Confidence sorting method and system for salt iodine content detection data

Abstract

The invention relates to the field of food safety monitoring, in particular to a confidence sorting method and system for salt iodine content detection data. The method comprises the following steps: performing correlation analysis on a preset influence index set according to a detection process type to obtain an associated influence index set; receiving an association influence index constraint interval; during salt sample detection, collecting time sequence information of correlation influence index monitoring values and comparing constraint intervals to obtain deviation vector time sequence information; counting an error mode value of the detection sample set meeting the deviation vector time sequence information; when the error mode value is smaller than the error threshold value, credible identification is carried out on the detection value; and counting the detection value set with the credible identifier to obtain an iodine content confidence value, and sending the iodine content confidence value to a user side. By comprehensively considering the multi-dimensional association influence indexes, a systematic data confidence sorting mechanism is established, and the reliability and precision of the detection result of the iodine content of the salt are effectively improved.

Description

Technical Field

[0001] The present invention relates to the field of food safety monitoring, and particularly to a confidence sorting method and system for salt iodine content detection data. Background Art

[0002] At present, the detection of salt iodine content mainly adopts the spectrophotometry method. By measuring the absorbance value of the sample at a specific wavelength and combining with the standard curve, the iodine content is determined. However, in the actual detection process, due to the influence of various factors such as the instrument performance of the spectrophotometer, the sample characteristics, the environmental conditions, and the operation specifications, the detection results of salt iodine content often have certain fluctuations and errors, resulting in low reliability and accuracy of the detection results. Summary of the Invention

[0003] Aiming at the technical problem of low reliability and accuracy of the salt iodine content detection results in the prior art, the present invention achieves the technical effect of improving the reliability and accuracy of the salt iodine content detection results, and provides a confidence sorting method and system for salt iodine content detection data to solve this problem.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] In the first aspect, the present invention provides a confidence sorting method for salt iodine content detection data, including: performing a correlation analysis on a preset influence index set according to the salt iodine content detection process type to obtain an associated influence index set; receiving the associated influence index constraint interval of the associated influence index set; when detecting a salt sample, collecting the time series information of the associated influence index monitoring values, and comparing it with the associated influence index constraint interval to obtain the time series information of the associated influence index deviation vector; statistically analyzing the mode value of the iodine content detection error of the iodine content detection sample set that satisfies the time series information of the associated influence index deviation vector; when the mode value of the iodine content detection error is less than the error threshold, performing a credible identification on the iodine content detection value of the salt sample; statistically analyzing the set of iodine content detection values of the salt samples with credible identification to obtain the salt iodine content confidence value and sending it to the user terminal.

[0006] In a second aspect, the present invention provides a confidence sorting system for salt iodine content detection data, comprising: a correlation analysis module for performing a correlation analysis on a preset set of influencing indicators according to the type of salt iodine content detection process to obtain an associated set of influencing indicators; a constraint interval receiving module for receiving the associated influencing indicator constraint intervals of the associated set of influencing indicators; a deviation vector analysis module for collecting the time series information of the monitored values of the associated influencing indicators when detecting a salt sample, and comparing it with the associated influencing indicator constraint intervals to obtain the time series information of the deviation vectors of the associated influencing indicators; an error mode statistics module for statistically calculating the iodine content detection error mode value of the set of iodine content detection samples that satisfy the time series information of the deviation vectors of the associated influencing indicators; a detection value identification module for credibly identifying the iodine content detection value of the salt sample when the iodine content detection error mode value is less than the error threshold; and a confidence value statistics module for statistically calculating the confidence value of the salt iodine content for the set of iodine content detection values of the salt samples with credible identification and sending it to the user terminal.

[0007] The beneficial effects of the present invention are as follows:

[0008] According to the type of salt iodine content detection process, a correlation analysis is performed on a preset set of influencing indicators to obtain an associated set of influencing indicators. For a specific detection process type, various factors that may affect the detection results are analyzed, such as instrument factors (light source stability, detector performance, monochromator performance, wavelength accuracy), sample factors (uniformity, stability, concentration range), environmental factors (temperature stability, humidity stability, electromagnetic interference), and operation factors (measurement parameter settings, sample placement and operation specifications, data processing methods), etc. Through correlation analysis, the indicators that have a significant correlation with the detection results are screened out to form an associated set of influencing indicators, laying a foundation for subsequent data processing. Receive the associated influencing indicator constraint intervals of the associated set of influencing indicators. After determining the associated set of influencing indicators, a constraint interval is set for each associated influencing indicator, and these constraint intervals represent the reasonable fluctuation range of each indicator under normal detection conditions. The setting of the constraint intervals can be based on theoretical analysis, historical data statistics, or expert experience, providing a basis for subsequent judgment of the stability of the detection process. During the actual salt sample detection process, the values of each associated influencing indicator are monitored in real time, and the time series information of their changes over time is recorded to obtain the time series information of the monitored values of the associated influencing indicators. The time series information of the monitored values of the associated influencing indicators is compared with the associated influencing indicator constraint intervals, and the deviation degree and direction of each indicator are calculated to form the time series information of the deviation vectors, reflecting the stability status of each influencing factor during the detection process.

[0009] Based on the obtained deviation vector time series information, a set of detection samples under a specific pattern of the deviation vector time series information of the associated influence index is screened, and the detection errors of these samples are statistically analyzed to obtain the mode value of the error distribution, thereby identifying the typical error pattern under specific perturbation conditions and providing a quantitative index for evaluating the reliability of the detection results. According to the obtained mode value of the iodine content detection error, it is compared with the preset error threshold to determine whether the detection result is reliable. When the mode value of the iodine content detection error is less than the error threshold, it indicates that under the current detection conditions, despite the fluctuations of various influencing factors, the detection result still has sufficient accuracy, so the detection value is marked as reliable data. Statistical processing is performed on the set of iodine content detection values of the salt samples that have passed the credibility assessment to calculate the confidence value of the iodine content in salt with high reliability, and the result is transmitted to the user terminal. This statistical method based on confidence sorting effectively eliminates the interference of unreliable data and improves the accuracy and reliability of the final result.

[0010] Through the above technical solution, the confidence sorting of the iodine content detection data of salt is realized, effectively solving the problems of insufficient detection accuracy and reliability caused by ignoring the quality assessment of the detection process in the traditional method, and improving the reliability and accuracy of the iodine content detection result of salt. Brief Description of the Drawings

[0011] Figure 1 It is a schematic flow chart of the confidence sorting method for the iodine content detection data of salt provided by the present invention;

[0012] Figure 2 It is a schematic structural diagram of the confidence sorting system for the iodine content detection data of salt provided by the present invention.

[0013] In the drawings, the components represented by each reference numeral are as follows:

[0014] Correlation analysis module 11, constraint interval receiving module 12, deviation vector analysis module 13, error mode statistics module 14, detection value identification module 15, confidence value statistics module 16. Detailed Embodiments

[0015] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0016] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the described features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0017] In the description of the present invention, the term "for example" is used to mean "serving as an example, illustration, or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as more preferred or more advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the present invention. In the following description, details are set forth for purposes of explanation. It should be understood that those of ordinary skill in the art can recognize that the present invention can be practiced without these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

[0018] Embodiment 1, as Figure 1 shown, the embodiment of the present invention provides a confidence sorting method for salt iodine content detection data, including:

[0019] S1. Perform a correlation analysis on a preset influence index set according to the salt iodine content detection process type to obtain an associated influence index set.

[0020] Specifically, during the salt iodine content detection process, the detection results are affected by various factors. The preset influence index set includes the indexes that may affect the salt iodine content detection results, and these indexes can be divided into four categories: instrument factors, sample factors, environmental factors, and operation factors. Instrument factors include light source stability, detector performance, monochromator performance, and instrument wavelength accuracy, etc.; sample factors include sample uniformity, sample stability, and sample concentration range, etc.; environmental factors include temperature stability, humidity stability, and electromagnetic interference, etc.; operation factors include measurement parameter settings, sample placement, operation specifications, and data processing methods, etc. Different salt iodine content detection process types (such as spectrophotometry, potentiometric titration method, etc.) are interfered by different influence indexes to different degrees. Therefore, for a specific salt iodine content detection process type, a correlation analysis is performed on the preset influence index set to screen out the indexes that have a high correlation with the detection results of this specific process type and form an associated influence index set.

[0021] Correlation analysis evaluates the degree of correlation between each preset influencing indicator and the test result through statistical methods, and only retains the indicators whose correlation reaches a certain threshold as associated influencing indicators. This screening method can effectively reduce the complexity of subsequent analysis, avoid interference introduced by irrelevant indicators, and improve the analysis efficiency and accuracy. The specific correlation analysis method will be described in detail later to comprehensively and systematically evaluate the influence degree of each indicator and ensure the reliability of the obtained set of associated influencing indicators.

[0022] By obtaining the set of associated influencing indicators, it lays a foundation for the subsequent confidence sorting process of salt iodine content detection data, clarifies the influencing factors that need to be monitored key, and effectively improves the accuracy and reliability of salt iodine content detection.

[0023] S2. Receive the associated influencing indicator constraint intervals of the set of associated influencing indicators.

[0024] Specifically, the associated influencing indicator constraint intervals are comprehensively determined by an expert group based on the standards for salt iodine content detection, industry specifications, technical parameters of instrument equipment, and statistical analysis of a large amount of historical detection data. The expert group consists of senior experts in the field of food detection, professional technical personnel for analytical instruments, and statistical experts to ensure the authority and scientific nature of the setting of the constraint intervals. Specifically, the determination process of the associated influencing indicator constraint intervals includes the following aspects: First, refer to the regulatory requirements for the detection environment, instrument status, and sample handling in relevant standards; second, consider the technical parameters and their stability indicators of detection equipment such as spectrophotometers; third, through statistical analysis of a large amount of historical detection data, determine the normal distribution intervals of each associated influencing indicator; subsequently, combined with the actual operating conditions and technical feasibility of the laboratory, the expert group comprehensively evaluates the reasonable constraint intervals of each indicator. For example, for the associated influencing indicator of light source stability, the expert group determines its constraint interval as the light intensity fluctuation range not exceeding ±2% based on the technical specifications of the spectrophotometer and historical data analysis; for the associated influencing indicator of temperature stability, the expert group determines its constraint interval as 20±2°C based on the laboratory environmental control ability and detection method requirements.

[0025] The received associated influencing indicator constraint intervals are the key judgment criteria for the confidence sorting of salt iodine content detection data. The reasonable setting of the constraint intervals affects the ability to identify abnormal states, and thus affects the evaluation results of the credibility of the detection data. By receiving the associated influencing indicator constraint intervals of the set of associated influencing indicators determined by the expert group, a scientific basis for judging whether the indicators are in a normal state is obtained, providing a reliable reference standard for the deviation state analysis in the subsequent steps.

[0026] By receiving the associated influencing indicator constraint intervals determined by the expert group, the definition of the normal state of the associated influencing indicators is established, laying a foundation for the subsequent confidence sorting process of salt iodine content detection data.

[0027] S3. When detecting the salt sample, collect the time-series information of the monitoring values of the associated impact indicators, compare it with the constraint intervals of the associated impact indicators, and obtain the time-series information of the deviation vectors of the associated impact indicators.

[0028] Specifically, based on obtaining the set of associated impact indicators and their constraint intervals, the status of each associated impact indicator in the actual detection process is monitored and evaluated in real time to quantify the degree of deviation of the indicator from the normal state.

[0029] During the detection of the iodine content in salt, each associated impact indicator in the determined set of associated impact indicators is monitored in real time through a sensor network, and its monitoring value changing over time is recorded to form the time-series information of the monitoring values of the associated impact indicators. The acquisition frequency of the time-series information of the monitoring values of the associated impact indicators is set according to the characteristics of the detection equipment and the requirements of the detection process to ensure that the dynamic change process of the associated impact indicators can be captured. After obtaining the time-series information of the monitoring values of the associated impact indicators, it is compared and analyzed with the received constraint intervals of the associated impact indicators. In the comparison process, for each associated impact indicator, at each monitoring moment, the degree of deviation between the monitoring value and the constraint interval is calculated to generate a deviation vector. The deviation vector not only includes the direction of deviation (higher than the upper constraint limit or lower than the lower constraint limit), but also quantifies the degree of deviation, providing more abundant information for subsequent analysis. For example, assuming that the constraint interval for temperature stability is 20 ± 2°C, if the monitoring value at a certain moment is 23.5°C, the deviation vector at this moment is +1.5°C, indicating that it exceeds the upper constraint limit by 1.5°C; if the monitoring value is 17.8°C, the deviation vector is -0.2°C, indicating that it is 0.2°C lower than the lower constraint limit; if the monitoring value is 21°C, it is within the constraint interval and the deviation vector is 0. Arrange the deviation vectors at each moment in chronological order to form the time-series information of the deviation vectors of the associated impact indicators. This time-series information not only reflects whether each associated impact indicator deviates from the constraint interval, but also reflects the time pattern and dynamic change trend of the deviation, providing an important basis for subsequent evaluation of the credibility of the detection results.

[0030] By obtaining the time-series information of the deviation vectors of the associated impact indicators, the actual status of each associated impact indicator in the process of detecting the iodine content in salt can be comprehensively grasped, laying a foundation for subsequent result analysis and credibility evaluation based on similar detection conditions, and effectively improving the accuracy and reliability of the confidence sorting of the detection data of the iodine content in salt.

[0031] S4. Statistically obtain the mode value of the iodine content detection errors of the set of iodine content detection samples that satisfy the time-series information of the deviation vectors of the associated impact indicators.

[0032] Specifically, based on obtaining the time-series information of the deviation vector of the associated influence index, iodine content detection samples under similar influencing factor conditions are screened, and the distribution characteristics of the detection errors of these samples are analyzed to provide a basis for evaluating the credibility of the current detection results.

[0033] First, based on the obtained time-series information of the deviation vector of the associated influence index, a set of iodine content detection samples that meet the time-series information of the deviation vector of the associated influence index is screened from the historical detection database. Among them, meeting the time-series information of the deviation vector of the associated influence index means that in the historical detection process, the deviation state of its associated influence index has a high similarity with the deviation state of the current detection sample (i.e., the time-series information of the deviation vector of the associated influence index). This similarity takes into account not only the direction and degree of deviation of each associated influence index, but also the time pattern and dynamic change characteristics of the deviation.

[0034] For the screened set of iodine content detection samples, calculate the iodine content detection error of each iodine content detection sample, that is, the difference between the actual detection value and the true value of the sample. Then, statistical analysis is performed on these error values to determine their distribution characteristics, especially to find the mode of the error values, that is, the error value with the highest occurrence frequency, which is defined as the iodine content detection error mode value. The iodine content detection error mode value reflects the systematic error most likely to occur in the detection results under a specific deviation state of the associated influence index (i.e., the time-series information of the deviation vector of the associated influence index). This systematic error is mainly caused by the deviation of the associated influence index and has a certain regularity and predictability. By statistically analyzing the central tendency of this error, the reliability of the results under the current detection conditions can be evaluated, and a quantitative basis can be provided for judging the credibility of subsequent detection values.

[0035] The iodine content detection error mode value obtained through statistical analysis reflects the error characteristics of the detection system under specific influencing factor conditions, providing a reference for subsequent evaluation of the credibility of the detection results. This method based on historical data and analysis under similar conditions improves the scientificity and accuracy of the confidence sorting of salt iodine content detection data.

[0036] S5. When the iodine content detection error mode value is less than the error threshold, a credible label is given to the iodine content detection value of the salt sample.

[0037] Specifically, based on the obtained iodine content detection error mode value, the credibility of the iodine content detection value of the current salt sample is judged and labeled, providing a basis for realizing the confidence sorting of salt iodine content detection data.

[0038] First, compare the mode value of the iodine content detection error obtained statistically with a preset error threshold. The error threshold is a pre-set maximum acceptable error limit, and its setting is based on factors such as the standard requirements for salt iodine content detection, quality control requirements, and accuracy requirements of the actual application scenario. When the mode value of the iodine content detection error is less than the error threshold, it indicates that under the current deviation state of the associated influencing indicators, the systematic error generated by the detection system is within the acceptable range, and the detection result has a high reliability. At this time, the system makes a credible identification of the iodine content detection value of this salt sample and marks it as a credible result. The credible identification can be in the form of data flag bits, quality grades, or credibility scores, etc., so as to distinguish detection results with different credibility levels in subsequent data processing and applications. On the contrary, if the mode value of the iodine content detection error is greater than or equal to the error threshold, it indicates that under the current deviation state of the associated influencing indicators, the detection system may generate a large systematic error, and the reliability of the detection result is affected. At this time, no credible identification is made for the iodine content detection value of this salt sample, and it may be marked as a result to be verified or a result with low credibility, indicating that re-detection or other correction measures may be required.

[0039] Through the credibility judgment method based on error analysis, it is possible to effectively identify and screen out detection results with high reliability under various detection conditions, avoiding the influence of detection errors caused by the deviation of associated influencing indicators on the final result. This method is different from the traditional way of directly statistically analyzing multiple detection results. Instead, it considers the actual influence of various disturbance factors in the detection process and can more accurately evaluate the reliability of the detection result. Through the credible identification process, an effective screening and grading of the salt iodine content detection results are achieved, providing support for obtaining more accurate and reliable detection results, and effectively improving the accuracy and reliability of salt iodine content detection.

[0040] S6. Statistically analyze the set of iodine content detection values of salt samples with credible identification, and obtain the salt iodine content confidence value to be sent to the user terminal.

[0041] Specifically, first, salt samples marked with credible identifiers are screened out from all iodine content detection samples to form a set of iodine content detection values of salt samples. The iodine content detection values of the salt samples in the set of iodine content detection values of salt samples are suitable as the calculation basis for the final result because they meet the error control requirements and have high reliability. Through this screening, detection results with large errors caused by excessive disturbances are effectively excluded, improving the accuracy of the final result. Calculate the mean value of the iodine content detection values of the salt samples in the screened set of iodine content detection values of salt samples, and use the result of the mean value calculation as the confidence value of salt iodine content. Then, send the confidence value of salt iodine content to the user side through the communication network. The user side can be a data management system of a detection laboratory, a quality control system of a salt production enterprise, a supervision platform of a regulatory department, etc. Through the received confidence value of salt iodine content, users can obtain highly reliable detection results, providing a basis for subsequent quality evaluation, production adjustment, or regulatory decision-making.

[0042] Through statistical analysis and result feedback, the whole process of confidence sorting of salt iodine content detection data is completed, realizing the transformation from raw detection data to highly reliable final results. Compared with the traditional method of directly statistically analyzing multiple detection results, the present invention takes into account various influencing factors in the detection process. Through analysis and screening, the accuracy and reliability of the salt iodine content detection results are effectively improved, providing support for the supervision and management of salt iodization.

[0043] Furthermore, according to the salt iodine content detection process type, perform a correlation analysis on the preset set of influencing indicators to obtain an associated set of influencing indicators, including:

[0044] S11. Send the salt iodine content detection process type and the preset set of influencing indicators to the first distributed node to perform a correlation analysis and obtain a first associated set of influencing indicators;

[0045] S12. Until the salt iodine content detection process type and the preset set of influencing indicators are sent to the Nth distributed node to perform a correlation analysis and obtain the Nth associated set of influencing indicators;

[0046] S13. Perform trigger frequency sorting based on the first associated set of influencing indicators until the Nth associated set of influencing indicators to obtain the associated set of influencing indicators.

[0047] In a feasible implementation manner, a correlation analysis method based on distributed computing is proposed, which improves the accuracy and reliability of the correlation analysis through multi-node collaborative analysis.

[0048] First, send the salt iodine content detection process type and the preset impact index set to the first distributed node to perform correlation analysis and obtain the first associated impact index set. A distributed node is a processing unit dispersed in different geographical locations or different computing resources, and each distributed node has independent computing capabilities and data resources; the first distributed node is any one of the N distributed nodes. After receiving the salt iodine content detection process type and the preset impact index set, the first distributed node performs correlation analysis based on the historical detection data and analysis model stored locally, evaluates the correlation degree between each preset impact index and the detection result, screens out the indexes with significant correlation, and forms the first associated impact index set.

[0049] Similarly, send the salt iodine content detection process type and the preset impact index set to the second distributed node until the Nth distributed node in a similar manner. Each node independently performs correlation analysis to obtain the second associated impact index set until the Nth associated impact index set respectively. Among them, N represents the total number of nodes participating in the calculation in the distributed system, and the specific value can be configured according to the data scale and calculation requirements. Through parallel computing of multiple distributed nodes, the analysis efficiency is effectively improved. At the same time, by using the data resources and analysis perspectives of different nodes, the comprehensiveness and representativeness of the analysis results are enhanced. After that, perform trigger frequency sorting based on the first associated impact index set until the Nth associated impact index set to obtain the associated impact index set. Trigger frequency sorting means counting the occurrence frequency of each preset impact index in the N associated impact index sets, and screening out the indexes with higher occurrence frequencies as the final associated impact indexes. This analysis method based on multi-node collaboration and trigger frequency makes full use of the data and computing advantages of the distributed system and can more comprehensively and accurately identify the indexes that have a significant impact on a specific salt iodine content detection process.

[0050] Through distributed correlation analysis, the accuracy and reliability of identifying associated impact indexes are effectively improved, laying a solid foundation for subsequent confidence sorting of salt iodine content detection data. It is applicable to the analysis and processing of large-scale detection data, can fully explore the regular information contained in historical data, and improve the accuracy and stability of salt iodine content detection.

[0051] Further, performing trigger frequency sorting based on the first associated impact index set until the Nth associated impact index set to obtain the associated impact index set includes:

[0052] S131. Perform the same-attribute trigger frequency statistics on the first associated impact index set until the Nth associated impact index set to obtain the first index trigger frequency until the Qth index trigger frequency;

[0053] S132. Calculate the ratio of the first index trigger frequency to N, and set it as the first index correlation degree;

[0054] S133. Until the ratio of the triggering frequency of the Qth indicator to N is calculated, which is set as the Qth indicator correlation degree.

[0055] S134. Extract the indicators whose first indicator correlation degree to the Qth indicator correlation degree is greater than or equal to the correlation degree threshold, and add them to the associated influence indicator set.

[0056] In a preferred embodiment, first, the same-attribute triggering frequency statistics are performed on the first associated influence indicator set to the Nth associated influence indicator set to obtain the first indicator triggering frequency to the Qth indicator triggering frequency. The so-called same-attribute triggering frequency statistics refer to, for each indicator in the preset influence indicator set, counting the number of times it appears in the associated influence indicator sets generated by N distributed nodes. Here, Q represents the total number of indicators in the preset influence indicator set, and each indicator has a corresponding triggering frequency, which reflects the degree of consistency of different distributed nodes' judgments on the relevance of this indicator. For example, assume that a preset influence indicator "light source stability" appears m times in N associated influence indicator sets, then its triggering frequency is m; assume that another preset influence indicator "sample uniformity" appears n times in N associated influence indicator sets, then its triggering frequency is n. Through this kind of statistics, the first indicator triggering frequency to the Qth indicator triggering frequency is obtained, comprehensively reflecting the selection situation of each indicator in the multi-node analysis. Then, calculate the ratio of the first indicator triggering frequency to N, which is set as the first indicator correlation degree. The correlation degree is a value ranging from 0 to 1, representing the probability or credibility that this indicator is judged as a relevant indicator. For example, if the triggering frequency of the first indicator is m and the total number of distributed nodes is N, then the correlation degree of the first indicator is m / N. The higher the correlation degree, the more distributed nodes consider this indicator to have a significant correlation with the detection result, and the higher the credibility of this judgment.

[0057] Subsequently, in a similar manner, calculate the ratios of the triggering frequencies of the second indicator to the Qth indicator to N in sequence, which are respectively set as the second indicator correlation degree to the Qth indicator correlation degree. Through this calculation, the correlation degrees of all indicators in the preset influence indicator set are obtained, providing a quantitative basis for subsequent screening. After that, extract the indicators whose first indicator correlation degree to the Qth indicator correlation degree is greater than or equal to the correlation degree threshold, and add them to the associated influence indicator set. The correlation degree threshold is a preset judgment criterion, indicating the minimum correlation degree required to accept an indicator as an associated influence indicator. The setting of the correlation degree threshold needs to balance the analysis accuracy and the screening strictness, and can be adjusted according to the actual application requirements and system performance. For example, if the correlation degree threshold is set to 0.7, only the indicators whose triggering frequencies exceed 70% of the total number of nodes will be included in the final associated influence indicator set.

[0058] Through a quantitative screening method based on trigger frequency and correlation degree, it is possible to extract correlation impact indicators with highly consistent recognition from the multi-node analysis results, effectively avoiding the biases and limitations that may be brought about by single-node analysis. This method gives full play to the advantages of distributed computing, can more reliably identify the factors that have a significant impact on the salt iodine content detection results, and provides a basis for subsequent confidence sorting of detection data.

[0059] Further, send the salt iodine content detection process type and the preset impact indicator set to the first distributed node to perform correlation analysis and obtain a first set of correlation impact indicators, including:

[0060] S111. Send the salt iodine content detection process type and the preset impact indicator set to the first distributed node, and collect the historical data of salt iodine content detection. Among them, the historical data of salt iodine content detection includes salt production record parameters, preset impact indicator record characteristic values, and salt iodine content detection record values;

[0061] S112. Perform cluster analysis on the historical data of salt iodine content detection according to the salt production record parameters to obtain multiple clusters of historical data of salt iodine content detection;

[0062] S113. Based on the preset impact indicator record characteristic values and the salt iodine content detection record values, traverse the multiple clusters of historical data of salt iodine content detection to perform grey correlation degree analysis, perform correlation analysis, obtain multiple initial sets of correlation impact indicators, take the union, and add them to the first set of correlation impact indicators.

[0063] In a preferred embodiment, first, send the salt iodine content detection process type and the preset impact indicator set to the first distributed node to collect the historical data of salt iodine content detection. Among them, the historical data of salt iodine content detection includes salt production record parameters, preset impact indicator record characteristic values, and salt iodine content detection record values. The salt production record parameters include information such as production batch, production date, raw material source, and process parameters, which are used to identify the sample characteristics under different production conditions; the preset impact indicator record characteristic values refer to the actual monitored values of each preset impact indicator during the historical detection, such as the specific values of factors such as light source stability, temperature, and humidity; the salt iodine content detection record values are the salt iodine content measurement results obtained in the historical detection. The first distributed node obtains relevant historical detection data from its local database or from the central database through a data interface according to the received salt iodine content detection process type and preset impact indicator set, providing data support for subsequent analysis.

[0064] Then, based on the salt production record parameters, cluster analysis is performed on the historical data of salt iodine content detection to obtain multiple clusters of historical data of salt iodine content detection. Cluster analysis is an unsupervised learning method aimed at classifying data samples with similar characteristics into one category. Based on the key characteristics in the salt production record parameters, such as raw material source, production process, production equipment, etc., algorithms such as K-means, hierarchical clustering, or density clustering are used to divide the historical data of salt iodine content detection into multiple clusters. The samples within each cluster have high similarity in salt production conditions, and this classification method helps to identify the key factors affecting the detection results under specific production conditions, improving the pertinence and accuracy of correlation analysis.

[0065] Subsequently, based on the recorded characteristic values of the preset influencing indicators and the recorded values of the detected salt iodine content, traverse the multiple clusters of historical data of salt iodine content detection for grey relational analysis, perform correlation analysis, obtain multiple initial associated influencing indicator sets, take the union, and add it to the first associated influencing indicator set. Grey relational analysis is an effective method for dealing with uncertainty problems, especially suitable for situations with a small sample size and incomplete information. Specifically, for each cluster of historical data of salt iodine content detection, extract the recorded characteristic values of the preset influencing indicators as the comparison sequence, and the recorded values of the detected salt iodine content as the reference sequence, calculate the grey relational degree between each preset influencing indicator and the detection result, and screen out the indicators with a relational degree higher than the threshold to form the initial associated influencing indicator set. After traversing all data clusters, multiple initial associated influencing indicator sets are obtained, and then the union of these sets is taken to form the first associated influencing indicator set of the first distributed node.

[0066] Through this method based on data clustering and grey relational analysis, the regular information contained in the historical detection data can be fully mined, and the factors that have a significant impact on the salt iodine content detection results under different production conditions can be identified. This method comprehensively considers the diversity of production conditions and the complexity of influencing factors, can more comprehensively and accurately determine the associated influencing indicators, and provides a basis for the subsequent confidence sorting of salt iodine content detection data. By adopting this historical data analysis method, the accurate identification of associated influencing indicators is realized on the distributed computing architecture, improving the accuracy and adaptability of the confidence sorting of salt iodine content detection data, and providing technical guarantee for obtaining high-quality detection results.

[0067] Furthermore, based on the recorded characteristic values of the preset influencing indicators and the recorded values of the detected salt iodine content, traverse the multiple clusters of historical data of salt iodine content detection for grey relational analysis, perform correlation analysis, and obtain multiple initial associated influencing indicator sets, including:

[0068] S1131. Extract the first cluster of historical data of salt iodine content detection from the multiple clusters of historical data of salt iodine content detection;

[0069] S1132. Extract the first cluster of preset impact index record eigenvalue sets from the first cluster of salt iodine content detection historical data. After dimensionless processing, set it as the comparison sequence;

[0070] S1133. Extract the first cluster of salt iodine content detection record value sets from the first cluster of salt iodine content detection historical data. After dimensionless processing, set it as the reference sequence;

[0071] S1134. According to the comparison sequence and the reference sequence, perform grey relational analysis to obtain the first cluster of relational degree sets;

[0072] S1135. Extract the preset impact indexes in the first cluster of relational degree sets that are greater than or equal to the relational degree threshold, and add them to the first initial associated impact index set;

[0073] S1136. Add the first initial associated impact index set to the multiple initial associated impact index sets.

[0074] In a preferred embodiment, first, extract the first cluster of salt iodine content detection historical data from multiple clusters of salt iodine content detection historical data. Herein, the first cluster refers to one of the multiple data clusters formed after the clustering analysis in step S112. Each cluster of data represents a set of detection records under specific salt production conditions and has similar production backgrounds and process characteristics. First, process the first cluster of salt iodine content detection historical data, and then sequentially process other clusters of data to ensure a comprehensive analysis of the influencing factors under different production conditions.

[0075] Then, extract the first cluster of preset impact index record eigenvalue sets from the historical data of salt iodine content detection. After dimensionless processing, it is set as the comparison sequence. The first cluster of preset impact index record eigenvalue sets contains the actual values of each preset impact index recorded during the historical detection, such as the specific measurement values of factors like light source stability, temperature, humidity, etc. Since there may be significant differences in the dimensions and orders of magnitude of different indicators, dimensionless processing is performed on these eigenvalues, such as methods like initialization, mean normalization, or interval normalization, to eliminate the influence of dimensional differences on the analysis results. The data after dimensionless processing is used as the comparison sequence for grey relational analysis, which is used to compare with the reference sequence to evaluate its correlation. At the same time, extract the first cluster of salt iodine content detection record value sets from the historical data of salt iodine content detection. After dimensionless processing, it is set as the reference sequence. The first cluster of salt iodine content detection record value sets contains the salt iodine content measurement results obtained during the historical detection. Similarly, dimensionless processing is performed on these detected values to ensure comparability with the comparison sequence. The processed data is used as the reference sequence for grey relational analysis and is the reference standard for evaluating the correlation of each preset impact index. Subsequently, based on the comparison sequence and the reference sequence, perform grey relational analysis to obtain the first cluster of correlation degree sets. Specifically, first calculate the difference sequence between the reference sequence and each comparison sequence, then calculate the correlation coefficient at each moment based on the minimum difference, maximum difference, and resolution coefficient, and finally obtain the grey relational degree by taking the average value. Through this calculation, the similarity and correlation degree between each preset impact index and the salt iodine content detection result are quantified, forming the first cluster of correlation degree sets containing the correlation degrees of all preset impact indexes.

[0076] After that, extract the preset impact indexes in the first cluster of correlation degree sets that are greater than or equal to the correlation degree threshold and add them to the first initial associated impact index set. The correlation degree threshold is a preset judgment criterion, which represents the minimum correlation degree required to accept an index as an associated impact index. Screen out the preset impact indexes with a correlation degree higher than the correlation degree threshold. These indexes have a significant correlation with the salt iodine content detection result and are included in the first initial associated impact index set. Subsequently, add the first initial associated impact index set to multiple initial associated impact index sets, thereby saving the analysis results of the first cluster of salt iodine content detection historical data to the overall result set, preparing for the subsequent integration of the analysis results of each cluster. In a similar manner, process the second cluster until the last cluster of data in turn, obtain multiple initial associated impact index sets, and finally take their union to form the associated impact index set of the distributed node.

[0077] Through the method based on grey relational analysis, it is possible to accurately evaluate the correlation between each preset influence index and the detection result of iodine content in salt under the conditions of limited data and incomplete information, provide a basis for determining the associated influence index, and provide technical support for improving the reliability and accuracy of detection data. By using the grey relational analysis method, the accurate identification of the associated influence index is realized. On the basis of fully considering different production conditions, the influence degree of each preset influence index is comprehensively evaluated, providing a reliable technical means for the confidence sorting of the detection data of iodine content in salt.

[0078] Further, statistically count the set of iodine content detection samples that satisfy the time-series information of the deviation vector of the associated influence index, including:

[0079] S41. Based on the time-series information of the deviation vector of the associated influence index, construct a deviation state similarity evaluation function:

[0080] ,

[0081] ,

[0082] Among them, represents the deviation vector similarity of the i-th attribute associated influence index, represents the total number of monitoring times of the i-th attribute associated influence index, t represents the t-th moment of the monitoring time, represents the time-series information of the deviation vector of the i-th attribute associated influence index, represents the time-series information of the deviation vector of the i-th attribute associated influence index of the sample, represents the deviation vector of the i-th attribute associated influence index at the t-th moment, represents the deviation vector of the i-th attribute associated influence index of the sample at the t-th moment, represents the time-series information of the deviation vectors of all attribute associated influence indexes, represents the time-series information of the deviation vectors of all attribute associated influence indexes of the sample, M represents the total number of attributes, , represents a small constant;

[0083] S42. According to the deviation state similarity evaluation function, evaluate the deviation state similarity between the time-series information of the deviation vector of the associated influence index of the candidate salt sample and the time-series information of the deviation vector of the associated influence index;

[0084] S43. When the deviation state similarity is greater than or equal to the similarity threshold, add the candidate salt sample to the set of iodine content detection samples.

[0085] Specifically, a sample screening method based on the evaluation of the similarity of deviation states is proposed. By quantitatively evaluating the similarity of deviation states between samples, a set of iodine content detection samples with similar detection conditions is screened out, providing a reliable data basis for subsequent error analysis.

[0086] First, based on the time-series information of the deviation vectors of the associated influence indicators, a deviation state similarity evaluation function is constructed as:

[0087] , .

[0088] This evaluation function adopts the design idea of the structural similarity index and can effectively quantify the similarity degree of different samples in the deviation state of the associated influence indicators. The function is divided into two levels: first, calculate the deviation vector similarity of the associated influence indicators of each attribute, and then determine the overall similarity based on the similarity of each attribute.

[0089] At the attribute level, an improved cosine similarity formula is used to calculate the deviation vector similarity of the associated influence indicators of the i-th attribute . This formula takes into account the dynamic characteristics of the time-series information and quantifies the direction consistency and amplitude similarity of the two time-series information through the ratio of the product and the sum of squares of the deviation vectors at each moment. represents the time-series information of the deviation vector of the associated influence indicator of the i-th attribute of the current detection sample, which is a vector sequence that changes with time; represents the time-series information of the deviation vector of the associated influence indicator of the i-th attribute of the candidate sample in the historical database, which is also a time-series vector sequence. These two sequences record the change of the deviation state of different samples on the same associated influence indicator over time. represents the total number of monitoring moments of the associated influence indicator of the i-th attribute, that is, the number of time points at which the indicator is monitored during the entire detection process; t represents the specific moment during the monitoring process, and its value range is from 1 to T, representing different points in the time series. For each moment t, represents the deviation vector of the associated influence indicator of the i-th attribute of the current detection sample at the t-th moment, that is, the degree of deviation of the monitored value at this moment from the constraint interval; similarly, represents the deviation vector of the associated influence indicator of the i-th attribute of the candidate sample at the t-th moment. These deviation vectors are the basic units for calculating the similarity and reflect the deviation state of the indicator at each moment. is a small constant used to avoid the denominator being zero and improve the stability of the algorithm. By averaging the similarities at each moment, the comprehensive similarity degree of the entire time series is obtained.

[0090] At the overall level, the minimum value of the similarities of each attribute is taken as the overall similarity . Among them, The deviation vector time series information characterizing the associated influence indicators of all attributes of the current detection sample is a set containing all attribute deviation information; similarly, The deviation vector time series information characterizing the associated influence indicators of all attributes of the sample to be selected. M represents the total number of attributes, that is, the number of associated influence indicators. The value with the lowest similarity is found among all M attributes as the overall evaluation result. The relational expression indicates that the deviation vectors at each moment are components of the corresponding time series information, and the time series information of each attribute is in turn a component of the overall deviation information, clearly describing the hierarchical relationship between the data.

[0091] Subsequently, according to the deviation state similarity evaluation function, the deviation state similarity between the deviation vector time series information of the associated influence indicators of the sample to be selected for salt and the deviation vector time series information of the associated influence indicators is evaluated. Specifically, the deviation vector time series information of the associated influence indicators of the current detection sample is used as a reference and compared with the deviation vector time series information of the associated influence indicators of the sample to be selected for salt to calculate the deviation state similarity between the two. This comparison not only considers the direction and magnitude of the deviation but also the dynamic change characteristics in the time series, enabling a comprehensive evaluation of the comparability of the detection conditions between samples. When the deviation state similarity is greater than or equal to the similarity threshold, the sample to be selected for salt is added to the set of iodine content detection samples. The similarity threshold is a preset judgment criterion, indicating the minimum similarity required to accept a sample as a reference sample. Through this screening, a sample set with high similarity in the deviation state of the associated influence indicators is formed. These samples are comparable to the current detection sample in terms of being affected by interfering factors and are suitable as a reference basis for evaluating systematic errors.

[0092] Through the sample screening method based on the deviation state similarity evaluation, samples similar to the current detection conditions can be accurately screened out from historical data, providing effective data support for subsequent error analysis. This method fully considers the time series characteristics and multi-attribute characteristics of the deviation of the associated influence indicators, can more comprehensively and accurately evaluate the similarity between samples, and improves the accuracy and reliability of the confidence sorting of salt iodine content detection data. This deviation state similarity evaluation method realizes the accurate screening of detection samples, ensures that the data basis for error analysis has high comparability and representativeness, provides a guarantee for obtaining accurate iodine content detection error median values, and further improves the accuracy and reliability of salt iodine content detection.

[0093] Furthermore, the iodine content detection value set of the salt samples with credible identifiers is statistically analyzed, and the iodine content confidence value of the salt is obtained and sent to the user terminal, including:

[0094] S61. Perform box plot analysis on the iodine content detection value set of the salt samples to obtain the box body iodine content detection value set of the salt samples;

[0095] S62. Conduct a central tendency assessment on the set of iodine content test values of the salt samples in the box to obtain the distribution interval of the salt iodine content, which is set as the confidence value of the salt iodine content and sent to the user terminal.

[0096] In a preferred embodiment, first, perform box plot analysis on the set of iodine content test values of the salt samples to obtain the set of iodine content test values of the salt samples in the box. Specifically, first calculate the quartiles of the set of iodine content test values of the salt samples, including the first quartile Q1 (25% quantile), the second quartile Q2 (median, 50% quantile), and the third quartile Q3 (75% quantile), and determine the interquartile range IQR (i.e., Q3 - Q1). Based on the quartiles and the interquartile range, define the upper and lower boundaries: the upper boundary is Q3 + 1.5 * IQR, and the lower boundary is Q1 - 1.5 * IQR. The samples with iodine content test values of the salt samples exceeding these boundaries are regarded as potential outliers or extreme points and excluded from subsequent analysis. The test values between the upper and lower boundaries constitute the set of iodine content test values of the salt samples in the box, which is suitable for subsequent central tendency analysis. Through box plot analysis, extreme values that may be affected by unknown factors or measurement errors can be effectively identified and excluded, improving the accuracy and reliability of subsequent statistical analysis.

[0097] Subsequently, conduct a central tendency assessment on the set of iodine content test values of the salt samples in the box to obtain the distribution interval of the salt iodine content, which is set as the confidence value of the salt iodine content and sent to the user terminal. The central tendency assessment aims to extract the statistical quantity that can represent the overall characteristics from the filtered set of iodine content test values of the salt samples in the box, providing a basis for determining the final test result. First, calculate the basic statistical quantities of the set of iodine content test values of the salt samples in the box, such as the mean, median, standard deviation, or variance, etc., to comprehensively grasp the distribution characteristics of the data. On this basis, select a statistical method based on the distribution type of the data to determine the distribution interval of the salt iodine content. If the data approximately follows a normal distribution, the distribution interval can be defined in the form of mean ± K times the standard deviation, where K is a coefficient determined according to the confidence level; if the data distribution does not conform to the normal assumption, a quantile-based method can be used, such as taking the median as the center point and extending to specific quantiles on both sides to form a non-parametric distribution interval. The determined distribution interval of the salt iodine content is used as the confidence value of the salt iodine content and sent to the user terminal through the communication network.

[0098] Through the above steps, outliers are effectively excluded from the set of detection values with trusted identifiers, representative central tendency features are extracted, and a confidence value for the iodine content in salt with high reliability is formed. This method based on box plot analysis and central tendency evaluation fully considers the distribution characteristics and possible uncertainties of the data, provides accurate and reliable detection results for users, and effectively supports the quality control and supervision management of the iodine content in salt. By adopting the above method, a scientific transformation from trusted detection results to final confidence values is achieved, the accuracy and reliability of the iodine content detection in salt are improved, and high-quality detection results are provided for users.

[0099] Embodiment 2, as Figure 2 shown, based on the same inventive concept as the confidence sorting method for the iodine content detection data of salt provided in Embodiment 1, the embodiment of the present invention further provides a confidence sorting system for the iodine content detection data of salt, including:

[0100] A correlation analysis module 11, configured to perform a correlation analysis on a preset set of influence indicators according to the type of the iodine content detection process of salt to obtain an associated set of influence indicators;

[0101] A constraint interval receiving module 12, configured to receive the associated influence indicator constraint intervals of the associated set of influence indicators;

[0102] A deviation vector analysis module 13, configured to collect the time series information of the associated influence indicator monitoring values when detecting a salt sample, and compare it with the associated influence indicator constraint intervals to obtain the time series information of the associated influence indicator deviation vectors;

[0103] An error mode statistics module 14, configured to statistically calculate the iodine content detection error mode value of the set of iodine content detection samples that satisfy the time series information of the associated influence indicator deviation vectors;

[0104] A detection value identification module 15, configured to perform a trusted identification on the iodine content detection value of the salt sample when the iodine content detection error mode value is less than an error threshold;

[0105] A confidence value statistics module 16, configured to statistically calculate the confidence value of the iodine content in salt for the set of iodine content detection values of the salt samples with trusted identifiers and send it to the user terminal.

[0106] Further, the correlation analysis module 11 includes the following execution steps:

[0107] Send the type of the iodine content detection process of salt and the preset set of influence indicators to the first distributed node to perform a correlation analysis to obtain a first associated set of influence indicators;

[0108] Until the salt iodine content detection process type and the preset influence index set are sent to the Nth distributed node, perform correlation analysis to obtain the Nth associated influence index set;

[0109] Based on the first associated influence index set until the Nth associated influence index set, perform trigger frequency sorting to obtain the associated influence index set.

[0110] Further, the correlation analysis module 11 further includes the following execution steps:

[0111] Perform the same-attribute trigger frequency statistics on the first associated influence index set until the Nth associated influence index set to obtain the first index trigger frequency until the Qth index trigger frequency;

[0112] Calculate the ratio of the first index trigger frequency to N, and set it as the first index correlation degree;

[0113] Until calculating the ratio of the Qth index trigger frequency to N, and set it as the Qth index correlation degree;

[0114] Extract the indexes in the first index correlation degree until the Qth index correlation degree that are greater than or equal to the correlation degree threshold, and add them to the associated influence index set.

[0115] Further, the correlation analysis module 11 further includes the following execution steps:

[0116] Send the salt iodine content detection process type and the preset influence index set to the first distributed node, and collect the historical data of salt iodine content detection. Among them, the historical data of salt iodine content detection includes salt production record parameters, preset influence index record characteristic values, and salt iodine content detection record values;

[0117] According to the salt production record parameters, perform cluster analysis on the historical data of salt iodine content detection to obtain multiple clusters of historical data of salt iodine content detection;

[0118] Based on the preset influence index record characteristic values and the salt iodine content detection record values, traverse the multiple clusters of historical data of salt iodine content detection to perform grey correlation degree analysis, perform correlation analysis, obtain multiple initial associated influence index sets, take the union, and add them to the first associated influence index set.

[0119] Further, the correlation analysis module 11 further includes the following execution steps:

[0120] Extract the first cluster of historical data of salt iodine content detection from the multiple clusters of historical data of salt iodine content detection;

[0121] Extract the first cluster of preset impact index record eigenvalue sets from the historical data of the iodine content detection of the first cluster of table salt. After dimensionless processing, set it as the comparison sequence;

[0122] Extract the first cluster of iodine content detection record value sets of the first cluster of table salt from the historical data of the iodine content detection of the first cluster of table salt. After dimensionless processing, set it as the reference sequence;

[0123] According to the comparison sequence and the reference sequence, perform grey relational analysis to obtain the first cluster of relational degree sets;

[0124] Extract the preset impact indexes in the first cluster of relational degree sets that are greater than or equal to the relational degree threshold, and add them to the first initial associated impact index set;

[0125] Add the first initial associated impact index set to the multiple initial associated impact index sets.

[0126] Further, the error mode statistics module 14 includes the following execution steps:

[0127] Based on the deviation vector time series information of the associated impact index, construct a deviation state similarity evaluation function:

[0128] ,

[0129] ,

[0130] Among them, Characterize the deviation vector similarity of the i-th attribute associated impact index, Characterize the total number of monitoring times of the i-th attribute associated impact index, t represents the t-th moment of the monitoring time, Characterize the deviation vector time series information of the i-th attribute associated impact index, Characterize the deviation vector time series information of the i-th attribute associated impact index of the sample, Characterize the deviation vector of the i-th attribute associated impact index at the t-th moment, Characterize the deviation vector of the i-th attribute associated impact index of the sample at the t-th moment, Characterize the deviation vector time series information of all attribute associated impact indexes, Characterize the deviation vector time series information of all attribute associated impact indexes of the sample, M represents the total number of attributes, , Characterize a small constant;

[0131] According to the deviation state similarity evaluation function, evaluate the deviation state similarity between the deviation vector time series information of the associated impact index of the to-be-selected table salt sample and the deviation vector time series information of the associated impact index;

[0132] When the similarity of the deviation state is greater than or equal to the similarity threshold, add the candidate salt sample to the iodine content detection sample set.

[0133] Further, the confidence value statistics module 16 includes the following execution steps:

[0134] Perform box plot analysis on the iodine content detection value set of the salt sample to obtain the boxed iodine content detection value set of the salt sample;

[0135] Evaluate the central tendency of the boxed iodine content detection value set of the salt sample to obtain the distribution interval of the salt iodine content, and set it as the confidence value of the salt iodine content to be sent to the user terminal.

[0136] It should be noted that in the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0137] Those skilled in the art should understand that the embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program code.

[0138] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0139] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0140] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus, so that a series of operation steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions for implementing the process Figure 1 one process or a plurality of processes and / or blocks Figure 1 steps for the functions specified in one block or a plurality of blocks.

[0141] Although the preferred embodiments of the present invention have been described, additional changes and modifications can be made by those skilled in the art once they learn the basic inventive concept.

[0142] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalent technologies, the present invention is also intended to include these modifications and variations.

Claims

1. Confidence sorting method for detecting iodine content in table salt, characterized in that Including: Performing a correlation analysis on a preset set of influencing indicators according to the type of salt iodine content detection process to obtain an associated set of influencing indicators; Receiving the constraint intervals of the associated influencing indicators in the associated set of influencing indicators; When detecting a salt sample, collecting the time series information of the monitored values of the associated influencing indicators, comparing it with the constraint intervals of the associated influencing indicators, and obtaining the time series information of the deviation vectors of the associated influencing indicators; Statistically obtaining the mode value of the iodine content detection error of the set of iodine content detection samples that satisfy the time series information of the deviation vectors of the associated influencing indicators; When the mode value of the iodine content detection error is less than the error threshold, performing a credible identification on the detected value of the salt sample iodine content; Statistically analyzing the set of detected values of the salt sample iodine content with credible identification, obtaining the confidence value of the salt iodine content, and sending it to the user terminal.

2. The confidence sorting method for salt iodine content detection data according to claim 1, characterized in that, Performing a correlation analysis on a preset set of influencing indicators according to the type of salt iodine content detection process to obtain an associated set of influencing indicators, including: Sending the type of salt iodine content detection process and the preset set of influencing indicators to the first distributed node to perform a correlation analysis and obtain a first associated set of influencing indicators; Until the type of salt iodine content detection process and the preset set of influencing indicators are sent to the Nth distributed node to perform a correlation analysis and obtain the Nth associated set of influencing indicators; Performing trigger frequency sorting based on the first associated set of influencing indicators to the Nth associated set of influencing indicators to obtain the associated set of influencing indicators.

3. The confidence sorting method for salt iodine content detection data according to claim 2, characterized in that Performing trigger frequency sorting based on the first associated set of influencing indicators to the Nth associated set of influencing indicators to obtain the associated set of influencing indicators, including: Performing a same-attribute trigger frequency statistics on the first associated set of influencing indicators to the Nth associated set of influencing indicators to obtain the trigger frequency of the first indicator to the trigger frequency of the Qth indicator; Calculating the ratio of the trigger frequency of the first indicator to N, and setting it as the first indicator correlation degree; Until calculating the ratio of the trigger frequency of the Qth indicator to N, and setting it as the Qth indicator correlation degree; Extracting the indicators in the first indicator correlation degree to the Qth indicator correlation degree that are greater than or equal to the correlation degree threshold, and adding them to the associated set of influencing indicators.

4. The confidence sorting method for salt iodine content detection data according to claim 2, wherein Sending the type of salt iodine content detection process and the preset set of influencing indicators to the first distributed node to perform a correlation analysis and obtain a first associated set of influencing indicators, including: Sending the type of salt iodine content detection process and the preset set of influencing indicators to the first distributed node, and collecting the historical data of salt iodine content detection, where the historical data of salt iodine content detection includes salt production record parameters, preset influencing indicator record characteristic values, and detected record values of salt iodine content; Performing a cluster analysis on the historical data of salt iodine content detection according to the salt production record parameters to obtain multiple clusters of historical data of salt iodine content detection; Based on the preset influencing indicator record characteristic values and the detected record values of salt iodine content, traversing the multiple clusters of historical data of salt iodine content detection to perform a grey correlation degree analysis, performing a correlation analysis, obtaining multiple initial associated sets of influencing indicators, taking the union, and adding them to the first associated set of influencing indicators.

5. The confidence sorting method for salt iodine content detection data according to claim 4, characterized in that Based on the preset influence index record feature values and the detected record values of the iodine content in salt, traverse the multi-cluster historical data of iodine content detection in salt for grey relational analysis, perform correlation analysis, and obtain multiple initial associated influence index sets, including: Extract the first cluster of historical data of iodine content detection in salt from the multi-cluster historical data of iodine content detection in salt; Extract the set of feature values of the first cluster of preset influence index records from the first cluster of historical data of iodine content detection in salt. After dimensionless processing, it is set as the comparison sequence; Extract the set of detected record values of the iodine content in the first cluster of salt from the first cluster of historical data of iodine content detection in salt. After dimensionless processing, it is set as the reference sequence; Perform grey relational analysis according to the comparison sequence and the reference sequence to obtain the first cluster of correlation degree sets; Extract the preset influence indexes in the first cluster of correlation degree sets that are greater than or equal to the correlation degree threshold, and add them to the first initial associated influence index set; Add the first initial associated influence index set to the multiple initial associated influence index sets.

6. The confidence sorting method for salt iodine content detection data according to claim 1, characterized in that Statistically analyze the set of iodine content detection samples that satisfy the time series information of the deviation vector of the associated influence index, including: Based on the time series information of the deviation vector of the associated influence index, construct a deviation state similarity evaluation function: ， ， Among them, Characterize the deviation vector similarity of the i-th attribute correlation impact index, Characterize the total number of monitoring times of the i-th attribute correlation impact index, t represents the t-th moment of the monitoring time, Characterize the deviation vector time series information of the i-th attribute correlation impact index, Characterize the deviation vector time series information of the i-th attribute correlation impact index of the sample, Characterize the deviation vector of the i-th attribute correlation impact index at the t-th moment, Characterize the deviation vector of the i-th attribute correlation impact index of the sample at the t-th moment, Characterize the deviation vector time series information of all attribute correlation impact indexes, Characterize the deviation vector time series information of all attribute correlation impact indexes of the sample, M represents the total number of attributes, , Characterize a small constant; According to the deviation state similarity evaluation function, evaluate the deviation state similarity between the time series information of the deviation vector of the associated influence index of the candidate salt sample and the time series information of the deviation vector of the associated influence index; When the deviation state similarity is greater than or equal to the similarity threshold, add the candidate salt sample to the set of iodine content detection samples.

7. The confidence sorting method for salt iodine content detection data according to claim 1, characterized in that Statistically analyze the set of detected values of the iodine content in salt samples with a credible label, and obtain the confidence value of the iodine content in salt and send it to the user terminal, including: Perform box plot analysis on the set of detected values of the iodine content in the salt samples to obtain the set of detected values of the iodine content in the boxed salt samples; Perform a central tendency evaluation on the set of detected values of the iodine content in the boxed salt samples to obtain the distribution interval of the iodine content in salt, which is set as the confidence value of the iodine content in salt and sent to the user terminal.

8. Confidence sorting system for detecting iodine content in table salt, characterized in that, For implementing the method according to any one of claims 1 to 7, the system includes: A correlation analysis module, configured to perform correlation analysis on a preset influence index set according to the type of iodine content detection process in salt to obtain an associated influence index set; A constraint interval receiving module, configured to receive the associated influence index constraint interval of the associated influence index set; A deviation vector analysis module, configured to collect the time series information of the monitored values of the associated influence index when detecting a salt sample, compare it with the associated influence index constraint interval, and obtain the time series information of the deviation vector of the associated influence index; An error mode statistical module, configured to statistically analyze the mode value of the iodine content detection error of the set of iodine content detection samples that satisfy the time series information of the deviation vector of the associated influence index; A detected value identification module, configured to perform a credible identification on the detected value of the iodine content in the salt sample when the mode value of the iodine content detection error is less than the error threshold; A confidence value statistical module, configured to statistically analyze the set of detected values of the iodine content in salt samples with a credible label, and obtain the confidence value of the iodine content in salt and send it to the user terminal.

Images