Supplier quality evaluation method based on big data analysis

Abstract

The invention discloses a supplier quality evaluation method based on big data analysis, and relates to the technical field of supplier quality data evaluation management. The method comprises the following steps of collecting original heterogeneous data, performing preprocessing, generating a data quality report, performing quantitative evaluation on multi-source heterogeneity and real-time performance of supplier quality data, and triggering an optimization adjustment mode. According to the method, after original heterogeneous data is collected from a supplier historical transaction database and preprocessed, a data quality report is generated, the multi-source isomerism and the real-time performance of the supplier quality data are quantitatively evaluated, the data are compared with judgment values in the database and analyzed, a data integration optimization mode and a resource scheduling optimization mode are triggered, and the quality of the supplier is optimized. The real-time performance of data in supplier quality evaluation is improved, and the problem that the real-time performance of a supplier quality evaluation method is insufficient due to data multi-source heterogeneous in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of supplier quality data evaluation and management, and particularly to a quality evaluation method for suppliers based on big data analysis. Background Art

[0002] With the increasing complexity and globalization trend of supply chain management, the quality evaluation of suppliers has become a key link for enterprises to ensure the stable supply of raw materials, components or services and product consistency. Due to problems such as single data dimension, low evaluation efficiency, and insufficient dynamic adaptability in traditional evaluation methods, it is difficult to meet the comprehensive requirements of modern enterprises for supplier quality, delivery ability, cost control, and risk prediction. The supplier quality evaluation method based on big data analysis has emerged as the times require. By integrating multi-source heterogeneous data such as historical transaction records, production data, and market feedback, and combining algorithms such as machine learning and natural language processing, it can comprehensively depict the operation status of suppliers, quickly identify potential risks, so as to achieve more accurate and efficient evaluation, help enterprises optimize supply chain management, and enhance overall competitiveness.

[0003] Existing supplier quality evaluation methods classify and link suppliers by using techniques such as natural language fuzzy matching algorithms, clustering analysis, and discriminant analysis; through big data analysis, enterprises can quickly identify high-risk or low-quality suppliers and optimize the supply chain cooperation network; multi-source data is integrated with multi-dimensional data such as the basic information of suppliers, historical supply records, after-sales service quality, and market evaluations.

[0004] For example, a supplier quality evaluation method disclosed in the invention patent with the publication number of CN114049038A includes: establishing a weight factor value matrix according to the severity and impact degree of quality problems of supplied products, and calculating the supplied S-PPM value by using a formula; calculating the after-sales S-PPM of products according to a formula, setting a score conversion algorithm for the supplied S-PPM value to convert the supplied S-PPM value into a value under a 100-point system; setting a score conversion algorithm for the after-sales S-PPM value to convert the after-sales S-PPM value into a value under a 100-point system; setting deduction items and corresponding deduction scores according to audits and inspections, and obtaining the quality assurance ability score of the product by removing the deduction items under a 100-point system; setting a first weight number for the supplied S-PPM, a second weight number for the after-sales S-PPM, and a third weight number for the quality assurance ability score, and calculating the total score.

[0005] For example, the supplier quality assessment method disclosed in the invention patent with the publication number: CN119417314A includes: obtaining the initial quality evaluation labels of each supplier's data; counting the comprehensive data quality evaluation values of the uploaded information quality processing of each first-class supplier; determining the initial quality inspection constraint period of each first-class supplier and conducting quality tracking inspections; obtaining the quality tracking inspection data for analysis and processing to obtain the quality tracking monitoring results; dynamically adjusting the initial quality inspection constraint period based on the quality tracking monitoring results of each first-class supplier. By analyzing the data quality of the materials uploaded by the supplier, the present invention ensures the availability of the quality assessment materials, screens out the suppliers with unqualified initial quality evaluations, focuses the quality assessment resources on the suppliers with qualified initial quality evaluations, conducts customized quality tracking inspections on each first-class supplier, improves the efficiency and results of the quality tracking inspections, and timely and dynamically adjusts the quality inspection constraint period.

[0006] However, in the process of implementing the technical solution of the invention in the embodiments of the present application, it is found that the above technology has at least the following technical problems: When this method is applied to the enterprise cross-regional supply chain, there is a problem of insufficient real-time evaluation. When data is transmitted across regions, heterogeneous data processing requires a large amount of CPU / memory resources, which exacerbates the server load, causes backpressure and queue accumulation, resulting in data integration delays. At the same time, insufficient network bandwidth leads to a slow data synchronization speed, the data writing speed is lower than the reading speed, the system has backpressure, resulting in data transmission delays. And in high-concurrency scenarios, insufficient server processing power will also cause synchronization delays, and there is a problem of insufficient evaluation of data multi-source heterogeneity; in supply chain management, factors such as market demand, production conditions, and logistics conditions change at any time, resulting in rapid changes in data distribution. The data adjustment speed cannot keep up with the actual data change speed, resulting in the inability to accurately identify new abnormal patterns, and there is a problem of insufficient real-time performance of the supplier quality assessment method due to data multi-source heterogeneity. Summary of the Invention

[0007] The embodiments of the present application provide a quality assessment method for suppliers based on big data analysis, which solves the problem of insufficient real-time performance of the supplier quality assessment method due to data multi-source heterogeneity in the prior art and improves the real-time performance of data in supplier quality assessment.

[0008] An embodiment of the present application provides a quality evaluation method for suppliers based on big data analysis, including the following steps: collecting original heterogeneous data from the historical transaction database of suppliers, preprocessing the collected original heterogeneous data to obtain heterogeneous data; generating a data quality report based on the preprocessed heterogeneous data, and respectively quantitatively evaluating the multi-source heterogeneity of supplier quality data and the real-time performance of the supplier quality evaluation method through the data quality report to obtain a multi-source heterogeneity quantification value and a data real-time quantification value; comparing and analyzing the multi-source heterogeneity quantification value and the data real-time quantification value with the first determination value of the multi-source heterogeneity quantification value and the second determination value of the data real-time quantification value in the supplier quality evaluation database respectively to obtain an optimization adjustment mode, and the optimization adjustment mode includes a data integration optimization mode and a resource scheduling optimization mode; generating an optimization efficiency heat map through the Grafana visualization platform, inputting the heterogeneous data after data integration optimization and resource scheduling optimization and integrated into the supplier quality evaluation module, and the supplier quality evaluation module is used to evaluate the overall quality of the supplier, and the optimization efficiency heat map shows the improvement of efficiency after data integration optimization and resource scheduling optimization.

[0009] One or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages: 1. By obtaining the multi-source heterogeneity quantification value and the data real-time quantification value, the adaptive optimization of the supplier quality evaluation method is realized. Based on the data integration optimization mode and the resource scheduling optimization mode, it can automatically identify data integration delays or resource shortages, and achieve the optimal allocation of resources through edge cache acceleration and load balancing algorithms. Combining multiple indicators such as abnormal pattern recognition delay and cross-region round-trip time can avoid misjudgment of a single indicator.

[0010] 2. Through the weighted least connection number algorithm and the three-level degradation strategy, the degradation rate is reduced in high-concurrency scenarios, and the priority processing of core services such as abnormal pattern recognition is guaranteed. It supports the automatic parsing and conversion of diverse data formats such as XML, JSON, and CSV, and combines the Snappy compression algorithm to reduce the amount of data transmitted across regions and reduce network latency.

[0011] 3. Through the calculation of the multi-source heterogeneity deviation degree, the data real-time deviation degree and the comprehensive urgency degree, the health status of the system is intuitively displayed. When the urgency exceeds the threshold, a pulse warning is triggered and automated operation and maintenance operations such as expansion and degradation are linked. The improvement of efficiency before and after optimization is compared through the heat map to quantitatively verify the technical improvement effect. Description of the Drawings

[0012] Figure 1 It is a flowchart of a quality evaluation method for suppliers based on big data analysis provided by an embodiment of the present application; Figure 2The heat map for optimizing performance provided by the embodiments of the present application; Figure 3 The specific flowchart of a quality assessment method for suppliers based on big data analysis provided by the embodiments of the present application. Detailed implementation manners

[0013] The embodiments of the present application provide a quality assessment method for suppliers based on big data analysis, which solves the problem in the prior art that the quality assessment method for suppliers has insufficient real-time performance due to multi-source heterogeneous data. After collecting and preprocessing the original heterogeneous data from the supplier's historical transaction database, a data quality report is generated to quantitatively evaluate the multi-source heterogeneity and real-time performance of the supplier quality data, and compare and analyze it with the judgment value in the database to trigger the data integration optimization mode and the resource scheduling optimization mode, improving the real-time performance of the data in the supplier quality assessment.

[0014] The technical solutions in the embodiments of the present application are to solve the above problem that the quality assessment method for suppliers has insufficient real-time performance due to multi-source heterogeneous data, and the general idea is as follows: After collecting and preprocessing the original heterogeneous data from the supplier's historical transaction database, a data quality report is generated to quantitatively evaluate the multi-source heterogeneity and real-time performance of the supplier quality data, and compare and analyze it with the judgment value in the database to trigger the data integration optimization mode and the resource scheduling optimization mode, improving the real-time performance of the data in the supplier quality assessment.

[0015] To better understand the above technical solutions, the above technical solutions will be described in detail below in conjunction with the accompanying drawings of the specification and specific implementation manners.

[0016] As Figure 1As shown in the figure, it is a flowchart of a quality assessment method for suppliers based on big data analysis provided by an embodiment of the present application. The method includes the following steps: Collect original heterogeneous data from the historical transaction database of the supplier, preprocess the collected original heterogeneous data to obtain heterogeneous data; Based on the preprocessed heterogeneous data, generate a data quality report, and respectively conduct quantitative evaluations on the multi-source heterogeneity of supplier quality data and the real-time performance of the supplier quality assessment method through the data quality report to obtain a quantitative value of multi-source heterogeneity of data and a quantitative value of data real-time performance; Compare and analyze the quantitative value of multi-source heterogeneity of data and the quantitative value of data real-time performance with the first determination value of the quantitative value of multi-source heterogeneity of data and the second determination value of the quantitative value of data real-time performance in the supplier quality assessment database respectively to obtain an optimization and adjustment mode, and the optimization and adjustment mode includes a data integration optimization mode and a resource scheduling optimization mode; Generate an optimization efficiency heat map through the Grafana visualization platform, input the heterogeneous data after data integration optimization and resource scheduling optimization and integrated into the supplier quality assessment module. The supplier quality assessment module is used to evaluate the overall quality of the supplier, and the optimization efficiency heat map shows the improvement of efficiency after data integration optimization and resource scheduling optimization.

[0017] In this embodiment, collect original heterogeneous data from the historical transaction database of the supplier, the customer feedback system, the financial statement repository, and the market monitoring platform. The original heterogeneous data includes structured transaction records, unstructured text evaluations, tabular financial indicators, and real-time market dynamic data. The heterogeneous data after connection and integration can be supplier quality data. For example, the order quantity, order amount, etc., and according to the weights assigned to the corresponding data by the preset personnel, correct the heterogeneous data after connection and integration. The correction is the weight assignment process, and add the corrected heterogeneous data after connection and integration to obtain the final supplier quality assessment score. Classify the suppliers proportionally according to the supplier quality assessment score. For example, sort the supplier quality assessment scores in descending order and divide them into 4 levels according to the preset ratio, namely A, B, C, and D levels. The supplier quality assessment module is a part of the supply chain management system. The supplier quality assessment module is used to comprehensively evaluate the quality of the supplier. As a part of the supply chain management system, the supplier quality assessment method is closely integrated with modules such as supply chain management, quality management, data analysis, and reporting to jointly ensure the efficient operation of the enterprise supply chain and the stable improvement of product quality.

[0018] Further, the specific steps for obtaining heterogeneous data are as follows: Collect the original heterogeneous data from the historical transaction database of the supplier. The original heterogeneous data represents multi-dimensional dynamic feature data obtained from different supplier data sources for describing and reflecting the quality of the supplier. The original heterogeneous data comes from different data sources and has different formats and structures; perform standardization processing on the collected original heterogeneous data to eliminate the dimensional differences between the data, and obtain heterogeneous data.

[0019] Further, the specific steps for generating a data quality report are as follows: Connect and integrate the preprocessed heterogeneous data through a data quality management tool; Define data quality rules in the data quality management tool, and evaluate the connected and integrated heterogeneous data according to the data quality rules defined in the data quality management tool to generate a data quality report. The data quality report includes a data quality score; Analyze the data quality report to obtain the quantification value of data multi-source heterogeneity and the quantification value of data real-time performance.

[0020] In this embodiment, the data quality management tool is the IBM data quality tool, the data quality management tool provided by SAP, etc.

[0021] Further, the specific steps to obtain the data multi-source heterogeneity quantification value are as follows: Obtain the data multi-source heterogeneity quantification value threshold and the data multi-source heterogeneity quantification value importance score from the supplier quality assessment database; the data multi-source heterogeneity quantification value threshold includes the anomaly pattern recognition delay threshold, the cross-region round-trip time threshold, the metadata alignment time threshold, and the heterogeneous data coverage threshold; the data multi-source heterogeneity quantification value importance score includes the importance score of the anomaly pattern recognition delay, the importance score of the cross-region round-trip time, the importance score of the metadata alignment time, and the importance score of the heterogeneous data coverage; perform a ratio analysis of the anomaly pattern recognition delay and the anomaly pattern recognition delay threshold, average the results of the ratio analysis, and correct the results of the averaging process using the importance score of the anomaly pattern recognition delay, which is denoted as the first component of the data multi-source heterogeneity quantification value; perform a ratio analysis of the cross-region round-trip time and the cross-region round-trip time threshold, average the results of the ratio analysis, and correct the results of the averaging process using the importance score of the cross-region round-trip time, which is denoted as the second component of the data multi-source heterogeneity quantification value; perform a ratio analysis of the metadata alignment time and the metadata alignment time threshold, average the results of the ratio analysis, and correct the results of the averaging process using the importance score of the metadata alignment time, which is denoted as the third component of the data multi-source heterogeneity quantification value; perform a ratio analysis of the heterogeneous data coverage threshold and the heterogeneous data coverage, and correct the results of the ratio analysis using the importance score of the heterogeneous data coverage, which is denoted as the fourth component of the data multi-source heterogeneity quantification value; perform a coupling analysis on the first component of the data multi-source heterogeneity quantification value, the second component of the data multi-source heterogeneity quantification value, the third component of the data multi-source heterogeneity quantification value, and the fourth component of the data multi-source heterogeneity quantification value to obtain the data multi-source heterogeneity quantification value; the data multi-source heterogeneity quantification value represents a measure of the degree of multi-source and heterogeneity of the supplier data.

[0022] In this embodiment, the specific method for analyzing and obtaining the data multi-source heterogeneity quantification value is as follows: ; ; Number the preset supplier quality data multi-source heterogeneity detection points in sequence. Denote as the number of the supplier quality data multi-source heterogeneity detection point under the -th supplier quality data multi-source heterogeneity detection segment. Denote

[0023] as the total number of the supplier quality data multi-source heterogeneity detection point numbers. Represents the serial number of the multi-source heterogeneous detection segment of supplier quality data. , Represents the total number of serial numbers of the multi-source heterogeneous detection segments of supplier quality data.

[0024] Represents the quantified value of multi-source heterogeneity of data at the multi-source heterogeneous detection point of the supplier quality data.

[0025] Represents the abnormal pattern recognition delay under the multi-source heterogeneous detection segment of the supplier quality data, which refers to the time from the generation of supplier data to the system's recognition of abnormal patterns such as format mismatch, data missing, etc.

[0026] Represents the abnormal pattern recognition delay threshold, which is a preset abnormal pattern recognition delay threshold obtained from the supplier quality assessment database and can be the average value of the abnormal pattern recognition delays under the historical multi-source heterogeneous detection segments of supplier quality data in the supplier quality assessment database.

[0027] Represents the cross-regional round-trip time under the multi-source heterogeneous detection segment of the supplier quality data, which represents the time for supplier data to be transmitted from one region to another and then back.

[0028] Represents the cross-regional round-trip time threshold, which is a preset cross-regional round-trip time threshold obtained from the supplier quality assessment database and can be the average value of the cross-regional round-trip times under the historical multi-source heterogeneous detection segments of supplier quality data in the supplier quality assessment database.

[0029] Represents the metadata alignment time under the multi-source heterogeneous detection segment of the supplier quality data, which is the time required to standardize and align metadata from different supplier data sources, such as field names, data types, data formats, etc.

[0030] Represents the metadata alignment time threshold, which is a preset metadata alignment time threshold obtained from the supplier quality assessment database and can be the average value of the metadata alignment times under the historical multi-source heterogeneous detection segments of supplier quality data in the supplier quality assessment database.

[0031] Represents the The heterogeneous data coverage rate under multiple heterogeneous data sources of supplier quality data represents the ratio of the number of data sources of different types and formats that the system can actually process and integrate to the total number of supplier data sources of different types and formats that need to be processed. Count the total number of supplier data sources of different types and formats that need to be processed. For example, XML files, JSON files, CSV files, database tables, JSON data returned by API interfaces, etc. Each type or format counts as one data source. Count the number of supplier data sources of different types and formats that the system can successfully process and integrate.

[0032] Represents the heterogeneous data coverage rate threshold, which is a preset heterogeneous data coverage rate threshold obtained from the supplier quality assessment database. It can be the average value of the heterogeneous data coverage rate under the multi-source heterogeneous data detection points of historical supplier quality data in the supplier quality assessment database.

[0033] Is the preset importance score for abnormal pattern recognition latency obtained from the supplier quality assessment database.

[0034] Is the preset importance score for cross-region round-trip time obtained from the supplier quality assessment database.

[0035] Is the preset importance score for metadata alignment time obtained from the supplier quality assessment database.

[0036] Is the preset importance score for heterogeneous data coverage rate obtained from the supplier quality assessment database.

[0037] Through the preset importance score mapping table in the database, the rapid association between abnormal pattern recognition latency, cross-region round-trip time, metadata alignment time, and heterogeneous data coverage rate and their corresponding importance scores is realized. This mapping table defines a set of clear association rules, which can convert the specific values of each parameter into the corresponding importance scores.

[0038] The longer the cross-region round-trip time means the increase in supplier data transmission time, which leads to an increase in the latency of abnormal pattern recognition, and the longer the abnormal pattern recognition latency; the longer the metadata alignment time, the metadata used is not the latest, which affects abnormal pattern recognition, and the longer the abnormal pattern recognition latency; the lower the heterogeneous data coverage rate means that some data formats cannot be effectively parsed and recognized, increasing the complexity and latency of abnormal pattern recognition, and the longer the abnormal pattern recognition latency; the longer the cross-region round-trip time, the metadata needs to be synchronized between different systems across regions, and the longer the metadata alignment time.

[0039] There is a positive correlation between the anomaly pattern recognition delay and the quantification value of data multi-source heterogeneity. The longer the anomaly pattern recognition delay, the more likely it is to miss real-time data anomalies, and the larger the quantification value of data multi-source heterogeneity. There is a positive correlation between the cross-region round-trip time and the quantification value of data multi-source heterogeneity. The longer the cross-region round-trip time, the greater the data transmission delay of the supplier, and the larger the quantification value of data multi-source heterogeneity. There is a positive correlation between the metadata alignment time and the quantification value of data multi-source heterogeneity. The longer the metadata alignment time, the more likely it is to result in inconsistent supplier data versions, and the larger the quantification value of data multi-source heterogeneity. There is a negative correlation between the heterogeneous data coverage rate and the quantification value of data multi-source heterogeneity. The higher the heterogeneous data coverage rate, the more supplier data from different sources and formats the system can handle, and the smaller the quantification value of data multi-source heterogeneity.

[0040] Further, the specific steps to obtain the data real-time quantification value are as follows: Obtain the data real-time quantification value threshold and the data real-time quantification value importance score from the supplier quality assessment database; The data real-time quantification value threshold includes the high-concurrency degradation rate threshold, the node processing request response time threshold, the cross-region round-trip time threshold, and the metadata alignment time threshold; The data real-time quantification value importance score includes the importance score of the high-concurrency degradation rate, the importance score of the node processing request response time, the importance score of the cross-region round-trip time, and the importance score of the metadata alignment time; Conduct a ratio analysis of the high-concurrency degradation rate threshold and the high-concurrency degradation rate, and use the importance score of the high-concurrency degradation rate to correct the result of the ratio analysis, denoted as the first component of the data real-time quantification value; Conduct a ratio analysis of the node processing request response time threshold and the node processing request response time, average the result of the ratio analysis, and use the importance score of the node processing request response time to correct the averaged result, denoted as the second component of the data real-time quantification value; Conduct a ratio analysis of the cross-region round-trip time threshold and the cross-region round-trip time, average the result of the ratio analysis, and use the importance score of the cross-region round-trip time to correct the averaged result, denoted as the third component of the data real-time quantification value; Conduct a ratio analysis of the metadata alignment time threshold and the metadata alignment time, average the result of the ratio analysis, and use the importance score of the metadata alignment time to correct the averaged result, denoted as the fourth component of the data real-time quantification value; Conduct a coupling analysis on the coupling results of the first component, the second component, the third component, and the fourth component of the data real-time quantification value to obtain the data real-time quantification value; The data real-time quantification value represents the real-time performance of the supplier quality assessment method.

[0041] In this embodiment, the specific method for obtaining the data real-time quantification value through analysis is as follows: ; ; Number the preset data real-time detection points in sequence, indicating the number of the data real-time detection point under the th data real-time detection segment, , indicating the total number of the data real-time detection point numbers.

[0042] Divide the preset data real-time time into data real-time detection segments of the same time length, indicating the number of the data real-time detection segment, , indicating the total number of the data real-time detection segment numbers.

[0043] indicating the data real-time quantization value of the th data real-time detection point.

[0044] indicating the high-concurrency degradation rate under the th data real-time detection point, which represents the ratio of the number of degradation processing occurrences to the total number of requests in a high-concurrency scenario, and records degradation processing such as service rejection, function simplification, response time extension, etc.

[0045] indicating a constant term, which is a number set to avoid meaningless points in the data.

[0046] indicating the high-concurrency degradation rate threshold, which is a preset high-concurrency degradation rate threshold obtained from the supplier quality evaluation database, and can be the average value of the high-concurrency degradation rates under the historical data real-time detection points in the supplier quality evaluation database.

[0047] indicating the node processing request response time under the th data real-time detection segment, which represents the total time elapsed from when the client sends a request to when the node starts to process the request and then returns the processing result to the client.

[0048] indicating the node processing request response time threshold, which is a preset node processing request response time threshold obtained from the supplier quality evaluation database, and can be the average value of the node processing request response times under the historical data real-time detection segments in the supplier quality evaluation database.

[0049] indicating the cross-region round-trip time under the th data real-time detection segment.

[0050] Indicates the cross-region round-trip time threshold, which is a preset cross-region round-trip time threshold obtained from the supplier quality assessment database and can be the average cross-region round-trip time in the real-time detection section of historical data in the supplier quality assessment database.

[0051] Indicates the metadata alignment time under the nth data real-time detection section.

[0052] Indicates the metadata alignment time threshold, which is a preset metadata alignment time threshold obtained from the supplier quality assessment database and can be the average metadata alignment time in the real-time detection section of historical data in the supplier quality assessment database.

[0053] Is the preset high-concurrency degradation rate importance score obtained from the supplier quality assessment database.

[0054] Is the preset node processing request response time importance score obtained from the supplier quality assessment database.

[0055] Is the preset cross-region round-trip time importance score obtained from the supplier quality assessment database.

[0056] Is the preset metadata alignment time importance score obtained from the supplier quality assessment database.

[0057] Through the preset importance score mapping table in the database, the quick association between the high-concurrency degradation rate, node processing request response time, cross-region round-trip time, and metadata alignment time and their corresponding importance scores is realized. This mapping table defines a set of clear association rules that can convert the specific values of each parameter into the corresponding importance scores.

[0058] The longer the node processing request response time, the higher the system load or the more insufficient the resources, and the degradation strategy is started, such as rejecting some requests or providing simplified services, and the higher the high-concurrency degradation rate; the longer the cross-region round-trip time, the longer the overall request processing time, thus increasing the system pressure in the case of high concurrency and triggering the degradation strategy, and the higher the high-concurrency degradation rate; the longer the metadata alignment time, the higher the data consistency problem and the higher the high-concurrency degradation rate; the increase in the cross-region round-trip time will lead to an increase in the node processing request response time; the longer the metadata alignment time, the longer the node needs to wait for data synchronization to complete when processing requests, thus increasing the node processing request response time.

[0059] There is a negative correlation between the high-concurrency degradation rate and the quantitative value of data real-time performance. The higher the high-concurrency degradation rate, it means that the system cannot fully process all requests during peak hours, and it is necessary to reduce the service level or reject some requests, and the lower the quantitative value of data real-time performance; there is a negative correlation between the request response time of the node and the quantitative value of data real-time performance. The longer the request response time of the node, it indicates that the system processes requests more slowly, and the lower the quantitative value of data real-time performance; there is a negative correlation between the cross-region round-trip time and the quantitative value of data real-time performance. The longer the cross-region round-trip time, the greater the delay in data synchronization and request processing, and the lower the quantitative value of data real-time performance; there is a negative correlation between the metadata alignment time and the quantitative value of data real-time performance. The longer the metadata alignment time, it leads to an increase in the delay of data processing and request response, and the lower the quantitative value of data real-time performance.

[0060] Furthermore, the specific steps of the data integration optimization mode are as follows: obtain the first determination value of the quantitative value of data multi-source heterogeneity and the second determination value of the quantitative value of data real-time performance from the supplier quality assessment database; if the quantitative value of data multi-source heterogeneity is lower than or equal to the first determination value of the quantitative value of data multi-source heterogeneity, there is no need to trigger the data integration optimization mode; if the quantitative value of data multi-source heterogeneity is greater than the first determination value of the quantitative value of data multi-source heterogeneity, trigger the data integration optimization mode, and the triggered data integration optimization mode is used to improve the efficiency of data integration.

[0061] In this embodiment, the data integration optimization mode is used to improve the efficiency of data integration, process heterogeneous data formats from different suppliers (such as XML, JSON, CSV, etc.), and preset parsing rules. Based on regular expression and Schema mapping technology, it can convert supplier data in different formats into a unified internal format. For example, for supplier data in JSON format, JSON Schema can be used to define the data structure; for supplier data in XML format, XML Schema can be used for validation and parsing. A central template library is created through a distributed file system to store and manage all parsing templates. The template library should support operations such as adding, deleting, updating, and querying templates. Serverless functions (such as AWS Lambda) are introduced to test the real-time parsing ability of newly connected supplier data sources. Based on the test results, parsing templates are dynamically generated and stored in the template library. When a new supplier data source is received, its format is first identified (such as through file extensions, MIME types, etc.). According to the identified data format, the corresponding parsing template is searched from the template library, and the supplier data is parsed using the matched template and converted into a unified internal format. If an exception occurs during the parsing process (such as format mismatch, data missing, etc.), the exception information is recorded and corresponding processing is performed (such as notifying the administrator, discarding the data, etc.). When a new supplier data source is received, its metadata (such as field names, data types, description information, etc.) is extracted. Using a semantic matching engine, the metadata of the new supplier data source is matched with the metadata of the existing supplier data sources to automatically identify the field mapping relationship and generate a metadata mapping table for subsequent data integration. Edge computing nodes are deployed at the regional boundaries (such as locations geographically close to suppliers or data sources). Redis Cluster is deployed on each edge node to cache hot supplier data, and a configuration such as the LRU (Least Recently Used) policy is used. Key values with an access frequency exceeding the preset threshold are automatically cached to ensure the effective utilization of the cache space. Data synchronization between the edge nodes and the central node is achieved to ensure data consistency. In the archiving and storage of supplier data, a high compression ratio is required to reduce the storage space. Data compression is performed through the Snappy data compression algorithm. If the compression ratio is greater than the preset threshold, an alarm is triggered. During the data transmission process, non-real-time supplier data is compressed to reduce the amount of transmitted data. At the data receiving end, the compressed data is decompressed to restore the original data. Through dynamic resource scheduling, the utilization rate of computing resources is improved. An automatic scaling policy is configured. If the CPU utilization rate is greater than the preset threshold, automatic expansion is performed; when the CPU utilization rate is lower than the preset threshold, automatic scaling is performed to reduce the high concurrency degradation rate.

[0062] Further, the specific steps of the resource scheduling optimization mode are as follows: If the data real-time quantization value is greater than or equal to the second determination value of the data real-time quantization value, the resource scheduling optimization mode does not need to be triggered; if the data real-time quantization value is lower than the second determination value of the data real-time quantization value, the resource scheduling optimization mode is triggered, and the triggered resource scheduling optimization mode is used to improve resource utilization.

[0063] In this embodiment, the weighted least connection number algorithm is used for load balancing. Weights are assigned according to the processing capabilities of each node (such as CPU performance, memory size, etc.) to ensure that the supplier data load is more evenly distributed. Corresponding weights are configured for each node according to the processing capabilities of the nodes. The higher the weight, the stronger the processing ability of the node, and more supplier data connections can be allocated. Requests are dynamically allocated to different nodes according to the weights of the nodes and the current number of supplier data connections. Through real-time monitoring and adjustment, it is ensured that the load between nodes is less than the standard value to achieve load balancing. A resource reservation queue is built for high-priority tasks (such as supplier anomaly mode recognition) to ensure that high-priority tasks can still be processed in a timely manner when resources are scarce. Through mechanisms such as cgroup, the resource occupancy of non-critical tasks is restricted, and sufficient resources can be reserved for high-priority tasks to ensure that their performance is not affected. A queue scheduling strategy is configured to ensure that high-priority tasks can obtain processing resources first, and the scheduling strategy can be dynamically adjusted according to the urgency and importance of the tasks. A three-level degradation strategy is implemented, and two-level degradation thresholds of CPU utilization are preset, and the secondary degradation threshold is greater than the primary degradation threshold. If the CPU utilization is less than the primary degradation threshold, non-core functions (such as supplier quality trend visualization reports) are turned off to release resources for core services; if the CPU utilization is greater than the primary degradation threshold and less than the secondary degradation threshold, only one type of supplier data is processed to further reduce resource consumption; otherwise, it indicates a network interruption, and the edge node cache data is enabled to provide read-only services to ensure that some functions are still available. Each supplier area independently records the data change timestamp, and after the network is restored, the data is merged according to the latest timestamp to ensure data consistency and integrity.

[0064] Further, the specific steps for generating the optimization efficiency heat map are as follows: Generate the optimization efficiency heat map through the Grafana visualization platform, and set the two-dimensional coloring rule. The abscissa of the optimization efficiency heat map is the deviation degree of the data multi-source heterogeneity quantization value, and the vertical coordinate axis is the deviation degree of the data real-time quantization value; Based on the deviation degree of the data multi-source heterogeneity quantization value and the deviation degree of the data real-time quantization value, obtain the comprehensive optimization urgency; Obtain the first-level urgency threshold and the second-level urgency threshold from the supplier quality assessment database; If the comprehensive optimization urgency is less than or equal to the first-level urgency threshold, the corresponding area in the heat map is light green, indicating that the system status is qualified and no immediate intervention is required; If the comprehensive optimization urgency is greater than the first-level urgency threshold and less than the second-level urgency threshold, the corresponding area in the heat map is orange, indicating that potential risks need to be concerned and the warning log record is started; If the comprehensive optimization urgency is greater than or equal to the second-level urgency threshold, set it to dark red and trigger a pulsating flashing effect, requiring immediate handling.

[0065] Further, the specific steps for obtaining the comprehensive optimization urgency are as follows: Obtain the comprehensive optimization urgency threshold and the comprehensive optimization urgency importance score from the supplier quality assessment database; The comprehensive optimization urgency threshold includes the first determination value of the data multi-source heterogeneity quantization value and the second determination value of the data real-time quantization value; The comprehensive optimization urgency importance score includes the importance score of the data multi-source heterogeneity quantization value and the importance score of the data real-time quantization value; Analyze the ratio of the deviation between the data multi-source heterogeneity quantization value and the first determination value of the data multi-source heterogeneity quantization value to the first determination value of the data multi-source heterogeneity quantization value, perform averaging processing on the results of the ratio analysis, and use the importance score of the data multi-source heterogeneity quantization value to correct the results of the averaging processing, denoted as the deviation degree of the data multi-source heterogeneity quantization value; Analyze the ratio of the deviation between the data real-time quantization value and the second determination value of the data real-time quantization value to the second determination value of the data real-time quantization value, perform averaging processing on the results of the ratio analysis, and use the importance score of the data real-time quantization value to correct the results of the averaging processing, denoted as the deviation degree of the data real-time quantization value; Perform coupling analysis on the deviation degree of the data multi-source heterogeneity quantization value and the deviation degree of the data real-time quantization value to obtain the comprehensive optimization urgency.

[0066] In this embodiment, the specific method for obtaining the comprehensive optimization urgency through analysis is as follows: ; ; Indicates the comprehensive optimization urgency.

[0067] Indicates the data multi-source heterogeneity quantization value of the data multi-source heterogeneity detection point of the quality data of the th supplier.

[0068] It is expressed as the first determination value of the quantization value of the multi-source heterogeneity of the data obtained from the supplier quality assessment database.

[0069] It represents the quantization value of the data timeliness of the

[0070] It is expressed as the second determination value of the quantization value of the data timeliness of the data obtained from the supplier quality assessment database.

[0071] It represents the constant term. If is not 0, then is 0. If is 0, then is a relatively small constant term, which is a number set to avoid meaningless points in the data, such as 0.001.

[0072] It is the importance score of the preset multi-source heterogeneity quantization value of the data obtained from the supplier quality assessment database.

[0073] It is the importance score of the preset data timeliness quantization value of the data obtained from the supplier quality assessment database.

[0074] Through the preset importance score mapping table in the database, the quick association between the quantization value of data multi-source heterogeneity and the quantization value of data timeliness and the corresponding importance scores is realized. This mapping table defines a set of clear association rules, which can convert the specific numerical values of each parameter into the corresponding importance scores.

[0075] As Figure 2 shown, the horizontal axis represents the deviation degree of the multi-source heterogeneity evaluation value, the vertical axis is the deviation degree of the data timeliness quantization value, and the color depth is positively correlated with the comprehensive optimization urgency. The light green area represents low urgency, the system status is qualified, and no immediate intervention is required; the orange area represents medium urgency, potential risks need to be concerned, and the warning log record is started; the dark red area represents high urgency, and relevant personnel are notified to handle it immediately.

[0076] Furthermore, to trigger the pulse flashing effect and the specific steps for immediate handling are as follows: when the comprehensive optimization urgency is greater than the three-level urgency threshold, the corresponding area of the heat map displays a red pulse warning, and a work order is sent to the operation and maintenance personnel of the supplier operation and maintenance management platform, with the priority set to urgent; if the waiting time after sending the work order with the urgent priority is greater than the preset waiting time, the system will automatically start the three-level downgrading strategy and close the non-core functions.

[0077] In this embodiment, when the comprehensive optimization urgency of a certain area in the enterprise's cross-regional supply chain is detected to exceed the preset third-level urgency threshold, the corresponding area in the heat map will issue a warning in the form of a red pulse and send a work order with an emergency priority to the operation and maintenance personnel of the supplier operation and maintenance management platform. If the waiting time after sending the work order with an emergency priority is greater than the preset waiting time, the system will automatically activate the third-level downgrading strategy. The waiting time represents the time difference between the time point when the work order with an emergency priority is sent and the current time point. There are a preset warning response time reference value and a preset warning response time threshold. The preset warning response time reference value represents the expected warning response time, that is, the ideal time interval between the issuance of the warning signal and the adoption of corresponding measures. The preset warning response time threshold represents a critical value used to determine whether the warning response exceeds the acceptable range. If the warning response time is less than the preset warning response time reference value, the first-level downgrading strategy is activated. The first-level downgrading is to turn off non-core functions, suspend background tasks such as user behavior analysis and visualization report generation, and terminate non-real-time data batch processing jobs (such as historical data archiving); if the warning response time is greater than the preset warning response time reference value and less than the preset warning response time threshold, the second-level downgrading strategy is activated. The second-level downgrading service scope shrinks to limit the data processing object to only process data of a certain type of suppliers, such as the top 10 core suppliers in terms of contract amount, suspend data interaction with non-critical suppliers, interrupt their API query responses, and stop receiving their data streams (Kafka consumers suspend subscriptions); if the warning response time is greater than the preset warning response time threshold, the third-level downgrading strategy is activated. The third-level downgrading is edge cache disaster tolerance. When the network is interrupted, the edge node cache data is enabled to provide read-only services, the cache data based on Redis Cluster responds to query requests, and all data write operations (such as order submission and data update) are rejected. A dedicated resource pool is reserved for tasks such as abnormal mode recognition and core supplier data processing. Priority tags are added to critical requests in the load balancer (such as Nginx) to ensure preferential routing to idle nodes, aiming to timely detect and handle potential performance problems, prevent problem escalation, and ensure the normal operation of core services.

[0078] In summary, as Figure 3As shown, collect the original heterogeneous data from the historical transaction database of the supplier, and preprocess the collected original heterogeneous data, including standardization processing to eliminate the dimensional differences between the data, to obtain heterogeneous data. Based on the preprocessed heterogeneous data, generate a data quality report. Through the analysis of the data quality report, quantitatively evaluate the multi-source heterogeneity and real-time nature of the supplier quality data respectively to obtain the quantitative value of data multi-source heterogeneity and the quantitative value of data real-time nature. Compare and analyze the quantitative value of data multi-source heterogeneity and the quantitative value of data real-time nature with the first determination value and the second determination value in the supplier quality evaluation database respectively to obtain optimization adjustment modes, including data integration optimization mode and resource scheduling optimization mode. Generate an optimization efficiency heat map through the Grafana visualization platform. Input the data after data integration optimization and resource scheduling optimization into the supplier quality evaluation module to evaluate the overall quality of the supplier and display the improvement of efficiency after data integration optimization and resource scheduling optimization. The optimization efficiency heat map sets a two-dimensional coloring rule. The corresponding area in the heat map is light green, indicating a normal state; the corresponding area in the heat map is orange, and the warning log record is started; if the corresponding area in the heat map is dark red, a work order will be sent to the operation and maintenance personnel of the supplier operation and maintenance management platform, and the system will automatically start a three-level downgrading strategy, such as closing non-core functions, edge cache disaster recovery, and providing read-only services.

[0079] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. 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 storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0080] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of processes and / or blocks in the flowchart and / or block diagram can also 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 processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the specified functions in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0081] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction device that implements the functions specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in the flowchart.

[0082] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus, such that a series of operational steps are performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing steps for implementing the functions specified in one or more processes and / or blocks Figure 1 in one or more processes and / or blocks Figure 1 specified in the flowchart.

[0083] 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 of the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present invention.

[0084] It is obvious that those skilled in the art can make various changes and deformations to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and deformations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these modifications and deformations.

Claims

1. A supplier quality assessment method based on big data analysis, characterized in that: The following steps are involved: Collecting original heterogeneous data from the supplier's historical transaction database, and preprocessing the collected original heterogeneous data to obtain heterogeneous data; Based on the pre-processed heterogeneous data, a data quality report is generated. Through the data quality report, the multi-source heterogeneity of supplier quality data and the real-time performance of supplier quality assessment methods are quantitatively evaluated, and the quantitative values ​​of data multi-source heterogeneity and data real-time performance are obtained; Compare and analyze the data multi-source heterogeneity quantification value and the data real-time quantification value with the first judgment value of the data multi-source heterogeneity quantification value and the second judgment value of the data real-time quantification value in the supplier quality evaluation database, respectively, to obtain an optimization adjustment mode, which includes a data integration optimization mode and a resource scheduling optimization mode; An optimization efficiency heat map is generated through the Grafana visualization platform, and the heterogeneous data that has been connected and integrated after data integration optimization and resource scheduling optimization is input into the supplier quality evaluation module. The supplier quality evaluation module is used to evaluate the overall quality of suppliers. The optimization efficiency heat map shows the efficiency improvement after data integration optimization and resource scheduling optimization.

2. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of obtaining heterogeneous data are: Collecting original heterogeneous data from the supplier's historical transaction database. The original heterogeneous data refers to multi-dimensional dynamic feature data obtained from different supplier data sources to describe the supplier's quality. The collected original heterogeneous data are standardized to eliminate the dimensional differences between the data and obtain heterogeneous data.

3. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of generating the data quality report are: Connect and integrate pre-processed heterogeneous data through data quality management tools; Evaluate the connected and integrated heterogeneous data according to the data quality rules defined in the data quality management tool and generate a data quality report; Analyze the data quality report to obtain the quantitative values ​​of data multi-source heterogeneity and data real-time performance.

4. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of obtaining the quantitative value of data multi-source heterogeneity are: Obtaining a data multi-source heterogeneity quantitative value threshold and a data multi-source heterogeneity quantitative value importance score from a supplier quality assessment database; The data multi-source heterogeneity quantification value thresholds include an abnormal pattern recognition delay threshold, a cross-region round-trip time threshold, a metadata alignment time threshold, and a heterogeneous data coverage threshold; The importance scores of the data multi-source heterogeneity quantification values ​​include the importance scores of abnormal pattern recognition delay, cross-region round-trip time, metadata alignment time, and heterogeneous data coverage; Perform a ratio analysis on the abnormal pattern recognition delay and the abnormal pattern recognition delay threshold, average the results of the ratio analysis, and use the importance score of the abnormal pattern recognition delay to correct the average result, which is recorded as the first component of the data multi-source heterogeneity quantification value; Perform a ratio analysis on the cross-region round-trip time and the cross-region round-trip time threshold, average the results of the ratio analysis, and use the importance score of the cross-region round-trip time to correct the average result, which is recorded as the second component of the data multi-source heterogeneity quantification value; Perform a ratio analysis on the metadata alignment time and the metadata alignment time threshold, average the results of the ratio analysis, and use the importance score of the metadata alignment time to correct the average result, which is recorded as the third component of the data multi-source heterogeneity quantification value; Perform a ratio analysis on the heterogeneous data coverage threshold and the heterogeneous data coverage, and use the importance score of the heterogeneous data coverage to correct the result of the ratio analysis, which is recorded as the fourth component of the data multi-source heterogeneity quantification value; The first component of the data multi-source heterogeneity quantization value, the second component of the data multi-source heterogeneity quantization value, the third component of the data multi-source heterogeneity quantization value and the fourth component of the data multi-source heterogeneity quantization value are coupled and analyzed to obtain the data multi-source heterogeneity quantization value; The data multi-source heterogeneity quantitative value represents a measure of the multi-source and heterogeneity of supplier data.

5. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of obtaining the real-time quantitative value of the data are: Obtaining a data real-time quantitative value threshold and a data real-time quantitative value importance score from a supplier quality assessment database; The data real-time quantitative value thresholds include a high concurrency degradation rate threshold, a node processing request response time threshold, a cross-region round-trip time threshold, and a metadata alignment time threshold; The importance score of the data real-time quantitative value includes the importance score of high concurrency degradation rate, the importance score of node processing request response time, the importance score of cross-region round-trip time and the importance score of metadata alignment time; Perform a ratio analysis between the high concurrency degradation rate threshold and the high concurrency degradation rate, and use the importance score of the high concurrency degradation rate to correct the result of the ratio analysis, which is recorded as the first component of the data real-time quantization value; Perform a ratio analysis on the node processing request response time threshold and the node processing request response time, average the result of the ratio analysis, and use the importance score of the node processing request response time to correct the average result, which is recorded as the second component of the data real-time quantization value; Perform a ratio analysis on the cross-region round-trip time threshold and the cross-region round-trip time, average the results of the ratio analysis, and use the importance score of the cross-region round-trip time to correct the average result, which is recorded as the third component of the data real-time quantitative value; Perform a ratio analysis on the metadata alignment time threshold and the metadata alignment time, average the results of the ratio analysis, and use the importance score of the metadata alignment time to correct the average result, which is recorded as the fourth component of the data real-time quantization value; The coupling results of the first component of the data real-time quantization value, the second component of the data real-time quantization value, the third component of the data real-time quantization value and the fourth component of the data real-time quantization value are coupled and analyzed to obtain the data real-time quantization value, which is used to measure the real-time performance of the supplier quality evaluation method.

6. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of the data integration optimization model are: Obtaining a first determination value of a data multi-source heterogeneity quantification value and a second determination value of a data real-time quantification value from a supplier quality evaluation database; If the data multi-source heterogeneity quantization value is lower than or equal to the data multi-source heterogeneity quantization value first determination value, there is no need to trigger the data integration optimization mode; If the data multi-source heterogeneity quantization value is greater than the data multi-source heterogeneity quantization value first determination value, a data integration optimization mode is triggered, and the triggered data integration optimization mode is used to improve the efficiency of data integration.

7. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of the resource scheduling optimization mode are: If the data real-time quantization value is greater than or equal to the second data real-time quantization value determination value, there is no need to trigger the resource scheduling optimization mode; If the data real-time quantization value is lower than the second data real-time quantization value determination value, a resource scheduling optimization mode is triggered, and the triggered resource scheduling optimization mode is used to improve resource utilization.

8. A supplier quality assessment method based on big data analysis as claimed in claim 1, characterized in that: The specific steps of generating the optimization efficiency heat map are: Generate an optimization efficiency heat map through the Grafana visualization platform and set a two-dimensional coloring rule. The horizontal axis of the optimization efficiency heat map is the deviation of the quantitative value of multi-source heterogeneity of data, and the vertical axis is the deviation of the quantitative value of real-time data. Based on the deviation of the quantitative value of data multi-source heterogeneity and the deviation of the quantitative value of data real-time, the comprehensive optimization urgency is obtained; Obtaining a first-level urgency threshold and a second-level urgency threshold from a supplier quality assessment database; If the comprehensive optimization urgency is less than or equal to the first-level urgency threshold, the corresponding area in the heat map is light green, indicating that the system status is qualified and no immediate intervention is required; If the comprehensive optimization urgency is greater than the first-level urgency threshold and the comprehensive optimization urgency is less than the second-level urgency threshold, the corresponding area in the heat map will be orange, indicating that potential risks need to be paid attention to and early warning log recording will be started; If the comprehensive optimization urgency is greater than or equal to the secondary urgency threshold, a dark red color is set and a pulse flashing effect is triggered, requiring immediate processing.

9. A supplier quality assessment method based on big data analysis as claimed in claim 8, characterized in that: The specific steps of obtaining the comprehensive optimization urgency are: Obtaining a comprehensive optimization urgency threshold and a comprehensive optimization urgency importance score from a supplier quality assessment database; The comprehensive optimization urgency threshold includes a first determination value of the data multi-source heterogeneity quantification value and a second determination value of the data real-time quantification value; The comprehensive optimization urgency importance score includes the importance score of the data multi-source heterogeneity quantification value and the importance score of the data real-time quantification value; The deviation between the data multi-source heterogeneity quantization value and the first judgment value of the data multi-source heterogeneity quantization value is analyzed with the first judgment value of the data multi-source heterogeneity quantization value, the result of the proportion analysis is averaged, and the result of the averaged processing is corrected using the importance score of the data multi-source heterogeneity quantization value, which is recorded as the deviation of the data multi-source heterogeneity quantization value; Perform a ratio analysis on the deviation between the data real-time quantization value and the second determination value of the data real-time quantization value, average the result of the ratio analysis, and use the importance score of the data real-time quantization value to correct the result of the average processing, which is recorded as the deviation of the data real-time quantization value; The deviation of the quantified value of data multi-source heterogeneity and the deviation of the quantified value of data real-time are coupled and analyzed to obtain the comprehensive optimization urgency.

10. A supplier quality assessment method based on big data analysis as claimed in claim 8, characterized in that: The trigger pulse flickering effect requires immediate processing of the following specific steps: When the comprehensive optimization urgency is greater than the third-level urgency threshold, the corresponding area of ​​the heat map displays a red pulse warning, and a work order is sent to the operation and maintenance personnel of the supplier operation and maintenance management platform, with the priority set to emergency; If the waiting time after sending an urgent priority work order is greater than the preset waiting time, the system will automatically start the three-level downgrade strategy and shut down non-core functions.

Images