Machine tool processing monitoring method and device based on industrial internet of things, terminal and medium

Abstract

The application discloses a machine tool processing monitoring method and device based on an industrial Internet of Things, a terminal and a medium. The method comprises the following steps: acquiring machine tool operation data and processing process data; performing feature extraction processing on the machine tool operation data and the processing process data to obtain a global feature vector; obtaining an abnormality attribution prediction result according to the global feature vector through a preset abnormality attribution prediction model; generating a customized monitoring result framework based on the abnormality attribution prediction result and a preset post; obtaining a target monitoring result corresponding to a target post through the customized monitoring result framework, and sending the target monitoring result to a terminal corresponding to a monitoring result receiving post in the preset post. The application aims to provide customized monitoring results for different levels and different post staff, so as to improve the understanding efficiency and decision response speed of monitoring information for different post roles.

Description

Technical Field

[0001] This application relates to the field of data processing technology, and in particular to a machine tool processing monitoring method, device, terminal and medium based on the Industrial Internet of Things. Background Technology

[0002] As the core production equipment in discrete manufacturing, the operational stability and machining accuracy of CNC machine tools directly determine product quality and production efficiency. Traditional machine tool machining monitoring mainly relies on manual inspection, offline quality testing, and simple threshold alarms.

[0003] However, the shortcomings of existing machine tool processing monitoring technologies are mainly reflected in the following aspects: most of them only use a single type of sensor data (such as spindle vibration data) for analysis, ignoring the inherent correlation between processing data (such as tool usage time, processing materials, and preset parameters) and machine tool operation data; existing methods all output monitoring results in a uniform format, without considering the differences in information needs of staff at different levels and in different positions; the early warning mechanisms of existing methods usually only trigger a uniform alarm based on the severity of the anomaly, and cannot provide differentiated pushes based on job responsibilities and early warning levels. Summary of the Invention

[0004] The main purpose of this application is to provide a machine tool processing monitoring method, device, terminal and medium based on the Industrial Internet of Things, which aims to provide customized monitoring results for workers at different levels and in different positions, and improve the efficiency of understanding monitoring information and decision-making response speed for different roles while ensuring consistent monitoring accuracy.

[0005] To achieve the above objectives, this application provides a machine tool processing monitoring method based on the Industrial Internet of Things (IIoT), the method comprising: Acquire machine tool operation data and machining process data; The machine tool operation data and machining process data are subjected to feature extraction processing to obtain a global feature vector; By using a pre-defined anomaly attribution prediction model, the anomaly attribution prediction result is obtained based on the global feature vector. Based on the anomaly attribution prediction results and the preset job positions, a customized monitoring result framework is generated. Using the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction results, the target monitoring result corresponding to the target position is obtained, and the target monitoring result is sent to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0006] Specifically, the step of performing feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector includes: The machine tool operation data and machining process data are preprocessed to obtain a standardized input tensor; The input tensor is subjected to spatial local feature extraction processing to obtain a spatial feature map; Long-distance temporal dependency feature extraction processing is performed on the spatial feature map to obtain a bidirectional temporal feature map; The bidirectional temporal feature map is subjected to global average pooling to obtain the global feature vector.

[0007] Specifically, the step of performing spatial local feature extraction on the input tensor to obtain a spatial feature map includes: By using a pre-defined first convolutional network, a first output feature map is obtained based on the input tensor; By using a pre-defined second convolutional network, a second output feature map is obtained based on the input tensor; A third output feature map is obtained by using a preset third convolutional network based on the input tensor, wherein the ratio of the kernel size of the preset third convolutional network, the kernel size of the preset second convolutional network, and the kernel size of the preset first convolutional network is 5:3:1. The spatial feature map is obtained by using a preset feature splicing network based on the first output feature map, the second output feature map, and the third output feature map.

[0008] Specifically, the step of performing long-distance temporal dependency feature extraction processing on the spatial feature map to obtain a bidirectional temporal feature map includes: The spatial feature maps are sequentially input into a preset forward long short-term memory unit in ascending time order to obtain a forward hidden state vector. The spatial feature map is input into a preset reverse long short-term memory unit in reverse chronological order to obtain the reverse hidden state vector. The forward hidden state vector and the backward hidden state vector are concatenated to obtain the bidirectional temporal feature map.

[0009] Specifically, the preset anomaly attribution prediction model includes a shared feature layer, an anomaly classification task branch layer, an attribution localization task branch layer, and an output layer. The step of obtaining the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model includes: The shared feature vector is obtained through the shared feature layer based on the global feature vector; Through the anomaly classification task branch layer, an anomaly type probability vector is obtained based on the shared feature vector; Through the attribution localization task branch layer, attribution parameter weight vector is obtained based on the shared feature vector, wherein the attribution parameter weight vector is used to characterize the contribution of machine tool operation data and machining process data to the current abnormal state of machine tool machining. The output layer obtains the anomaly attribution prediction result based on the anomaly type probability vector and the attribution parameter weight vector. The anomaly attribution prediction result includes the anomaly type identification result, the confidence level corresponding to the anomaly type identification result, and the attribution parameter weight vector.

[0010] Specifically, the preset positions include front-line operation positions, equipment maintenance positions, production management positions, and enterprise decision-making positions, and the customized monitoring result framework includes an early warning level classification framework and a preset position reference data framework. The process of generating a customized monitoring result framework based on the anomaly attribution prediction results and preset job positions includes: Based on the confidence level corresponding to the anomaly type identification result, the early warning level classification framework is constructed. The early warning level classification framework includes a first-level early warning, a second-level early warning, a third-level early warning, and a normal state level. The first-level early warning is used to characterize serious anomalies that require immediate response, the second-level early warning is used to characterize general anomalies that require timely attention, the third-level early warning is used to characterize minor anomalies that require continuous attention, and the normal state level is used to characterize normal machine tool operation. Based on the preset job positions, the anomaly attribution prediction results are subjected to data filtering processing to obtain the preset job position reference data framework, wherein the preset job position reference data framework is used to characterize the mapping relationship between the preset job positions and the processed anomaly attribution prediction results.

[0011] Specifically, sending the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset positions includes: If the target monitoring results include a Level 1 warning, then the target monitoring results will be sent to the preset post. If the target monitoring results include a level-two warning, then the target monitoring results will be sent to the front-line operator, the equipment maintenance operator, and the production management operator. If the target monitoring results include a Level 3 warning, then the target monitoring results will be sent to the front-line operation post and the equipment maintenance post.

[0012] To achieve the above objectives, this application also provides a machine tool processing monitoring device based on the Industrial Internet of Things, the device comprising: The first unit is used to acquire machine tool operation data and machining process data; The second unit is used to perform feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector. The third unit is used to obtain the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model. The fourth unit is used to generate a customized monitoring result framework based on the anomaly attribution prediction results and the preset job positions. The fifth unit is used to obtain the target monitoring result corresponding to the target position through the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction result, and send the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0013] To achieve the above objectives, this application also provides a terminal, including a memory storing multiple instructions; the processor loads instructions from the memory to execute the steps in any of the methods provided in this application.

[0014] To achieve the above objectives, this application also provides a medium storing a plurality of instructions adapted for loading by a processor to execute the steps in any of the methods provided in this application.

[0015] This application provides a machine tool processing monitoring method, device, terminal, and medium based on the Industrial Internet of Things (IIoT). The method first preprocesses the raw sensor data to obtain equipment anomaly data; then analyzes the equipment anomaly data to obtain anomaly analysis results; using a preset multi-branch convolutional model, based on the anomaly analysis results, it performs anomaly attribution and localization processing on the equipment in the factory workshop to obtain anomaly attribution and localization results; finally, using a preset sequence-to-sequence model, based on the anomaly attribution and localization results and the anomaly analysis results, it generates target anomaly conclusion information and target anomaly suggestion information to improve the accuracy and efficiency of anomaly handling by personnel in various positions. Attached Figure Description

[0016] Figure 1 A flowchart illustrating the method provided in the embodiments of this application; Figure 2 A schematic diagram illustrating the application implementation of the preset anomaly attribution prediction model provided in this application embodiment; Figure 3 This is a schematic diagram of the device provided in the embodiments of this application; Figure 4 This is a schematic diagram of the terminal structure provided in an embodiment of this application. Detailed Implementation

[0017] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0018] The shortcomings of existing machine tool processing monitoring technologies are mainly reflected in the following aspects: most of them only use single-type sensor data (such as spindle vibration data) for analysis, ignoring the inherent correlation between processing data (such as tool usage time, processing materials, preset parameters) and machine tool operation data; existing methods all output monitoring results in a uniform format, without considering the differences in information needs of staff at different levels and in different positions; the early warning mechanisms of existing methods usually only trigger a uniform alarm based on the severity of the anomaly, and cannot provide differentiated pushes based on job responsibilities and early warning levels.

[0019] Therefore, this application provides a machine tool processing monitoring method, device, terminal, and medium based on the Industrial Internet of Things to solve practical technical problems.

[0020] In some embodiments, the device may be integrated into an electronic device, such as a terminal or server.

[0021] In some embodiments, the server may also be implemented as a terminal.

[0022] The server can be a standalone physical server, a server cluster or distributed system consisting of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms.

[0023] The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and the server can be connected directly or indirectly through wired or wireless communication, which is not limited herein.

[0024] The following sections provide detailed descriptions of each example. It should be noted that the sequence numbers of the following embodiments are not intended to limit the preferred order of the embodiments.

[0025] This application provides a machine tool processing monitoring method based on the Industrial Internet of Things. The method can provide customized monitoring results for workers at different levels and in different positions. While ensuring consistent monitoring accuracy, it can improve the efficiency of different roles in understanding monitoring information and the speed of decision-making response.

[0026] In some embodiments, the Industrial Internet of Things (IIoT) upon which the method is based provides high-quality data input for neural network models and provides a running and interactive environment for logic algorithms through a five-layer architecture of "sensing-transmission-edge-cloud-application".

[0027] Specifically, the perception layer synchronously collects two types of core data—machine tool operating status (vibration, current, temperature, etc.) and machining process (process parameters, tool information, workpiece information, etc.)—through industrial sensors and standard interfaces of the CNC system. It uses the OPCUA protocol to achieve standardized access for machine tools from different brands, ensuring multi-source data timestamp accuracy within 1ms. The network layer employs a hybrid network architecture, primarily using industrial Ethernet and secondarily using a 5G industrial private network, providing low-latency, highly reliable encrypted data transmission. The edge computing layer is deployed on-site in the workshop, completing data preprocessing and millisecond-level real-time inference of neural network models, supporting independent operation without network access and local early warning. The cloud platform layer is responsible for centrally storing historical data, periodically training and optimizing models, and pushing updates via OTA, while also realizing full lifecycle management of equipment. The application layer corresponds to the logic algorithm module, transforming the unified model output into customized monitoring results and early warning notifications for four types of positions: front-line operators, equipment maintenance personnel, production management personnel, and enterprise decision-makers.

[0028] like Figure 1 The specific process of the method can be as follows: S110: Acquire machine tool operation data and machining process data.

[0029] In some embodiments, machine tool operation data and machining process data can be collected through the Industrial Internet of Things (IIoT). The machine tool operation data includes parameters such as spindle speed, feed rate, cutting power, spindle vibration acceleration, X / Y / Z axis displacement, motor current, hydraulic system pressure, and oil temperature. The machining process data includes parameters such as workpiece number, machining process number, tool number, tool usage time, machining material type, and preset machining parameters.

[0030] S120. Perform feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector.

[0031] In some embodiments, the step of performing feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector includes the steps A1 to A4 shown below: A1. Preprocess the machine tool operation data and machining process data to obtain the standardized input tensor.

[0032] In some embodiments, the machine tool operation data and machining process data can first be processed into a standardized two-dimensional data matrix, where the rows of the two-dimensional data matrix represent different acquisition timestamps and the columns represent different monitoring parameters, and then a standardized input tensor with the dimension of time step multiplied by the feature dimension is obtained.

[0033] A2. Perform spatial local feature extraction processing on the input tensor to obtain a spatial feature map.

[0034] In some embodiments, the spatial local feature extraction process on the input tensor to obtain a spatial feature map includes the steps A21 to A24 shown below: A21. A first output feature map is obtained by using a preset first convolutional network based on the input tensor.

[0035] In some embodiments, the preset first convolutional network is a one-dimensional convolutional layer with a kernel size of 1. The input tensor X is subjected to sliding convolution operation with a convolution kernel of size 1 multiplied by the feature dimension. Each convolution kernel performs a weighted summation of all feature dimensions at each time step, and adds a bias term to obtain the first output feature map F1.

[0036] A22. A second output feature map is obtained by using a pre-set second convolutional network based on the input tensor.

[0037] In some embodiments, the preset second convolutional network is a one-dimensional convolutional layer with a kernel size of 3. The input tensor X is subjected to sliding convolution operation with a convolutional kernel of size 3 multiplied by the feature dimension. Each convolutional kernel performs a weighted summation of all feature dimensions at three consecutive time steps, and a bias term is added to obtain the second output feature map F2.

[0038] A23. A third output feature map is obtained by using a preset third convolutional network based on the input tensor, wherein the ratio of the kernel size of the preset third convolutional network, the kernel size of the preset second convolutional network, and the kernel size of the preset first convolutional network is 5:3:1.

[0039] In some embodiments, the preset third convolutional network is a one-dimensional convolutional layer with a kernel size of 5. The input tensor X is subjected to sliding convolution operation with a convolutional kernel of size 5 multiplied by the feature dimension. Each convolutional kernel performs a weighted summation of all feature dimensions over 5 consecutive time steps, and a bias term is added to obtain the third output feature map F3.

[0040] A24. The spatial feature map is obtained by using a preset feature splicing network based on the first output feature map, the second output feature map, and the third output feature map.

[0041] In some embodiments, the output feature maps F1, F2, and F3 of the three branches are concatenated along the feature dimension to obtain a multi-scale spatial feature map F_conv, whose dimension = time step * (number of output channels of the first branch + number of output channels of the second branch + number of output channels of the third branch).

[0042] A3. Perform long-distance time-dependent feature extraction processing on the spatial feature map to obtain a bidirectional time feature map.

[0043] In some embodiments, the long-distance temporal dependency feature extraction process on the spatial feature map to obtain a bidirectional temporal feature map includes the steps A31 to A33 as shown below: A31. Input the spatial feature map into the preset forward long short-term memory unit in ascending time order to obtain the forward hidden state vector.

[0044] In some embodiments, the multi-scale spatial feature map F_conv is sequentially input into the forward long short-term memory unit in chronological order. Each unit at each time step receives the input features at the current time step and the hidden state at the previous time step. The cell state at the current time step and the forward hidden state vector H_forward are calculated through the gating mechanism of input gate, forget gate, and output gate.

[0045] A32. Input the spatial feature map into a preset reverse long short-term memory unit in reverse chronological order to obtain the reverse hidden state vector.

[0046] In some embodiments, the multi-scale spatial feature map F_conv is input into the inverse long short-term memory unit in reverse chronological order. Each unit at each time step receives the input features at the current time step and the hidden state at the previous time step. The cell state at the current time step and the inverse hidden state vector H_backward are calculated through the same gating mechanism.

[0047] A33. The forward hidden state vector and the reverse hidden state vector are concatenated to obtain the bidirectional temporal feature map.

[0048] In some embodiments, the forward hidden state vector H_forward and the backward hidden state vector H_backward with the same time step are concatenated along the feature dimension to obtain a bidirectional temporal feature map F_lstm, whose dimension = time step * (forward hidden state dimension + backward hidden state dimension).

[0049] A4. Perform global average pooling on the bidirectional time feature map to obtain the global feature vector.

[0050] In some embodiments, a global average pooling operation is performed on the bidirectional temporal feature map F_lstm in the time dimension. The calculation process is as follows: the feature values ​​of all time steps in each feature dimension are arithmetically averaged to obtain the global feature vector F_global, whose dimension = 1 * (forward hidden state dimension + backward hidden state dimension), which contains the comprehensive depth features of the machine tool operation and processing data.

[0051] S130. By using a preset anomaly attribution prediction model, the anomaly attribution prediction result is obtained based on the global feature vector.

[0052] In some embodiments, such as Figure 2 The preset anomaly attribution prediction model includes a shared feature layer, an anomaly classification task branch layer, an attribution localization task branch layer, and an output layer.

[0053] Specifically, the step of obtaining the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model includes the steps B1 to B4 shown below: B1. Through the shared feature layer, a shared feature vector is obtained based on the global feature vector.

[0054] In some embodiments, the shared feature layer includes two fully connected networks for mapping the global feature vector F_global to a shared feature space suitable for multi-task learning.

[0055] The first fully connected layer: performs matrix multiplication between the global feature vector F_global and the first layer weight matrix, adds the first layer bias vector, and then performs a non-linear transformation through the ReLU activation function to obtain the intermediate shared feature vector F_shared1.

[0056] The second fully connected layer: The intermediate shared feature vector F_shared1 is multiplied by the weight matrix of the second layer, and then the bias vector of the second layer is added. Then, a nonlinear transformation is performed through the ReLU activation function to obtain the shared feature vector F_shared2. The shared feature vector is the common input of the two task branches described below.

[0057] B2. Through the anomaly classification task branch layer, an anomaly type probability vector is obtained based on the shared feature vector.

[0058] In some embodiments, the shared feature vector F_shared2 is multiplied by the anomaly classification weight matrix, and then an anomaly classification bias vector is added. The output is then converted into a probability distribution using the Softmax activation function to obtain the anomaly type probability vector P_class. This vector has a dimension of 1 multiplied by the total number of anomaly types. Each element represents the probability of occurrence of the corresponding anomaly type, including normal state, tool wear anomaly, spindle vibration anomaly, feed system anomaly, hydraulic system anomaly, electrical system anomaly, etc. The anomaly type corresponding to the element with the highest probability value in the anomaly type probability vector P_class can be taken as the final anomaly type identification result Y_class, and this maximum probability value is recorded as the anomaly confidence score Score_conf.

[0059] B3. Through the attribution localization task branch layer, attribution parameter weight vector is obtained based on the shared feature vector, wherein the attribution parameter weight vector is used to characterize the contribution of machine tool operation data and machining process data to the current abnormal state of machine tool machining.

[0060] In some embodiments, the shared feature vector F_shared2 is multiplied by the attribution localization weight matrix, and then the attribution localization bias vector is added. The output is then converted to weight values ​​between 0 and 1 using a sigmoid activation function, resulting in the attribution parameter weight vector W_root. This vector has a dimension of 1 multiplied by the total number of original monitored parameters. Each element represents the contribution of the corresponding monitored parameter to the current abnormal state; a larger weight value indicates a higher probability that the parameter is attributable to the abnormality.

[0061] B4. Through the output layer, the anomaly attribution prediction result is obtained based on the anomaly type probability vector and the attribution parameter weight vector, wherein the anomaly attribution prediction result includes the anomaly type identification result, the confidence level corresponding to the anomaly type identification result, and the attribution parameter weight vector.

[0062] In some embodiments, the anomaly attribution prediction results obtained through the output layer include: anomaly type identification result Y_class, confidence score Score_conf, and attribution parameter weight vector W_root.

[0063] S140. Based on the anomaly attribution prediction results and the preset job positions, generate a customized monitoring result framework.

[0064] In some embodiments, the preset positions include front-line operators, equipment maintenance personnel, production management personnel, and enterprise decision-making personnel, and the customized monitoring result framework includes an early warning level classification framework and a preset position reference data framework.

[0065] Specifically, the step of generating a customized monitoring result framework based on the anomaly attribution prediction results and preset job positions includes the following steps C1 to C2: C1. Based on the confidence level corresponding to the anomaly type identification result, construct the early warning level classification framework, wherein the early warning level classification framework includes a first-level early warning, a second-level early warning, a third-level early warning, and a normal state level, wherein the first-level early warning is used to characterize serious anomalies that require immediate response, the second-level early warning is used to characterize general anomalies that require timely attention, the third-level early warning is used to characterize minor anomalies that require continuous attention, and the normal state level is used to characterize normal machine tool operation.

[0066] In some embodiments, the monitoring results are divided into three warning levels based on the confidence level Score_conf: Level 1 Warning: A confidence level of 0.9 or higher indicates a high probability of a serious anomaly, requiring immediate response. Level 2 alert: Confidence level greater than or equal to 0.7 and less than 0.9 indicates a moderate probability of a general anomaly, requiring immediate attention. Level 3 alert: Confidence level greater than or equal to 0.5 and less than 0.7 indicates a low probability of a minor anomaly, requiring continued monitoring. Normal state: A confidence level of less than 0.5 indicates that the machine tool is operating normally.

[0067] C2. Based on the preset job position, perform data filtering processing on the anomaly attribution prediction results to obtain the preset job position reference data framework, wherein the preset job position reference data framework is used to characterize the mapping relationship between the preset job position and the processed anomaly attribution prediction results.

[0068] In some embodiments, the anomaly attribution prediction results are subjected to differentiated data filtering processing based on the responsibilities and information needs of different positions: For frontline operators: Only retain anomaly type information and the most direct operational suggestions directly related to the current processing procedure, filtering out technical details and root cause analysis. If the status is normal, only output the message "Equipment is operating normally"; if the status is warning, only output a simple description of the warning level, anomaly type, and corresponding emergency operation steps.

[0069] For equipment maintenance personnel: retain complete anomaly types, confidence levels, core root cause parameters with weights greater than 0.6 in the attribution parameter weight vector, and corresponding maintenance recommendations. Filter out production management information irrelevant to equipment maintenance.

[0070] For production management positions: retain information such as the type of anomaly, confidence level, time of occurrence, affected workpiece and process numbers, estimated downtime, and impact assessment on the production schedule. Filter out overly technical parameters and maintenance details.

[0071] For enterprise decision-making positions: only retain statistical information on the occurrence of anomalies, their impact on overall production progress, equipment failure rate trends, and corresponding cost loss estimates. Filter out specific technical details and detailed information on individual anomalies.

[0072] S150. Using the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction results, obtain the target monitoring result corresponding to the target position, and send the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0073] In some embodiments, the target monitoring results may include a visualization interface for different preset positions.

[0074] Specifically, for frontline operators: a simple and clear indicator light-style visual interface is used. In normal operation, a green indicator light and the text "Operating Normally" are displayed; in warning situations, yellow (Level 3), orange (Level 2), and red (Level 1) indicator lights are displayed according to the level. A simple description of the abnormality type and emergency operating procedures are displayed in large font, avoiding the use of technical jargon.

[0075] Equipment maintenance staff: Employs a technical parameter-based visual interface. It displays real-time values ​​and normal ranges of core root cause parameters in a dashboard format, historical trends of these parameters in a line graph format, and anomaly details and maintenance suggestions in a list format. It also supports viewing raw data and historical anomaly records.

[0076] Production Management Position: A production flow visualization interface is used. The machine locations where anomalies occur and the affected processes are marked on the production flow chart. Suggestions for adjusting the production plan are displayed in Gantt chart format, and information on affected workpieces and their estimated completion times are displayed in tabular format.

[0077] For enterprise decision-making positions: A statistical report-style visual interface is used. The failure rate statistics for each workshop and type of equipment are displayed in bar chart format, the distribution of anomaly types is displayed in pie chart format, and the monthly trend of equipment failure rate is displayed in line chart format. At the same time, the direct and indirect cost losses caused by anomalies are estimated.

[0078] In some embodiments, sending the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset position includes the following specific implementation process: If the target monitoring results include a Level 1 warning, then the target monitoring results will be sent to the preset post. If the target monitoring results include a level-two warning, then the target monitoring results will be sent to the front-line operator, the equipment maintenance operator, and the production management operator. If the target monitoring results include a Level 3 warning, then the target monitoring results will be sent to the front-line operation post and the equipment maintenance post.

[0079] Specifically, Level 1 warning: Immediately send a warning notification to all relevant positions. Send audible and visual alarms and handheld terminal pop-up notifications to front-line operators; send telephone voice alarms and SMS notifications to equipment maintenance personnel; send emergency messages and email notifications to production management personnel; and send key points highlighted in the daily anomaly summary report to enterprise decision-making personnel.

[0080] Level 2 Warning: Sends warning notifications to frontline operators, equipment maintenance personnel, and production management personnel. Frontline operators receive a pop-up notification on their handheld terminals; equipment maintenance personnel receive a message notification; production management personnel receive an email notification; enterprise decision-making personnel are only mentioned in the weekly anomaly summary report.

[0081] Level 3 Early Warning: Early warning notifications are sent only to frontline operators and equipment maintenance personnel. Handheld terminal notifications are sent to frontline operators; reminders for pending tasks within the system are sent to equipment maintenance personnel; production management and enterprise decision-making personnel are not notified separately, but are only reflected in monthly statistical reports.

[0082] In summary, this application provides a machine tool processing monitoring method based on the Industrial Internet of Things, which can provide customized monitoring results for workers at different levels and in different positions. While ensuring consistent monitoring accuracy, it can improve the efficiency of different roles in understanding monitoring information and the speed of decision-making response.

[0083] To better implement the above methods, this application also provides a machine tool processing monitoring device based on the Industrial Internet of Things (IIoT). This device can be integrated into an electronic device, such as a terminal or server. The terminal can be a mobile phone, tablet, smart Bluetooth device, laptop, personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers.

[0084] For example, in this embodiment, the method of this application embodiment will be described in detail by taking the integration of a machine tool processing monitoring device based on the Industrial Internet of Things into a terminal as an example.

[0085] For example, such as Figure 3 As shown, the machine tool processing monitoring device 300 based on the Industrial Internet of Things may include a first unit 301, a second unit 302, a third unit 303, a fourth unit 304, and a fifth unit 305. The device includes: Unit 301 is used to acquire machine tool operation data and machining process data; The second unit 302 is used to perform feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector. The third unit 303 is used to obtain the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model; The fourth unit 304 is used to generate a customized monitoring result framework based on the anomaly attribution prediction results and the preset job positions. The fifth unit 305 is used to obtain the target monitoring result corresponding to the target position through the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction result, and send the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0086] In practice, each of the above units can be implemented as an independent entity or can be arbitrarily combined to be implemented as the same or several entities. For the specific implementation of each of the above units, please refer to the previous method embodiments, which will not be repeated here.

[0087] As can be seen from the above, the embodiments of this application can provide customized monitoring results for staff at different levels and in different positions, and improve the efficiency of different roles in understanding monitoring information and the speed of decision-making response while ensuring consistent monitoring accuracy.

[0088] This application also provides an electronic device, which can be a terminal, a server, or other similar device. The terminal can be a mobile phone, tablet computer, smart Bluetooth device, laptop computer, personal computer, etc.; the server can be a single server or a server cluster composed of multiple servers, etc.

[0089] In some embodiments, the product processing device may also be integrated into multiple electronic devices, such as multiple servers, with multiple servers implementing the machine tool processing monitoring method based on the Industrial Internet of Things of this application.

[0090] In this embodiment, the electronic device will be described in detail as a terminal, for example, such as... Figure 4 As shown, it illustrates a structural schematic diagram of the terminal 400 involved in an embodiment of this application. Specifically: The terminal 400 may include components such as a processor 401 with one or more processing cores, a memory 402 with one or more media, a power supply 403, an input module 404, and a communication module 405. Those skilled in the art will understand that... Figure 4The terminal 400 structure shown does not constitute a limitation on the terminal 400, and may include more or fewer components than shown, or combine certain components, or have different component arrangements. Wherein: The processor 401 is the control center of the terminal 400. It connects various parts of the terminal 400 via various interfaces and lines, and performs various functions and processes data by running or executing software programs and / or modules stored in the memory 402, and by calling data stored in the memory 402, thereby providing overall monitoring of the terminal 400. In some embodiments, the processor 401 may include one or more processing cores; in some embodiments, the processor 401 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may also not be integrated into the processor 401.

[0091] The memory 402 can be used to store software programs and modules. The processor 401 executes various functional applications and data processing by running the software programs and modules stored in the memory 402. The memory 402 may mainly include a program storage area and a data storage area. The program storage area may store the operating system, application programs required for at least one function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the terminal 400, etc. In addition, the memory 402 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device. Accordingly, the memory 402 may also include a memory controller to provide the processor 401 with access to the memory 402.

[0092] The terminal 400 also includes a power supply 403 that supplies power to the various components. In some embodiments, the power supply 403 can be logically connected to the processor 401 through a power management system, thereby enabling functions such as charging, discharging, and power consumption management through the power management system. The power supply 403 may also include one or more DC or AC power supplies, recharging systems, power fault detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0093] The terminal 400 may also include an input module 404, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0094] The terminal 400 may also include a communication module 405. In some embodiments, the communication module 405 may include a wireless module. The terminal 400 can perform short-range wireless transmission through the wireless module of the communication module 405, thereby providing users with wireless broadband Internet access. For example, the communication module 405 can be used to help users send and receive emails, browse web pages, and access streaming media.

[0095] Although not shown, terminal 400 may also include a display unit, etc., which will not be described in detail here. Specifically, in this embodiment, processor 401 in terminal 400 loads the executable files corresponding to the processes of one or more applications into memory 402 according to the following instructions, and processor 401 runs the applications stored in memory 402 to realize various functions, as follows: Acquire machine tool operation data and machining process data; The machine tool operation data and machining process data are subjected to feature extraction processing to obtain a global feature vector; By using a pre-defined anomaly attribution prediction model, the anomaly attribution prediction result is obtained based on the global feature vector. Based on the anomaly attribution prediction results and the preset job positions, a customized monitoring result framework is generated. Using the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction results, the target monitoring result corresponding to the target position is obtained, and the target monitoring result is sent to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0096] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0097] As can be seen from the above, the embodiments of this application can provide customized monitoring results for staff at different levels and in different positions, and improve the efficiency of different roles in understanding monitoring information and the speed of decision-making response while ensuring consistent monitoring accuracy.

[0098] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be accomplished by instructions, or by instructions controlling related hardware. These instructions can be stored in a medium and loaded and executed by a processor.

[0099] To this end, embodiments of this application provide a medium storing multiple instructions that can be loaded by a processor to execute steps in any of the machine tool processing monitoring methods based on the Industrial Internet of Things provided in embodiments of this application. For example, the instructions can execute the following steps: Acquire machine tool operation data and machining process data; The machine tool operation data and machining process data are subjected to feature extraction processing to obtain a global feature vector; By using a pre-defined anomaly attribution prediction model, the anomaly attribution prediction result is obtained based on the global feature vector. Based on the anomaly attribution prediction results and the preset job positions, a customized monitoring result framework is generated. Using the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction results, the target monitoring result corresponding to the target position is obtained, and the target monitoring result is sent to the terminal corresponding to the monitoring result receiving position in the preset positions.

[0100] The medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0101] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a medium. A processor of a computer device reads the computer instructions from the medium and executes the computer instructions, causing the computer device to perform the methods provided in the various optional implementations of the above embodiments.

[0102] Since the instructions stored in the medium can execute the steps in any of the machine tool processing monitoring methods based on the Industrial Internet of Things provided in the embodiments of this application, the beneficial effects that any of the machine tool processing monitoring methods based on the Industrial Internet of Things provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0103] The above provides a detailed description of a machine tool processing monitoring method, device, terminal, and medium based on the Industrial Internet of Things (IIoT) provided in the embodiments of this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A machine tool processing monitoring method based on the Industrial Internet of Things, characterized in that, The method includes: Acquire machine tool operation data and machining process data; The machine tool operation data and machining process data are subjected to feature extraction processing to obtain a global feature vector; By using a pre-defined anomaly attribution prediction model, the anomaly attribution prediction result is obtained based on the global feature vector. Based on the anomaly attribution prediction results and the preset job positions, a customized monitoring result framework is generated. Using the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction results, the target monitoring result corresponding to the target position is obtained, and the target monitoring result is sent to the terminal corresponding to the monitoring result receiving position in the preset positions.

2. The method as described in claim 1, characterized in that, The step of performing feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector includes: The machine tool operation data and machining process data are preprocessed to obtain a standardized input tensor; The input tensor is subjected to spatial local feature extraction processing to obtain a spatial feature map; Long-distance temporal dependency feature extraction processing is performed on the spatial feature map to obtain a bidirectional temporal feature map; The bidirectional temporal feature map is subjected to global average pooling to obtain the global feature vector.

3. The method as described in claim 2, characterized in that, The step of performing spatial local feature extraction on the input tensor to obtain a spatial feature map includes: By using a pre-defined first convolutional network, a first output feature map is obtained based on the input tensor; By using a pre-defined second convolutional network, a second output feature map is obtained based on the input tensor; A third output feature map is obtained by using a preset third convolutional network based on the input tensor, wherein the ratio of the kernel size of the preset third convolutional network, the kernel size of the preset second convolutional network, and the kernel size of the preset first convolutional network is 5:3:

1. The spatial feature map is obtained by using a preset feature splicing network based on the first output feature map, the second output feature map, and the third output feature map.

4. The method as described in claim 2, characterized in that, The step of performing long-distance temporal dependency feature extraction on the spatial feature map to obtain a bidirectional temporal feature map includes: The spatial feature maps are sequentially input into a preset forward long short-term memory unit in ascending time order to obtain a forward hidden state vector. The spatial feature map is input into a preset reverse long short-term memory unit in reverse chronological order to obtain the reverse hidden state vector. The forward hidden state vector and the backward hidden state vector are concatenated to obtain the bidirectional temporal feature map.

5. The method as described in claim 1, characterized in that, The preset anomaly attribution prediction model includes a shared feature layer, an anomaly classification task branch layer, an attribution localization task branch layer, and an output layer. The step of obtaining the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model includes: The shared feature vector is obtained through the shared feature layer based on the global feature vector; Through the anomaly classification task branch layer, an anomaly type probability vector is obtained based on the shared feature vector; Through the attribution localization task branch layer, attribution parameter weight vector is obtained based on the shared feature vector, wherein the attribution parameter weight vector is used to characterize the contribution of machine tool operation data and machining process data to the current abnormal state of machine tool machining. The output layer obtains the anomaly attribution prediction result based on the anomaly type probability vector and the attribution parameter weight vector. The anomaly attribution prediction result includes the anomaly type identification result, the confidence level corresponding to the anomaly type identification result, and the attribution parameter weight vector.

6. The method as described in claim 5, characterized in that, The preset positions include front-line operators, equipment maintenance personnel, production management personnel, and enterprise decision-making personnel. The customized monitoring result framework includes an early warning level classification framework and a preset position reference data framework. The process of generating a customized monitoring result framework based on the anomaly attribution prediction results and preset job positions includes: Based on the confidence level corresponding to the anomaly type identification result, the early warning level classification framework is constructed. The early warning level classification framework includes a first-level early warning, a second-level early warning, a third-level early warning, and a normal state level. The first-level early warning is used to characterize serious anomalies that require immediate response, the second-level early warning is used to characterize general anomalies that require timely attention, the third-level early warning is used to characterize minor anomalies that require continuous attention, and the normal state level is used to characterize normal machine tool operation. Based on the preset job positions, the anomaly attribution prediction results are subjected to data filtering processing to obtain the preset job position reference data framework, wherein the preset job position reference data framework is used to characterize the mapping relationship between the preset job positions and the processed anomaly attribution prediction results.

7. The method as described in claim 6, characterized in that, The step of sending the target monitoring results to the terminal corresponding to the monitoring result receiving position in the preset positions includes: If the target monitoring results include a Level 1 warning, then the target monitoring results will be sent to the preset post. If the target monitoring results include a level-two warning, then the target monitoring results will be sent to the front-line operator, the equipment maintenance operator, and the production management operator. If the target monitoring results include a Level 3 warning, then the target monitoring results will be sent to the front-line operation post and the equipment maintenance post.

8. A machine tool processing monitoring device based on the Industrial Internet of Things, characterized in that, The device includes: The first unit is used to acquire machine tool operation data and machining process data; The second unit is used to perform feature extraction processing on the machine tool operation data and machining process data to obtain a global feature vector. The third unit is used to obtain the anomaly attribution prediction result based on the global feature vector through a preset anomaly attribution prediction model. The fourth unit is used to generate a customized monitoring result framework based on the anomaly attribution prediction results and the preset job positions. The fifth unit is used to obtain the target monitoring result corresponding to the target position through the customized monitoring result framework, based on the target position in the preset positions and the anomaly attribution prediction result, and send the target monitoring result to the terminal corresponding to the monitoring result receiving position in the preset positions.

9. A terminal, characterized in that, The method includes a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to perform the steps of the method as described in any one of claims 1 to 7.

10. A medium, characterized in that, The medium stores a plurality of instructions adapted for loading by a processor to execute the steps of the method according to any one of claims 1 to 7.

Images