned multi-source heterogeneous time series data after the unified time base is calculated, and the corresponding multi-source heterogeneous time series data is time-aligned according to the time after the unified time base.

[0030] Specifically, to eliminate trigger delays and time misalignments from different data sources, all multi-source heterogeneous time-series data are uniformly time-mapped according to a reference time to achieve time alignment: In the formula, This represents the time in a unified time base after mapping to the reference time. This represents the time offset of the data source s corresponding to the u-th manufacturing object.

[0031] In this preferred embodiment, time alignment of the time-series data is achieved based on a set reference time.

[0032] Step S103: Divide the above-mentioned slowly changing process time series data into several first slices according to the preset time window length; Specifically, after time alignment, on a unified time axis, for time series data of a slowly changing process, a sliding window of a preset time window length is used for alignment and segmentation. This sliding window can be represented as: In the formula, This indicates that the preset time window length is... sliding window, This indicates the end time of the sliding time window on the unified time axis. Indicates the preset time window length.

[0033] Step S104: Construct a time window based on the trigger time corresponding to the short-term event time series data, and divide the corresponding short-term event time series data according to the time window to obtain several second slices; Specifically, on a unified timeline, based on the trigger time of the events corresponding to the short-term event time series data, a corresponding time window is established as follows: In the formula, Indicates the trigger time The corresponding time window, This indicates the trigger time of the k-th event; Indicates the forward extension duration before the triggering time. Indicates the duration of the extension after the triggering time. and All are specified by the process and testing procedures.

[0034] Specifically, the entire partitioning process can be uniformly represented as: In the formula, This represents the unified slicing operator, which ultimately outputs a mixed set of slices consisting of the first and second slices. This indicates the limitation of the function within a certain time window.

[0035] Preferably, the time alignment and slicing method of the present invention enables data with different sampling rates and different trigger calibers to be aligned under a unified time base.

[0036] Step S105: Based on all the first slices and all the second slices, calculate several statistical characteristic quantities of the above-mentioned device to be evaluated; Specifically, based on all first slices and all second slices, their respective statistical characteristics are calculated.

[0037] In a preferred embodiment, the statistical characteristics of the device to be evaluated calculated based on all first slices and all second slices include: Obtain the preset standard value and allowable deviation range corresponding to each multi-source heterogeneous time series data; Specifically, multi-source heterogeneous time-series data actually contains several key parameters for quality assessment. These key parameters include process parameters and test time-series parameters. Process parameters include continuous sampling parameters such as temperature, pressure, humidity, torque, speed, and micro-water. Test time-series parameters include sealing pressure curves, impact records, partial discharge counts, and other time-series detection data that can characterize the manufacturing quality and test status of the equipment.

[0038] For each key parameter, based on relevant technical standards or enterprise internal control requirements, its preset standard value and tolerance band width (i.e., the allowable deviation range mentioned above) are determined in advance. These two data can characterize the specification band of the corresponding parameter.

[0039] Based on the preset standard values ​​and allowable deviation ranges corresponding to all slices, the deviation characteristics of the above-mentioned equipment to be evaluated are calculated; wherein, the above-mentioned slices include: the first slice and the second slice; Specifically, first calculate the dimensionless deviation ratio of each slice: In the formula, This represents the dimensionless deviation ratio of the key parameter i' at time t. This represents the value of the key parameter i' at time t (i.e., the data point at time t within the slice). This represents the preset standard value of the key parameter i'. This indicates the allowable deviation of the key parameter i' when centered on a preset standard value. This indicates that the value of the numerical stability term is greater than 0. Its function is to prevent the dimensionless deviation ratio from diverging due to extremely small allowable deviations.

[0040] Subsequently, for the maximum dimensionless deviation of each parameter in each slice, a deviation feature group for each parameter is formed. Then, by summing up the deviation feature groups of all parameters, the deviation characteristics of the equipment to be evaluated can be obtained.

[0041] Based on the median of the current slice, normalize the current slice to obtain the data robustness strength value; Specifically, to suppress the effects of missing data, jumps, and local outliers, the data points within the slice are normalized using the median and MAD (Median Absolute Deviation) to obtain the data robustness value: In the formula, This represents the data robustness strength value of the key parameter i' at time t. Representing data points The median of the corresponding slice, This represents the median absolute deviation of the key parameter i' from the median of all data points within the current slice S. This represents a numerically stable term greater than 0, used to prevent the denominator from being too small or equal to 0.

[0042] Preferably, in the above formula, the coefficient 1.4826 is used to... The standard deviation is uniformly estimated under a normal distribution.

[0043] Based on the robustness strength values ​​of the above data, the abnormal data in the corresponding slices are determined, and the abnormal data is removed to obtain the preprocessed slices. Specifically, in each slice, there may be extreme values ​​such as jumps or outliers caused by false alarms from the sensor. In order to avoid these anomalies affecting the subsequent fitting process, it is necessary to use data robustness strength value alignment to detect abnormal data. If the data robustness strength value of a certain data point exceeds the corresponding set threshold, it is determined to be abnormal data and is removed.

[0044] Preferably, for outlier data, the preprocessing can also be achieved by reducing its weight in the least squares fitting.

[0045] Least squares fitting is performed on each preprocessed slice to calculate the stationarity deviation of the current preprocessed slice. Specifically, least squares fitting is performed on each preprocessed slice to obtain the trend slope and goodness of fit: In the formula, Representing data The corresponding trend slope of the slice, Representing data The goodness of fit of the corresponding slice, where a represents the linear fit intercept and b represents the linear fit slope parameter; Represents a moment on a unified timeline; Representing data The predicted value obtained by linear fitting at time t.

[0046] The residuals are then calculated based on the predicted values, and the stationarity deviation, which characterizes the short-term volatility of the data around the trend line, is calculated from the residuals. In the formula, This represents the smoothed residual value at time t. This represents the smoothing coefficient, and its value is... , Represents the residual at time t. This represents the degree of stationarity deviation at time t. This represents the standard deviation of the slice residuals.

[0047] Extract the third slice corresponding to the pressure test time series data of the pressure holding section from all slices, and perform linear regression on the third slice to obtain the steady-state measure value. Specifically, linear regression was performed on the pressure holding section pressure test time series data obtained under approximately isothermal and steady-state operating conditions: In the formula, Indicates time The stress value after linear regression fitting. Indicates the regression intercept. The regression coefficient representing the time term. This represents the regression coefficient that characterizes the effect of temperature compensation. Indicates time Temperature measurement value.

[0048] Subsequently, based on the pressure fitting values ​​within the third slice corresponding to the pressure holding stage... The steady-state measure used to characterize the degree of pressure fluctuation and the existence of significant and continuous drift during the pressure holding phase is calculated: In the formula, Represents a steady-state performance measure. This represents the root mean square error of the third slice. This represents the range of the third slice.

[0049] Specifically, the closer the above steady-state measure value is to 1, the more stable the state.

[0050] Extract the fourth slice from all slices to represent the mechanical property test results, and obtain the first speed threshold for determining the start of the action and the second speed threshold for determining the stability of the action from the fourth slice. Specifically, the aforementioned mechanical characteristic test includes: the fourth slice corresponding to displacement and velocity; firstly, for the fourth slice corresponding to velocity, the velocity threshold used to determine the start and stability of the action is obtained.

[0051] Based on the aforementioned fourth slice, first speed threshold, and second speed threshold, the mechanical quality factor of the aforementioned equipment to be evaluated is determined. Specifically, the aforementioned mechanical quality factors refer to the opening or closing time, the bounce duration, and the overshoot ratio. First, based on two speed thresholds, the threshold cross-validation method is used to determine the start time of the action and the time when the action reaches a stable state. Then, combining the fourth slice corresponding to the displacement, and considering the start and end intervals of the action, the duration of the oscillation before stabilization, and the deviation of the peak value from the steady-state position, the opening or closing time, the bounce duration, and the overshoot ratio are determined. Among these, the key moments are: In the formula, Indicates the moment the action begins; This indicates the velocity value of the fourth slice at time t; This represents the first velocity threshold used to determine the start of an action; This represents the second velocity threshold used to determine whether an action has entered a steady state. This indicates the moment when the action reaches a stable state.

[0052] The aforementioned deviation characteristics, data robustness strength value, stationarity deviation, steady-state measurement value, and mechanical quality factor are used as the above statistical characteristic quantities.

[0053] In this preferred embodiment, statistical characteristics of the device to be evaluated are calculated for each slice.

[0054] Step S106: Determine the quality assessment result of the above-mentioned equipment to be evaluated based on all statistical characteristics.

[0055] Specifically, the above quality assessment results include two parts. The first is the compliance verification result, which indicates whether the equipment meets the most basic qualification requirements (i.e., meets the relevant compliance clauses). The second is the operational risk assessment result, which indicates whether the equipment is prone to triggering risks during subsequent operation.

[0056] In a preferred embodiment, the above quality assessment results include: compliance verification results and operational risk assessment results; Based on all statistical characteristics, the quality assessment results for the equipment to be evaluated are determined, including: Obtain the historical mean and historical standard deviation of the above-mentioned deviation characteristics, the historical mean and historical standard deviation of the above-mentioned data robustness strength values, and the historical mean and historical standard deviation of the above-mentioned stationarity deviation. Specifically, based on relevant data from the historical manufacturing of equipment of the same model or process, the historical mean and historical standard deviation of each statistical characteristic are obtained.

[0057] Based on the corresponding historical mean and historical standard deviation, the deviation characteristics, data robustness strength value and stationarity deviation are standardized to obtain the corresponding standardized characteristic values. Specifically, consistency determination and standardization of each statistical characteristic are performed using the following formula: In the formula, This represents the k-th standardized statistical feature. This represents the statistical characteristic before the k-th standardization. This represents the k-th historical mean. This represents the k-th historical standard deviation.

[0058] Preferably, the obtained standardized feature values ​​can be combined with the maximum dimensionless deviation and feature concatenation can be performed using an aggregation operator to obtain a vector representing the uniform quality characteristics of the equipment: In the formula, Indicates equipment A vector of uniform quality characteristics. Represents the aggregation operator, This represents a vector concatenation operation. Indicates key parameters The maximum dimensionless deviation, This represents the statistical feature vector or representative value corresponding to the key parameter i'. Indicates equipment The corresponding anomaly inspection report text features.

[0059] Specifically, this vector can be used as input data for the model when it is necessary to use relevant production quality assessment and risk determination models to evaluate and predict equipment.

[0060] The standardized feature values ​​corresponding to the above deviation characteristics are compared with the preset deviation threshold; If the standardized feature value corresponding to the above-mentioned deviation characteristics exceeds the corresponding preset deviation threshold, the compliance verification result of the above-mentioned equipment to be evaluated is determined to be passed; otherwise, the compliance verification result of the above-mentioned equipment to be evaluated is determined to be failed. The robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the above data are compared with the corresponding preset safety threshold ranges. If the robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the above data all exceed the corresponding preset safety threshold range, the operational risk assessment result of the above-mentioned equipment to be evaluated is determined to be that there is an operational risk.

[0061] Specifically, the aforementioned preset deviation thresholds and preset safety threshold ranges are all based on the process requirement standards given for the corresponding equipment.

[0062] Preferably, the aforementioned preset deviation threshold can be further divided into three thresholds to indicate whether the data is qualified, requires review, or is unqualified. Based on these three thresholds, it is determined which specific processing is required for the equipment. For example, if its standardized feature value is within the threshold range corresponding to the requirement for review, it indicates that there are doubts about the data of the equipment, and it needs to be reviewed.

[0063] Preferably, the preset safety threshold range used for judging operational risks can be further subdivided into ranges representing different risk levels, thereby further determining which level of risk the device has triggered. If the risk level is very high, it needs to be dealt with first.

[0064] Preferably, the present invention fuses and analyzes multi-source quantized signals such as relative deviation features, robust deviation features, and stationary deviation features, thereby achieving a unified characterization of slow drift and sudden anomalies.

[0065] In this preferred embodiment, the quality assessment result of the device to be evaluated is determined based on all statistical characteristics.

[0066] In another preferred embodiment, it further includes: Obtain the defect count and abnormal inspection report text of the above-mentioned equipment to be evaluated during the manufacturing process; Specifically, the aforementioned defect count is a discrete defect category count obtained through image detection of appearance, welds, and markings during the manufacturing process, represented using vectors. The aforementioned anomaly inspection report text comprises both the inspection report and the anomaly text. Illustratively, the vector used to represent the defect count is as follows: In the formula, The vector representing the defect count. These represent the count values ​​corresponding to each defect category, and P represents the total number of defect categories.

[0067] The above-mentioned abnormal inspection report text is compared with the preset abnormal clauses to obtain the abnormal clauses triggered by the above-mentioned equipment to be evaluated. Specifically, the triggered abnormal clauses can be determined by extracting each clause from the preset abnormal clauses and comparing them with the content recorded in the abnormal inspection report text, and vector representation is also used.

[0068] After obtaining the quality assessment results of the aforementioned equipment to be evaluated, feedback is provided based on the aforementioned defect count and the triggered anomaly clauses as evidence for the quality assessment.

[0069] Specifically, the final quality assessment evidence is presented as follows: In the formula, This represents the vector corresponding to the quality assessment evidence. This represents the clause trigger count vector corresponding to the abnormal clause.

[0070] Specifically, this quality assessment evidence, along with the compliance verification results, is fed back to the staff after the compliance verification is completed.

[0071] In this preferred embodiment, a vector representing evidence of quality assessment is constructed for feedback to staff based on defect counts and abnormal inspection report texts of the equipment to be evaluated during the manufacturing process.

[0072] In another preferred embodiment, before determining the quality assessment result of the device to be evaluated based on all statistical characteristics, the method further includes: Based on the above reference time and multi-source heterogeneous time series data, the data source time deviation is calculated; Based on the time deviation of the above data sources, the time synchronization reliability is calculated; Specifically, firstly, for each slice, Welford incremental online updates are used to update the mean and cumulative sum of squared deviations of the data in each slice to obtain the statistical characteristics of the sample distribution within the slice and identify abnormal fluctuations; based on this, the time synchronization reliability is calculated by combining the time deviation of the data source. In the formula, Let represent the mean at time t. This represents the mean at time t-1. This represents the sampled data point at time t. Indicates the reliability of time synchronization. This indicates the maximum permissible time deviation set in advance. This represents the number of samples included in the statistics up to time t. This represents the cumulative sum of squared deviations up to time t. This represents the cumulative sum of squared deviations up to time t-1. This represents the time offset of the data source s in the u-th manufacturing object relative to the reference time.

[0073] Among them, the Welford incremental method is used to obtain and This is used for online statistical characterization of data distribution within slices, and provides a statistical basis for subsequent slice validity assessment, coverage evaluation, and outlier identification; the above-mentioned time-synchronization reliability It is further calculated from the time deviation of the data source.

[0074] Specifically, the smaller the time discrepancy, the closer the time synchronization reliability is to 1.

[0075] Missing and invalid values ​​are detected for data points within each slice to obtain the missing rate and coverage. Specifically, coverage reflects whether the data within the current window has been completely collected, while the missing rate reflects the amount of missing data. Let the theoretical total number of sampling points for the current slice under the preset sampling rules be... The number of valid sampling points is The number of data points corresponding to missing or invalid values ​​is Then the missing rate and coverage can be expressed as follows: In the formula, This indicates the missing rate of the current slice. Indicates the coverage of the current slice. This represents the theoretical total number of sampling points for the current slice. This indicates the number of data points corresponding to missing and invalid values ​​within the current slice. This indicates the number of valid sampling points within the current slice.

[0076] The time synchronization reliability, missing rate, and coverage are compared with the corresponding preset normal threshold ranges. If the time synchronization reliability, missing rate, and coverage all meet the corresponding preset normal threshold ranges, the quality assessment result of the device to be evaluated is determined based on all statistical features. Otherwise, supplementary sampling or verification of the corresponding multi-source heterogeneous time series data should be performed.

[0077] Specifically, a trusted gating system can be used to compare the missing rate, time synchronization reliability, and coverage with the corresponding preset normal threshold ranges. If the missing rate, time synchronization reliability, and coverage all meet the requirements, then subsequent quality assessment can be performed. Otherwise, the original multi-source heterogeneous time series data needs to be supplemented or manually reviewed, and subsequent quality assessment operations should be stopped.

[0078] In this preferred embodiment, before determining the quality assessment results of the equipment based on all statistical characteristics, data detection is performed using the equipment's missing rate, time synchronization reliability, and coverage to promptly identify data that needs to be reviewed and supplemented, thereby improving the accuracy of the final quality assessment results.

[0079] Based on the above method embodiments, the present invention provides corresponding apparatus embodiments.

[0080] like Figure 2 As shown, an embodiment of the present invention provides a device quality assessment apparatus based on multi-source heterogeneous data, comprising: The system includes a data acquisition module, a data type classification module, a time series data segmentation module for slowly changing processes, a time series data segmentation module for short-term events, a statistical feature calculation module, and a quality assessment module. The aforementioned data acquisition module is used to acquire multi-source heterogeneous time-series data representing the manufacturing process of the equipment to be evaluated; Specifically, the equipment to be evaluated refers to main power grid equipment (such as GIS, gas-insulated switchgear). Therefore, the aforementioned multi-source heterogeneous time-series data consists of process time-series data (including process and environmental data) and test time-series data collected during the entire manufacturing process of the main power grid equipment, including manufacturing and performance measurement. For illustrative purposes, the aforementioned process time-series data includes temperature, pressure, humidity, torque, etc., while the aforementioned test time-series data includes sealing pressure data, impact records, partial discharge counts, speed, displacement, etc.

[0081] The aforementioned data type segmentation module is used to segment the multi-source heterogeneous time series data according to the temporal change pattern, thereby obtaining slowly changing process time series data and short-term event time series data. Specifically, the aforementioned multi-source heterogeneous time-series data includes both long-cycle smooth change segments such as drying, aging, and pressure holding, the core value of which is trend change (e.g., pressure slowly decreasing, temperature gradually increasing). Simultaneously, the data also includes short-term events such as circuit breaker opening and closing actions, the start and end of withstand voltage and partial discharge tests, equipment start-up and shutdown, and alarms. The core value of these short-term events is instantaneous details (e.g., speed change at the moment of circuit breaker opening, partial discharge signal at the start of withstand voltage test). Therefore, it is necessary to classify these two types of time series according to their change patterns into "slow-change processes" and "event-type processes (i.e., the processes corresponding to short-term event time-series data)" before proceeding with subsequent operations.

[0082] The aforementioned time series data segmentation module for the gradual change process is used to segment the aforementioned time series data for the gradual change process according to a preset time window length to obtain several first slices; Specifically, after time alignment, on a unified time axis, for time series data of a slowly changing process, a sliding window of a preset time window length is used for alignment and segmentation. This sliding window can be represented as: In the formula, This indicates that the preset time window length is... sliding window, This indicates the end time of the sliding time window on the unified time axis. Indicates the preset time window length.

[0083] The aforementioned short-term event time-series data segmentation module is used to construct a time window based on the trigger time corresponding to the aforementioned short-term event time-series data, and to segment the corresponding short-term event time-series data according to the aforementioned time window to obtain several second slices; Specifically, on a unified timeline, based on the trigger time of the events corresponding to the short-term event time series data, a corresponding time window is established as follows: In the formula, Indicates the trigger time The corresponding time window, This indicates the trigger time of the k-th event; Indicates the forward extension duration before the triggering time. Indicates the duration of the extension after the triggering time. and All are specified by the process and testing procedures.

[0084] Specifically, the entire partitioning process can be uniformly represented as: In the formula, This represents the unified slicing operator, which ultimately outputs a mixed set of slices consisting of the first and second slices. This indicates the limitation of the function within a certain time window.

[0085] Preferably, the time alignment and slicing method of the present invention enables data with different sampling rates and different trigger calibers to be aligned under a unified time base.

[0086] The aforementioned statistical characteristic calculation module is used to calculate several statistical characteristic quantities of the device to be evaluated based on all first slices and all second slices. Specifically, the statistical feature calculation module calculates the statistical features of each of the first slices and the second slices respectively.

[0087] The aforementioned quality assessment module is used to determine the quality assessment result of the equipment to be assessed based on all statistical characteristics.

[0088] Specifically, in the quality assessment module, compliance verification results are evaluated to indicate whether the equipment meets the most basic qualification requirements (i.e., meets the relevant compliance clauses), and operational risk assessment results are evaluated to indicate whether the equipment is likely to trigger risks during subsequent operation.

[0089] In a preferred embodiment, it further includes: a time alignment module; The aforementioned time alignment module is used to obtain a reference time and the corresponding time offset of the aforementioned multi-source heterogeneous time series data before dividing the aforementioned multi-source heterogeneous time series data according to the time series change pattern to obtain the slowly changing process time series data and the short-term event time series data; based on the aforementioned reference time and time offset, calculate the time of the aforementioned multi-source heterogeneous time series data after the unified time base, and perform time alignment of the corresponding multi-source heterogeneous time series data according to the time after the unified time base.

[0090] In another preferred embodiment, the statistical characteristic calculation module includes: The system includes a data feature acquisition unit, a deviation feature calculation unit, a data robustness strength value calculation unit, a slice preprocessing unit, a stationarity deviation calculation unit, a steady-state measurement value calculation unit, a velocity threshold acquisition unit, a mechanical quality factor determination unit, and a statistical feature quantity generation unit. The aforementioned data feature acquisition unit is used to acquire the preset standard value and allowable deviation range corresponding to each multi-source heterogeneous time series data. The aforementioned deviation characteristic calculation unit is used to calculate the deviation characteristics of the device to be evaluated based on the preset standard values ​​and allowable deviation ranges corresponding to all slices; wherein, the slices include: a first slice and a second slice; The aforementioned data robustness strength calculation unit is used to normalize the current slice based on the median of the current slice to obtain the data robustness strength value. The aforementioned slice preprocessing unit is used to determine the abnormal data in the corresponding slice based on the aforementioned data robustness strength value, and to remove the aforementioned abnormal data to obtain the preprocessed slice. The aforementioned stationarity deviation calculation unit is used to perform least squares fitting on each preprocessed slice to calculate the stationarity deviation corresponding to the current preprocessed slice. The aforementioned steady-state metric calculation unit is used to extract the third slice corresponding to the pressure test time series data of the pressure holding section from all slices, and to perform linear regression on the aforementioned third slice to obtain the steady-state metric value. The speed threshold acquisition unit described above is used to extract a fourth slice representing the mechanical characteristic test results from all slices, and to acquire a first speed threshold for determining the start of the action and a second speed threshold for determining the stability of the action from the fourth slice. The aforementioned mechanical quality factor determination unit is used to determine the mechanical quality factor of the equipment to be evaluated based on the aforementioned fourth slice, the first speed threshold, and the second speed threshold. The aforementioned statistical feature generation unit is used to take the aforementioned deviation characteristics, data robustness strength value, stationarity deviation degree, steady-state measurement value, and mechanical quality factor as the aforementioned statistical feature quantities.

[0091] In another preferred embodiment, the above quality assessment results include: compliance verification results and operational risk assessment results; The aforementioned quality assessment module includes: The system includes a standard deviation data acquisition unit, a feature standardization unit, a threshold comparison unit, a compliance verification unit, a safety threshold range comparison unit, and an operational risk assessment unit. The aforementioned standard deviation data acquisition unit is used to acquire the historical mean and historical standard deviation of the aforementioned deviation characteristics, the historical mean and historical standard deviation of the aforementioned data robustness strength value, and the historical mean and historical standard deviation of the aforementioned stationarity deviation. The aforementioned feature standardization unit is used to standardize the deviation feature, data robustness strength value, and stationarity deviation based on the corresponding historical mean and historical standard deviation, respectively, to obtain the corresponding standardized feature values. The threshold comparison unit described above is used to compare the standardized feature value corresponding to the above deviation feature with a preset deviation threshold. The aforementioned compliance verification unit is used to determine that the compliance verification result of the device under evaluation is passed if the standardized feature value corresponding to the aforementioned deviation feature exceeds the corresponding preset deviation threshold; otherwise, it determines that the compliance verification result of the device under evaluation is failed. The aforementioned safety threshold range comparison unit is used to compare the aforementioned robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor with the corresponding preset safety threshold ranges, respectively. The aforementioned operational risk assessment unit is used to determine that the operational risk assessment result of the equipment to be assessed is that there is an operational risk when the robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the aforementioned data all exceed the corresponding preset safety threshold range.

[0092] In another preferred embodiment, it further includes: Feedback module; The aforementioned feedback module is used to obtain the defect count and abnormal inspection report text of the aforementioned equipment to be evaluated during the manufacturing process; The above-mentioned abnormal inspection report text is compared with the preset abnormal clauses to obtain the abnormal clauses triggered by the above-mentioned equipment to be evaluated. After obtaining the quality assessment results of the aforementioned equipment to be evaluated, feedback is provided based on the aforementioned defect count and the triggered anomaly clauses as evidence for the quality assessment.

[0093] In another preferred embodiment, it further includes: Quality assessment result determination module; The aforementioned quality assessment result determination module is used to calculate the data source time deviation based on the aforementioned reference time and multi-source heterogeneous time series data before determining the quality assessment result of the aforementioned equipment to be assessed based on all statistical characteristics. Based on the time deviation of the above data sources, the time synchronization reliability is calculated; Missing and invalid values ​​are detected for data points within each slice to obtain the missing rate and coverage. The time synchronization reliability, missing rate, and coverage are compared with the corresponding preset normal threshold ranges. If the time synchronization reliability, missing rate, and coverage all meet the corresponding preset normal threshold ranges, the quality assessment result of the device to be evaluated is determined based on all statistical features. Otherwise, supplementary sampling or verification of the corresponding multi-source heterogeneous time series data should be performed.

[0094] It should be noted that the device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without creative effort. The above schematic diagram is merely an example of a device quality assessment device based on multi-source heterogeneous data and does not constitute a limitation on a device quality assessment device based on multi-source heterogeneous data. It may include more or fewer components than illustrated, or combine certain components, or use different components.

[0095] Based on the above method embodiments, the present invention provides corresponding terminal device embodiments.

[0096] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the device quality assessment method based on multi-source heterogeneous data described in any embodiment of the present invention.

[0097] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the device. The aforementioned terminal devices may be computing devices such as desktop computers, laptops, handheld computers, and cloud servers. These devices may include, but are not limited to, processors and memory. The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. This processor is the control center of the device, connecting various parts of the device via various interfaces and lines. The aforementioned memory can be used to store the aforementioned computer programs and / or modules. The aforementioned processor implements various functions of the aforementioned device by running or executing the computer programs and / or modules stored in the aforementioned memory, and by calling data stored in the memory. The aforementioned memory may mainly include a program storage area and a data storage area, wherein the program storage area may store the operating system, at least one application program required for a function, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0098] Based on the above method embodiments, the present invention provides corresponding storage medium embodiments.

[0099] Another embodiment of the present invention provides a storage medium including a stored computer program, wherein, when the computer program is running, it controls the device where the storage medium is located to execute the device quality assessment method based on multi-source heterogeneous data described in any embodiment of the present invention.

[0100] In this embodiment, the storage medium is a computer-readable storage medium, and the computer program includes computer program code, which may be in the form of source code, object code, executable file, or some intermediate form. The computer-readable medium may include any entity or device capable of carrying the computer program code, recording media, USB flash drive, portable hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0101] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for evaluating equipment quality based on multi-source heterogeneous data, characterized in that, include: Acquire multi-source heterogeneous time-series data to represent the manufacturing process of the equipment to be evaluated; The multi-source heterogeneous time-series data are divided according to the temporal change pattern to obtain slowly changing process time-series data and short-time event time-series data; The time series data of the slowly changing process is divided according to a preset time window length to obtain several first slices; A time window is constructed based on the trigger time corresponding to the short-term event time series data, and the corresponding short-term event time series data is segmented according to the time window to obtain several second slices; Based on all the first slices and all the second slices, several statistical characteristic quantities of the device to be evaluated are calculated; The quality assessment result of the device to be evaluated is determined based on all statistical characteristics.

2. The equipment quality assessment method based on multi-source heterogeneous data according to claim 1, characterized in that, Before dividing the multi-source heterogeneous time-series data according to the temporal variation pattern to obtain slowly varying process time-series data and short-time event time-series data, the method further includes: Obtain the reference time and the time offset corresponding to the multi-source heterogeneous time series data; Based on the reference time and the time offset, the time of the multi-source heterogeneous time series data after the unified time base is calculated, and the corresponding multi-source heterogeneous time series data is time-aligned according to the time after the unified time base.

3. The equipment quality assessment method based on multi-source heterogeneous data according to claim 2, characterized in that, The calculation of several statistical characteristics of the device to be evaluated based on all first slices and all second slices includes: Obtain the preset standard value and allowable deviation range corresponding to each multi-source heterogeneous time series data; Based on the preset standard values ​​and allowable deviation ranges corresponding to all slices, the deviation characteristics of the device to be evaluated are calculated; wherein, the slices include: a first slice and a second slice; Based on the median of the current slice, normalize the current slice to obtain the data robustness strength value; Based on the data robustness strength value, abnormal data in the corresponding slice is determined, and after removing the abnormal data, a preprocessed slice is obtained. Least squares fitting is performed on each preprocessed slice to calculate the stationarity deviation of the current preprocessed slice. Extract the third slice corresponding to the pressure test time series data of the pressure holding section from all slices, and perform linear regression on the third slice to obtain the steady-state metric value; Extract a fourth slice from all slices to represent the mechanical property test results, and obtain a first speed threshold for determining the start of the action and a second speed threshold for determining the stability of the action from the fourth slice. The mechanical quality factor of the device to be evaluated is determined based on the fourth slice, the first speed threshold, and the second speed threshold. The deviation characteristics, data robustness strength value, stationarity deviation, steady-state measure value, and mechanical quality factor are used as the statistical feature quantities.

4. The equipment quality assessment method based on multi-source heterogeneous data according to claim 3, characterized in that, The quality assessment results include: compliance verification results and operational risk assessment results; The process of determining the quality assessment result of the equipment to be evaluated based on all statistical characteristics includes: Obtain the historical mean and historical standard deviation of the deviation feature, the historical mean and historical standard deviation of the data robustness strength value, and the historical mean and historical standard deviation of the stationarity deviation. Based on the corresponding historical mean and historical standard deviation, the deviation characteristics, data robustness strength value and stationarity deviation are standardized to obtain the corresponding standardized characteristic values. The standardized feature value corresponding to the deviation feature is compared with a preset deviation threshold. If the standardized feature value corresponding to the deviation feature exceeds the corresponding preset deviation threshold, the compliance verification result of the device to be evaluated is determined to be passed; otherwise, the compliance verification result of the device to be evaluated is determined to be failed. The robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor of the data are compared with the corresponding preset safety threshold ranges. If the data robustness strength value, stability deviation, steady-state measurement value, and mechanical quality factor all exceed the corresponding preset safety threshold range, the operational risk assessment result of the equipment to be evaluated is determined to be that there is an operational risk.

5. The equipment quality assessment method based on multi-source heterogeneous data according to claim 4, characterized in that, Also includes: Obtain the defect count and abnormal inspection report text of the equipment to be evaluated during the manufacturing process; The anomaly inspection report text is compared with the preset anomaly clauses to obtain the anomaly clauses triggered by the device to be evaluated. After obtaining the quality assessment results of the equipment to be evaluated, feedback is provided based on the defect count and the triggered exception clauses as evidence of the quality assessment.

6. The equipment quality assessment method based on multi-source heterogeneous data according to claim 5, characterized in that, Before determining the quality assessment result of the device to be evaluated based on all statistical characteristics, the process also includes: Based on the reference time and the multi-source heterogeneous time series data, the data source time deviation is calculated; The time synchronization reliability is calculated based on the time deviation of the data source. Missing and invalid values ​​are detected for data points within each slice to obtain the missing rate and coverage. The time synchronization reliability, missing rate, and coverage are compared with the corresponding preset normal threshold ranges. If the time synchronization reliability, missing rate, and coverage all meet the corresponding preset normal threshold ranges, the quality assessment result of the device to be evaluated is determined based on all statistical features. Otherwise, supplementary sampling or verification of the corresponding multi-source heterogeneous time series data should be performed.

7. A device for evaluating equipment quality based on multi-source heterogeneous data, characterized in that, include: The system includes a data acquisition module, a data type classification module, a time series data segmentation module for slowly changing processes, a time series data segmentation module for short-term events, a statistical feature calculation module, and a quality assessment module. The data acquisition module is used to acquire multi-source heterogeneous time-series data representing the manufacturing process of the equipment to be evaluated; The data type division module is used to divide the multi-source heterogeneous time series data according to the temporal change pattern to obtain slowly changing process time series data and short-term event time series data. The gradual change process time series data segmentation module is used to segment the gradual change process time series data according to a preset time window length to obtain several first slices; The short-term event time series data segmentation module is used to construct a time window based on the trigger time corresponding to the short-term event time series data, and segment the corresponding short-term event time series data according to the time window to obtain several second slices; The statistical feature calculation module is used to calculate several statistical features of the device to be evaluated based on all first slices and all second slices. The quality assessment module is used to determine the quality assessment result of the device to be assessed based on all statistical characteristics.

8. The equipment quality assessment device based on multi-source heterogeneous data according to claim 7, characterized in that, Also includes: Time alignment module; The time alignment module is used to obtain a reference time and the time offset corresponding to the multi-source heterogeneous time series data before dividing the multi-source heterogeneous time series data according to the time series change pattern to obtain the slowly changing process time series data and the short-time event time series data; calculate the time of the multi-source heterogeneous time series data after unifying the time base according to the reference time and the time offset; and perform time alignment on the corresponding multi-source heterogeneous time series data according to the time after unifying the time base.

9. A terminal device, characterized in that, The device includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements a device quality assessment method based on multi-source heterogeneous data as described in any one of claims 1 to 6.

10. A storage medium, characterized in that, The storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the storage medium is located to perform a device quality assessment method based on multi-source heterogeneous data as described in any one of claims 1 to 6.

Images