A product quality sampling inspection method, system and medium based on industrial Internet of Things

Abstract

The present invention provides a product quality sampling method, system and medium based on the industrial Internet of Things, and relates to the technical field of product quality management. The method includes: obtaining model data, production data and equipment data, determining a first preset condition, a first sampling parameter, a second preset condition and a second sampling parameter; obtaining production feature data, in response to the production feature data satisfying the first preset condition: generating a first sampling instruction based on the first sampling parameter, and obtaining a first quality inspection result; in response to the historical sampling result and / or the historical model data satisfying the second preset condition: generating a second sampling instruction based on the second sampling parameter, and obtaining a second quality inspection result; based on the first quality inspection result and the second quality inspection result, adjusting the first preset condition, the first sampling parameter and the second preset condition and the second sampling parameter. The present invention can intelligently adjust the conditions and / or intensity of the sampling inspection to improve the quality inspection efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of product quality management, and in particular to a product quality sampling method, system and medium based on industrial Internet of Things. Background Art

[0002] In complex production and / or assembly processes, the production and assembly of parts of various types, specifications and materials are involved. These parts have different functions in the overall product and have different requirements for quality control standards. In view of this situation, differentiated sampling standards and strengths need to be implemented for different parts.

[0003] Therefore, it is necessary to provide a product quality sampling method, system and medium based on the industrial Internet of Things to conduct targeted quality inspections on products or workpieces on the production line, intelligently adjust the conditions and / or intensity of the sampling inspections, and improve the quality inspection efficiency. Summary of the invention

[0004] One or more embodiments of the present specification provide a product quality sampling method based on the Industrial Internet of Things, the method is executed by an Industrial Internet of Things management platform of a product quality sampling system based on the Industrial Internet of Things, the system includes an Industrial Internet of Things user platform, an Industrial Internet of Things service platform, the Industrial Internet of Things management platform, an Industrial Internet of Things sensor network platform and an Industrial Internet of Things perception control platform; the Industrial Internet of Things management platform is communicatively connected with the Industrial Internet of Things sensor network platform and the Industrial Internet of Things service platform; the Industrial Internet of Things perception control platform includes production equipment of each production link on the production line, sampling devices and quality inspection equipment deployed on the production line, and realizes data interaction with the Industrial Internet of Things management platform through the Industrial Internet of Things sensor network platform. The method includes: obtaining model data through a data center, collecting production data and equipment data through the industrial Internet of Things perception control platform and obtaining them through the industrial Internet of Things sensor network platform; determining a first preset condition, a first sampling parameter, a second preset condition and a second sampling parameter based on the model data, the production data and the equipment data; collecting production feature data through the industrial Internet of Things perception control platform and obtaining them through the industrial Internet of Things sensor network platform, and in response to the production feature data satisfying the first preset condition: generating a first sampling instruction based on the first sampling parameter, and sending the first sampling instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform, so as to control the sampling device to extract and transport the first sample to the quality inspection equipment based on the first sampling instruction. ; Obtain a first quality inspection result from the quality inspection equipment through the industrial Internet of Things sensor network platform; Obtain historical sampling results and historical model data through the data center, and in response to the historical sampling results and / or the historical model data satisfying the second preset condition: Based on the second sampling parameter, generate a second sampling instruction, and send the second sampling instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the second sample to the quality inspection equipment based on the second sampling instruction; Obtain a second quality inspection result from the quality inspection equipment through the industrial Internet of Things sensor network platform; and Based on the first quality inspection result and the second quality inspection result, adjust the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter.

[0005] One or more embodiments of the present specification provide a product quality sampling system based on the Industrial Internet of Things, the system comprising an Industrial Internet of Things user platform, an Industrial Internet of Things service platform, an Industrial Internet of Things management platform, an Industrial Internet of Things sensor network platform and an Industrial Internet of Things perception control platform, the Industrial Internet of Things management platform comprising a data center; the Industrial Internet of Things management platform is communicatively connected with the Industrial Internet of Things sensor network platform and the Industrial Internet of Things service platform; the Industrial Internet of Things perception control platform comprises production equipment of each production link on the production line, a sampling device and quality inspection equipment deployed on the production line, and realizes data interaction with the Industrial Internet of Things management platform through the Industrial Internet of Things sensor network platform; the sampling device is configured to extract a workpiece as a sample and transmit the sample to the quality inspection equipment, and the quality inspection equipment is configured to perform quality inspection on the sample; the Industrial Internet of Things management platform is configured to execute the above-mentioned product quality sampling method based on the Industrial Internet of Things.

[0006] One or more embodiments of the present specification provide a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the method described in the above embodiments.

[0007] The beneficial effects of the present invention include but are not limited to: (1) By acquiring a large amount of production-related data, targeted sampling conditions and sampling parameters are generated. By judging whether different sampling conditions are met, the workpiece can be more comprehensively sampled, and based on the quality inspection results, the sampling conditions and sampling parameters can be intelligently adjusted to effectively improve the quality inspection efficiency and reduce labor costs; (2) By judging whether the monitoring data meets the evaluation parameters, the content of the second preset condition can be enriched, and then a more comprehensive judgment can be made on whether to conduct sampling. At the same time, the factor determination model is used to determine the influencing factor data, and then the evaluation parameters are determined, so that more accurate evaluation parameters can be determined; (3) The influencing factor data can be used to determine the monitoring device that needs to be turned on, the monitoring frequency of the monitoring device, and the amount of data collected, so that more monitoring data that has a more significant impact on the quality of the workpiece can be collected, thereby improving the effectiveness of the monitoring data and facilitating an accurate judgment on whether to conduct sampling. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] This specification will be further described in the form of exemplary embodiments, which will be described in detail by the accompanying drawings. These embodiments are not restrictive, and in these embodiments, the same number represents the same structure, wherein:

[0009] Figure 1 It is a schematic diagram of the platform structure of a product quality sampling inspection system based on the industrial Internet of Things according to some embodiments of this specification;

[0010] Figure 2is an exemplary flow chart of a product quality sampling method based on the industrial Internet of Things according to some embodiments of this specification;

[0011] Figure 3 is an exemplary schematic diagram of determining a model according to factors shown in some embodiments of this specification;

[0012] Figure 4 is an exemplary flow chart for determining a first quality inspection instruction and a second quality inspection instruction according to some embodiments of this specification;

[0013] Figure 5 It is an exemplary schematic diagram of the coverage model shown in some embodiments of this specification. DETAILED DESCRIPTION

[0014] In order to more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly introduces the drawings required for describing the embodiments. The drawings do not represent all implementation methods.

[0015] It should be understood that the "system", "device", "unit" and / or "module" used herein are a method for distinguishing different components, elements, parts, portions or assemblies at different levels. If other words can achieve the same purpose, the words can be replaced by other expressions.

[0016] Unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0017] When the operations are performed according to the step description in the embodiments of this specification, unless otherwise specified, the order of the steps is interchangeable, the steps may be omitted, and other steps may be included in the operation process.

[0018] Figure 1 It is a schematic diagram of the platform structure of a product quality sampling inspection system based on the industrial Internet of Things as shown in some embodiments of this specification.

[0019] In some embodiments, Figure 1 As shown, the product quality sampling inspection system 100 based on the industrial Internet of Things includes an industrial Internet of Things user platform 110, an industrial Internet of Things service platform 120, an industrial Internet of Things management platform 130, an industrial Internet of Things sensor network platform 140 and an industrial Internet of Things perception control platform 150.

[0020] The industrial Internet of Things user platform 110 refers to a platform for interacting with users. In some embodiments, the industrial Internet of Things user platform is configured as a terminal device, etc. The terminal device includes a tablet computer, a mobile phone, etc.

[0021] The industrial Internet of Things service platform 120 refers to a platform that provides information and services to users. In some embodiments, the industrial Internet of Things service platform is configured on a server. The industrial Internet of Things service platform can interact with the industrial Internet of Things user platform.

[0022] The industrial Internet of Things management platform 130 refers to a platform for managing information and / or data related to the industrial Internet of Things-based product quality sampling inspection system 100. In some embodiments, the industrial Internet of Things management platform is connected to the industrial Internet of Things sensor network platform and the industrial Internet of Things service platform. The industrial Internet of Things management platform includes a processor and a storage device, etc. The storage device stores a data center.

[0023] The data center stores information and / or data related to the product quality sampling inspection system 100 based on the industrial Internet of Things in advance. For example, model data, etc. The data center includes a variety of databases. The database refers to a database management system that supports high concurrent access.

[0024] The processor includes a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction processor (ASIP), etc. or any combination thereof.

[0025] In some embodiments, the industrial Internet of Things management platform is configured to perform a product quality sampling method based on the industrial Internet of Things. For more information, see Figures 2 to 5 Related description.

[0026] The IIoT sensor network platform 140 is a platform for integrated management of sensor information. In some embodiments, the IIoT sensor network platform is configured as a communication network or a gateway, etc. The IIoT sensor network platform can interact with the IIoT management platform and the IIoT perception control platform.

[0027] The IIoT perception control platform 150 refers to a functional platform for generating perception information and executing control information. In some embodiments, the IIoT perception control platform includes production equipment at each production link on the production line, sampling devices deployed on the production line, multiple monitoring devices, and quality inspection equipment, and realizes data interaction with the IIoT management platform through the IIoT sensor network platform.

[0028] The monitoring device refers to a device used to obtain monitoring data. In some embodiments, the monitoring device includes a thermometer, a hygrometer, an air quality monitor, and a pressure gauge. The monitoring device is deployed at any feasible position on the production line.

[0029] Production equipment refers to equipment for performing production activities. Production activities include producing or assembling workpieces, etc. In some embodiments, the workpiece includes a single part or a component assembled from multiple parts, etc. Production equipment includes automatic assembly machines, etc.

[0030] For the sake of convenience, the following description takes assembly as an example of production activity.

[0031] The sampling device refers to a device for extracting a workpiece as a sample and transmitting the sample to a quality inspection device. In some embodiments, the sampling device includes a robotic arm, a conveyor belt, and the like.

[0032] Quality inspection equipment refers to equipment used to perform quality inspection on samples. In some embodiments, the quality inspection equipment includes a material testing machine, etc.

[0033] For detailed description of the above, please refer to Figures 2 to 5 Related description.

[0034] In some embodiments of this specification, a product quality sampling inspection system based on the industrial Internet of Things can form an information operation closed loop between various functional platforms, coordinate and operate regularly, and realize the informatization and intelligence of product quality sampling inspection.

[0035] Figure 2 This is an exemplary flow chart of a product quality sampling method based on the industrial Internet of Things according to some embodiments of this specification. Figure 2 As shown, the process 200 includes the following steps. In some embodiments, the process 200 can be executed by an industrial Internet of Things management platform (hereinafter referred to as the management platform).

[0036] Step 210, obtain model data through the data center, collect production data and equipment data through the industrial Internet of Things perception control platform and obtain production data and equipment data through the industrial Internet of Things sensor network platform.

[0037] For more information about the product quality sampling system based on the Industrial Internet of Things and various platforms, see Figure 1 and its related description.

[0038] Model data refers to data related to the model of the workpiece. In some embodiments, the model data includes the type, model, etc. of the workpiece.

[0039] Production data refers to data related to the operation of the production process of the production line. In some embodiments, the production data includes the production rate of the production line, etc. The production rate refers to the rate at which a workpiece is assembled by the production line.

[0040] Equipment data refers to data related to production equipment. In some embodiments, the equipment data includes at least one of equipment type, current operating power, rated parameters, equipment operating time, equipment maintenance cycle, and equipment failure records. Rated parameters include the maximum operating power and minimum operating power of production equipment. Equipment maintenance cycle and equipment failure records are entered into the industrial Internet of Things perception control platform by technicians after they perform maintenance and troubleshooting on production equipment.

[0041] Step 220, based on the model data, the production data and the equipment data, determine the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter.

[0042] The first preset condition is used to determine whether to perform a first type of random inspection. The first type of random inspection refers to a random inspection performed by the sampling device controlled by the management platform in response to the first preset condition being met.

[0043] In some embodiments, the first preset condition includes that the accumulated production time meets the sampling inspection period and / or the accumulated production quantity reaches the sampling inspection threshold, etc.

[0044] Sampling parameters refer to parameters related to sample extraction. Samples refer to workpieces used for quality inspection. In some embodiments, sampling parameters include sampling quantity and sampling interval, etc. Among them, sampling quantity refers to the number of samples extracted during a single round of sampling inspection. Sampling interval refers to the number of workpieces between adjacent samples during a single round of sampling inspection.

[0045] The first sampling parameters refer to the sampling parameters corresponding to the first type of random inspection.

[0046] The second preset condition is used to determine whether to conduct the second type of random inspection.

[0047] The second type of random inspection refers to the random inspection performed by the sampling device controlled by the management platform in response to the second preset condition being met.

[0048] In some embodiments, the second preset condition includes at least one of the number of unqualified products in the historical sampling results reaching a quantity threshold, and / or the historical model data is different from the model data.

[0049] In some embodiments, the second preset condition also includes multiple feasible conditions such as the monitoring data not satisfying the evaluation parameters. In response to one or more of the multiple second preset conditions being satisfied, the management platform determines to perform the second type of random inspection. For an explanation of the monitoring data not satisfying the evaluation parameters, see Figure 3 and its related description.

[0050] The second sampling parameters refer to the sampling parameters corresponding to the second type of random inspection.

[0051] In some embodiments, the management platform obtains the first preset condition and the first sampling parameter by querying the initial first condition and the initial first parameter corresponding to the model data in the model table based on the model data, production data and equipment data, and then adjusting the initial first condition and the initial first parameter according to the first adjustment direction based on the production data and the equipment data.

[0052] The model table is pre-set based on historical experience, and includes a plurality of model data and an initial first condition, an initial first parameter, an initial second condition, and an initial second parameter corresponding to each model data.

[0053] The first adjustment direction is preset based on historical experience. For example, the faster the production rate, the longer the equipment operation time or the longer the equipment maintenance cycle, the larger the adjustment sampling quantity and the smaller the sampling interval, the shorter the adjustment sampling cycle and the smaller the sampling threshold and quantity threshold.

[0054] In some embodiments, the method by which the management platform determines the second preset condition and the second sampling parameter is similar to the method for determining the first preset condition and the first sampling parameter described above.

[0055] Step 230: Collect production feature data through the industrial Internet of Things perception control platform and obtain production feature data through the industrial Internet of Things sensor network platform.

[0056] Production characteristic data refers to data related to the operation of the production line.

[0057] In some embodiments, the production feature data includes the cumulative production time and cumulative production quantity of the production line. The cumulative production time refers to the total time for the production line to produce. The cumulative production quantity refers to the total number of workpieces produced by the production line.

[0058] Step 240, in response to the production characteristic data satisfying the first preset condition: based on the first sampling parameter, generate a first sampling instruction, and obtain a first quality inspection result from the quality inspection equipment through the industrial Internet of Things sensor network platform.

[0059] The first sampling instruction is an instruction for instructing the sampling device to perform the first type of sampling. In some embodiments, the first sampling instruction includes a robot arm instruction and a conveyor belt instruction. The robot arm instruction is used to instruct the robot arm to extract the sample. The conveyor belt instruction is used to instruct the conveyor belt to transport the sample to the quality inspection device.

[0060] The robot arm instructions include sampling interval and sampling quantity, etc. The conveyor belt instructions include the conveyor belt transportation rate, etc. The conveyor belt transportation rate is preset based on historical experience.

[0061] In some embodiments, the management platform sends the first sampling instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the first sample to the quality inspection equipment based on the first sampling instruction.

[0062] The first sample refers to the sample extracted by the extraction device based on the first sampling instruction.

[0063] For more information about the first sampling instruction, please refer to Figure 4 Related instructions.

[0064] The first quality inspection result refers to the result obtained by the quality inspection device for the first sample. In some embodiments, the first quality inspection result includes the qualified rate of the first sample, etc.

[0065] In some embodiments, the management platform obtains the first quality inspection result from the quality inspection equipment through the industrial Internet of Things sensor network platform.

[0066] Step 250, obtaining historical sampling results and historical model data through the data center.

[0067] The historical sampling results refer to the quality inspection results in the historical sampling process. The historical quality inspection results include the historical first quality inspection results and the historical second quality inspection results. For the description of the second quality inspection result, please refer to step 260 and its related description.

[0068] Historical model data refers to the model data during historical sampling.

[0069] Step 260, in response to the historical sampling results and / or historical model data satisfying the second preset condition: based on the second sampling parameter, generate a second sampling instruction, and obtain a second quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform.

[0070] The second sampling instruction is an instruction for instructing the sampling device to perform a second type of sampling. In some embodiments, the second sampling instruction contains the same data type as the first sampling instruction, and the process of determining the second sampling instruction is similar to the process of determining the first sampling instruction.

[0071] In some embodiments, the management platform sends the second sampling instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the second sample to the quality inspection equipment based on the second sampling instruction.

[0072] The second sample refers to the sample extracted by the extraction device based on the second sampling instruction.

[0073] For more information about the second sampling instruction, please refer to Figure 4 Related instructions.

[0074] The second quality inspection result refers to the result obtained by the quality inspection device for the second sample. In some embodiments, the second quality inspection result includes the qualified rate of the second sample, etc.

[0075] In some embodiments, the management platform obtains the second quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform.

[0076] In some embodiments, the management platform independently determines whether the first preset condition or the second preset condition is met. To prevent the time interval between multiple rounds of random inspections from being too short, the management platform can determine whether the time interval between two adjacent rounds of random inspections is greater than or equal to a time threshold when performing multiple rounds of random inspections. If the time interval is less than the time threshold, the management platform does not perform the random inspection with a later execution order in the two adjacent rounds of random inspections. The time threshold is positively correlated with the number of samples or the sampling interval.

[0077] Step 270: Adjust the first preset condition, the first sampling parameter, and the second preset condition and the second sampling parameter based on the first quality inspection result and the second quality inspection result.

[0078] In some embodiments, the management platform adjusts the first preset condition, the first sampling parameter, and the second preset condition and the second sampling parameter according to the second adjustment direction based on the first quality inspection result and the second quality inspection result. For example, in response to the quality inspection result of a single round of random inspection being lower than the first threshold or the quality inspection results of multiple rounds of random inspection being lower than the second threshold, the management platform makes adjustments according to the second adjustment direction. In response to the quality inspection result of a single round of random inspection being higher than the third threshold or the quality inspection results of multiple rounds of random inspection being higher than the fourth threshold, the management platform makes adjustments according to the third adjustment direction. The quality inspection result includes the first quality inspection result or the second quality inspection result. The above four thresholds are pre-set based on historical experience, for example, the size relationship is the first threshold < the second threshold < the fourth threshold < the third threshold.

[0079] The second adjustment direction is preset based on historical experience. For example, in response to the quality inspection result of a single round of random inspection being lower than the first threshold or the quality inspection results of multiple rounds of random inspection being lower than the second threshold, the number of samples is increased, the sampling interval is reduced, the random inspection cycle is shortened, and the random inspection threshold and quantity threshold are reduced.

[0080] The third adjustment direction is preset based on historical experience. For example, in response to the quality inspection result of a single round of random inspection being higher than the third threshold or the quality inspection results of multiple rounds of random inspection being higher than the fourth threshold, the number of samples is adjusted to be reduced, the sampling interval is increased, the random inspection cycle is extended, and the random inspection threshold and quantity threshold are increased. The adjustment range is negatively correlated to the size of the quality inspection result.

[0081] In some embodiments, the management platform may also adjust the above threshold each time it makes adjustments according to the second adjustment direction or the third adjustment direction. The management platform obtains the threshold adjusted accordingly by the technician. The method for adjusting the above threshold may be preset.

[0082] In some embodiments, the management platform may preset the first preset condition, the first sampling parameter, and the upper and lower limits of the second preset condition and the second sampling parameter. If the adjustment range of the management platform has reached the upper and lower limits, the adjustment is stopped.

[0083] In some embodiments of this specification, a large amount of production-related data is obtained to generate targeted sampling conditions and sampling parameters. By judging whether different sampling conditions are met, a more comprehensive sampling inspection can be performed on the workpiece, and based on the quality inspection results, the sampling conditions and sampling parameters can be intelligently adjusted to effectively improve the quality inspection efficiency and reduce labor costs.

[0084] Figure 3 It is an exemplary schematic diagram of determining a model based on the factors shown in some embodiments of this specification.

[0085] In some embodiments, the management platform obtains monitoring data through multiple monitoring devices of the industrial Internet of Things perception control platform. The second preset condition also includes that the monitoring data does not meet the evaluation parameters.

[0086] Monitoring data refers to data related to the production environment. In some embodiments, the monitoring data includes at least one of ambient temperature, ambient humidity, air impurity, and ambient air pressure. In some embodiments, the monitoring data is obtained through multiple monitoring devices and uploaded to the industrial Internet of Things perception control platform by the monitoring devices.

[0087] The evaluation parameter refers to the numerical range of the monitoring data used to determine whether to conduct a random inspection. The monitoring data does not meet the evaluation parameter means that the monitoring data exceeds the numerical range of the evaluation parameter. In some embodiments, different types of monitoring data correspond to different evaluation parameters. The evaluation parameters are pre-set according to actual needs.

[0088] In some embodiments, the management platform determines the influencing factor data 360 of the workpiece quality through the factor determination model 350 based on the workpiece feature data 310 , model data 320 , equipment data 330 and link process parameters 340 , and determines the evaluation parameters 370 based on the influencing factor data 360 .

[0089] For more information about the model data and device data, see the related description of step 210 .

[0090] Workpiece feature data refers to data related to the features of the workpiece itself. In some embodiments, the workpiece feature data includes material features, shape features, manufacturing methods, etc.

[0091] Material characteristics include the physical properties and chemical composition of the workpiece material.

[0092] Shape features include the shape and size of the workpiece.

[0093] Manufacturing methods include machining or forging.

[0094] The process parameters refer to data related to the production process of the manufacturing method. In some embodiments, the process parameters include multiple production links for implementing the manufacturing method and parameters related to the production links. For example, if the manufacturing method is forging, the process parameters include production links such as blanking and heating, as well as the amount of blanking and the heating temperature.

[0095] In some embodiments, the management platform obtains historical workpiece feature data and historical link process parameters corresponding to the model data as the workpiece feature data and link process parameters through the data center.

[0096] The influencing factor data refers to data related to factors that affect the quality of the workpiece. In some embodiments, the influencing factor data includes an influencing factor and an influencing score corresponding to the influencing factor.

[0097] In some embodiments, the influencing factors include multiple types, and each influencing factor corresponds to a type of monitoring data. It should be noted that since it may not be necessary to obtain all types of monitoring data in actual production, the type of monitoring data corresponding to the influencing factor may not exist in the currently obtained monitoring data.

[0098] The impact score is used to characterize the degree of influence of the impact factor on the workpiece quality. In some embodiments, the impact score is represented by a numerical value (0-1, etc.). The larger the numerical value, the greater the degree of influence of the impact factor on the workpiece quality.

[0099] The factor determination model refers to a model used to determine the influencing factor data. In some embodiments, the factor determination model is a machine learning model, for example, any one or combination of a multilayer perceptron (MLP) model or other custom model structures.

[0100] In some embodiments, the management platform trains the factor determination model based on the training data set by using a gradient descent method, etc. The training data set includes training samples and training labels corresponding to the training samples.

[0101] The training samples include sample workpiece feature data, sample model data, sample equipment data and sample link process parameters, and the training labels include actual influencing factor data corresponding to the training samples.

[0102] In some embodiments, the factor determination model is trained by: inputting a plurality of training samples with training labels into the initial factor determination model, constructing a loss function through the training labels and the prediction results of the initial factor determination model, iteratively updating the initial factor determination model based on the loss function, and completing the factor determination model training when the loss function of the initial factor determination model satisfies a preset condition. The preset conditions include the convergence of the loss function, the number of iterations reaching a set value, etc.

[0103] In some embodiments, the number of training samples corresponding to each sample link process parameter meets the sample quantity condition.

[0104] In some embodiments, the sample quantity condition includes that the number of training samples corresponding to each sample link process parameter is not less than the corresponding sample quantity threshold. The sample quantity threshold corresponding to each sample quantity threshold is pre-set based on the link process difficulty. The greater the link process difficulty, the greater the sample quantity threshold. For an explanation of the link process difficulty, see the relevant description below.

[0105] In some embodiments, the quality inspection results may also include the failure rate of the sample.

[0106] In some embodiments, the management platform determines the historical impact factor data as training labels based on the correlation between the historical monitoring data and the unqualified rate in the historical quality inspection results within the sample time period. The sample time period refers to the time period for collecting training samples.

[0107] In some embodiments, the management platform determines the correlation between the historical monitoring data and the unqualified rate in the historical quality inspection results through a correlation algorithm based on the fluctuation of the historical monitoring data in the sample time period and the change of the unqualified rate in the historical quality inspection results. The correlation algorithm includes SPSS, SAS, etc.

[0108] In some embodiments, in response to a correlation between a type of historical monitoring data and a non-conforming rate in a historical quality inspection result, the management platform determines that the type of historical monitoring data is a historical impact factor, and the impact score of the historical impact factor is positively correlated with the correlation. The management platform determines multiple historical impact factors and impact scores through the above method, and then obtains historical impact factor data, and uses the historical impact factor data as training labels.

[0109] In some embodiments of the present specification, the efficiency of model training can be improved by training the factor determination model with a large amount of historical data. At the same time, the versatility of the factor determination model can be improved by training the model with training samples corresponding to process parameters of different sample links.

[0110] In some embodiments, the management platform determines the evaluation parameters based on the impact factor data. For example, the management platform identifies the impact factor with an impact score greater than an impact threshold as a relevant impact factor based on the impact factor data. The impact threshold is negatively correlated to the pass rate of historical quality inspection results.

[0111] In some embodiments, the management platform determines the span of the numerical range of the evaluation parameter based on the impact score corresponding to the relevant impact factor. For example, the span of the numerical range of the evaluation parameter is negatively correlated to the impact score corresponding to the relevant impact factor. The span of the numerical range is represented by the difference between the maximum value and the minimum value of the numerical range.

[0112] In some embodiments of this specification, by judging whether the monitoring data meets the evaluation parameters, the content of the second preset condition can be enriched, thereby more comprehensively judging whether to conduct a random inspection. At the same time, the factor determination model is used to determine the influencing factor data, thereby determining the evaluation parameters, so that more accurate evaluation parameters can be determined.

[0113] In some embodiments, the management platform adjusts the evaluation parameters based on the second quality inspection result and the process difficulty of the link.

[0114] The process difficulty of a link is used to characterize the complexity of a production link. In some embodiments, the process difficulty of a link is represented by a numerical value, etc. The larger the numerical value, the higher the process difficulty of the link. For more information about the process difficulty of a link, see the relevant description of step 410.

[0115] In some embodiments, the management platform adjusts the evaluation parameters in a variety of ways based on the second quality inspection results and the process difficulty of the link. For example, in response to the second quality inspection results of a single round of random inspection being higher than the third threshold or the second quality inspection results of multiple rounds of random inspections being higher than the fourth threshold, the management platform expands the numerical range of the evaluation parameters.

[0116] For another example, the span of the numerical range of the evaluation parameter is negatively correlated to the process difficulty of the link. For the description of the third threshold and the fourth threshold, please refer to the relevant description of step 220.

[0117] In some embodiments of the present specification, the evaluation parameters are adjusted according to the second quality inspection results and the process difficulty of each link, so that the determined evaluation parameters can be more in line with the actual production situation.

[0118] In some embodiments, the management platform determines monitoring parameters based on influencing factor data and monitoring device status; and generates monitoring instructions based on the monitoring parameters, and sends the monitoring instructions to the industrial Internet of Things perception control platform to control multiple monitoring devices to perform monitoring based on the monitoring parameters.

[0119] The monitoring device status refers to the operating status of the monitoring device. For example, whether the monitoring device is turned on or off. The management platform collects the monitoring device status through the industrial Internet of Things perception control platform and obtains the monitoring device status through the industrial Internet of Things sensor network platform.

[0120] In some embodiments, the monitoring parameters include at least one of a monitoring device that needs to be turned on and a monitoring frequency and a data collection amount of the monitoring device.

[0121] In some embodiments, the management platform determines the monitoring device corresponding to the influencing factor as the monitoring device that needs to be turned on based on the influencing factor data and the monitoring device status, and determines the monitoring frequency and the amount of collected data based on the impact score of the influencing factor, wherein the monitoring frequency and the amount of collected data are positively correlated with the impact score.

[0122] In some embodiments of the present specification, the influencing factor data can be used to determine the monitoring devices that need to be turned on, the monitoring frequency of the monitoring devices, and the amount of data collected, so that more monitoring data that has a more significant impact on the quality of the workpiece can be collected, thereby improving the effectiveness of the monitoring data and facilitating accurate judgment on whether to conduct random inspections.

[0123] Figure 4 FIG. 1 is an exemplary flow chart of determining a first quality inspection instruction and a second quality inspection instruction according to some embodiments of this specification. Figure 4 As shown, the process 400 includes the following steps. In some embodiments, the process 400 may be executed by a management platform.

[0124] Step 410: Determine a first detection parameter corresponding to the first preset condition and a second detection parameter corresponding to the second preset condition.

[0125] For the first and second preconditions, see Figure 2 and its related description.

[0126] The first detection parameter refers to the parameter when the quality inspection device detects the first sample. In some embodiments, the first detection parameter includes a quality inspection plan and software and hardware allocation.

[0127] The quality inspection plan includes quality inspection content and quality inspection algorithm. The quality inspection content includes detecting whether the appearance of the workpiece is worn, whether the mechanical properties meet the standards, etc. The quality inspection content is pre-set based on actual needs.

[0128] The quality inspection algorithm refers to the algorithm used by the quality inspection equipment when conducting inspections. The quality inspection algorithm also includes the amount of data required to apply the quality inspection algorithm.

[0129] The amount of hardware and software allocated refers to the amount of hardware and software allocated to the quality inspection solution. The amount of hardware includes the number of quality inspection equipment and / or computing equipment, etc. The amount of software includes data transmission bandwidth, etc. Among them, the computing equipment is configured to perform the calculation of the quality inspection algorithm.

[0130] In some embodiments, the management platform determines the fragility of the workpiece based on the workpiece feature data, and determines the first detection parameter based on the workpiece fragility. Figure 3 Related description.

[0131] Workpiece fragility is used to characterize the degree to which a workpiece is susceptible to damage.

[0132] In some embodiments, the management platform determines the reference fragility as the workpiece fragility by querying the reference fragility corresponding to the workpiece feature data in the fragility data table based on the workpiece feature data. The fragility data table is pre-set based on historical data and includes multiple workpiece feature data and the reference fragility corresponding to each workpiece feature data. The reference fragility is obtained by a technician performing a damage experiment on the workpiece corresponding to the workpiece feature data. The damage experiment includes a stress fatigue experiment, etc.

[0133] In some embodiments, the management platform queries the reference first detection parameter corresponding to the workpiece fragility in the historical parameter table based on the workpiece fragility, and determines the reference first detection parameter as the first detection parameter. The historical parameter table is pre-set based on historical data, including multiple workpiece fragility and the reference first detection parameter and reference second detection parameter corresponding to each workpiece fragility. The management platform selects the historical production process corresponding to the workpiece fragility in the historical data, and counts the first detection parameter and the second detection parameter actually used in the historical production process into the historical parameter table.

[0134] The second detection parameter refers to the parameter used by the quality inspection equipment when detecting the second sample.

[0135] In some embodiments, the second detection parameter is similar to the first detection parameter. The method of determining the second detection parameter is similar to the method of determining the first detection parameter.

[0136] In some embodiments, the first detection parameter and the second detection parameter are positively correlated with the process difficulty of the link. For an explanation of the process difficulty of the link, see Figure 3 and its related description.

[0137] In some embodiments, the management platform determines the process difficulty of a link based on the number of link processes. For example, the more link processes there are, the higher the process difficulty of the link.

[0138] The number of process steps refers to the number of production steps for producing a workpiece. In some embodiments, the management platform obtains the number of historical process steps corresponding to the model data through the data center, and uses the number of historical process steps as the number of process steps.

[0139] In some embodiments of the present specification, the process difficulty of each link can be determined more accurately by the number of link processes, thereby improving the accuracy of determining the first detection parameter and the second detection parameter.

[0140] Step 420 , in response to the production characteristic data satisfying the first preset condition, generating a first quality inspection instruction based on the first detection parameter.

[0141] In some embodiments, the first sampling inspection instruction also includes a first quality inspection instruction.

[0142] The first quality inspection instruction refers to an instruction for controlling the quality inspection device to perform quality inspection on the first sample based on the first detection parameter.

[0143] In some embodiments, the management platform sends the first quality inspection instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform to control the quality inspection equipment to perform quality inspection on the first sample based on the first quality inspection instruction.

[0144] Step 430 , in response to the historical sampling results and / or the historical model data satisfying a second preset condition, a second quality inspection instruction is generated based on a second detection parameter.

[0145] For historical sampling results and historical model data, see Figure 2 and its related description.

[0146] The second sampling inspection instruction also includes a second quality inspection instruction.

[0147] The second quality inspection instruction refers to a control instruction for controlling the quality inspection device to perform quality inspection on the second sample based on the second detection parameter.

[0148] In some embodiments, the management platform sends the second quality inspection instruction to the industrial Internet of Things perception control platform through the industrial Internet of Things sensor network platform to control the quality inspection equipment to perform quality inspection on the second sample based on the second quality inspection instruction.

[0149] In some embodiments of the present specification, the first quality inspection instruction and the second quality inspection instruction can be used to specifically control the quality inspection equipment to perform quality inspection on workpieces corresponding to different sampling conditions, which is beneficial to improving the efficiency and accuracy of quality inspection.

[0150] Figure 5 It is an exemplary schematic diagram of the coverage model shown in some embodiments of this specification.

[0151] In some embodiments, the management platform determines a first detection accuracy based on a first detection parameter, and determines a second detection accuracy based on a second detection parameter.

[0152] The first detection accuracy refers to the accuracy of the quality inspection equipment in performing quality inspection on the first sample.

[0153] In some embodiments, the management platform constructs a quality inspection feature vector based on the first detection parameter, model data, and quality inspection equipment data, matches a reference quality inspection vector that meets the matching condition in the vector database, and determines the label corresponding to the reference quality inspection vector as the first detection accuracy. The matching condition includes the highest similarity with the quality inspection feature vector. The vector similarity is negatively correlated with the vector distance. The vector distance includes the Euclidean distance, etc.

[0154] The vector database is pre-set based on historical data, including multiple reference quality inspection vectors and labels corresponding to the reference quality inspection vectors. The management platform constructs a reference quality inspection vector based on the historical first detection parameters, historical model data, and historical quality inspection equipment data in the historical data, and uses the actual first detection accuracy as the label of the reference quality inspection vector. Among them, the actual first detection accuracy is the accuracy of the historical quality inspection results corresponding to the historical first detection parameters, and the accuracy of the historical quality inspection results is determined by manually checking the quality inspection results. For an explanation of the model data, see step 210 and its related description.

[0155] The quality inspection equipment data includes the type of quality inspection equipment, equipment maintenance cycle, etc. In some embodiments, the management platform obtains the quality inspection equipment data through a data center.

[0156] The second detection accuracy refers to the accuracy of the quality inspection of the second sample by the quality inspection device. In some embodiments, the second detection accuracy is similar to the first detection accuracy, and the method for determining the second detection accuracy is similar to the method for determining the first detection accuracy.

[0157] In some embodiments, the management platform may determine first fluctuation data based on the first quality inspection result, and determine second fluctuation data based on the second quality inspection result.

[0158] The first fluctuation data refers to data used to characterize the fluctuation of the first quality inspection result. In some embodiments, the first fluctuation data is represented by the variance of multiple first quality inspection results.

[0159] In some embodiments, the management platform calculates the first fluctuation data based on a plurality of first quality inspection results.

[0160] The second fluctuation data refers to data used to characterize the fluctuation of the second quality inspection result. In some embodiments, the second fluctuation data is similar to the first fluctuation data. The management platform calculates the second fluctuation data based on multiple second quality inspection results.

[0161] It is understandable that during a production process, multiple rounds of random inspections may occur and multiple quality inspection results may be obtained.

[0162] In some embodiments, in response to the first fluctuation data, the second fluctuation data, the first detection accuracy, and the second detection accuracy all satisfying the correction conditions, the management platform generates a plurality of candidate sampling parameters based on the first detection accuracy and the second detection accuracy.

[0163] The correction condition refers to the condition for determining whether to update the sampling parameters and the detection parameters. In some embodiments, the correction condition includes that the first fluctuation data and the second fluctuation data are both greater than the fluctuation threshold, and the first detection accuracy and the second detection accuracy are both less than the accuracy threshold. In some embodiments, the fluctuation threshold is negatively correlated with the fragility of the workpiece and the difficulty of the process. The accuracy threshold is positively correlated with the fragility of the workpiece and the difficulty of the process.

[0164] For an explanation of the fragility of the workpiece and the difficulty of the process, please refer to the relevant description of step 410.

[0165] The candidate sampling parameters refer to parameters to be determined that are related to the sampling process. In some embodiments, the candidate sampling parameters include a set of first sampling parameters and second sampling parameters and first detection parameters and second detection parameters.

[0166] In some embodiments, the management platform determines multiple candidate sampling parameters based on the first detection accuracy and the second detection accuracy. For example, the first sampling parameter and the second sampling parameter remain unchanged, the management platform randomly adjusts the first detection accuracy and the second detection accuracy, and reversely determines the first detection parameter and the second detection parameter corresponding to the adjusted first detection accuracy and the second detection accuracy through the second vector database, and determines the obtained first detection parameter and the second detection parameter and the unchanged first sampling parameter and the second sampling parameter as a candidate sampling parameter.

[0167] For another example, the management platform reversely determines the first detection parameter and the second detection parameter corresponding to the adjusted first detection accuracy and the second detection accuracy through the second vector database, and adjusts the first sampling parameter and the second sampling parameter based on the obtained first detection parameter and the second detection parameter. The management platform adjusts the sampling parameter based on the adjustment direction of the detection parameter.

[0168] Exemplarily, the adjustment direction of the detection parameters is to increase the allocation of software and hardware, and the sampling parameters are correspondingly adjusted to shorten the sampling interval and increase the number of samples.

[0169] In some embodiments, the management platform may preset an upper adjustment limit and a lower adjustment limit for the first detection parameter and the second detection parameter. If the content adjusted by the management platform has reached the upper adjustment limit or the lower adjustment limit, the adjustment is stopped.

[0170] In some embodiments, the management platform also determines a first adjustment range based on the first quality inspection result and the first fluctuation data; determines a second adjustment range based on the second quality inspection result and the second fluctuation data; and adjusts the first detection accuracy and the second detection accuracy based on the first adjustment range and the second adjustment range.

[0171] The first adjustment range refers to the range for adjusting the first detection accuracy. The second adjustment range refers to the range for adjusting the second detection accuracy.

[0172] In some embodiments, the first adjustment amplitude and the second adjustment amplitude are positively correlated with the corresponding fluctuation data and negatively correlated with the corresponding quality inspection result.

[0173] In some embodiments of the present specification, a more reasonable adjustment range can be determined through fluctuation data of quality inspection results and historical quality inspection results, thereby accurately adjusting the detection accuracy.

[0174] In some embodiments, the management platform determines the quality inspection coverage 540 of multiple candidate sampling parameters based on multiple candidate sampling parameters 510 and production process data 520 through a coverage model 530.

[0175] In some embodiments, the production process data includes the first preset condition, the second preset condition, the first quality inspection result, the second quality inspection result, the historical first quality inspection result, the historical second quality inspection result, the workpiece feature data, the model data and the equipment data. For the description of the above content, see Figure 2 and its related description.

[0176] Quality inspection coverage is used to characterize the quality inspection accuracy of the overall quality inspection results of the workpiece after random inspection based on candidate sampling parameters.

[0177] The coverage model refers to a model used to determine the quality inspection coverage. In some embodiments, the coverage model is a machine learning model, for example, any one or combination of a recurrent neural network (RNN) or other custom model structures.

[0178] In some embodiments, the management platform can train the coverage model based on a large number of quality inspection training samples and corresponding quality inspection labels by gradient descent method, etc. The quality inspection training samples include sample sampling parameters and sample production process data. The quality inspection labels include the actual quality inspection coverage corresponding to the quality inspection training samples.

[0179] In some embodiments, the management platform determines the quality inspection training samples and their quality inspection labels based on the historical data. For example, the management platform uses the historical sampling parameters and historical production process data in the historical data as the quality inspection training samples, and uses the historical quality inspection coverage as the quality inspection label.

[0180] In some embodiments, the management platform controls the quality inspection equipment to re-inspect each workpiece corresponding to the quality inspection training sample in a single historical time period of the historical production process to obtain an overall quality inspection result, compares the overall quality inspection result with the historical quality inspection result of the same historical time period in the historical production process, calculates the difference between the pass rate corresponding to the overall quality inspection result and the pass rate corresponding to the historical quality inspection result, and determines the actual quality inspection coverage corresponding to the quality inspection training sample in the historical time period based on the difference. The smaller the difference, the higher the actual quality inspection coverage corresponding to the quality inspection training sample. The management platform obtains the quality inspection coverage corresponding to multiple historical time periods, and uses the average of the quality inspection coverage as the actual quality inspection coverage corresponding to the quality inspection training sample.

[0181] In some embodiments, the training process of the coverage model is similar to the training process of the factor determination model. For instructions on training the coverage model, see Figure 3 Related description.

[0182] In some embodiments, the management platform determines a comprehensive sampling score for the candidate sampling parameters based on the quality inspection coverage.

[0183] The comprehensive score of random inspection refers to the comprehensive score of the candidate sampling parameters. The higher the comprehensive score of random inspection, the higher the efficiency or accuracy of random inspection based on the candidate sampling parameters.

[0184] In some embodiments, the management platform determines the reference comprehensive score corresponding to the quality inspection coverage as the sampling comprehensive score of the candidate sampling parameters by querying the scoring table based on the quality inspection coverage. The scoring table is pre-set based on historical data, including multiple quality inspection coverages and the reference comprehensive score corresponding to each quality inspection coverage. The reference comprehensive score is manually marked by the technician based on the actual sampling situation.

[0185] In some embodiments, the management platform performs a weighted summation of the quality inspection coverage, the single-piece quality inspection time, and the number of candidate samples in the candidate sampling parameters, and uses the summation result as the comprehensive score of the random inspection. The weight of each data in the weighted summation is pre-set based on historical experience. The management platform obtains the single-piece quality inspection time corresponding to the model data through the data center.

[0186] In some embodiments, the management platform updates the first sampling parameter and the first detection parameter as well as the second sampling parameter and the second detection parameter based on the quality inspection coverage, the comprehensive score of the random inspection and multiple candidate sampling parameters. For example, the management platform uses the candidate sampling parameter with the highest comprehensive score of the random inspection and a quality inspection coverage greater than the first coverage threshold as the updated sampling parameter, and uses each parameter in the updated sampling parameter as the new detection parameter and sampling parameter. The first coverage threshold is preset based on historical experience.

[0187] In some embodiments of the present specification, multiple candidate sampling parameters are determined, and the candidate sampling parameters are comprehensively evaluated based on their comprehensive sampling scores and quality inspection coverage, so as to determine the candidate sampling parameters that can best ensure the quality of the sampling, and update the existing sampling parameters and detection parameters to make the quality inspection results more accurate.

[0188] In some embodiments, in response to the quality inspection coverage of multiple candidate sampling parameters satisfying the screening conditions, the management platform screens multiple candidate sampling parameters to obtain target sampling parameters; and adjusts the first preset condition and the second preset condition until the target sampling parameters meet the update conditions.

[0189] The screening condition refers to a condition used to screen the target sampling parameter. In some embodiments, the quality inspection coverage of the plurality of candidate sampling parameters is not greater than the first coverage threshold.

[0190] In some embodiments, in response to the quality inspection coverage of multiple candidate sampling parameters satisfying the screening condition, the management platform removes the candidate sampling parameters whose quality inspection coverage is less than the second coverage threshold, and uses the remaining candidate sampling parameters as target sampling parameters. The first coverage threshold is greater than the second coverage threshold.

[0191] In some embodiments, the first coverage threshold and the second coverage threshold are positively correlated to the workpiece cost and the process difficulty of the link. The workpiece cost is positively correlated to the workpiece fragility.

[0192] In some embodiments, the management platform Figure 2 The method for adjusting the first preset condition and the second preset condition in the method for adjusting the first preset condition and the second preset condition is to adjust the first preset condition and the second preset condition, and re-input the adjusted first preset condition and the second preset condition and the target sampling parameter into the coverage model to determine the new quality inspection coverage corresponding to the multiple target sampling parameters respectively. If the new quality inspection coverage still meets the screening condition, the first preset condition and the second preset condition are adjusted again until the multiple target sampling parameters meet the update condition. Among them, the update condition includes that the quality inspection coverage of at least one target sampling parameter among the multiple target sampling parameters is greater than the first coverage threshold. The management platform uses the target sampling parameter with the highest comprehensive inspection score among the target sampling parameters whose quality inspection coverage is greater than the first coverage threshold as the updated sampling parameter.

[0193] In some embodiments of the present specification, by updating the first preset condition and the second preset condition, more reasonable sampling can be performed when the workpiece is subsequently sampled, which is beneficial to reducing the workload of quality inspection and reducing redundant data generated by quality inspection.

[0194] Some embodiments of the present specification also provide a computer-readable storage medium, which stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes any of the methods in the above embodiments.

[0195] Furthermore, certain features, structures or characteristics in one or more embodiments of this specification may be appropriately combined.

[0196] In some embodiments, numbers describing the number of components and attributes are used. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise specified, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, which may change according to the required features of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining the digits. Although the numerical domains and parameters used to confirm the breadth of the range in some embodiments of this specification are approximate values, in specific embodiments, the setting of such numerical values ​​is as accurate as possible within the feasible range.

[0197] If there is any inconsistency or conflict between the descriptions, definitions, and / or usage of terms in the materials cited in this specification and the contents described in this specification, the descriptions, definitions, and / or usage of terms in this specification shall prevail.

Claims

1. A product quality sampling method based on industrial Internet of Things, characterized in that: The method is performed by an industrial Internet of Things management platform of a product quality sampling inspection system based on the industrial Internet of Things, and the system includes an industrial Internet of Things user platform, an industrial Internet of Things service platform, the industrial Internet of Things management platform, an industrial Internet of Things sensor network platform and an industrial Internet of Things perception control platform; The industrial Internet of Things management platform is communicatively connected with the industrial Internet of Things sensor network platform and the industrial Internet of Things service platform; The industrial Internet of Things perception control platform includes production equipment in each production link on the production line, sampling devices and quality inspection equipment deployed on the production line, and realizes data interaction with the industrial Internet of Things management platform through the industrial Internet of Things sensor network platform; The method comprises: Acquire model data through the data center, collect production data and equipment data through the industrial Internet of Things perception control platform and obtain them through the industrial Internet of Things sensor network platform; Based on the model data, the production data and the equipment data, determine a first preset condition, a first sampling parameter, a second preset condition and a second sampling parameter; wherein the first preset condition is used to determine whether to perform a first type of sampling inspection, and the first sampling parameter refers to a sampling parameter corresponding to the first type of sampling inspection; the second preset condition is used to determine whether to perform a second type of sampling inspection, and the second sampling parameter refers to the sampling parameter corresponding to the second type of sampling inspection; The production characteristic data is collected by the industrial Internet of Things perception control platform and obtained via the industrial Internet of Things sensor network platform, and in response to the production characteristic data satisfying the first preset condition: Based on the first sampling parameter, a first sampling inspection instruction is generated, and the first sampling inspection instruction is sent to the industrial Internet of Things sensing control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the first sample to the quality inspection equipment based on the first sampling inspection instruction; Obtaining a first quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform; Acquiring historical sampling results and historical model data through the data center, in response to the historical sampling results and / or the historical model data satisfying the second preset condition: Based on the second sampling parameter, a second sampling inspection instruction is generated, and the second sampling inspection instruction is sent to the industrial Internet of Things sensing control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the second sample to the quality inspection equipment based on the second sampling inspection instruction; Obtaining a second quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform; and Adjusting the first preset condition, the first sampling parameter, the second preset condition, and the second sampling parameter based on the first quality inspection result and the second quality inspection result includes: In response to a quality inspection result of a single round of random inspection being lower than a first threshold or the quality inspection results of multiple rounds of random inspection being lower than a second threshold, the industrial Internet of Things management platform adjusts the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter according to a second adjustment direction; In response to the quality inspection result of the single round of random inspection being higher than a third threshold or the quality inspection results of multiple rounds of random inspection being higher than a fourth threshold, the industrial Internet of Things management platform adjusts the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter according to a third adjustment direction; wherein the quality inspection result includes the first quality inspection result or the second quality inspection result, and the second adjustment direction and the third adjustment direction are pre-set based on historical experience.

2. The method according to claim 1, characterized in that The industrial Internet of Things perception control platform includes multiple monitoring devices; the industrial Internet of Things management platform obtains monitoring data through the multiple monitoring devices of the industrial Internet of Things perception control platform; The second preset condition includes that the monitoring data does not satisfy the evaluation parameter; The method further comprises: Based on the workpiece feature data, the model data, the equipment data and the process parameters of each link, the influencing factor data of the workpiece quality is determined by a factor determination model, wherein the factor determination model is a machine learning model; and Based on the impact factor data, the evaluation parameters are determined.

3. The method according to claim 2, characterized in that The method further comprises: Determining monitoring parameters based on the influencing factor data and the status of the monitoring device; and, Based on the monitoring parameters, a monitoring instruction is generated, and the monitoring instruction is sent to the industrial Internet of Things perception control platform to control the multiple monitoring devices to perform monitoring based on the monitoring parameters.

4. The method according to claim 1, characterized in that The first sampling inspection instruction includes a first quality inspection instruction, the second sampling inspection instruction includes a second quality inspection instruction, and the method further includes: Determine a first detection parameter corresponding to the first preset condition and a second detection parameter corresponding to the second preset condition; In response to the production characteristic data satisfying the first preset condition: Based on the first detection parameter, generate the first quality inspection instruction, and send the first quality inspection instruction to the industrial Internet of Things perception control platform to control the quality inspection device to perform quality inspection on the first sample based on the first quality inspection instruction; and, In response to the historical sampling result and / or the historical model data satisfying the second preset condition: Based on the second detection parameter, the second quality inspection instruction is generated, and the second quality inspection instruction is sent to the industrial Internet of Things perception control platform to control the quality inspection equipment to perform quality inspection on the second sample based on the second quality inspection instruction.

5. The method according to claim 4, characterized in that The method further comprises: Determining a first detection accuracy based on the first detection parameter, and determining a second detection accuracy based on the second detection parameter; Determining first fluctuation data based on the first quality inspection result, and determining second fluctuation data based on the second quality inspection result; In response to the first fluctuation data, the second fluctuation data, the first detection accuracy, and the second detection accuracy all satisfying a correction condition: Based on the first detection accuracy and the second detection accuracy, generating a plurality of candidate sampling parameters; Based on the multiple candidate sampling parameters and the production process data, determining the quality inspection coverage of the multiple candidate sampling parameters through a coverage model, wherein the coverage model is a machine learning model; Determine a comprehensive sampling score based on the quality inspection coverage; and, Based on the quality inspection coverage, the comprehensive score of the random inspection and the multiple candidate sampling parameters, the first sampling parameters and the first detection parameters as well as the second sampling parameters and the second detection parameters are updated.

6. A product quality sampling inspection system based on industrial Internet of Things, characterized in that: The system includes an industrial Internet of Things user platform, an industrial Internet of Things service platform, an industrial Internet of Things management platform, an industrial Internet of Things sensor network platform and an industrial Internet of Things perception control platform, and the industrial Internet of Things management platform includes a data center; The industrial Internet of Things management platform is communicatively connected with the industrial Internet of Things sensor network platform and the industrial Internet of Things service platform; The industrial Internet of Things perception control platform includes production equipment in each production link on the production line, sampling devices and quality inspection equipment deployed on the production line, and realizes data interaction with the industrial Internet of Things management platform through the industrial Internet of Things sensor network platform; The sampling device is configured to extract a workpiece as a sample and transmit the sample to the quality inspection device, and the quality inspection device is configured to perform quality inspection on the sample; The industrial Internet of Things management platform is configured to: Acquire model data through the data center, collect production data and equipment data through the industrial Internet of Things perception control platform and obtain them through the industrial Internet of Things sensor network platform; Based on the model data, the production data and the equipment data, determine a first preset condition, a first sampling parameter, a second preset condition and a second sampling parameter; wherein the first preset condition is used to determine whether to perform a first type of sampling inspection, and the first sampling parameter refers to a sampling parameter corresponding to the first type of sampling inspection; the second preset condition is used to determine whether to perform a second type of sampling inspection, and the second sampling parameter refers to the sampling parameter corresponding to the second type of sampling inspection; The production characteristic data is collected by the industrial Internet of Things perception control platform and obtained via the industrial Internet of Things sensor network platform, and in response to the production characteristic data satisfying the first preset condition: Based on the first sampling parameter, a first sampling inspection instruction is generated, and the first sampling inspection instruction is sent to the industrial Internet of Things sensing control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the first sample to the quality inspection equipment based on the first sampling inspection instruction; Obtaining a first quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform; Acquiring historical sampling results and historical model data through the data center, in response to the historical sampling results and / or the historical model data satisfying the second preset condition: Based on the second sampling parameter, a second sampling inspection instruction is generated, and the second sampling inspection instruction is sent to the industrial Internet of Things sensing control platform through the industrial Internet of Things sensor network platform to control the sampling device to extract and transport the second sample to the quality inspection equipment based on the second sampling inspection instruction; Obtaining a second quality inspection result from the quality inspection device through the industrial Internet of Things sensor network platform; and Adjusting the first preset condition, the first sampling parameter, the second preset condition, and the second sampling parameter based on the first quality inspection result and the second quality inspection result includes: In response to a quality inspection result of a single round of random inspection being lower than a first threshold or the quality inspection results of multiple rounds of random inspection being lower than a second threshold, the industrial Internet of Things management platform adjusts the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter according to a second adjustment direction; In response to the quality inspection result of the single round of random inspection being higher than a third threshold or the quality inspection results of multiple rounds of random inspection being higher than a fourth threshold, the industrial Internet of Things management platform adjusts the first preset condition, the first sampling parameter, the second preset condition and the second sampling parameter according to a third adjustment direction; wherein the quality inspection result includes the first quality inspection result or the second quality inspection result, and the second adjustment direction and the third adjustment direction are pre-set based on historical experience.

7. The system according to claim 6, characterized in that The industrial Internet of Things perception control platform includes multiple monitoring devices; the industrial Internet of Things management platform obtains monitoring data through the multiple monitoring devices of the industrial Internet of Things perception control platform; the second preset condition includes that the monitoring data does not meet the evaluation parameters; The industrial Internet of Things management platform is further configured to: Based on the workpiece feature data, the model data, the equipment data and the process parameters of each link, the influencing factor data of the workpiece quality is determined by a factor determination model, wherein the factor determination model is a machine learning model; and Based on the impact factor data, the evaluation parameters are determined.

8. The system according to claim 6, characterized in that The first sampling inspection instruction includes a first quality inspection instruction, the second sampling inspection instruction includes a second quality inspection instruction, and the industrial Internet of Things management platform is further configured as follows: Determine a first detection parameter corresponding to the first preset condition and a second detection parameter corresponding to the second preset condition; In response to the production characteristic data satisfying the first preset condition: Based on the first detection parameter, generate the first quality inspection instruction, and send the first quality inspection instruction to the industrial Internet of Things perception control platform to control the quality inspection device to perform quality inspection on the first sample based on the first quality inspection instruction; as well as, In response to the historical sampling result and / or the historical model data satisfying the second preset condition: Based on the second detection parameter, the second quality inspection instruction is generated, and the second quality inspection instruction is sent to the industrial Internet of Things perception control platform to control the quality inspection equipment to perform quality inspection on the second sample based on the second quality inspection instruction.

9. The system according to claim 8, characterized in that The industrial Internet of Things management platform is further configured to: Determining a first detection accuracy based on the first detection parameter, and determining a second detection accuracy based on the second detection parameter; Determining first fluctuation data based on the first quality inspection result, and determining second fluctuation data based on the second quality inspection result; In response to the first fluctuation data, the second fluctuation data, the first detection accuracy, and the second detection accuracy all satisfying a correction condition: Based on the first detection accuracy and the second detection accuracy, generating a plurality of candidate sampling parameters; Based on the multiple candidate sampling parameters and the production process data, determining the quality inspection coverage of the multiple candidate sampling parameters through a coverage model, wherein the coverage model is a machine learning model; Based on the quality inspection coverage, determine the comprehensive score of the random inspection; as well as, Based on the quality inspection coverage, the comprehensive score of the random inspection and the multiple candidate sampling parameters, the first sampling parameters and the first detection parameters as well as the second sampling parameters and the second detection parameters are updated.

10. A computer-readable storage medium, characterized in that: The storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the method as claimed in claim 1.

Images