Workpiece Quality Detection Method, System, Device and Medium Based on Industrial Internet of Things

Abstract

The present application discloses a workpiece quality detection method, system, device and medium for the industrial Internet of Things, relating to the technical field of the Internet of Things. The method includes: obtaining surface temperature data of a workpiece according to a preset acquisition frequency, and obtaining a real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency, wherein the workpiece is placed at a preset temperature, and the surface temperature data includes real-time temperatures of surface feature points at multiple acquisition moments; determining internal temperature data of the workpiece according to the surface temperature data; determining temperature characteristic values of the workpiece at the multiple acquisition moments according to the surface temperature data and the internal temperature data, and determining current characteristic values of the workpiece according to the real-time current; and judging whether the workpiece is qualified based on a preset workpiece quality evaluation model and according to the temperature characteristic values and the current characteristic values. The present application has the effect of improving the accuracy of workpiece quality detection.

Description

Technical Field

[0001] This application relates to the technical field of industrial Internet of Things, and particularly to a workpiece quality detection method, system, device and medium based on industrial Internet of Things. Background Art

[0002] Workpiece quality detection is a crucial part in manufacturing, and its quality directly affects the performance, reliability and service life of products. In modern production processes, workpieces often need to go through multiple inspection processes to ensure they meet the design specifications and usage requirements. These processes include appearance inspection, dimensional measurement, material composition analysis, mechanical property testing, etc. By using high-precision measuring instruments and advanced detection technologies, such as optical measurement, ultrasonic testing, X-ray imaging, etc., manufacturing enterprises can effectively identify potential quality problems and reduce the risk of unqualified products.

[0003] However, in actual operation, workpiece quality detection faces many challenges. For example, factors such as the experience level of operators, the calibration status of instruments, and the stability of the detection environment may all lead to fluctuations in detection results. In addition, with the increasing complexity of product design, traditional detection methods are sometimes difficult to meet the needs of new workpieces. This limitation makes manual detection vulnerable to subjective factors and there is a possibility of missing defects.

[0004] In addition, although automated detection and intelligent detection technologies have been applied in recent years, in some cases, the application of these technologies is still restricted by equipment investment and technological maturity, resulting in the detection accuracy not reaching the expected level. Therefore, the current accuracy of workpiece quality detection is still relatively low. How to improve the reliability and precision of detection technology and reduce the interference of human factors will be the key issues that need to be focused on in workpiece quality detection. Summary of the Invention

[0005] In order to improve the accuracy of workpiece quality detection, this application provides a workpiece quality detection method, system, device and medium based on industrial Internet of Things.

[0006] In the first aspect, this application provides a workpiece quality detection method based on industrial Internet of Things, adopting the following technical solutions:

[0007] A workpiece quality detection method based on industrial Internet of Things is applied to an industrial Internet of Things system. The industrial Internet of Things system includes a management platform, a sensing network platform and an object platform that are sequentially communicatively connected. The method is executed by the management platform and includes:

[0008] Obtain the surface temperature data of the workpiece according to a preset acquisition frequency, and obtain the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency, wherein the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition times;

[0009] Determine the internal temperature data of the workpiece according to the surface temperature data, wherein the internal temperature data includes the predicted temperatures of internal feature points at the multiple acquisition times;

[0010] Determine the temperature characteristic values of the workpiece at the multiple acquisition times according to the surface temperature data and the internal temperature data, and determine the current characteristic values of the workpiece according to the real-time current, wherein the temperature characteristic values include at least one of internal average temperature, surface average temperature, overall average temperature, maximum surface temperature deviation value, minimum surface temperature deviation value, maximum internal temperature deviation value, minimum internal temperature deviation value, maximum overall temperature deviation value, minimum overall temperature deviation value, and the current characteristic values include at least one of average current, maximum current deviation value, minimum current deviation value;

[0011] Based on a preset workpiece quality assessment model, determine whether the workpiece is qualified according to the temperature characteristic values and the current characteristic values.

[0012] By adopting the above technical solution, first obtain the surface temperature data of the workpiece according to the preset acquisition frequency, and obtain the real-time current flowing through the workpiece under the preset voltage according to the preset acquisition frequency. Wherein, the workpiece is placed at the preset temperature, and the surface temperature data includes the real-time temperatures of the surface feature points at multiple acquisition times. Then determine the internal temperature data of the workpiece according to the surface temperature data. Wherein, the internal temperature data includes the predicted temperatures of the internal feature points at multiple acquisition times. Then determine the temperature characteristic values of the workpiece at multiple acquisition times according to the surface temperature data and the internal temperature data, and determine the current characteristic values of the workpiece according to the real-time current. Wherein, the temperature characteristic values include at least one of the internal average temperature, the surface average temperature, the overall average temperature, the maximum internal temperature deviation value, the minimum internal temperature deviation value, the maximum overall temperature deviation value, and the minimum overall temperature deviation value. The current characteristic values include at least one of the average current, the maximum current deviation value, and the minimum current deviation value. Finally, based on the preset workpiece quality evaluation model, determine whether the workpiece is qualified according to the temperature characteristic values and the current characteristic values; through the above method, the accuracy and intelligent level of workpiece quality detection are improved. By collecting surface temperature and current data in real time, the thermal behavior and electrical performance of the workpiece under preset conditions can be comprehensively reflected, ensuring the accuracy of the detection results. The comprehensive analysis of the temperature characteristic values includes the thermal uniformity of the interior and the surface, as well as the overall temperature deviation, which can effectively identify potential defects and non-uniformities, reducing the misjudgment rate. At the same time, the analysis of the current characteristic values provides a necessary basis for evaluating the electrical performance, strengthening the multi-dimensional characteristics of the detection. Overall, by implementing this detection scheme, not only the accuracy of workpiece quality evaluation is improved, but also data support is provided for intelligent manufacturing, contributing to a more efficient production process and higher-quality product output.

[0013] Optionally, the step of determining the internal temperature data of the workpiece according to the surface temperature data includes:

[0014] Obtain a historical temperature data set, and divide the historical temperature data set into a training set and a test set according to a preset ratio. Wherein, the historical temperature data set includes historical surface temperature data and historical internal temperature data corresponding to the historical surface temperature data;

[0015] Set the hyperparameters of the pre-constructed internal temperature generation data model according to the grid search algorithm, and use the root mean square error RMSE and the coefficient of determination R 2 as evaluation indicators;

[0016] Train the pre-constructed internal temperature generation data model according to the training set to obtain a trained internal temperature generation data model;

[0017] Test the trained internal temperature generation data model according to the test set, and judge whether the error is within a preset range according to the evaluation index. If so, use the trained internal temperature generation data model as the preset internal temperature generation data model;

[0018] Based on the preset internal temperature generation data model and according to the surface temperature data, obtain the internal temperature data of the workpiece.

[0019] By adopting the above technical solution, a historical temperature data set is obtained, and the historical temperature data set is divided into a training set and a test set according to a preset ratio. Among them, the historical temperature data set includes historical surface temperature data and historical internal temperature data corresponding to the historical surface temperature data. Then, the hyperparameters of the pre-constructed internal temperature generation data model are set according to the grid search algorithm, and the root mean square error RMSE and the coefficient of determination R 2 are used as evaluation indexes. Then, the pre-constructed internal temperature generation data model is trained according to the training set to obtain a trained internal temperature generation data model. Then, the trained internal temperature generation data model is tested according to the test set, and it is judged whether the error is within the preset range according to the evaluation index. If the error is not within the preset range, the trained internal temperature generation data model is used as the preset internal temperature generation data model. Finally, based on the preset internal temperature generation data model and according to the surface temperature data, the internal temperature data of the workpiece is obtained.

[0020] Optionally, the preset internal temperature generation data model includes a first input layer, a hidden layer, and a first output layer. The step of obtaining the internal temperature data of the workpiece based on the preset internal temperature generation data model and according to the surface temperature data includes:

[0021] Through the first input layer and according to the surface temperature data, generate a first input feature vector;

[0022] Through the hidden layer and according to the input feature vector, generate a feature extraction vector;

[0023] Through the first output layer and according to the feature extraction vector, generate a second output feature vector;

[0024] Based on the output second feature vector, generate the internal temperature data of the workpiece.

[0025] By adopting the above technical solution, in order to obtain the internal temperature data of the workpiece, through the first input layer, and based on the surface temperature data, a first input feature vector is generated. Then, through the hidden layer, and based on the input feature vector, a feature extraction vector is generated. Then, through the first output layer, and based on the feature extraction vector, a second output feature vector is generated. Finally, based on the output second feature vector, the internal temperature data of the workpiece is generated.

[0026] Optionally, the hidden layer includes a first sub-hidden layer, a second sub-hidden layer, and a third sub-hidden layer. The ratio of the number of neurons in the first sub-hidden layer, the second sub-hidden layer, and the third sub-hidden layer is 5:3:1. The step of passing through the hidden layer and generating a feature extraction vector based on the input feature vector includes:

[0027] Input the input feature vector into the first sub-hidden layer to obtain a first intermediate feature vector;

[0028] Input the first intermediate feature vector into the second sub-hidden layer to obtain a second intermediate feature vector;

[0029] Input the second intermediate feature vector into the third sub-hidden layer to obtain a feature extraction vector.

[0030] By adopting the above technical solution, in order to generate a feature extraction vector, input the input feature vector into the first sub-hidden layer to obtain a first intermediate feature vector, then input the first intermediate feature vector into the second sub-hidden layer to obtain a second intermediate feature vector, and finally input the second intermediate feature vector into the third sub-hidden layer to obtain a feature extraction vector.

[0031] Optionally, the preset workpiece quality evaluation model includes a second input layer, a first feature extraction layer, a second feature extraction layer, a splicing layer, and a second output layer. The step of judging whether the workpiece is qualified based on the preset workpiece quality evaluation model and according to the temperature feature value and the current feature value includes:

[0032] Through the second input layer, and based on the temperature feature value, generate a second input feature vector;

[0033] Through the second input layer, and based on the current feature value, generate a third input feature vector;

[0034] Through the first feature extraction layer, and based on the second input feature vector, generate a first intermediate feature vector;

[0035] Through the second feature extraction layer, and based on the third input feature vector, generate a second intermediate feature vector;

[0036] Generate a data feature tensor through the splicing layer and based on the first intermediate feature vector and the second intermediate feature vector;

[0037] Generate a second output feature vector through the second output layer and based on the data feature tensor;

[0038] Generate the qualified probability of the workpiece based on the second output feature vector, and determine whether the workpiece is qualified according to the qualified probability.

[0039] By adopting the above technical solution, in order to further determine whether the workpiece is qualified, through the second input layer, generate a second input feature vector according to the temperature feature value, then through the second input layer, generate a third input feature vector according to the current feature value, then through the first feature extraction layer, generate a first intermediate feature vector according to the second input feature vector, then through the second feature extraction layer, generate a second intermediate feature vector according to the third input feature vector, then through the splicing layer, generate a data feature tensor according to the first intermediate feature vector and the second intermediate feature vector, then through the second output layer, generate a second output feature vector according to the data feature tensor, and finally generate the qualified probability of the workpiece based on the second output feature vector, and determine whether the workpiece is qualified according to the qualified probability.

[0040] Optionally, the step of determining whether the workpiece is qualified according to the qualified probability includes:

[0041] Judge whether the qualified probability is within a preset range according to the qualified probability;

[0042] If so, obtain the number of the workpiece, and update the preset workpiece table according to the number, where the preset workpiece table includes the detected workpieces and the qualified information of the detected workpieces, and the preset workpiece table includes the mapping relationship between the detected workpieces and the qualified information of the detected workpieces;

[0043] If not, obtain the scan data of the workpiece, and judge whether the workpiece needs to be reprocessed according to the scan data.

[0044] By adopting the above technical solution, in order to perform subsequent processing according to the qualified probability, first judge whether the qualified probability is within a preset range according to the qualified probability. If the qualified probability is within the preset range, obtain the number of the workpiece, and update the preset workpiece table according to the number, where the preset workpiece table includes the detected workpieces and the qualified information of the detected workpieces, and the preset workpiece table includes the mapping relationship between the detected workpieces and the qualified information of the detected workpieces. If the qualified probability is not within the preset range, obtain the scan data of the workpiece, and judge whether the workpiece needs to be reprocessed according to the scan data.

[0045] Optionally, the step of determining whether secondary processing of the workpiece is required based on the scanning data includes:

[0046] Extract features from the scanning data according to the scanning data to obtain corresponding point cloud feature data;

[0047] Construct a three-dimensional model of the workpiece based on the point cloud feature data, and compare the three-dimensional model with a preset model according to the Meshlab tool to obtain comparison data;

[0048] Calculate the similarity between the three-dimensional model and the preset model according to the comparison data, and determine whether the similarity is greater than a preset value. If not, perform secondary processing on the workpiece.

[0049] By adopting the above technical solution, in order to determine whether secondary processing of the workpiece is required, first extract features from the scanning data according to the scanning data to obtain corresponding point cloud feature data, then construct a three-dimensional model of the workpiece based on the point cloud feature data, and compare the three-dimensional model with a preset model according to the Meshlab tool to obtain comparison data, then calculate the similarity between the three-dimensional model and the preset model according to the comparison data, and determine whether the similarity is greater than a preset value. If the similarity is greater than the preset value, perform secondary processing on the workpiece.

[0050] In a second aspect, the present application also provides a workpiece quality detection system based on the industrial Internet of Things, adopting the following technical solution:

[0051] A workpiece quality detection system based on the industrial Internet of Things includes a management platform, a sensing network platform, and an object platform that are communicatively connected in sequence. The management platform is configured with:

[0052] A data acquisition module, configured to place the workpiece at a preset temperature, acquire the surface temperature data of the workpiece according to a preset acquisition frequency, and acquire the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency, where the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition times;

[0053] An internal temperature determination module, configured to determine the internal temperature data of the workpiece according to the surface temperature data, where the internal temperature data includes the predicted temperatures of internal feature points at the multiple acquisition times;

[0054] An eigenvalue determination module is configured to determine temperature eigenvalues of the workpiece at the multiple acquisition moments according to the surface temperature and the internal temperature, and determine current eigenvalues of the workpiece according to the real-time current, where the temperature eigenvalues include at least one of an internal average temperature, a surface average temperature, an overall average temperature, a maximum internal temperature deviation value, a minimum internal temperature deviation value, a maximum overall temperature deviation value, and a minimum overall temperature deviation value, and the current eigenvalues include at least one of an average current, a maximum current deviation value, and a minimum current deviation value;

[0055] A quality assessment module is configured to determine whether the workpiece is qualified based on a preset workpiece quality assessment model and according to the temperature eigenvalues and the current eigenvalues.

[0056] In a third aspect, the present application further provides a computer device, adopting the following technical solution:

[0057] A computer device includes a memory and a processor. A computer program that can run on the processor is stored on the memory. When the processor executes the computer program, the method described in the first aspect is implemented.

[0058] In a fourth aspect, the present application further provides a computer-readable storage medium, adopting the following technical solution:

[0059] A computer-readable storage medium stores a computer program that can be loaded and executed by a processor to implement the method described in the first aspect.

[0060] In summary, the present application at least includes the following beneficial technical effects: First, obtain the surface temperature data of the workpiece according to a preset acquisition frequency, and obtain the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency. Here, the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition times. Then, determine the internal temperature data of the workpiece according to the surface temperature data, where the internal temperature data includes the predicted temperatures of internal feature points at multiple acquisition times. Then, determine the temperature characteristic values of the workpiece at multiple acquisition times according to the surface temperature data and the internal temperature data, and determine the current characteristic values of the workpiece according to the real-time current. Here, the temperature characteristic values include at least one of the internal average temperature, surface average temperature, overall average temperature, maximum internal temperature deviation value, minimum internal temperature deviation value, maximum overall temperature deviation value, and minimum overall temperature deviation value, and the current characteristic values include at least one of the average current, maximum current deviation value, and minimum current deviation value. Finally, based on a preset workpiece quality evaluation model, determine whether the workpiece is qualified according to the temperature characteristic values and the current characteristic values; Through the above method, the accuracy and intelligent level of workpiece quality detection are improved. By collecting surface temperature and current data in real time, the thermal behavior and electrical performance of the workpiece under preset conditions can be comprehensively reflected, ensuring the accuracy of the detection results. The comprehensive analysis of the temperature characteristic values includes the thermal uniformity of the internal and surface parts, as well as the overall temperature deviation, which can effectively identify potential defects and non-uniformities, reducing the misjudgment rate. At the same time, the analysis of the current characteristic values provides a necessary basis for evaluating the electrical performance, strengthening the multi-dimensional characteristics of the detection. Overall, by implementing this detection scheme, not only the accuracy of workpiece quality evaluation is improved, but also data support is provided for intelligent manufacturing, contributing to a more efficient production process and higher-quality product output. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 is the overall flow schematic diagram of an embodiment of the present application.

[0062] Figure 2 is the structural schematic diagram of one application scenario of the system of an embodiment of the present application.

[0063] Figure 3 is the structural schematic diagram of another application scenario of the system of an embodiment of the present application.

[0064] Figure 4 is the structural block diagram of the computer device of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0065] In order to make the objectives, technical solutions and advantages of the present application clearer and more understandable, the following further describes the present application in detail with reference to the attached Figures 1-4 drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0066] The embodiment of the present application discloses a workpiece quality detection method based on the industrial Internet of Things.

[0067] Referring to Figure 1 , a workpiece quality detection method based on the industrial Internet of Things, characterized in that it is applied to an industrial Internet of Things system, the industrial Internet of Things system includes a management platform, a sensing network platform and an object platform that are sequentially communicatively connected, and the method is executed by the management platform, including:

[0068] Step S11, obtaining the surface temperature data of the workpiece according to a preset acquisition frequency, and obtaining the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency.

[0069] Wherein, the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition moments.

[0070] It should be noted that during the acquisition process of the surface temperature data and the real-time current according to the preset acquisition frequency, several data acquisition moments will be generated. The starting acquisition moments of the surface temperature data and the real-time current in this application are the same. Therefore, the real-time temperature of the surface feature points and the corresponding real-time current at the current moment will be acquired at each data acquisition moment.

[0071] It should be further noted that the surface feature points can be the center, vertex, intersection point or equal division point of the diagonal of each surface, the equal division point of the edge line or other representative points.

[0072] Step S12, determining the internal temperature data of the workpiece according to the surface temperature data.

[0073] Wherein, the internal temperature data includes the predicted temperatures of internal feature points at multiple acquisition moments;

[0074] It should be noted that the internal feature points can be the center point of the workpiece. When the internal feature point is the center point of the workpiece, the internal temperature data is the center temperature, that is, the temperature of the center point. The internal feature points can also be other representative points. For example, when the workpiece is a cuboid, the equal division point of the center line connecting the centers of two parallel surfaces.

[0075] Step S13, determining the temperature characteristic values of the workpiece at multiple acquisition moments according to the surface temperature data and the internal temperature data, and determining the current characteristic value of the workpiece according to the real-time current.

[0076] Among them, the temperature characteristic values include at least one of the internal average temperature, surface average temperature, overall average temperature, maximum surface temperature deviation value, minimum surface temperature deviation value, maximum internal temperature deviation value, minimum internal temperature deviation value, maximum overall temperature deviation value, and minimum overall temperature deviation value; the current characteristic values include at least one of the average current, maximum current deviation value, and minimum current deviation value;

[0077] It can be understood that the overall average temperature is the average temperature calculated based on the surface temperature and the internal temperature; for the maximum surface temperature deviation value and the minimum surface temperature deviation value, if the workpiece has n surface characteristic points, the real-time temperatures of the n surface characteristic points at any one acquisition moment can be expressed as:

[0078] T = {T(1), T(2), T(3), …, T(n)};

[0079] Among them, T(1) represents the temperature of the 1st surface characteristic point, T(2) represents the temperature of the 2nd surface characteristic point, T(3) represents the temperature of the 3rd surface characteristic point, and T(n) represents the temperature of the nth surface characteristic point.

[0080] The maximum value Tm of the real-time temperatures respectively corresponding to the n surface characteristic points is:

[0081] Tm = MAX{T(1), T(2), T(3), …, T(n)};

[0082] The calculation formula for the surface average temperature Ta of the workpiece is:

[0083] Ta = (T(1) + T(2) + T(3) + … + T(n)) / n;

[0084] The calculation formula for the surface temperature deviation value ΔT of the workpiece is:

[0085] ΔT(1) = T(1) - Ta;

[0086] ΔT(2) = T(2) - Ta;

[0087] ΔT(3) = T(3) - Ta;

[0088] ……

[0089] ΔT(n) = T(n) - Ta;

[0090] Among them, ΔT(1) is the surface temperature deviation value of the 1st surface characteristic point, ΔT(2) is the surface temperature deviation value of the 2nd surface characteristic point, ΔT(3) is the surface temperature deviation value of the 3rd surface characteristic point, and ΔT(n) is the surface temperature deviation value of the nth surface characteristic point;

[0091] The maximum deviation value ΔTmax of the surface temperature of the workpiece is:

[0092] ΔTmax = MAX{ΔT(1), ΔT(2), ΔT(3)…, ΔT(n)};

[0093] The minimum deviation value ΔTmin of the surface temperature of the workpiece is:

[0094] ΔTmin = MIN{ΔT(1), ΔT(2), ΔT(3)…, ΔT(n)};

[0095] Similarly, the calculation methods of the maximum deviation value of the internal temperature, the minimum deviation value of the internal temperature, the maximum deviation value of the overall temperature, the minimum deviation value of the overall temperature, the maximum deviation value of the current, and the minimum deviation value of the current can all refer to the maximum deviation value and the minimum deviation value of the surface temperature.

[0096] Step S14, based on a preset workpiece quality evaluation model, and determine whether the workpiece is qualified according to the temperature characteristic value and the current characteristic value.

[0097] In the above embodiment, first obtain the surface temperature data of the workpiece according to a preset acquisition frequency, and obtain the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency. Among them, the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface characteristic points at multiple acquisition moments. Then determine the internal temperature data of the workpiece according to the surface temperature data. Among them, the internal temperature data includes the predicted temperatures of internal characteristic points at multiple acquisition moments. Then determine the temperature characteristic values of the workpiece at multiple acquisition moments according to the surface temperature data and the internal temperature data, and determine the current characteristic value of the workpiece according to the real-time current. Among them, the temperature characteristic values include at least one of the internal average temperature, the surface average temperature, the overall average temperature, the maximum deviation value of the internal temperature, the minimum deviation value of the internal temperature, the maximum deviation value of the overall temperature, and the minimum deviation value of the overall temperature. The current characteristic values include at least one of the average current, the maximum deviation value of the current, and the minimum deviation value of the current. Finally, based on a preset workpiece quality evaluation model, and determine whether the workpiece is qualified according to the temperature characteristic value and the current characteristic value; through the above method, the accuracy and intelligent level of workpiece quality detection are improved. By collecting surface temperature and current data in real time, the thermal behavior and electrical performance of the workpiece under preset conditions can be comprehensively reflected, ensuring the accuracy of the detection results. The comprehensive analysis of the temperature characteristic values includes the thermal uniformity of the internal and surface, as well as the overall temperature deviation, which can effectively identify potential defects and non-uniformities, reducing the misjudgment rate. At the same time, the analysis of the current characteristic values provides a necessary basis for evaluating the electrical performance, strengthening the multi-dimensional characteristics of the detection. Overall, by implementing this detection scheme, not only the accuracy of workpiece quality evaluation is improved, but also data support is provided for intelligent manufacturing, contributing to a more efficient production process and higher-quality product output.

[0098] As a further implementation of the method, the step of determining the internal temperature data of the workpiece based on the surface temperature data includes:

[0099] Step S21, obtaining a historical temperature data set and dividing the historical temperature data set into a training set and a test set according to a preset ratio.

[0100] Among them, the historical temperature data set includes historical surface temperature data and historical internal temperature data corresponding to the historical surface temperature data.

[0101] Step S22, setting the hyperparameters of the pre-constructed internal temperature generation data model according to the grid search algorithm, and taking the root mean square error RMSE and the coefficient of determination R 2 as evaluation indicators.

[0102] Step S23, training the pre-constructed internal temperature generation data model according to the training set to obtain a trained internal temperature generation data model.

[0103] Step S24, testing the trained internal temperature generation data model according to the test set, and judging whether the error is within a preset range according to the evaluation indicators. If so, taking the trained internal temperature generation data model as the preset internal temperature generation data model.

[0104] Specifically, testing the trained internal temperature generation data model according to the test set, and judging whether the error is within a preset range according to the evaluation indicators. If the error is within the preset range, taking the trained internal temperature generation data model as the preset internal temperature generation data model. If the error is not within the preset range, continue training and optimization.

[0105] Step S25, based on the preset internal temperature generation data model and according to the surface temperature data, obtaining the internal temperature data of the workpiece.

[0106] In the above implementation, obtaining a historical temperature data set and dividing the historical temperature data set into a training set and a test set according to a preset ratio. Among them, the historical temperature data set includes historical surface temperature data and historical internal temperature data corresponding to the historical surface temperature data. Then, setting the hyperparameters of the pre-constructed internal temperature generation data model according to the grid search algorithm, and taking the root mean square error RMSE and the coefficient of determination R 2As an evaluation index, then train the pre-constructed internal temperature generation data model according to the training set to obtain the trained internal temperature generation data model. Then, test the trained internal temperature generation data model according to the test set, and judge whether the error is within the preset range according to the evaluation index. If the error is not within the preset range, use the trained internal temperature generation data model as the preset internal temperature generation data model. Finally, based on the preset internal temperature generation data model and according to the surface temperature data, obtain the internal temperature data of the workpiece.

[0107] As a further implementation of the method, the preset internal temperature generation data model includes a first input layer, a hidden layer, and a first output layer. The step of obtaining the internal temperature data of the workpiece based on the preset internal temperature generation data model and according to the surface temperature data includes:

[0108] Step S31, through the first input layer, and according to the surface temperature data, generate a first input feature vector.

[0109] It can be understood that the input layer is the starting point of the preset internal temperature generation data model, mainly responsible for receiving external data. Each neuron in the input layer corresponds to an input feature, and the original data is passed to the subsequent part; the input data usually needs to be standardized or normalized to improve the performance and convergence speed of the network.

[0110] Step S32, through the hidden layer, and according to the input feature vector, generate a feature extraction vector.

[0111] It can be understood that the hidden layer is responsible for in-depth feature extraction and transformation of the input data, introducing non-linearity using activation functions (such as ReLU, Sigmoid, or Tanh) so that the network can capture complex patterns and relationships; according to the complexity of the task, there may be one or more hidden layers, and the number of layers and the number of neurons in each layer can significantly affect the expressive ability and learning effect of the model.

[0112] Step S33, through the first output layer, and according to the feature extraction vector, generate a second output feature vector.

[0113] It can be understood that the output layer is responsible for converting the calculation results of the hidden layer into the final output form. The number of neurons in the output layer is usually directly related to the requirements of the task. For example, in a binary classification problem, there may be only one output neuron, while in a multi-classification problem, the number of output neurons is the same as the number of classification categories; the output layer also uses specific activation functions, such as Softmax (for multi-classification) or linear activation (for regression), to ensure that the output results meet the requirements of the task.

[0114] Step S34: Generate the internal temperature data of the workpiece based on the output second feature vector.

[0115] In the above embodiment, in order to obtain the internal temperature data of the workpiece, through the first input layer, and based on the surface temperature data, generate the first input feature vector, then through the hidden layer, and based on the input feature vector, generate the feature extraction vector, then through the first output layer, and based on the feature extraction vector, generate the second output feature vector, and finally generate the internal temperature data of the workpiece based on the output second feature vector.

[0116] As a further embodiment of the method, the hidden layer includes a first sub-hidden layer, a second sub-hidden layer, and a third sub-hidden layer. The neuron number ratio of the first sub-hidden layer, the second sub-hidden layer, and the third sub-hidden layer is 5:3:1. The step of generating the feature extraction vector through the hidden layer and based on the input feature vector includes:

[0117] Step S41: Input the input feature vector into the first sub-hidden layer to obtain the first intermediate feature vector.

[0118] Step S42: Input the first intermediate feature vector into the second sub-hidden layer to obtain the second intermediate feature vector.

[0119] Step S43: Input the second intermediate feature vector into the third sub-hidden layer to obtain the feature extraction vector.

[0120] In the above embodiment, in order to generate the feature extraction vector, input the input feature vector into the first sub-hidden layer to obtain the first intermediate feature vector, then input the first intermediate feature vector into the second sub-hidden layer to obtain the second intermediate feature vector, and finally input the second intermediate feature vector into the third sub-hidden layer to obtain the feature extraction vector.

[0121] As a further embodiment of the method, the preset workpiece quality evaluation model includes a second input layer, a first feature extraction layer, a second feature extraction layer, a splicing layer, and a second output layer. The step of judging whether the workpiece is qualified based on the preset workpiece quality evaluation model and according to the temperature feature value and the current feature value includes:

[0122] Step S51: Through the second input layer, and based on the temperature feature value, generate the second input feature vector.

[0123] Step S52: Through the second input layer, and based on the current feature value, generate the third input feature vector.

[0124] Step S53: Through the first feature extraction layer, and based on the second input feature vector, generate the first intermediate feature vector.

[0125] Step S54: Generate a second intermediate feature vector through the second feature extraction layer based on the third input feature vector.

[0126] Step S55: Generate a data feature tensor through the concatenation layer based on the first intermediate feature vector and the second intermediate feature vector.

[0127] It can be understood that the concatenation layer is a deep learning network layer mainly used to connect feature vectors from multiple inputs along a specific dimension to generate a more comprehensive and rich feature tensor. In the architecture of a neural network, the concatenation layer can build bridges between different neurons or feature maps, allowing the model to combine feature information from different layers or different network paths to enhance the overall learning ability.

[0128] Step S56: Generate a second output feature vector through the second output layer based on the data feature tensor.

[0129] Step S57: Generate the qualified probability of the workpiece based on the second output feature vector and determine whether the workpiece is qualified according to the qualified probability.

[0130] In the above embodiment, to further determine whether the workpiece is qualified, a second input feature vector is generated through the second input layer based on the temperature feature value, then a third input feature vector is generated through the second input layer based on the current feature value, then a first intermediate feature vector is generated through the first feature extraction layer based on the second input feature vector, then a second intermediate feature vector is generated through the second feature extraction layer based on the third input feature vector, then a data feature tensor is generated through the concatenation layer based on the first intermediate feature vector and the second intermediate feature vector, then a second output feature vector is generated through the second output layer based on the data feature tensor, and finally the qualified probability of the workpiece is generated based on the second output feature vector and it is determined whether the workpiece is qualified according to the qualified probability.

[0131] As a further embodiment of the method, the step of determining whether the workpiece is qualified according to the qualified probability includes:

[0132] Step S61: Determine whether the qualified probability is within a preset range according to the qualified probability.

[0133] Step S62: If so, obtain the number of the workpiece and update the preset workpiece table according to the number.

[0134] Among them, the preset workpiece table includes the detected workpieces and the qualified information of the detected workpieces, and the preset workpiece table includes the mapping relationship between the detected workpieces and the qualified information of the detected workpieces;

[0135] Step S63, if not, obtain the scan data of the workpiece, and determine whether secondary processing of the workpiece is required according to the scan data.

[0136] In the above embodiment, in order to perform subsequent processing according to the pass probability, first determine whether the pass probability is within a preset range according to the pass probability. If the pass probability is within the preset range, obtain the number of the workpiece, and update the preset workpiece table according to the number. The preset workpiece table includes the detected workpieces and the pass information of the detected workpieces. The preset workpiece table includes the mapping relationship between the detected workpieces and the pass information of the detected workpieces. If the pass probability is not within the preset range, obtain the scan data of the workpiece, and determine whether secondary processing of the workpiece is required according to the scan data.

[0137] As a further embodiment of the method, the step of determining whether secondary processing of the workpiece is required according to the scan data includes:

[0138] Step S71, extract features from the scan data according to the scan data to obtain corresponding point cloud feature data.

[0139] Step S72, construct a three-dimensional model of the workpiece according to the point cloud feature data, and compare the three-dimensional model with a preset model according to the Meshlab tool to obtain comparison data.

[0140] Step S73, calculate the similarity between the three-dimensional model and the preset model according to the comparison data, and determine whether the similarity is greater than a preset value. If not, perform secondary processing on the workpiece.

[0141] Specifically, calculate the similarity between the three-dimensional model and the preset model according to the comparison data, and determine whether the similarity is greater than a preset value. If the similarity is not greater than the preset value, perform secondary processing on the workpiece. If the similarity is greater than the preset value, mark the workpiece as a scrap.

[0142] In the above embodiment, in order to determine whether secondary processing of the workpiece is required, first extract features from the scan data according to the scan data to obtain corresponding point cloud feature data, then construct a three-dimensional model of the workpiece according to the point cloud feature data, and compare the three-dimensional model with a preset model according to the Meshlab tool to obtain comparison data, and then calculate the similarity between the three-dimensional model and the preset model according to the comparison data, and determine whether the similarity is greater than a preset value. If the similarity is greater than the preset value, perform secondary processing on the workpiece.

[0143] The embodiment of the present application also discloses a workpiece quality detection system based on the industrial Internet of Things.

[0144] Reference Figure 2, A workpiece quality detection system based on the industrial Internet of Things, including a management platform, a sensing network platform, and an object platform that are communicatively connected in sequence. The management platform is configured with:

[0145] A data acquisition module, configured to acquire the surface temperature data of the workpiece according to a preset acquisition frequency, and acquire the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency. Wherein, the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition moments;

[0146] An internal temperature determination module, configured to determine the internal temperature data of the workpiece according to the surface temperature data. Wherein, the internal temperature data includes the predicted temperatures of internal feature points at multiple acquisition moments;

[0147] A feature value determination module, configured to determine the temperature feature values of the workpiece at multiple acquisition moments according to the surface temperature data and the internal temperature data, and determine the current feature value of the workpiece according to the real-time current. Wherein, the temperature feature values include at least one of the internal average temperature, the surface average temperature, the overall average temperature, the maximum internal temperature deviation value, the minimum internal temperature deviation value, the maximum overall temperature deviation value, and the minimum overall temperature deviation value; the current feature values include at least one of the average current, the maximum current deviation value, and the minimum current deviation value;

[0148] A quality assessment module, configured to judge whether the workpiece is qualified based on a preset workpiece quality assessment model and according to the temperature feature values and the current feature values.

[0149] The overall framework of another application scenario of the workpiece quality detection system based on the industrial Internet of Things in this application is as Figure 3 shown, and may include a user platform, a service platform, a management platform, a sensing network platform, and an object platform that interact in sequence, forming a five-platform architecture based on the industrial Internet of Things. Among them, the service platform consists of a service general database, multiple service sub-platforms, and multiple service sub-databases; the management platform includes a data acquisition module, an internal temperature determination module, a feature value determination module, and a quality assessment module; the management platform can interact with the sensing network platform and the service platform; the sensing network platform can interact with the object platform. The sensing network platform includes n sensing network sub-platforms, and a sensing sub-database is set on each sensing network sub-platform.

[0150] Specifically, in the above-mentioned another application scenario, the workpiece quality detection system based on the industrial Internet of Things includes a management platform, which is configured to: obtain the surface temperature data of the workpiece according to a preset acquisition frequency, and obtain the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency, where the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition moments; determine the internal temperature data of the workpiece according to the surface temperature data, where the internal temperature data includes the predicted temperatures of internal feature points at multiple acquisition moments; determine the temperature characteristic values of the workpiece at multiple acquisition moments according to the surface temperature data and the internal temperature data, and determine the current characteristic values of the workpiece according to the real-time current, where the temperature characteristic values include at least one of the internal average temperature, surface average temperature, overall average temperature, maximum surface temperature deviation value, minimum surface temperature deviation value, maximum internal temperature deviation value, minimum internal temperature deviation value, maximum overall temperature deviation value, and minimum overall temperature deviation value, and the current characteristic values include at least one of the average current, maximum current deviation value, and minimum current deviation value; based on a preset workpiece quality evaluation model, and determine whether the workpiece is qualified according to the temperature characteristic values and the current characteristic values.

[0151] Through the interaction between the various functional platforms of the workpiece quality detection system based on the industrial Internet of Things based on the above three-platform or five-platform, a perfect closed-loop information operation logic is established, ensuring the orderly operation of the perception information and the control information, and realizing the intelligent management of the equipment.

[0152] The workpiece quality detection system based on the industrial Internet of Things of the present invention can implement any one of the methods in the workpiece quality detection system based on the industrial Internet of Things, and the specific working process of the workpiece quality detection method based on the industrial Internet of Things of the present invention can refer to the corresponding process in the above-mentioned workpiece quality detection method based on the industrial Internet of Things.

[0153] The embodiments of the present application also disclose a computer device.

[0154] Reference Figure 4 , a computer device includes a memory and a processor, a computer program is stored on the memory and can run on the processor, and when the processor executes the computer program, it implements any one of the above-mentioned workpiece quality detection methods based on the industrial Internet of Things.

[0155] The embodiments of the present application also disclose a computer-readable storage medium.

[0156] A computer-readable storage medium stores a computer program that can be loaded and executed by a processor to implement any one of the above-mentioned workpiece quality detection methods based on the industrial Internet of Things.

[0157] Among them, the computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, apparatus, or device; the program code contained on the computer-readable medium can be transmitted by any suitable medium, including but not limited to wireless, wire, optical fiber cable, RF, etc., or any suitable combination of the above.

[0158] The above are all preferred embodiments of this application. The protection scope of this application is not limited accordingly. Any feature disclosed in this specification (including the abstract and drawings), unless specifically described, can be replaced by other equivalent or similar-purpose alternative features. That is, unless specifically described, each feature is only an example of a series of equivalent or similar features.

Claims

1. An industrial Internet of Things-based workpiece quality detection method, characterized in that Applied to an industrial Internet of Things system, the industrial Internet of Things system includes a management platform, a sensing network platform, and an object platform that are communicatively connected in sequence. The method is executed by the management platform and includes: Obtain the surface temperature data of the workpiece according to a preset acquisition frequency, and obtain the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency. Wherein, the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition times; Determine the internal temperature data of the workpiece according to the surface temperature data, where the internal temperature data includes the predicted temperatures of internal feature points at the multiple acquisition times; Determine the temperature characteristic values of the workpiece at the multiple acquisition times according to the surface temperature data and the internal temperature data, and determine the current characteristic values of the workpiece according to the real-time current. Wherein, the temperature characteristic values include at least one of internal average temperature, surface average temperature, overall average temperature, maximum surface temperature deviation value, minimum surface temperature deviation value, maximum internal temperature deviation value, minimum internal temperature deviation value, maximum overall temperature deviation value, and minimum overall temperature deviation value. The current characteristic values include at least one of average current, maximum current deviation value, and minimum current deviation value; Based on a preset workpiece quality assessment model, determine whether the workpiece is qualified according to the temperature characteristic values and the current characteristic values.

2. The workpiece quality detection method based on industrial Internet of Things according to claim 1, wherein The step of determining the internal temperature data of the workpiece according to the surface temperature data includes: Obtain a historical temperature data set, and divide the historical temperature data set into a training set and a test set according to a preset ratio. Wherein, the historical temperature data set includes historical surface temperature data and historical internal temperature data corresponding to the historical surface temperature data; Set the hyperparameters of the pre-constructed internal temperature generation data model according to the grid search algorithm, and use the root mean square error RMSE and the coefficient of determination R 2 as evaluation metrics; Train the pre-constructed internal temperature generation data model according to the training set to obtain a trained internal temperature generation data model; Test the trained internal temperature generation data model according to the test set, and judge whether the error is within a preset range according to the evaluation index. If so, use the trained internal temperature generation data model as the preset internal temperature generation data model; Based on the preset internal temperature generation data model, obtain the internal temperature data of the workpiece according to the surface temperature data.

3. The workpiece quality detection method based on industrial Internet of Things according to claim 2, wherein The preset internal temperature generation data model includes a first input layer, a hidden layer, and a first output layer. The step of obtaining the internal temperature data of the workpiece based on the preset internal temperature generation data model and according to the surface temperature data includes: Pass through the first input layer and generate a first input feature vector according to the surface temperature data; Pass through the hidden layer and generate a feature extraction vector according to the input feature vector; Pass through the first output layer and generate a second output feature vector according to the feature extraction vector; Generate the internal temperature data of the workpiece based on the output second feature vector.

4. The workpiece quality detection method based on industrial Internet of Things according to claim 3, characterized in that The hidden layer includes a first sub - hidden layer, a second sub - hidden layer, and a third sub - hidden layer. The ratio of the number of neurons in the first sub - hidden layer, the second sub - hidden layer, and the third sub - hidden layer is 5:3:

1. The step of passing through the hidden layer and generating a feature extraction vector according to the input feature vector includes: Inputting the input feature vector into the first sub - hidden layer to obtain a first intermediate feature vector; Inputting the first intermediate feature vector into the second sub - hidden layer to obtain a second intermediate feature vector; Inputting the second intermediate feature vector into the third sub - hidden layer to obtain a feature extraction vector.

5. The workpiece quality detection method based on the industrial Internet of Things according to claim 1, wherein The preset workpiece quality evaluation model includes a second input layer, a first feature extraction layer, a second feature extraction layer, a splicing layer, and a second output layer. The step of judging whether the workpiece is qualified based on the preset workpiece quality evaluation model and according to the temperature feature value and the current feature value includes: Passing through the second input layer and generating a second input feature vector according to the temperature feature value; Passing through the second input layer and generating a third input feature vector according to the current feature value; Passing through the first feature extraction layer and generating a first intermediate feature vector according to the second input feature vector; Passing through the second feature extraction layer and generating a second intermediate feature vector according to the third input feature vector; Passing through the splicing layer and generating a data feature tensor according to the first intermediate feature vector and the second intermediate feature vector; Passing through the second output layer and generating a second output feature vector according to the data feature tensor; Generating the qualified probability of the workpiece based on the second output feature vector and judging whether the workpiece is qualified according to the qualified probability.

6. The workpiece quality detection method based on industrial Internet of Things according to claim 5, characterized in that, The step of judging whether the workpiece is qualified according to the qualified probability includes: Judging whether the qualified probability is within a preset range according to the qualified probability; If so, obtaining the number of the workpiece and updating a preset workpiece table according to the number, where the preset workpiece table includes the detected workpieces and the qualified information of the detected workpieces, and the preset workpiece table includes the mapping relationship between the detected workpieces and the qualified information of the detected workpieces; If not, obtaining the scan data of the workpiece and judging whether the workpiece needs secondary processing according to the scan data.

7. The workpiece quality detection method based on the industrial Internet of Things according to claim 6, characterized in that, The step of judging whether the workpiece needs secondary processing according to the scan data includes: Performing feature extraction on the scan data according to the scan data to obtain corresponding point cloud feature data; Constructing a three - dimensional model of the workpiece according to the point cloud feature data and comparing the three - dimensional model with a preset model using the Meshlab tool to obtain comparison data; Calculating the similarity between the three - dimensional model and the preset model according to the comparison data and judging whether the similarity is greater than a preset value. If not, performing secondary processing on the workpiece.

8. An artifact quality detection system based on the industrial Internet of Things, characterized in that, Including a management platform, a sensor network platform, and an object platform that are communicatively connected in sequence. The management platform is configured with: A data acquisition module, configured to acquire the surface temperature data of a workpiece according to a preset acquisition frequency, and acquire the real-time current flowing through the workpiece under a preset voltage according to the preset acquisition frequency, wherein the workpiece is placed at a preset temperature, and the surface temperature data includes the real-time temperatures of surface feature points at multiple acquisition moments; An internal temperature determination module, configured to determine the internal temperature data of the workpiece according to the surface temperature data, wherein the internal temperature data includes the predicted temperatures of internal feature points at the multiple acquisition moments; A feature value determination module, configured to determine the temperature feature values of the workpiece at the multiple acquisition moments according to the surface temperature data and the internal temperature data, and determine the current feature values of the workpiece according to the real-time current, wherein the temperature feature values include at least one of an internal average temperature, a surface average temperature, an overall average temperature, a maximum surface temperature deviation value, a minimum surface temperature deviation value, a maximum internal temperature deviation value, a minimum internal temperature deviation value, a maximum overall temperature deviation value, and a minimum overall temperature deviation value, and the current feature values include at least one of an average current, a maximum current deviation value, and a minimum current deviation value; A quality assessment module, configured to judge whether the workpiece is qualified based on a preset workpiece quality assessment model and according to the temperature feature values and the current feature values.

9. A computer device, characterized in that, It includes a memory and a processor, and a computer program capable of running on the processor is stored on the memory. When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that, A computer program is stored that can be loaded and executed by a processor to implement the method according to any one of claims 1 to 7.

Images