Equipment drift degree estimation method and device, electronic equipment and storage medium

Abstract

The invention discloses an equipment drift degree estimation method and device, electronic equipment and a storage medium, and relates to the technical field of equipment calibration. The method comprises the steps that in response to a calibration event for instrument equipment, historical calibration data are acquired, and a multi-source calibration number generated in the calibration process is collected; fusing multi-source calibration data and the historical calibration data to obtain a current calibration data set; key features are extracted from the current calibration data set, redundant information removal and dimension optimization are carried out on the extracted key features through a principal component analysis method, and target feature vectors are obtained; inputting the target feature vector into a pre-trained calibration calculation model for processing, and outputting a calibration result; and determining the drift degree according to the calibration result and the last calibration result of the last calibration event included in the historical calibration data. Core key features are accurately extracted, redundant interference is eliminated, core factors influencing the calibration result are focused, and the accuracy of calibration result prediction and drift degree evaluation is guaranteed.

Description

Technical Field

[0001] This invention relates to the field of equipment calibration technology, and in particular to a method, apparatus, electronic device, and storage medium for predicting the degree of equipment drift. Background Technology

[0002] During long-term use, the measurement accuracy of instruments gradually decreases due to factors such as component aging, environmental interference, and operational differences. This phenomenon is called drift. The degree of drift directly determines the reliability of the instrument's measurement results; therefore, accurately assessing the degree of drift is a core requirement for instrument calibration.

[0003] Traditional drift assessments often require waiting until the entire calibration cycle is completed before performing post-calculation calculations based on a large amount of measured data. This process is cumbersome, lacks real-time performance, and fails to provide a quick basis for adjusting calibration strategies. Consequently, calibration frequency and method settings lack flexibility, and errors are prone to accumulation or low calibration efficiency. To address these issues, a method for predicting equipment drift is urgently needed. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, storage medium, and computer program product for predicting the degree of device drift.

[0005] According to one aspect of the present invention, a method for predicting the degree of device drift is provided, comprising:

[0006] In response to calibration events for instruments and equipment, historical calibration data is acquired and multi-source calibration data generated during the current calibration process is collected.

[0007] The multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset;

[0008] Key features are extracted from the current calibration dataset, and redundant information is removed and dimensions are optimized using principal component analysis to obtain the target feature vector.

[0009] The target feature vector is input into a pre-trained calibration calculation model for processing, and the calibration result is output.

[0010] The degree of drift is determined based on the calibration results and the previous calibration results of the previous calibration event included in the historical calibration data.

[0011] According to another aspect of the present invention, a device for predicting the degree of device drift is provided, comprising:

[0012] The data acquisition module is used to respond to calibration events for instruments and equipment, acquire historical calibration data, and collect multi-source calibration data generated during the current calibration process;

[0013] Data fusion processing is used to fuse the multi-source calibration data and the historical calibration data to obtain the current calibration dataset;

[0014] The dimension reduction and optimization processing module is used to extract key features from the current calibration dataset, and to perform redundant information removal and dimension optimization on the extracted key features using principal component analysis to obtain the target feature vector.

[0015] The calibration calculation module is used to input the target feature vector into a pre-trained calibration calculation model for processing and output the calibration result.

[0016] The drift calculation module is used to determine the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data.

[0017] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0018] At least one processor; and

[0019] A memory that is communicatively connected to at least one processor; wherein,

[0020] The memory stores a computer program that can be executed by at least one processor, such that the at least one processor is able to execute the device drift estimation method of the present invention.

[0021] According to another aspect of the present invention, a computer-readable storage medium is provided, which stores computer instructions for causing a processor to execute and implement the device drift estimation method of the embodiments of the present invention.

[0022] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the steps in the above-described method.

[0023] The technical solution of this invention breaks through the limitations of a single data source. By deeply integrating multi-source calibration data with historical calibration data, a comprehensive and unified calibration dataset is constructed, avoiding analytical biases caused by data bias and improving the comprehensiveness of calibration and drift assessment. Relying on time series analysis, trend analysis, anomaly detection, and principal component analysis, core key features are accurately extracted and redundant interference is eliminated, focusing on the core factors affecting calibration results and drift trends, ensuring the accuracy of calibration result prediction and drift degree assessment. The pre-trained calibration calculation model quickly processes the optimized target feature vector and outputs calibration results in real time, without relying on post-calibration statistical analysis of the entire calibration cycle, significantly improving the timeliness and efficiency of the calibration process. By quantitatively calculating the difference between the current and previous calibration results, the degree of drift is obtained, achieving an objective determination of the drift state. This provides a clear basis for subsequent adaptive measures such as model retraining, calibration frequency adjustment, and calibration method optimization, breaking the rigid limitations of traditional calibration strategies and balancing calibration accuracy and resource utilization efficiency.

[0024] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention, 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 the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a flowchart illustrating a method for predicting the degree of equipment drift provided in an embodiment of the present invention;

[0027] Figure 2 This is a flowchart illustrating another method for predicting the degree of device drift provided in an embodiment of the present invention;

[0028] Figure 3 This is a schematic diagram of the structure of a device for predicting the degree of equipment drift provided in an embodiment of the present invention;

[0029] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the device drift prediction method of the present invention. Detailed Implementation

[0030] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0031] Example 1

[0032] Figure 1 This is a flowchart of a device drift estimation method provided in an embodiment of the present invention. This embodiment is applicable to instrument and equipment calibration scenarios. The method can be executed by a device drift estimation device, which can be implemented in hardware and / or software and can be configured in an electronic device.

[0033] like Figure 1 As shown, the methods for predicting equipment drift include:

[0034] S101. In response to calibration events for instruments and equipment, acquire historical calibration data and collect multi-source calibration data generated during the current calibration process.

[0035] Calibration events refer to the triggering conditions that initiate instrument calibration, including scheduled calibration (e.g., monthly), manually triggered calibration, and calibration triggered by instrument accuracy warnings. Historical calibration data refers to instrument calibration-related records prior to the current calibration event, including calibration date, historical calibration results, calibration personnel, environmental conditions, traceability information of calibration standards, and uncertainties. Multi-source calibration data refers to the multi-dimensional data generated during the current calibration process, including data collected by the instrument's built-in sensors (usually time-series data), the sequence of measurement values ​​output during the instrument calibration process, calibration environment data, calibration personnel data, and calibration standards.

[0036] Specifically, when a calibration event for an instrument or meter is triggered, historical calibration data is retrieved from the cloud database, and valid calibration records within the specified time range are filtered to form a historical calibration dataset. Simultaneously, multi-source data for the current calibration is collected. This includes sensor data such as temperature, pressure, current, and voltage collected by internal instrument sensors, with the collection frequency and accuracy set according to sensor type and calibration requirements; instrument measurement data including the instrument's own measurement output values ​​during calibration, especially multiple measured values ​​at various calibration points across the entire range; calibration environment data including environmental factors affecting calibration results such as temperature, humidity, and air pressure, with the collection frequency determined based on the rate of environmental change; calibration personnel data including calibration time, operation steps, operator ID, and other operation-related information; and calibration standard data including traceability information, accurate calibration values, and uncertainties of the calibration standards used.

[0037] Understandably, this step enables comprehensive coverage of multi-source data, breaking the limitations of traditional single data sources and providing complete data support for subsequent accurate analysis.

[0038] S102. The multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset.

[0039] Data fusion refers to the process of using specific technologies to unify multi-source calibration data from different sources, in different formats, and with different dimensions, with historical calibration data to form a dataset with consistent structure and logical correlation. The current calibration dataset refers to the unified dataset after fusion, which includes real-time multi-source data from the current calibration and historical calibration benchmark data. It is the core input for subsequent feature extraction and model calculation.

[0040] In some embodiments, fusing the multi-source calibration data and the historical calibration data to obtain the current calibration dataset includes S1021-S1022:

[0041] S1021. The multi-source calibration data is cleaned and standardized.

[0042] Optionally, the collected multi-source calibration data and historical calibration data are first cleaned to remove outliers, filter noisy data, and fill in missing values ​​to ensure data accuracy. Then, variables of different dimensions are processed using a standardization formula to eliminate scale differences. The standardization formula is as follows: ;

[0043] in, For standardized data, This is the original data. For data The mean, For data The standard deviation.

[0044] S1022. Based on a preset data fusion algorithm, the processed multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset.

[0045] Optionally, based on the calibration operation time, all preprocessed data are timestamped and aligned. A high-dimensional feature vector is formed by concatenating multiple variables, and sensor data, environmental data, calibration personnel data, calibration standard data and historical calibration data are fully integrated to construct a unified current calibration dataset.

[0046] Understandably, preprocessing eliminates interference caused by differences in data format and units, achieving data uniformity and comparability, and avoiding analytical errors caused by data heterogeneity. The deep integration of multi-source data and historical data integrates real-time status and historical trend information, providing a comprehensive and rich data foundation for subsequent key feature extraction, and improving the completeness and effectiveness of feature extraction.

[0047] S103. Extract key features from the current calibration dataset, and use principal component analysis to remove redundant information and optimize the dimensions of the extracted key features to obtain the target feature vector.

[0048] Key features refer to quantitative indicators that reflect the core principles of instrument calibration accuracy, environmental impact, operational consistency, and drift trend, including calibration deviation, environmental compensation coefficient, uncertainty contribution component, drift rate, and operational consistency index.

[0049] When extracting key features from the current calibration dataset, initial key features covering calibration deviation, environmental impact, uncertainty, drift trend, and operating procedures can be comprehensively extracted based on the current calibration dataset using three techniques: time series analysis, trend analysis, and anomaly detection. The specific implementation is as follows:

[0050] Time series analysis extracts time-dimensional features. The data sources are sensor data, instrument measurement data, and calibration environment data collected at fixed intervals (e.g., every 1 minute) within the current calibration cycle, and calibration result sequences with corresponding timestamps in historical calibration data. The processing logic integrates the above data into a multi-channel time series including sensor sequences, measurement output sequences, environmental sequences, and historical reference sequences. It analyzes the numerical correlation of each sequence at different time points (e.g., the change in instrument measurement deviation when the ambient temperature rises at a certain time point), extracts the time-dimensional change pattern of instrument calibration, and provides a time series basis for subsequent feature calculations.

[0051] Trend analysis extracts quantitative features. For calibration deviation, the measured value of the instrument is directly calculated by subtracting the precise calibration value of the calibration standard, resulting in the single calibration deviation at each calibration point, which serves as the core basic feature. For the environmental compensation coefficient, a correlation model between environmental data (temperature / humidity) and calibration deviation is constructed through regression analysis (such as linear regression), and the regression coefficient of the model is solved to quantify the degree of interference of environmental factors on the calibration results. For the drift rate, the average value of historical calibration deviation data is statistically analyzed over time, and the slope of the mean change is calculated through trend modeling to obtain the rate of decrease in instrument accuracy over time. For the operational consistency index, the standard deviation of multiple calibration results at multiple calibration points by the same operator is calculated; the smaller the standard deviation, the stronger the operational standardization.

[0052] Anomaly detection and removal of interfering features can be achieved by applying machine learning algorithms (such as isolated forests) to identify irregular fluctuations in multi-channel time series data (such as sudden jumps in instrument measurements or abnormal spikes in sensor data). By combining environmental data and operator records to analyze the causes of anomalies (such as sudden changes in environmental temperature and humidity or omissions in operating procedures), the feature values ​​corresponding to the abnormal data can be removed to avoid interfering with the authenticity of the core features.

[0053] Principal Component Analysis (PCA) is a statistical analysis method that removes redundant information and reduces data dimensionality through steps such as data centralization, covariance matrix calculation, eigenvalue and eigenvector solving, and principal component selection. Based on this, PCA is used to remove redundant information and optimize the dimensionality of the extracted key features to obtain the target feature vector, including S1031-S1034:

[0054] S1031. Perform data centralization processing on each of the key features to obtain the centralized key feature vector.

[0055] Optionally, data can be data-processed according to the following formula: ;

[0056] in, For the centered data vector, is the extracted key feature vector, and μ is the mean vector of the current calibration dataset.

[0057] S1032. Determine the covariance matrix of the current calibration dataset based on the centered key feature vector.

[0058] The optional ones are calculated according to the following formula: ;

[0059] Where Cov(x) is the covariance matrix of the current calibration dataset, and n is the total number of current and historical calibration records. This is the transpose of the centered data matrix.

[0060] S1033. Calculate the eigenvalues ​​of the covariance matrix and the eigenvectors corresponding to the eigenvalues, and sort the eigenvectors in descending order according to the magnitude of the eigenvalues.

[0061] Calculate the eigenvalues ​​and eigenvectors of the covariance matrix using the following formula:

[0062] ;

[0063] Where V is the eigenvector matrix and Λ is the eigenvalue diagonal matrix.

[0064] S1034. Select the top K feature vectors to form a feature vector matrix, and project the feature vector matrix onto a new feature space to obtain the target feature vector after dimensionality reduction optimization.

[0065] The purpose of this step is to select principal components, which mainly involves sorting the eigenvectors according to the magnitude of their eigenvalues ​​and selecting the eigenvectors corresponding to the k largest eigenvalues ​​to form a new eigenvector matrix V. k The centered data is projected onto a new feature space to obtain the dimensionality-reduced target feature vector. .

[0066] Understandably, this step extracts core key features, focuses on the core factors affecting calibration results and drift trends, avoids interference from irrelevant information, and improves the relevance of subsequent model calculations. Principal component analysis is used to remove redundant data, reduce dimensionality, reduce the amount of model computation, improve computational efficiency, and at the same time retain key information to ensure the effectiveness and accuracy of model input.

[0067] S104. Input the target feature vector into the pre-trained calibration calculation model for processing, and output the calibration result.

[0068] The pre-trained calibration calculation model refers to a regression model that uses the target feature vector corresponding to historical calibration data as input and historical calibration results as supervision labels. After training and optimization, it saves the optimal parameters and has the ability to calculate calibration results. The calibration result refers to the calibration deviation of multiple calibration points across the entire calibration process output by the model, which is the core quantitative indicator reflecting the current calibration accuracy of the instrument.

[0069] The training process of the calibration computation model is as follows: the historical fusion dataset is divided into 80% training set and 20% validation set. The historical target feature vector is used as input and the historical calibration result is used as label. The model parameters are initialized. The loss function (such as MSE) is minimized by gradient descent or tree splitting algorithm. An early stopping mechanism is used on the validation set to prevent overfitting. After training is completed, the optimal parameters are saved.

[0070] The target feature vector obtained from S103 is input into the pre-trained model, and then processed by the regression function. Calculate and output the calibration results at multiple calibration points across the entire calibration range; where P represents the predicted calibration result. These are the parameters of the regression model, and f is the regression function.

[0071] This step utilizes a pre-trained model to quickly output calibration results without waiting for actual measurements after the complete calibration cycle, improving the real-time nature of calibration result acquisition and saving time for subsequent drift calculation and strategy adjustment. The model, trained based on multi-source features and historical data, has high prediction accuracy, and the output calibration results can accurately reflect the current calibration status of the instrument, providing reliable core data for drift assessment.

[0072] S105. Determine the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data.

[0073] The previous calibration result refers to the calibration result corresponding to the previous calibration event adjacent to the current calibration event in the historical calibration data, including the mean and standard deviation of the calibration deviation of the previous calibration. The drift degree refers to the degree of change in the instrument's calibration accuracy between the current and previous calibrations, obtained through quantitative calculations, and is a core indicator for judging the degradation of instrument performance.

[0074] In one embodiment, determining the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data includes S1051-S1053:

[0075] S1051. Calculate the mean and standard deviation of the calibration results for this calibration event, and calculate the mean and standard deviation of the calibration results for the previous calibration event.

[0076] Among them, the mean value of the calibration results generated in this calibration event and standard deviation The mean of the previous calibration result for the previous calibration event. and standard deviation .

[0077] S1052. Calculate the mean deviation between the mean of the current calibration result and the mean of the previous calibration result, and calculate the standard deviation deviation between the standard deviation of the current calibration result and the standard deviation of the previous calibration result.

[0078] Among them, mean offset Standard deviation offset .

[0079] S1053. Determine the degree of drift based on the mean offset and the standard deviation offset, combined with the weights of the mean offset and the standard deviation offset.

[0080] Among them, the degree of drift as follows: ;

[0081] in, This represents the degree of drift between the current calibration and the previous calibration. The weights are the differences in mean. The weight of the standard deviation; ; ; These are the historical expanded uncertainties of the mean and standard deviation, respectively.

[0082] This step achieves precise assessment of drift degree through quantitative calculation, breaking the ambiguity of traditional qualitative judgment and providing an objective and accurate basis for judging instrument performance status. Based on the direct comparison between the current and previous calibration results, it focuses on the performance changes in adjacent calibration cycles, can timely capture instrument drift trends, and provide clear triggering basis for subsequent model retraining, calibration frequency adjustment or method optimization, ensuring the pertinence and effectiveness of the calibration strategy.

[0083] In this embodiment of the invention, the limitations of a single data source are overcome. By deeply integrating multi-source calibration data with historical calibration data, a comprehensive and unified calibration dataset is constructed, avoiding analytical biases caused by data bias and improving the comprehensiveness of calibration and drift assessment. Relying on time series analysis, trend analysis, anomaly detection, and principal component analysis, core key features are accurately extracted and redundant interference is eliminated, focusing on the core factors affecting calibration results and drift trends, ensuring the accuracy of calibration result prediction and drift degree assessment. The pre-trained calibration calculation model is used to quickly process the optimized target feature vector and output calibration results in real time, without relying on post-calibration statistical analysis of the entire calibration cycle, significantly improving the timeliness and efficiency of the calibration process. By quantitatively calculating the difference between the current and previous calibration results, the degree of drift is obtained, achieving an objective determination of the drift state. This provides a clear basis for subsequent adaptive measures such as model retraining, calibration frequency adjustment, and calibration method optimization, breaking the rigid limitations of traditional calibration strategies and balancing calibration accuracy and resource utilization efficiency.

[0084] Example 2

[0085] Figure 2 A flowchart illustrating a method for predicting equipment drift degree is provided in this embodiment of the invention. See also... Figure 2 The method includes the following steps:

[0086] S201. In response to a calibration event for an instrument or meter, acquire historical calibration data and collect multi-source calibration data generated during the current calibration process.

[0087] S202. The multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset.

[0088] S203. Extract key features from the current calibration dataset, and use principal component analysis to remove redundant information and optimize the dimensions of the extracted key features to obtain the target feature vector.

[0089] S204. Input the target feature vector into the pre-trained calibration calculation model for processing, and output the calibration result.

[0090] S205. Determine the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data.

[0091] The specific implementation methods of S201-S205 can be found in the description of the above embodiments, and will not be repeated here.

[0092] S206. In response to the drift degree being greater than a preset threshold, calculate the mean drift contribution rate based on the drift degree, mean offset, and the weight of the mean offset.

[0093] The preset threshold includes a drift threshold for the mean difference. The drift threshold of the difference between the standard deviation and the standard deviation It consists of the mean drift threshold and the standard deviation drift threshold determined based on the expanded uncertainty of the GUM method.

[0094] ; MPE stands for Maximum Permissible Error. This represents the combined standard uncertainty.

[0095] Mean drift contribution rate :

[0096] ; The mean offset represents the degree of drift between the current calibration and the previous calibration. Standard deviation offset .

[0097] S207. Determine the cause of the drift based on the mean drift contribution rate, and perform adaptive calibration based on the cause of the drift.

[0098] Optionally, if the mean drift contribution rate is in the first interval (e.g., >0.7), it is determined that the drift is mainly caused by the mean, and calibration adaptation is achieved by increasing the calibration frequency; if the mean drift contribution rate is in the second interval (e.g., [0.4, 0.7]), it is determined that the drift is caused by both the mean and the standard deviation, and calibration adaptation is achieved by increasing the calibration frequency and using a high-precision standard; if the mean drift contribution rate is in the third interval (e.g., <0.4), it is determined that the drift is caused by the standard deviation, and calibration adaptation is achieved by adjusting the calibration method; wherein, the calibration method includes increasing the number of repeated measurements, using multi-point calibration, employing dual-person verification, and enabling real-time environmental compensation; the value range of the third interval is lower than the value range of the second interval, and the value range of the second interval is lower than the value range of the first interval.

[0099] In this embodiment, the drift contribution rate is accurately attributed to the dominant cause of the drift (mean-dominant / standard deviation-dominant / common-dominant). Targeted measures such as switching to a high-precision standard, adding environmental compensation, and multi-point calibration are adopted to avoid calibration deviations caused by traditional one-size-fits-all adjustments, ensuring that the instrument remains in a state of accurate measurement over a long period.

[0100] Example 3

[0101] Figure 3 This is a schematic diagram of a device for predicting the degree of equipment drift provided in an embodiment of the present invention. This device can execute any of the equipment drift prediction methods of the present invention. For example... Figure 3 As shown, the device for predicting equipment drift includes:

[0102] The data acquisition module 301 is used to respond to calibration events for instruments and equipment, acquire historical calibration data and collect multi-source calibration data generated during the current calibration process;

[0103] Data fusion processing 302 is used to fuse the multi-source calibration data and the historical calibration data to obtain the current calibration dataset;

[0104] The dimension reduction and optimization processing module 303 is used to extract key features from the current calibration dataset, and to perform redundant information removal and dimension optimization on the extracted key features using principal component analysis to obtain the target feature vector.

[0105] The calibration calculation module 304 is used to input the target feature vector into a pre-trained calibration calculation model for processing and output the calibration result;

[0106] The drift calculation module 305 is used to determine the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data.

[0107] In some embodiments, the key features include calibration bias, environmental compensation coefficient, uncertainty contribution component, drift rate, and operational consistency index;

[0108] In terms of removing redundant information and optimizing the dimensions of the extracted key features using principal component analysis to obtain the target feature vector, data fusion processing 302 is specifically used for:

[0109] Each of the key features is digitized to obtain a centered key feature vector.

[0110] Based on the centered key feature vectors, the covariance matrix of the current calibration dataset is determined;

[0111] Calculate the eigenvalues ​​of the covariance matrix and the corresponding eigenvectors, and sort the eigenvectors in descending order according to the magnitude of the eigenvalues;

[0112] Select the top K feature vectors to form a feature vector matrix, and project the feature vector matrix onto a new feature space to obtain the target feature vector after dimensionality reduction optimization.

[0113] In some embodiments, regarding determining the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data, the drift calculation module 305 is specifically configured to:

[0114] Calculate the mean and standard deviation of the calibration results for this calibration event, and calculate the mean and standard deviation of the calibration results for the previous calibration event.

[0115] Calculate the mean deviation of the mean of the current calibration result from the mean of the previous calibration result, and calculate the standard deviation deviation of the current calibration result from the standard deviation of the previous calibration result;

[0116] The degree of drift is determined based on the mean offset and the standard deviation offset, combined with the weights of the mean offset and the standard deviation offset.

[0117] In some embodiments, it also includes:

[0118] The contribution calculation module is used to calculate the mean drift contribution rate based on the drift degree, mean offset, and the weight of the mean offset when the drift degree is greater than a preset threshold; wherein the preset threshold is composed of a mean drift threshold and a standard deviation drift threshold determined based on the expanded uncertainty of the GUM method.

[0119] An adaptive module is used to determine the cause of the drift based on the mean drift contribution rate, and to perform adaptive calibration based on the cause of the drift.

[0120] In some embodiments, the adaptive module is specifically used for:

[0121] If the mean drift contribution rate is in the first interval, it is determined that the drift is mainly caused by the mean, and calibration adaptation is completed by increasing the calibration frequency;

[0122] If the mean drift contribution rate is in the second interval, it is determined that the drift is caused by both the mean and the standard deviation. The calibration adaptation is completed by increasing the calibration frequency and using a high-precision standard.

[0123] If the mean drift contribution rate is in the third interval, it is determined that the drift is caused by the standard deviation, and the calibration method is adjusted to complete the calibration adaptation.

[0124] The calibration method includes increasing the number of repeated measurements, using multi-point calibration, employing dual-person verification, and enabling real-time environmental compensation; the value range of the third interval is lower than the value range of the second interval, and the value range of the second interval is lower than the value range of the first interval.

[0125] In some embodiments, in fusing the multi-source calibration data and the historical calibration data to obtain the current calibration dataset, the data fusion processing 302 is specifically used for:

[0126] The multi-source calibration data is cleaned and standardized.

[0127] Based on a preset data fusion algorithm, the processed multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset.

[0128] In some embodiments, the multi-source calibration data includes time-series data collected by built-in sensors of the instrument, a sequence of measurement values ​​output during the calibration process of the instrument, calibration environment data, calibration personnel data, and calibration standards.

[0129] The device drift estimation device provided in this embodiment of the invention can execute the device drift estimation method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0130] According to embodiments of the present invention, the present invention also provides an electronic device, a readable storage medium, and a computer program product.

[0131] Example 4

[0132] Figure 4 A schematic diagram of the structure of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the invention described and / or claimed herein.

[0133] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory 12 or a random access memory 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 12 or loaded from storage unit 18 into the random access memory 13. The random access memory 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, read-only memory 12, and random access memory 13 are interconnected via a bus 14. An input / output interface 15 is also connected to the bus 14.

[0134] Multiple components in electronic device 10 are connected to input / output interface 15, including: input unit 16; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disks, optical disks, etc.; and communication unit 19, such as network interface cards, modems, wireless transceivers, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0135] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as performing device drift estimation methods.

[0136] In some embodiments, the device drift estimation method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via read-only memory 12 and / or communication unit 19. When the computer program is loaded into random access memory 13 and executed by processor 11, one or more steps of the device drift estimation method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the device drift estimation method by any other suitable means (e.g., by means of firmware).

[0137] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), complex programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0138] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to the processor of a general-purpose computer, a special-purpose computer, or other programmable device for drift estimation, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0139] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0140] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device or liquid crystal display for displaying information to a user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with a user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0141] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet. The computing system can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having client-server relationships with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, a host product within the cloud computing service system, addressing the shortcomings of traditional physical hosts and virtual private servers, such as high management difficulty and weak business scalability.

[0142] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0143] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for predicting the degree of equipment drift, characterized in that, include: In response to calibration events for instruments and equipment, historical calibration data is acquired and multi-source calibration data generated during the current calibration process is collected. The multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset; Key features are extracted from the current calibration dataset, and redundant information is removed and dimensions are optimized using principal component analysis to obtain the target feature vector. The target feature vector is input into a pre-trained calibration calculation model for processing, and the calibration result is output. The degree of drift is determined based on the calibration results and the previous calibration results of the previous calibration event included in the historical calibration data.

2. The method according to claim 1, characterized in that, The key features include calibration deviation, environmental compensation coefficient, uncertainty contribution component, drift rate, and operational consistency index. The process involves removing redundant information and optimizing the dimensions of the extracted key features using principal component analysis to obtain the target feature vector, including: Each of the key features is digitized to obtain a centered key feature vector. Based on the centered key feature vectors, the covariance matrix of the current calibration dataset is determined; Calculate the eigenvalues ​​of the covariance matrix and the corresponding eigenvectors, and sort the eigenvectors in descending order according to the magnitude of the eigenvalues; Select the top K feature vectors to form a feature vector matrix, and project the feature vector matrix onto a new feature space to obtain the target feature vector after dimensionality reduction optimization.

3. The method according to claim 1, characterized in that, Determining the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data includes: Calculate the mean and standard deviation of the calibration results for this calibration event, and calculate the mean and standard deviation of the calibration results for the previous calibration event. Calculate the mean deviation of the mean of the current calibration result from the mean of the previous calibration result, and calculate the standard deviation deviation of the current calibration result from the standard deviation of the previous calibration result; The degree of drift is determined based on the mean offset and the standard deviation offset, combined with the weights of the mean offset and the standard deviation offset.

4. The method according to claim 1 or 3, characterized in that, Also includes: In response to the drift degree being greater than a preset threshold, the mean drift contribution rate is calculated based on the drift degree, mean offset, and the weight of the mean offset; wherein, the preset threshold is composed of a mean drift threshold and a standard deviation drift threshold determined based on the expanded uncertainty of the GUM method. Based on the mean drift contribution rate, the cause of the drift is determined, and adaptive calibration is performed based on the cause of the drift.

5. The method according to claim 4, characterized in that, The step of determining the cause of drift based on the mean drift contribution rate and performing adaptive calibration based on the cause of drift includes: If the mean drift contribution rate is in the first interval, it is determined that the drift is mainly caused by the mean, and calibration adaptation is completed by increasing the calibration frequency; If the mean drift contribution rate is in the second interval, it is determined that the drift is caused by both the mean and the standard deviation. The calibration adaptation is completed by increasing the calibration frequency and using a high-precision standard. If the mean drift contribution rate is in the third interval, it is determined that the drift is caused by the standard deviation, and the calibration method is adjusted to complete the calibration adaptation. The calibration method includes increasing the number of repeated measurements, using multi-point calibration, employing dual-person verification, and enabling real-time environmental compensation; the value range of the third interval is lower than the value range of the second interval, and the value range of the second interval is lower than the value range of the first interval.

6. The method according to claim 1, characterized in that, The process of fusing the multi-source calibration data and the historical calibration data to obtain the current calibration dataset includes: The multi-source calibration data is cleaned and standardized. Based on a preset data fusion algorithm, the processed multi-source calibration data and the historical calibration data are fused to obtain the current calibration dataset.

7. The method according to claim 1, characterized in that, The multi-source calibration data includes time-series data collected by built-in sensors of instruments and equipment, sequence of measurement values ​​output during the calibration process of instruments and equipment, calibration environment data, calibration personnel data, and calibration standards.

8. A device for predicting the degree of equipment drift, characterized in that, include: The data acquisition module is used to respond to calibration events for instruments and equipment, acquire historical calibration data, and collect multi-source calibration data generated during the current calibration process; Data fusion processing is used to fuse the multi-source calibration data and the historical calibration data to obtain the current calibration dataset; The dimension reduction and optimization processing module is used to extract key features from the current calibration dataset, and to perform redundant information removal and dimension optimization on the extracted key features using principal component analysis to obtain the target feature vector. The calibration calculation module is used to input the target feature vector into a pre-trained calibration calculation model for processing and output the calibration result. The drift calculation module is used to determine the degree of drift based on the calibration result and the previous calibration result of the previous calibration event included in the historical calibration data.

9. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method of any one of claims 1-7.

Images