Quality control method for supply chain fusing cloud platform and industrial internet

Abstract

The invention provides a quality management and control method for a supply chain fusing a cloud platform and an industrial internet, and belongs to the technical field of the internet, and the method comprises the steps: deploying a collection network for a corresponding supply link according to the supply mode of each supply link of the supply chain through employing the industrial internet technology, and transmitting the collection data of each supply link to the cloud platform; constructing a new standard set according to the current supply demand of each supply link and the mapping relationship between the acquisition network and the preset risk item under the corresponding supply link; analyzing the collected data of the corresponding supply link according to the new standard set, constructing a quality chain of the corresponding supply link, and comparing and analyzing the quality chain with a standard chain to determine a quality problem; and based on the quality problem of each supply link, determining a management and control optimization strategy of the corresponding supply link for optimization deployment, and meanwhile, based on the cloud platform, feeding back the quality problem to a management end of each supply link for reminding. And comprehensive and real-time intelligent management and control of all supply links on the supply chain are ensured.

Description

Technical Field

[0001] The present invention relates to the field of Internet technologies, and particularly to a quality control method for a supply chain integrating a cloud platform and industrial Internet. Background Art

[0002] With the development of global economic integration, the supply chain has become increasingly complex, covering multiple links such as raw material procurement, manufacturing, product transportation, warehousing management, and sales. In this process, any quality problem in one link may have a serious impact on the normal operation and final quality of the entire supply chain. Traditional supply chain quality control methods often rely on manual sampling inspection, paper records, and decentralized quality management systems, suffering from problems such as untimely information transmission, poor data accuracy, and difficulty in achieving cross-link collaborative management. With the development of industrial Internet and cloud platform technologies, although some enterprises have begun to try to apply new technologies to supply chain management, current applications mostly stay at the level of simple data storage and transmission, failing to fully utilize the powerful computing and storage capabilities of the cloud platform and the advantages of efficient data collection and device interconnection of the industrial Internet, and unable to achieve comprehensive and real-time intelligent control of the supply chain quality.

[0003] Therefore, the present invention proposes a quality control method for a supply chain integrating a cloud platform and industrial Internet. Summary of the Invention

[0004] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, which is used to facilitate the determination of existing quality problems for control optimization by setting up a collection network and constructing a new standard set under supply and demand, and ensure comprehensive and real-time intelligent control of all supply links in the supply chain.

[0005] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, including: Step 1: According to the supply modes of each supply link in the supply chain and by using industrial Internet technologies, deploy a collection network to the corresponding supply link, and transmit the collection data of each supply link to the cloud platform; Step 2: Construct a new standard set according to the current supply demand of each supply link and the mapping relationship between the collection network and the preset risk items under the corresponding supply link; Step 3: Based on the cloud platform, analyze the collection data of the corresponding supply link according to the new standard set, construct a quality chain for the corresponding supply link, and compare and analyze it with the standard chain to determine quality problems, where the standard chain is constructed from the minimum requirements of each new standard in the corresponding new standard set; Step 4: Determine the control optimization strategies for the corresponding supply links based on the quality problems in each supply link for optimized deployment. At the same time, feedback the quality problems to the management terminals of each supply link based on the cloud platform for reminder.

[0006] Preferably, construct a new standard set according to the current supply demands of each supply link and the mapping relationship between the acquisition network and the preset risk items under the corresponding supply link, including: Extract the business environments of each supply link in the supply chain under different historical supply demands from the historical database, analyze the business environments to obtain the risk values of each preset risk item, and perform cluster analysis on the business environments to construct several risk matrices; Perform standardization and normalization processing on each risk matrix to obtain the eigenvectors of each processed matrix, where the eigenvectors contain the characteristic coefficients of each preset risk item; Sort the characteristic coefficients in each eigenvector from largest to smallest to obtain a magnitude vector; Based on the first element in the magnitude vector, sequentially and continuously select three elements to map to a preset coordinate system for three-point connection drawing to determine the drawing shape; Rely on the drawing shape and combine the center point of the drawing shape to assign a first label to the first element among the three continuously selected elements; Extract the labels under the same preset risk item from all the first labels as the second labels, and analyze the shape stability and numerical stability of the corresponding preset risk item based on all the second labels; According to the drawing shape and the center point corresponding to each preset risk item under different risk matrices, determine the first bias based on the shape stability and the second bias based on the numerical stability; When the set standard of the preset risk item under the corresponding supply demand does not match the first bias and the second bias, rely on the first bias and the second bias to perform a single optimization on the corresponding set standard to obtain a new standard; Match the current supply demand of each supply link with all the historical demands of the same supply link. If there is a completely matching historical demand, obtain a new standard set according to the mapping relationship between the acquisition network and the preset risk items under the corresponding supply link; Otherwise, regard the historical demand with the largest matching coefficient that is not completely matching as the reference demand; Rely on the demand difference between the reference demand and the current supply demand, and input the demand difference into the demand analysis model to obtain the bias coefficients of each preset risk item, and perform a secondary optimization on the new standard based on the reference demand to obtain a new standard set.

[0007] Preferably, a new standard is obtained by optimizing the corresponding set standard according to the first bias and the second bias, including: Obtain the first mismatch factor and the second mismatch factor between the first bias, the second bias and the set standard under the corresponding supply demand respectively; Determine the optimization direction of the corresponding set standard according to the first mismatch factor and the second mismatch factor, wherein the optimization direction is the standard improvement direction, the standard reduction direction and the standard unchanged direction; Adjust the corresponding data set standard according to the optimization direction.

[0008] Preferably, the new standard based on the reference demand is optimized twice, including: Regard the new standard of the reference demand as the first standard; Obtain the bias coefficient of each preset risk item, and optimize the corresponding first standard in combination with the item weight of the preset risk item based on the reference demand.

[0009] Preferably, a quality chain corresponding to the supply link is constructed, including: Compare and analyze each new standard in the new standard set with the corresponding sub-data in the collected data of the corresponding supply link to determine the initial abnormal factor, and combine the setting position of the corresponding new standard in the corresponding supply link to construct an initial vector, and input the initial vector into the vector analysis model to obtain the output representation based on the corresponding new standard, wherein the output representation includes: the abnormal distribution and the abnormal degree at each abnormal distribution position; Obtain the blank chain of the corresponding supply link from the chain database; Input the output representation of each new standard under the corresponding supply link into the blank chain in turn to obtain the quality chain.

[0010] Preferably, determining the initial abnormal factor includes:

[0011] Wherein, Represents the initial abnormal factor; Represents the quantization result of the corresponding sub-data; Represents the quantization threshold of the corresponding new standard; Represents the arctangent function; Represents the set weight of the corresponding new standard; Represents the Bernoulli random variable; Represents the sign function; Represents the comparison sensitivity coefficient.

[0012] Preferably, comparing and analyzing with the standard chain to determine the quality problem, including: Obtain an abnormal combination of abnormal distribution positions based on the quality chain; Extract the quality analysis indicators involved in the lowest standard combination corresponding to the abnormal distribution position from the standard chain, and extract the analysis model from the index-model database to analyze the abnormal combination, determine the clear sub-problems and fuzzy sub-problems corresponding to the abnormal distribution position, where the fuzzy sub-problems are between clear and unclear and there is a fuzzy probability; Obtain the first indicator compared with the clear sub-problem and the second indicator compared with the fuzzy sub-problem, perform extraction of the same indicators on the first indicator and the second indicator, and determine whether there is the same indicator; Trace back the corresponding clear sub-problem according to the first indicator under the same indicator to determine the first proportion problem, and at the same time, trace back the corresponding fuzzy sub-problem according to the second indicator under the same indicator to determine the second proportion problem; Conduct a correlation analysis on the first proportion problem and the second proportion problem to determine the dependency association relationship of the corresponding second indicator on the first indicator; Clarify the fuzzy sub-problems under the corresponding second indicator according to the dependency association relationship and in combination with the fuzzy probability, and combine the sub-problems under the remaining independent indicators to obtain the quality problems of the abnormal distribution position.

[0013] Preferably, determine the control and optimization strategies for the corresponding supply links based on the quality problems of each supply link, including: Aggregate the quality problems of the supply link, and perform a normal distribution on the influence range of the quality problems involved in each aggregation result to obtain the concentrated influence range; Extract the quality problems under any maximum influence range in each aggregation result to construct an auxiliary function to be optimized, and at the same time, extract the quality problems that best match the corresponding concentrated influence range from each aggregation result to construct the main function to be optimized; Solve the main function to be optimized and the auxiliary function to be optimized to obtain the quality optimization binary group of each aggregation result; Use the range variance group in the normal distribution in each aggregation result as a reference index, and match the analysis strategy corresponding to the aggregation result from the aggregation quality type-index-strategy comparison table; Input the binary group into the analysis strategy to obtain the optimal function, and transform the optimal function to obtain the control and optimization strategy.

[0014] Compared with the prior art, the beneficial effects of this application are as follows: By setting up the acquisition network and constructing a new standard set under the supply demand, it is convenient to determine the existing quality problems for control and optimization, and ensure the comprehensive and real-time intelligent control of all supply links on the supply chain.

[0015] Other features and advantages of the present invention will be described in the following specification, and in part will be obvious from the specification, or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained by the structure specifically pointed out in the written specification and the drawings.

[0016] The technical solution of the present invention will be further described in detail below through the drawings and embodiments. Description of the Drawings

[0017] The drawings are used to provide a further understanding of the present invention, and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention, and do not constitute a limitation to the present invention. In the drawings: Figure 1 It is a flowchart of a quality control method for a supply chain integrating a cloud platform and industrial Internet in an embodiment of the present invention; Figure 2 It is a structural diagram of a preset coordinate system in an embodiment of the present invention. Detailed Embodiments

[0018] The following describes the preferred embodiments of the present invention with reference to the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0019] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, as Figure 1 shown, including: Step 1: Deploy a collection network to the corresponding supply link according to the supply mode of each supply link in the supply chain and by using industrial Internet technology, and transmit the collection data of each supply link to the cloud platform; Step 2: Construct a new standard set according to the current supply demand of each supply link and the mapping relationship between the collection network and the preset risk items under the corresponding supply link; Step 3: Analyze the collection data of the corresponding supply link based on the new standard set on the cloud platform, construct a quality chain for the corresponding supply link, and compare it with the standard chain to determine quality problems, where the standard chain is constructed based on the minimum requirements of each new standard in the corresponding new standard set; Step 4: Determine the control optimization strategy for the corresponding supply link based on the quality problems of each supply link for optimized deployment. At the same time, feedback the quality problems to the management terminals of each supply link based on the cloud platform for reminder.

[0020] In this embodiment, the supply link of the supply chain refers to the business stages with independent functions in the supply chain, such as raw material procurement, production and processing, warehousing and logistics, sales and service, etc. Taking the supply chain of electronic devices as an example, the supply links therein include: "chip procurement", "mainboard production", "finished product warehousing", "e-commerce sales", etc.

[0021] The supply mode refers to the business operation methods of each link, including procurement mode (such as JIT just-in-time procurement), production mode (such as assembly line production), logistics mode (such as third-party logistics TPL), etc.

[0022] Industrial Internet technology refers to the interconnection of physical devices and the network through devices such as sensors, RFID, and embedded systems to collect real-time data. For example, temperature sensors on the production line and RFID shelf tags in the warehouse are all Industrial Internet devices.

[0023] The acquisition network is a network composed of various sensors and communication modules, which is used to collect data of each link in real time. And the network components involved in different links are different and are interdependent with the corresponding supply link itself. For example, the transportation status acquisition network in the raw material transportation link is constructed through GPS sensors and humidity sensors.

[0024] Specifically: Select sensors according to the characteristics of the link: For example, vibration sensors are deployed in the production link to monitor the operation status of equipment, and temperature and humidity sensors are deployed in the warehousing link to monitor the storage environment.

[0025] Build a communication network: Realize the wireless transmission of sensor data through technologies such as 5G and LoRa, or achieve wired connection through industrial Ethernet.

[0026] Data access to the cloud platform: Preprocess the collected data (such as denoising, normalization) through an edge computing gateway, and then transmit it to the database of cloud platforms such as Alibaba Cloud and AWS.

[0027] In this embodiment, the current supply demand refers to the specific requirements of the current business scenario, such as order volume, delivery cycle, quality standard, etc. For example, during the e-commerce big promotion period, "daily delivery order volume ≥ 100,000 orders" is the current supply demand for the logistics link.

[0028] The preset risk items refer to the potential quality risk points set according to historical experience or industry standards, such as "excessive raw material impurities", "transportation delay", etc. For example, the preset risk items in the procurement link of electronic components include "non-compliance of component withstand voltage value" and "packaging breakage rate".

[0029] The mapping relationship refers to the corresponding association between the acquisition network data and the risk items. For example, "temperature sensor data" maps to "risk of excessive temperature during transportation". For example, the humidity sensor data in the warehousing link directly maps to "risk of goods getting damp".

[0030] The new standard set refers to the set of quality standards dynamically optimized according to current requirements, such as raw material acceptance standards, production process parameter thresholds, etc. For example, the new standard set includes "screen touch sensitivity threshold ≥ 95%", "battery charging efficiency standard ≥ 98%", etc.

[0031] In this embodiment, the quality chain is a chain that visualizes quality data at each time point with the process of the corresponding link as the axis, and is used to display abnormal distributions. The standard chain is a theoretical quality baseline constructed based on the minimum requirements of the new standard set and is used to compare with the actual quality chain.

[0032] The initial abnormal factor is a quantified index calculated by a formula and is used to characterize the degree of deviation between data and standards.

[0033] In this embodiment, the abnormal distribution refers to the position and degree of abnormal points in the quality chain, such as "the yield of the 3rd process in the production link drops sharply".

[0034] In this embodiment, the control and optimization strategy is an improvement measure formulated for quality problems, such as equipment maintenance plans, process parameter adjustments, personnel training, etc. For example, "adjust the wire bonding machine pressure from 70g to 80g", "increase the lens cleaning frequency to 2 times a day".

[0035] The management - end reminder is to push warning information to the person in charge of each link through the cloud platform, such as APP notifications, email reminders. For example, the warehousing management end receives a reminder of "the cold storage temperature exceeds the standard", triggering emergency cooling measures.

[0036] The beneficial effects of the above - mentioned technical solution are: by setting up the acquisition network and constructing the new standard set under the supply - demand situation, it is convenient to determine existing quality problems for control and optimization, ensuring comprehensive and real - time intelligent control of all supply links in the supply chain.

[0037] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet. According to the current supply demand of each supply link and the mapping relationship between the acquisition network and the preset risk items under the corresponding supply link, a new standard set is constructed, including: Extract the business environment of each supply link in the supply chain under different historical supply demands from the historical database, analyze the business environment to obtain the risk values of each preset risk item, and perform clustering analysis on the business environment to construct several risk matrices; Perform standardization and normalization processing on each risk matrix to obtain the eigenvectors of each processed matrix, where the eigenvectors contain the characteristic coefficients of each preset risk item; Sort the characteristic coefficients in each eigenvector from large to small to obtain a magnitude vector; Based on the first element in the size vector, three elements are sequentially and continuously screened and mapped to a preset coordinate system for three-point connection drawing to determine the drawing shape; Depending on the drawing shape and in combination with the center point of the drawing shape, assign a first label to the first element among the three continuously screened elements; Extract the labels under the same preset risk item from all the first labels as the second labels, and analyze the shape stability and numerical stability of the corresponding preset risk item based on all the second labels; According to the drawing shape and center point corresponding to each preset risk item under different risk matrices, determine the first bias based on the shape stability and the second bias based on the numerical stability; When the set standard of the preset risk item under the corresponding supply demand does not match the first bias and the second bias, optimize the corresponding set standard once based on the first bias and the second bias to obtain a new standard; Match the current supply demand under each supply link with all the historical demands under the same supply link. If there is a completely matching historical demand, obtain a new standard set according to the mapping relationship between the acquisition network and the preset risk item under the corresponding supply link; Otherwise, regard the historical demand with the largest matching coefficient that is not completely matched as the reference demand; Depending on the demand difference between the reference demand and the current supply demand, and input the demand difference into the demand analysis model to obtain the bias coefficient of each preset risk item, and perform a secondary optimization on the new standard based on the reference demand to obtain a new standard set.

[0038] In this embodiment, the historical database stores a database of the past operation data of the supply chain, covering the business information of each supply link in different periods, such as raw material procurement prices, production outputs, transportation times, etc. For example, in the historical database of a certain automobile manufacturing enterprise, there are data such as the monthly parts procurement quantity, supplier delivery on-time rate, and the number of production line equipment failures in the past 5 years. In this embodiment, the historical supply demand is the business demand faced by the supply chain operation in a past period, such as a specific order quantity, delivery deadline, product quality requirements, etc. For example, a certain e-commerce platform put forward a historical supply demand of reaching 1 million orders per day for the logistics link in June last year. In this embodiment, the business environment is the internal and external conditions that affect the operation of the supply link, including market demand, supplier status, production equipment status, etc. The preset risk items are preset according to experience or industry standards, and are potential risks that may affect the quality or normal operation of the supply chain.

[0039] In this embodiment, the environmental risk analysis model is obtained by training a neural network model based on different business environments and the risk values set by experts for the preset risk items that may exist in the environment. Therefore, the risk values of each preset risk item can be directly obtained.

[0040] In this embodiment, the preset risk items are set in advance. Only the corresponding sub-risk values need to be filled into the corresponding preset risk item filling positions to obtain the matrix of the corresponding clustering results. The columns of the matrix correspond to the same preset risk item, and the rows correspond to all the preset risk items in the corresponding environmental scenario.

[0041] In this embodiment, standardization and normalization processes transform the data to have a unified scale or distribution. Standardization usually converts the data into a standard normal distribution with a mean of 0 and a standard deviation of 1; normalization maps the data to the interval [0,1].

[0042] In this embodiment, the eigenvector is a vector that can reflect the core characteristics of the risk matrix after processing, and contains the characteristic coefficients of each preset risk item. For example, in the eigenvector of a certain risk matrix, the characteristic coefficient of the "raw material price fluctuation" risk item is 0.7, indicating that it accounts for a relatively large proportion in the characteristics of this matrix. In this embodiment, the magnitude vector is a vector obtained by arranging the characteristic coefficients in the eigenvector in descending order. For example, after sorting a certain eigenvector, a magnitude vector [0.8, 0.6, 0.3, 0.2] is obtained, which respectively corresponds to the characteristic coefficients of different preset risk items. Select the first three elements (0.8, 0.6, 0.3) and draw a triangle in the preset coordinate system. Suppose the magnitude vector is [0.5, 0.5, 0.5, 0.2], and draw a straight line in the element and coordinate system. It should be noted that the abscissa of the preset coordinate system is the number of elements, and the ordinate is the characteristic coefficient, as Figure 2 shown.

[0043] In this embodiment, the first label is the identifier given to the first element among the continuously selected three elements according to the drawn shape and the center point. The second label is the label set belonging to the same preset risk item extracted from all the first labels. In this embodiment, the shape stability is to evaluate the similarity degree of the drawn shapes of the preset risk items under different risk matrices, reflecting its variation law. If the drawn shapes of the "transportation delay" risk item are similar under multiple risk matrices, it indicates that its shape stability is high. Specifically, the shape stability is evaluated by calculating the shape similarity (such as the Hausdorff distance), and the numerical stability is evaluated by calculating the variance index of the characteristic coefficients. The numerical stability is to measure the degree of fluctuation of the characteristic coefficients of the preset risk items under different risk matrices. For example, the characteristic coefficients of the "equipment failure" risk item fluctuate less in different risk matrices, indicating that its numerical stability is high.

[0044] The first deviation is the preset risk item standard adjustment direction determined based on shape stability, such as the standard being increased, decreased, or remaining unchanged. The second deviation is the preset risk item standard adjustment direction determined based on numerical stability. For example, the set standard of "product defective rate" is 5%. After analyzing its shape and numerical stability, it shows that the risk has an upward trend. Both the first deviation and the second deviation are in the direction of increasing the standard. According to the calculation of the mismatch factor, the standard is adjusted to 3% to obtain the new standard.

[0045] In this embodiment, the set standard is the quality or operation standard currently set for the preset risk item. In this embodiment, the new standard is the new quality or operation standard obtained by optimizing the set standard according to the first deviation and the second deviation. For example, adjusting the set standard of "transportation time" from 24 hours to 18 hours is the new standard. The matching coefficient is a value that measures the similarity between the current supply demand and the historical demand, that is, it is calculated through a similarity function.

[0046] The reference demand is the incomplete matching historical demand with the largest matching coefficient for the current supply demand. The demand difference is, for example, the differences between the current supply demand and the reference demand in terms of order quantity, delivery time, product specifications, etc. In this embodiment, a demand analysis model is constructed based on a neural network. The input samples are different demand differences, and the input samples are the bias coefficients of each preset risk item corresponding to the demand difference. The bias coefficient is a coefficient that reflects the influence degree of the demand difference on the preset risk item standard. The new standard set is a set of all new preset risk item standards obtained through comprehensive primary optimization and secondary optimization for each supply link. Through the demand analysis model, the bias coefficient of the "special raw material specification supply" risk item is calculated to be 0.2. Combining with the new standard of the reference demand, the standard of this risk item is further adjusted to form a new standard set applicable to the current demand. The beneficial effects of the above technical solutions are as follows: By analyzing historical data and current demands, the quality standards of each link in the supply chain can be quickly adjusted to make the standards more in line with the actual business situation. Optimizing the standards based on the risk matrix and stability analysis can effectively identify potential risks and take measures in advance to reduce the probability of risk occurrence and the impact degree. The formulation of the new standard set helps each supply link operate more efficiently, reduce resource waste, and provide data support for supply chain management.

[0047] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, which relies on the first deviation and the second deviation to perform a primary optimization on the corresponding set standard to obtain a new standard, including: Respectively obtain the first mismatch factor and the second mismatch factor between the first deviation, the second deviation and the set standard under the corresponding supply demand; Determine the optimization direction for the corresponding set standard according to the first mismatch factor and the second mismatch factor, where the optimization direction includes a standard improvement direction, a standard reduction direction, and a standard unchanged direction; Adjust the corresponding data set standard according to the optimization direction.

[0048] In this embodiment, the first mismatch factor is a value quantifying the mismatch degree between the first deviation and the set standard under the corresponding supply demand. For example, if the first deviation of the "transportation delay" risk item is in the standard improvement direction, and the current transportation delay rate is high under the set standard, the calculated first mismatch factor is 0.6 (the larger the value, the higher the mismatch degree). The second mismatch factor is a quantitative index measuring the mismatch degree between the second deviation and the set standard. For example, for the "product packaging damage" risk item, the second deviation indicates that the standard needs to be improved, but the existing set standard fails to effectively control the damage rate, and the calculated second mismatch factor is 0.5.

[0049] The optimization direction is the adjustment direction for the set standard determined according to the mismatch factor, including the standard improvement direction (raising the standard requirements), the standard reduction direction (lowering the standard requirements), and the standard unchanged direction (maintaining the existing standard).

[0050] For example, if the first and second mismatch factors of a risk item are both high, the optimization direction is the standard improvement direction; if the mismatch factor is close to 0, the optimization direction is the standard unchanged direction.

[0051] In this embodiment, the first mismatch factor F1 is calculated according to the following formula: F1 = (T1 - S1) / S1, where T1 is the risk value estimated according to the shape trend and S1 is the risk value corresponding to the set standard.

[0052] The second mismatch factor F2 is calculated according to the following formula: F2 = (T2 - S2) / S2, where T2 is the risk value estimated based on numerical stability and S2 is the risk value corresponding to the set standard.

[0053] When F1 ≥ 0.1 and F2 ≥ 0.1, determine the optimization direction as the standard improvement direction; when 0 < F1 < 0.1 and 0 < F2 < 0.1, determine the optimization direction as the standard unchanged direction; otherwise, determine the optimization direction as the standard reduction direction.

[0054] The beneficial effects of the above technical solution are: By quantifying the mismatch factor and accurately determining the optimization direction, the set standard can better fit the actual risk situation of the supply chain, effectively solving the problems of standard lag or unreasonableness.

[0055] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, which performs secondary optimization on a new standard based on the reference requirements, including: Regarding the new standard of the reference requirements as the first standard; Obtaining the bias coefficient of each preset risk item, and combining the item weight of the preset risk item based on the reference requirements to perform secondary optimization on the corresponding first standard.

[0056] In this embodiment, the item weight is a weight value assigned according to the importance of the preset risk item to the overall quality or operation of the supply chain, and is used to measure the relative importance of each risk item in the standard optimization.

[0057] In this embodiment, the secondary optimization is obtained by Sc×(1 + w0×b0), where Sc is the threshold corresponding to the first standard, w0 is the item weight, and b0 is the bias coefficient.

[0058] The beneficial effect of the above technical solution is that by combining the bias coefficient and the item weight to perform secondary optimization on the first standard, the quality standard can more accurately match the current complex and changeable supply demands, avoiding problems of resource waste or insufficient quality.

[0059] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet, which constructs a quality chain corresponding to the supply link, including: Comparing and analyzing each new standard in the new standard set with the corresponding sub-data in the collected data of the corresponding supply link to determine the initial abnormal factor, and combining the setting position of the corresponding new standard in the corresponding supply link to construct an initial vector, and inputting the initial vector into a vector analysis model to obtain an output representation based on the corresponding new standard, where the output representation includes: abnormal distribution and the degree of abnormality at each abnormal distribution position; Obtaining the blank chain corresponding to the supply link from the chain database; Sequentially inputting the output representation of each new standard under the corresponding supply link into the blank chain to obtain the quality chain.

[0060] Preferably, determining the initial abnormal factor includes:

[0061] Among them, represents the initial abnormal factor; represents the quantization result of the corresponding sub-data; represents the quantization threshold of the corresponding new standard; represents the arctangent function; represents the set weight of the corresponding new standard; represents the Bernoulli random variable; represents the sign function; Indicates the contrast sensitivity coefficient.

[0062] In this embodiment, The value threshold is , The value of is 1.

[0063] In this embodiment, The value of is ±1, .

[0064] In this embodiment, the arctangent function first rises rapidly (sensitive to small deviations) and then flattens out (saturates at large deviations), which is in line with the actual logic that small deviations in the supply chain need to be corrected in a timely manner, and large deviations are mostly systematic problems (requiring separate investigation). For example, the production temperature standard is 25°C. When R changes from 25°C to 26°C (small deviation), the output of tan−1 rises rapidly, and the abnormal factor Q gives a sensitive warning; when R changes from 26°C to 35°C (large deviation), tan−1 approaches saturation, indicating that it is necessary to stop the machine for maintenance rather than just adjusting the parameters, which meets the hierarchical response requirements of on-site management.

[0065] The output Q not only has a deviation quantization value (e.g., 0.3 indicates moderate abnormality), but also reflects the deviation direction (positive / negative, corresponding to "exceeding the standard" or "not meeting the standard") through and implies random fluctuation information through . These dimensions make the description of abnormalities at the nodes of the quality chain more comprehensive (e.g., "Process 3: Q = 0.2, positive deviation, including random fluctuations"). During subsequent analysis, it can be quickly determined whether it is a continuous process problem (negative deviation and no random term) or occasional equipment fluctuations (positive deviation and random term), improving the efficiency of quality traceability.

[0066] In this embodiment, the corresponding sub-data is the data in the collected data that is directly associated with a new standard.

[0067] The initial abnormal factor quantization is an index that reflects the deviation degree between the sub-data and the new standard.

[0068] In this embodiment, the set position is the corresponding node of the new standard in the supply chain process (e.g., "the 3rd process" in the production process).

[0069] The initial vector is a vector composed of the initial abnormal factor and the set position, which is used as the input for the model. The initial abnormal factor sets the position encoding (e.g., [0.33, 5,...], where "5" represents the 5th process).

[0070] The vector analysis model is obtained by training a neural network model with different vectors as outputs and the position distribution and abnormal degree of abnormal elements in the vectors as outputs, and can be directly obtained.

[0071] In this embodiment, the abnormal distribution refers to the spatial / process distribution of abnormal data in the supply chain. For example, in the production process, "the abnormal rates of the 3rd and 5th processes are high", showing a distribution of "scattered processes but concentrated key nodes".

[0072] The degree of abnormality is a quantitative description of the severity of the abnormality (ranging from 0 to 1, where 1 represents a severe abnormality).

[0073] The chain database is a database that stores the blank chain templates and historical quality chain data of each supply chain link. A blank chain is a chain structure without actual data and only containing the link process framework (such as an empty template arranged according to "process 1 - process 2 -... " in the production process). For example, in the automobile manufacturing chain database, there are blank chains for processes such as machining, painting, and final assembly.

[0074] In this embodiment, the quality chain is a chain that fully reflects the quality status of the supply chain link after filling the new standard output representation (abnormal distribution, degree of abnormality) in the blank chain.

[0075] The beneficial effects of the above technical solution are as follows: By combining the initial abnormal factors with the vector model to quantify the degree of abnormality and locate the distribution, it solves the ambiguity of traditional empirical judgment. The quality chain integrates scattered standards and data into a process view, enabling managers to quickly grasp the quality status of the link, realizing precise diagnosis and management of the supply chain quality, improving the efficiency of abnormal handling, and reducing quality risks.

[0076] The present invention provides a quality control method for a supply chain integrating a cloud platform and industrial Internet. By comparing and analyzing with a standard chain to determine quality problems, it includes: Obtaining an abnormal combination at the abnormal distribution position based on the quality chain; Extracting the quality analysis indicators involved in the lowest standard combination corresponding to the abnormal distribution position from the standard chain, and extracting an analysis model from the index - model database to analyze the abnormal combination, determining the clear sub - problems and fuzzy sub - problems corresponding to the abnormal distribution position, where the fuzzy sub - problems are between clear and unclear and there is a fuzzy probability; Obtaining a first indicator for comparison with the clear sub - problem and a second indicator for comparison with the fuzzy sub - problem, extracting the same indicators for the first indicator and the second indicator, and determining whether there is the same indicator; Tracing back the corresponding clear sub - problem according to the first indicator under the same indicator to determine the first proportion problem, and at the same time, tracing back the corresponding fuzzy sub - problem according to the second indicator under the same indicator to determine the second proportion problem; Conducting a correlation analysis on the first proportion problem and the second proportion problem to determine the dependency and correlation relationship of the corresponding second indicator on the first indicator; Based on the dependency relationship and combined with fuzzy probability, the fuzzy sub - problems under the corresponding second index are clarified, and the quality problems of the abnormal distribution positions are obtained by combining the sub - problems under the remaining independent indexes.

[0077] In this embodiment, an abnormal combination is a set composed of multiple abnormal factors at a specific abnormal distribution position (such as a certain process or node). For example, the abnormal combination in the tire assembly process may include "insufficient bolt torque" (abnormal factor A), "unequal tire pressure" (abnormal factor B), etc.

[0078] The standard chain is a structured data chain that stores the lowest quality standards for each link. Corresponding to each node of the quality chain, for example, the standard chain nodes in the tire assembly process include the lowest standards such as "bolt torque ≥ 100N·m" and "tire pressure 2.2 ± 0.1bar".

[0079] The quality analysis index is a quantitative parameter used to evaluate abnormalities, such as yield rate, failure rate, temperature deviation, etc.

[0080] A clear sub - problem is an abnormal problem whose cause can be directly determined by an index without ambiguity. For example, if the measured value of the bolt torque is 80N·m (lower than the standard of 100N·m), it is directly determined as "insufficient torque".

[0081] A fuzzy sub - problem is an abnormal problem with multiple possible causes that requires probability judgment, and the uncertainty is represented by fuzzy probability. For example, the tire pressure fluctuates between 2.0 - 2.4bar (standard 2.2 ± 0.1bar), and the possible causes include "inflation equipment error" (probability 0.6) or "tire airtightness problem" (probability 0.4).

[0082] The first index and the second index respectively correspond to the analysis indexes of clear sub - problems and fuzzy sub - problems. For example, the first index "bolt torque" corresponds to a clear sub - problem; the second index "pressure fluctuation range" corresponds to a fuzzy sub - problem.

[0083] Dependency relationship: The causal probability relationship between fuzzy sub - problems and clear sub - problems. For example, "inflation equipment error" (fuzzy sub - problem) depends on "pressure sensor calibration deviation" (clear sub - problem) by 70%.

[0084] The first proportion problem is the contribution proportion of each cause in the clear sub - problem. For example, in the problem of insufficient bolt torque, "tool calibration error" accounts for 60% and "worker operation error" accounts for 40%.

[0085] The second proportion problem is the probability distribution of each possible cause in the fuzzy sub - problem. For example, in the problem of tire pressure fluctuation, the probability of "inflation equipment error" is 0.6 and the probability of "tire airtightness problem" is 0.4.

[0086] In this embodiment, the index - model database contains analysis models under different index combinations. For example, the regression model is preset and stored in this database. For instance, by calling the "solder paste thickness - printing pressure" regression model (Y = 0.05X + 0.1) in the database, the relationship between thickness and pressure is analyzed.

[0087] In this embodiment, an index comparison table for the clear sub - problem (C) and the fuzzy sub - problem (F) is constructed, and common indices (Icommon = C ∩ F) are identified based on the comparison index set.

[0088] Clear sub - problem C: "Solder paste thickness deviation" (indices: thickness, printing pressure).

[0089] Fuzzy sub - problem F: "High solder joint void rate" (indices: thickness, reflow temperature).

[0090] Operation: By comparison, the common index "solder paste thickness" is obtained.

[0091] Result: It is determined that "solder paste thickness" is the same index, and its influence on the two problems needs to be further analyzed.

[0092] In this embodiment, a causal graph model (such as a structural equation model) is constructed to quantify the dependence relationship between indices.

[0093] Calculate the conditional probability P(F∣C) to determine the dependence strength of the fuzzy sub - problem on the clear sub - problem.

[0094] Collect 100 groups of data. It is found that the occurrence rate of F when C occurs is 80%, and the occurrence rate of F when C does not occur is 20%. Calculate P(F∣C)=0.8, that is, the solder joint void is 80% dependent on the solder paste thickness deviation.

[0095] Result: Establish a dependence association relationship F0.8C.

[0096] In this embodiment, based on the dependence relationship, the fuzzy sub - problem is decomposed into: Dependent part: Fdependent = P(F∣C)×F.

[0097] Independent part: Findependent = F−Fdependent.

[0098] For the dependent part, directly associate with the cause of the clear sub - problem; for the independent part, analyze it separately.

[0099] The beneficial effects of the above technical solution are as follows: Through index association and probability reasoning, fuzzy problems are transformed into traceable clear reasons, relying on association analysis to directly point to key indicators, reducing the trial-and-error cost, and the established index dependence relationship can be used to predict potential problems.

[0100] The present invention provides a quality control method for a supply chain integrating a cloud platform and an industrial Internet. Based on the quality problems in each supply link, control optimization strategies for the corresponding supply links are determined, including: Perform aggregation processing on the quality problems in the supply link, and perform a normal distribution on the influence range of the quality problems involved in each aggregation result to obtain a concentrated influence range; Extract the quality problems under any maximum influence range in each aggregation result to construct an auxiliary function to be optimized. At the same time, extract the quality problems that best match the corresponding concentrated influence range from each aggregation result to construct a main function to be optimized; Solve the main function to be optimized and the auxiliary function to be optimized to obtain a quality optimization binary group for each aggregation result; Use the range variance group under the normal distribution in each aggregation result as a reference index, and match the analysis strategy for the corresponding aggregation result from the aggregation quality type-index-strategy comparison table; Input the binary group into the analysis strategy to obtain an optimal function, and transform the optimal function to obtain a control optimization strategy.

[0101] In this embodiment, the K-Means clustering algorithm is used to aggregate quality problems according to "process + type", and the influence range of each type of problem (such as the number of product batches and customers affected) is statistically analyzed. Python is used to fit the normal distribution, and μ and σ are output. For example, 5 categories such as "screen problems" (200 pieces) and "main board problems" (300 pieces) are aggregated. The number of orders affected by the "screen problems" is fitted, and μ = 95, σ = 15, and the normal distribution R2 = 0.92 (high fitting degree). The result obtained is a clear distinction of the concentrated influence boundary of the problem (68% of the problems affect 80 - 110 orders).

[0102] In this embodiment, the auxiliary function is to screen the instance with the largest influence range in the aggregated problems (such as the screen bubbles affecting 300 orders in a certain batch), define the goal as "contain the spread within 12 hours", and the variables as "temporary sampling frequency" and "manual rework input", and construct a linear function.

[0103] The main function is to select a typical instance with an influence range close to μ (such as the regular bubbles affecting 95 orders), define the goal as "reduce the defective rate from 5% to 1% within 30 days", and the variables as "equipment upgrade budget" and "process adjustment range", and construct a non-linear function (including equipment cost and yield correlation terms).

[0104] In this embodiment, the genetic algorithm (GA) or the gradient descent method is used to solve two functions to find the optimal solution.

[0105] Calculate the range variance group of the aggregation problem (such as the screen problem variance σ2 = 225), and query the comparison table to match the strategy (large variance → select the "quick response + fluctuation control" strategy, such as the PDCA cycle).

[0106] Input the binary group into the policy model, and use the decision tree algorithm to generate the optimal execution path (such as "perform urgent sampling inspection on the 1st - 3rd day, start equipment evaluation on the 4th day, and determine the upgrade plan on the 10th day").

[0107] Input the binary group into the PDCA policy model, and the output is: P (Plan): Develop the sampling inspection process and equipment upgrade plan within 3 days.

[0108] D (Do): Start sampling inspection on the 4th day and pilot the new equipment on the 15th day.

[0109] C (Check): Evaluate the sampling inspection effect and equipment yield rate on the 20th day.

[0110] A (Act): Optimize the plan on the 25th day and promote it comprehensively on the 30th day.

[0111] Result: The control strategy can be quantified to days and has strong implementability.

[0112] In this embodiment, for the quality optimization binary group, for example, after solving two functions, the combination of "short - term response plan + long - term radical solution" is obtained (such as "increase the full - inspection link (auxiliary) + upgrade the laminating equipment (main)").

[0113] In this embodiment, the range variance group describes the influence dispersion degree of different aggregation problems in the normal distribution using the variance group (large variance → large influence fluctuation).

[0114] The aggregation quality type - index - strategy comparison table is a predefined rule library, which associates problem types (such as assembly type / welding type), indexes (variance group) with analysis strategies (such as Six Sigma, PDCA).

[0115] The beneficial effects of the above - mentioned technical solution are: Through the closed - loop of aggregation - quantification - function solving - strategy matching, the quality problems are upgraded from scattered responses to systematic optimization, providing a quantifiable and executable technical path for quality control.

[0116] 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 claims of the present invention and their equivalent technologies, the present invention also intends to include these changes and modifications.

Claims

1. A quality control method for a supply chain integrating a cloud platform and industrial Internet, characterized in that, Including: Step 1: According to the supply modes of each supply link in the supply chain and by using industrial Internet technology, deploy a collection network to the corresponding supply link, and transmit the collection data of each supply link to the cloud platform; Step 2: Construct a new standard set according to the current supply demand of each supply link and the mapping relationship between the collection network and the preset risk items under the corresponding supply link; Step 3: Based on the cloud platform, analyze the collection data of the corresponding supply link according to the new standard set, construct the quality chain of the corresponding supply link, and compare and analyze it with the standard chain to determine quality problems. Among them, the standard chain is constructed based on the minimum requirements of each new standard in the corresponding new standard set; Step 4: Determine the control optimization strategy for the corresponding supply link based on the quality problems of each supply link for optimized deployment. At the same time, based on the cloud platform, feedback the quality problems to the management terminals of each supply link for reminder.

2. The quality control method for the supply chain integrating the cloud platform and industrial Internet according to claim 1, wherein Construct a new standard set according to the current supply demand of each supply link and the mapping relationship between the collection network and the preset risk items under the corresponding supply link, including: Extract the business environment of each supply link in the supply chain under different historical supply demands from the historical database, analyze the business environment to obtain the risk values of each preset risk item, and perform cluster analysis on the business environment to construct several risk matrices; Perform standardization and normalization processing on each risk matrix to obtain the eigenvectors of each processed matrix, where the eigenvectors contain the characteristic coefficients of each preset risk item; Sort the characteristic coefficients in each eigenvector from large to small to obtain a magnitude vector; Based on the first element in the magnitude vector, sequentially and continuously select three elements to map to a preset coordinate system for three-point connection drawing to determine the drawing shape; Rely on the drawing shape and combine with the center point of the drawing shape to assign a first label to the first element among the three continuously selected elements; Extract the labels under the same preset risk item from all the first labels as the second labels, and analyze the shape stability and numerical stability of the corresponding preset risk item based on all the second labels; According to the drawing shape and center point corresponding to each preset risk item under different risk matrices, determine the first bias based on the shape stability and the second bias based on the numerical stability; When the set standard of the preset risk item under the corresponding supply demand does not match the first bias and the second bias, rely on the first bias and the second bias to optimize the corresponding set standard once to obtain a new standard; Match the current supply demand of each supply link with all the historical demands of the same supply link. If there is a completely matching historical demand, obtain the new standard set according to the mapping relationship between the collection network and the preset risk items under the corresponding supply link; Otherwise, regard the historical demand with the largest matching coefficient that is not completely matched as the reference demand; Rely on the demand difference between the reference demand and the current supply demand, input the demand difference into the demand analysis model to obtain the bias coefficient of each preset risk item, and perform secondary optimization on the new standard based on the reference demand to obtain the new standard set.

3. The quality control method of the supply chain integrating the cloud platform and industrial Internet according to claim 2, characterized in that Optimize the corresponding set standard once based on the first bias and the second bias, including: Obtain the first mismatch factor between the first bias and the set standard under the corresponding supply demand and the second mismatch factor between the second bias and the set standard respectively; Determine the optimization direction for the corresponding set standard according to the first mismatch factor and the second mismatch factor, wherein the optimization direction is the standard improvement direction, the standard reduction direction and the standard unchanged direction; Adjust the corresponding data set standard according to the optimization direction.

4. The quality control method of the supply chain integrating the cloud platform and industrial Internet according to claim 2, wherein Perform a secondary optimization on the new standard based on the reference demand, including: Regard the new standard of the reference demand as the first standard; Obtain the bias coefficient of each preset risk item, and perform a secondary optimization on the corresponding first standard in combination with the item weight of the preset risk item based on the reference demand.

5. The quality control method for the supply chain integrating the cloud platform and industrial Internet according to claim 1, characterized in that, Construct a quality chain for the corresponding supply link, including: Compare and analyze each new standard in the new standard set with the corresponding sub-data in the collected data of the corresponding supply link to determine the initial abnormal factor, and construct an initial vector in combination with the setting position of the corresponding new standard in the corresponding supply link, and input the initial vector into the vector analysis model to obtain the output representation based on the corresponding new standard, wherein the output representation includes: the abnormal distribution and the degree of abnormality at each abnormal distribution position; Obtain the blank chain of the corresponding supply link from the chain database; Input the output representation of each new standard under the corresponding supply link into the blank chain in turn to obtain the quality chain.

6. The quality control method for the supply chain integrating the cloud platform and industrial Internet according to claim 5, wherein, Determine the initial abnormal factor, including: ； Among them, represents the initial anomaly factor; represents the quantization result of the corresponding sub-data; represents the quantization threshold of the corresponding new standard; represents the arctangent function; represents the set weight of the corresponding new standard; represents the Bernoulli random variable; represents the sign function; represents the contrast sensitivity coefficient.

7. The quality control method for the supply chain integrating the cloud platform and industrial Internet according to claim 5, characterized in that, Compare and analyze with the standard chain to determine the quality problem, including: Obtain the abnormal combination at the abnormal distribution position based on the quality chain; Extract the quality analysis indicators involved in the lowest standard combination at the corresponding abnormal distribution position from the standard chain, and extract the analysis model from the index-model database to analyze the abnormal combination to determine the clear sub-problem and the fuzzy sub-problem at the corresponding abnormal distribution position, wherein the fuzzy sub-problem is between clear and unclear and there is a fuzzy probability; Obtain the first indicator corresponding to the clear sub-problem and the second indicator corresponding to the fuzzy sub-problem, perform the same-index extraction on the first indicator and the second indicator, and judge whether there is the same indicator; Trace back the corresponding clear sub-problem according to the first indicator under the same indicator to determine the first proportion problem, and at the same time, trace back the corresponding fuzzy sub-problem according to the second indicator under the same indicator to determine the second proportion problem; Perform a correlation analysis on the first proportion problem and the second proportion problem to determine the dependence and correlation relationship of the corresponding second indicator on the corresponding first indicator; Clarify the fuzzy sub-problem under the corresponding second indicator according to the dependence and correlation relationship and in combination with the fuzzy probability, and combine the sub-problems under the remaining independent indicators to obtain the quality problem at the abnormal distribution position.

8. The quality control method for the supply chain integrating the cloud platform and industrial Internet according to claim 1, characterized in that Determine the control and optimization strategy for the corresponding supply link based on the quality problems of each supply link, including: Perform an aggregation process on the quality problems of the supply link, and perform a normal distribution on the influence range of the quality problems involved in each aggregation result to obtain the concentrated influence range; Construct an auxiliary function to be optimized by extracting quality problems under any maximum influence range in each aggregation result. At the same time, extract the quality problems that best match the corresponding concentrated influence range from each aggregation result to construct the main function to be optimized; Solve the main function to be optimized and the auxiliary function to be optimized to obtain the quality optimization binary group of each aggregation result; Use the range variance group under the normal distribution in each aggregation result as a reference index, and match the analysis strategy corresponding to the aggregation result from the aggregation quality type-index-strategy comparison table; Input the binary group into the analysis strategy to obtain the optimal function, and transform the optimal function to obtain the control optimization strategy.

Images