Machine tool abnormal operation data edge calculation method and system

Abstract

The invention relates to the technical field of machine tool data processing, in particular to a machine tool abnormal operation data edge calculation method and system, and the method comprises the steps: deploying a sensor at a key part of a machine tool for collecting machine tool operation data and transmitting the machine tool operation data to an edge calculation device which has calculation capability and a local database; dynamically optimizing a data transmission path by adopting an SDN network technology in a data transmission process; after receiving the data, the edge computing device performs data preprocessing and abnormal data detection, stores the data and a detection result in a local database and regularly uploads the data and the detection result to a cloud server; and the cloud server analyzes the trend of abnormal operation of the machine tool by adopting a big data analysis algorithm, and feeds back optimization suggestions to the edge computing device according to an analysis result. According to the method, the data processing delay is reduced by combining the edge computing with artificial intelligence, the transmission path is optimized, the cooperation of the edge computing equipment and the cloud is realized, and the processing efficiency of the abnormal operation data of the machine tool is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of machine tool data processing, and particularly relates to a method and system for edge computing of abnormal operation data of machine tools. Background Technique

[0002] In modern manufacturing, machine tools, as key processing equipment, the stability of their operation is crucial for product quality and production efficiency. A large amount of data is generated during the operation of machine tools, and this data can reflect the real-time operation status of the machine tools. The traditional methods for processing machine tool data mostly involve centrally transmitting the data to the cloud or a remote data center for analysis. However, this method faces problems such as high latency, limited network bandwidth, and data privacy leakage. The emergence of edge computing technology provides a new way for the rapid and efficient processing of abnormal operation data of machine tools. By placing the computing device close to the machine tool end to capture and process abnormal data in a timely manner, it avoids problems such as high latency faced during data transmission, improves the reliability of the machine tool, and ensures the continuity of production. Summary of the Invention

[0003] In view of the above-mentioned existing technical deficiencies, the present invention provides a method and system for edge computing of abnormal operation data of machine tools to improve the processing efficiency of abnormal data of machine tools.

[0004] The present invention is achieved through the following technical solutions:

[0005] A method for edge computing of abnormal operation data of machine tools is provided. The method includes the following steps:

[0006] Step S10: Deploy sensors at key parts of the machine tool to collect machine tool operation data and transmit it to an edge computing device with computing capabilities and a local database;

[0007] Step S20: During the data transmission process, use SDN network technology to dynamically optimize the data transmission path, and allocate the optimal transmission path for different abnormal operation data of the machine tool according to the real-time status of the machine tool operation and the data priority;

[0008] Step S30: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and the detection results in the local database, and regularly uploads them to the cloud server;

[0009] Step S40: The cloud server uses big data analysis algorithms to analyze the trend of abnormal operation of the machine tool, find the correlations between different machine tool abnormalities, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device performs optimization and adjustment according to the feedback suggestions;

[0010] Among them, the sensors deployed at the key parts of the machine tool in step S10 include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time; the installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocol.

[0011] Preferably, in the data transmission process of step S20, the SDN network technology is used to dynamically optimize the data transmission path. The steps of allocating the optimal transmission path for different abnormal operation data of the machine tool according to the real-time state of the machine tool operation and the data priority include:

[0012] Construct the workshop edge computing network architecture: Deploy network devices supporting SDN technology inside the workshop where the machine tool is located, including switches and routers, to form a communication network covering the entire workshop;

[0013] Data priority classification setting: Divide the data into priorities according to the importance and urgency of various operation data generated by the machine tool, and preset the priority division rules in the edge computing device;

[0014] Real-time monitor the operation state of the machine tool: Real-time monitor the operation state of the machine tool through the installed sensors, and analyze the operation state of the machine tool and the change trend of the data under each operation state according to the machine tool operation data collected by the sensors;

[0015] Formulate the dynamic path allocation strategy: According to the set data priority and the analyzed operation state of the machine tool, formulate a dynamic data transmission path allocation strategy;

[0016] Network resource allocation and optimization: During the transmission process of the machine tool operation data, continuously monitor the occupancy of the network bandwidth and the transmission delay. When congestion occurs in some links, the SDN controller reallocates the network resources.

[0017] Preferably, the process of the edge computing device receiving the machine tool operation data and performing data preprocessing and abnormal data detection in step S30 includes:

[0018] The edge computing device receives the sensor data in real time and performs data preprocessing;

[0019] Set the abnormal data threshold for preliminary abnormal judgment. When the machine tool operation data is greater than the set abnormal data threshold, it is preliminarily determined as abnormal data, and a first-level abnormal warning signal is generated, and the relevant data and warning information are stored in the local database;

[0020] When no abnormality is found in the preliminary abnormality judgment, the machine tool operation data is further analyzed through a machine tool abnormality data detection model based on a machine learning algorithm. When the model determines that the machine tool operation data is abnormal, a secondary abnormality warning signal is generated, and the relevant data and warning information are stored in the local database;

[0021] The local database sorts and stores the abnormal operation data of the machine tool according to the time sequence and the abnormality level, which is convenient for subsequent query and analysis. For frequently occurring abnormal situations, the parameters in the abnormality detection algorithm are automatically updated to improve the detection accuracy.

[0022] Among them, the steps for the edge computing device to receive sensor data in real time and perform data preprocessing include:

[0023] Noise removal: According to the machine tool operation environment and the characteristics of the sensor itself, analyze the noise types and characteristics of various data, and use appropriate filters to remove noise respectively. For example, due to the large number of electrical equipment in the workshop, the noise generated by electromagnetic interference often appears as superimposed high-frequency spike signals. A low-pass filter is selected, and the cut-off frequency is set slightly higher than the highest frequency signal generated during the normal operation of the machine tool to filter out high-frequency noise. For example, if the highest vibration frequency during the normal operation of the machine tool is 500Hz, the cut-off frequency of the low-pass filter is set to 600Hz to allow the normal vibration signal to pass through and remove high-frequency electromagnetic interference; The mechanical vibration of the machine tool itself transmitted to the sensor installation site will cause low-frequency fluctuation noise. For low-frequency mechanical vibration noise, a high-pass filter is used, and the cut-off frequency is set at a position lower than the lowest frequency signal generated during the normal operation of the machine tool to remove low-frequency interference; For temperature data, the data corresponding to the sudden jump in the slow change trend is determined as noise and removed; For vibration data, the part exceeding the normal vibration frequency range in the high-frequency band is determined as noise and removed;

[0024] Set the standard format and convert: Convert the machine tool operation data into a unified standard data format. The standard format of the machine tool operation data is set as a decimal coding format, the data length is 4 bytes, the high bits are in the front and the low bits are in the back, and the floating-point number precision is reserved to two decimal places. For the original data formats of different sensors, mapping rules are formulated to convert them into the standard coding format.

[0025] Among them, the steps for constructing and training the machine tool abnormality data detection model based on a machine learning algorithm include:

[0026] Dataset Preparation: Obtain data samples of normal and abnormal operations from the historical operation data of the machine tool, including temperature data, vibration frequency data, and current data. After data preprocessing, divide them into a training set, a validation set, and a test set according to the ratio of 70%:15%:15% respectively. And label the data samples in the training set. Normal data is labeled as 0, and abnormal data is labeled with a unique identifier according to its type. For example, spindle overheating is labeled as 1, tool wear is labeled as 2, etc.;

[0027] Determine the Loss Function and Optimizer: Use multi-class cross-entropy as the loss function and the Adam optimization algorithm as the optimizer to adaptively adjust the learning rate and dynamically update the model parameters. The initial learning rate is set to 0.0015 and decays according to the number of training epochs;

[0028] Model Construction: Build a model using the SVM support vector machine combined with the LSTM long short-term memory network. The entire model includes an LSTM layer and an SVM classifier; The LSTM layer includes an input gate, a forget gate, a candidate state layer, a state update layer, and an output gate; The input of the input gate is the labeled training set; The forget gate generates a forgetting factor through the Sigmoid function to determine the retention and forgetting of the input information of the input gate. The weight matrix in the forget gate is continuously optimized during model training to maintain the balance between information retention and forgetting; The candidate state layer generates a candidate state vector through the tanh activation function based on the current input data; The state update layer updates the current state through element-wise multiplication and addition according to the information filtered by the input gate and the candidate state vector generated by the candidate state layer; The output gate calculates the output vector of the LSTM layer at the current moment through the sigmoid function and the tanh function based on the current input and the updated state; The SVM classifier includes an input layer, a kernel function layer, and a decision layer; The input layer receives the output vector of the LSTM layer, including the key information left after the analysis and processing of the machine tool operation data by the LSTM layer, such as the feature vector representation after the LSTM layer captures the trends of machine tool vibration and temperature changes, as the direct input for SVM classification; The kernel function layer uses the Gaussian kernel function to perform a non-linear transformation on the data received by the input layer, mapping the data that is linearly inseparable in the low-dimensional space to a high-dimensional space to enhance the separability of the data; The decision layer determines the decision boundary based on the principle of maximizing the margin and outputs the classification result, including whether the machine tool operation data is normal or abnormal. When the output indicates that the machine tool operation data is abnormal, the corresponding machine tool fault type is also output;

[0029] Model Training and Validation: After the model is constructed, set the model parameters and use the divided training set as the input to train the model. After each training epoch, use the divided validation set to validate the trained model;

[0030] Model evaluation: When overfitting occurs, stop the model training and validation operations, and use the divided test set for model evaluation;

[0031] Model optimization: Adjust the model parameters according to the model evaluation results and retrain until the optimal combination of model parameters is obtained, and determine the corresponding model version

[0032] Preferably, the steps of the cloud server using big data analysis algorithms to analyze the trend of abnormal machine tool operation in step S40 include:

[0033] Data aggregation and integration: The cloud server regularly collects abnormal machine tool operation data information from each edge computing device, including the model number, serial number, abnormal occurrence time, abnormal type, and abnormal data value of the machine tool;

[0034] Time series analysis: According to the collected abnormal machine tool data information, construct a time series model, and use the sliding window technique to select abnormal data windows at different time scales for analysis, including the past 1 hour, the past 1 day, and the past 1 week;

[0035] Association analysis: In the abnormal machine tool operation data information regularly collected by the cloud server from each edge computing device, use the Apriori algorithm to analyze the combinations of abnormal machine tool operations that occur simultaneously;

[0036] Data visualization: Convert the abnormal machine tool operation trend obtained from time series analysis and the abnormal machine tool association status obtained from association analysis into visual charts. Use bar charts to display the frequencies of different abnormal machine tool types, line charts to display the trend of abnormal machine tool operation over time, and network charts to display the association status between abnormal machine tool operations.

[0037] In addition, to achieve the above object, the present invention also proposes an edge computing system for abnormal machine tool operation data, and the edge computing system for abnormal machine tool operation data includes:

[0038] Machine tool operation data acquisition module: Deploy sensors at key parts of the machine tool to collect machine tool operation data and transmit it to an edge computing device with computing capabilities and a local database;

[0039] Machine tool operation data transmission module: During the data transmission process, use SDN network technology to dynamically optimize the data transmission path, and allocate the optimal transmission path for different abnormal machine tool operation data according to the real-time state of the machine tool operation and data priority;

[0040] Edge computing module: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and detection results in the local database, and regularly uploads them to the cloud server;

[0041] Collaboration and Optimization Module: The cloud server uses big data analysis algorithms to analyze the trend of abnormal operation of machine tools, find the correlations between abnormalities of different machine tools, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device makes optimization adjustments based on the feedback suggestions.

[0042] In the machine tool operation data acquisition module, the sensors deployed at key parts of the machine tool include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time. The installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocols.

[0043] In addition, to achieve the above object, the present invention also proposes an edge computing device for machine tool abnormal operation data. The device includes: a memory, a processor, and programs such as a machine learning-based machine tool abnormal data detection algorithm stored on the memory and executable on the processor. The programs such as the machine learning-based machine tool abnormal data detection algorithm are for implementing the steps of a method for edge computing of machine tool abnormal operation data as described above.

[0044] In addition, to achieve the above object, the present invention also provides a computer program product. The computer program product includes programs such as a machine learning-based machine tool abnormal data detection algorithm. When the programs such as the machine learning-based machine tool abnormal data detection algorithm are executed by a processor, they implement a method for edge computing of machine tool abnormal operation data as described above.

[0045] The advantages and effects of the present invention are:

[0046] A method and system for edge computing of machine tool abnormal operation data proposed by the present invention use the method of combining edge computing with artificial intelligence and big data analysis for data acquisition and control, reducing the processing delay of machine tool abnormal operation data. At the same time, the SDN network is used to allocate the data transmission path, optimizing the transmission path configuration, avoiding network congestion, and improving the processing efficiency of machine tool abnormal operation data. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0048] Figure 1 It is a flowchart of a method for edge computing of machine tool abnormal operation data of the present invention.

[0049] Figure 2 This is a schematic structural diagram of an edge computing system for abnormal operation data of a machine tool according to the present invention.

[0050] Figure 3 This is a schematic block diagram of an electronic device for edge computing of abnormal operation data of a machine tool according to the present invention. Specific embodiments

[0051] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0052] The present invention provides a method for edge computing of abnormal operation data of a machine tool, as Figure 1 shown, including the following steps:

[0053] Step S10: Deploy sensors at key parts of the machine tool to collect machine tool operation data and transmit it to an edge computing device with computing capabilities and a local database.

[0054] Among them, the sensors deployed at key parts of the machine tool in step S10 include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time; the installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocols.

[0055] Step S20: Use SDN network technology to dynamically optimize the data transmission path during the data transmission process, and allocate the optimal transmission path for different machine tool abnormal operation data according to the real-time state of the machine tool operation and data priority.

[0056] Specifically, the steps of using SDN network technology to dynamically optimize the data transmission path during the data transmission process in step S20 and allocating the optimal transmission path for different machine tool abnormal operation data according to the real-time state of the machine tool operation and data priority include:

[0057] Construct a workshop edge computing network architecture: Deploy network devices supporting SDN technology inside the workshop where the machine tool is located, including switches and routers, to form a communication network covering the entire workshop; enable each machine tool to access this network and perform data interaction with the edge computing device;

[0058] Data Priority Classification Setting: Classify the data according to the importance and urgency of various operation data generated by the machine tool, and preset the priority classification rules in the edge computing device; for example, set the data directly related to the safety of the machine tool equipment and causing serious failures, such as spindle temperature, motor current overload information, etc., as the highest priority; set the data used for equipment performance monitoring, such as conventional vibration amplitude, etc., as the medium priority; set the data used for daily status recording, such as standby duration, as the lowest priority;

[0059] Real-time Monitoring of Machine Tool Operating Status: Real-time monitor the operating status of the machine tool through the installed sensors, and analyze the operating status of the machine tool and the change trend of data under each operating status based on the machine tool operating data collected by the sensors;

[0060] Formulation of Dynamic Path Allocation Strategy: According to the set data priority and the analyzed machine tool operating status, formulate a dynamic data transmission path allocation strategy; when an abnormal situation occurs in a certain machine tool, such as a sharp rise in spindle temperature, which is the highest priority, the SDN controller immediately triggers an instruction to reconfigure the data flow at the network switch and router levels, and directly transmit the relevant data of the machine tool to the main control edge server in the workshop through a dedicated high-speed channel to ensure that the data can be processed in the shortest time; for the standby state data, which is the lowest priority, it automatically switches to a low-speed and energy-saving transmission path to avoid network congestion and ensure the reasonable utilization of network resources;

[0061] Network Resource Allocation and Optimization: During the transmission of machine tool operation data, continuously monitor the occupancy of network bandwidth and transmission delay. When congestion occurs in some links, the SDN controller reallocates network resources; for example, temporarily increase the bandwidth of the high-priority data transmission link, or adjust the transmission time slots of other low-priority data to maintain the efficient and stable operation of the entire edge network and ensure that all abnormal operation data of various machine tools can be transmitted according to the optimized path.

[0062] Step S30: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and the detection results in the local database, and uploads them to the cloud server regularly.

[0063] Specifically, the process of the edge computing device performing data preprocessing and abnormal data detection after receiving the machine tool operation data in step S30 includes:

[0064] The edge computing device receives sensor data in real time and performs data preprocessing;

[0065] Set an abnormal data threshold for preliminary abnormal judgment. When the machine tool operation data is greater than the set abnormal data threshold, it is preliminarily determined as abnormal data, and a first-level abnormal warning signal is generated, and the relevant data and warning information are stored in the local database;

[0066] When no abnormality is found in the preliminary abnormal judgment, the machine tool operation data is further analyzed through a machine tool abnormal data detection model based on a machine learning algorithm. When the model determines that the machine tool operation data is abnormal, a second-level abnormal warning signal is generated, and the relevant data and warning information are stored in the local database;

[0067] The local database sorts and stores the machine tool abnormal operation data according to the time sequence and abnormal level, which is convenient for subsequent query and analysis. For frequently occurring abnormal situations, the parameters in the abnormal detection algorithm are automatically updated to improve the detection accuracy.

[0068] Among them, the steps for the edge computing device to receive sensor data in real time and perform data preprocessing include:

[0069] Noise removal: According to the machine tool operation environment and the characteristics of the sensor itself, analyze the noise types and characteristics of various data, and use appropriate filters to remove noise respectively. For example, due to the large number of electrical equipment in the workshop, the noise generated by electromagnetic interference often appears as superimposed high-frequency spike signals. A low-pass filter is selected, and the cut-off frequency is set slightly higher than the highest frequency signal generated during normal operation of the machine tool to filter out high-frequency noise. For example, when the highest vibration frequency during normal operation of the machine tool is 500Hz, the cut-off frequency of the low-pass filter is set to 600Hz to allow normal vibration signals to pass through and remove high-frequency electromagnetic interference; The mechanical vibration of the machine tool itself transmitted to the sensor installation site will cause low-frequency fluctuation noise. For low-frequency mechanical vibration noise, a high-pass filter is used, and the cut-off frequency is set at a position lower than the lowest frequency signal generated during normal operation of the machine tool to remove low-frequency interference; For temperature data, the data corresponding to the sudden jump in the slow change trend is determined as noise and removed; For vibration data, the part of the high-frequency band that exceeds the normal vibration frequency range is determined as noise and removed;

[0070] Set the standard format and convert: Convert the machine tool operation data into a unified standard data format. The standard format of the machine tool operation data is set as a decimal encoding format, the data length is 4 bytes, the high-order bits are in front and the low-order bits are behind, and the floating-point precision is reserved to two decimal places. For the original data formats of different sensors, mapping rules are formulated to convert them into the standard encoding format. For example, for the hexadecimal temperature sensor data, a conversion function is set to convert the hexadecimal number into a decimal integer, and conversion is performed according to the range and precision of the temperature sensor to obtain the temperature value that meets the requirements of the standard format, ensuring the compatibility of the data in the subsequent processing process.

[0071] Among them, the steps for constructing and training the machine tool abnormal data detection model based on machine learning algorithms include:

[0072] Dataset preparation: Obtain the data samples of normal and abnormal operations in the historical operation data of the machine tool, including temperature data, vibration frequency data, and current data. After data preprocessing, they are divided into a training set, a validation set, and a test set according to the ratio of 70%:15%:15% respectively. And the data samples in the training set are labeled. Normal data is labeled as 0, and abnormal data is labeled with a unique identifier according to the type. For example, spindle overheating is labeled as 1, tool wear is labeled as 2, etc.;

[0073] Determine the loss function and optimizer: Use multi-class cross-entropy as the loss function, and the Adam optimization algorithm as the optimizer to adaptively adjust the learning rate and dynamically update the model parameters. The initial learning rate is set to 0.0015 and is attenuated and adjusted according to the number of training epochs;

[0074] Model construction: Use the SVM support vector machine combined with the LSTM long short-term memory network to construct the model. The entire model includes an LSTM layer and an SVM classifier; the LSTM layer includes an input gate, a forget gate, a candidate state layer, a state update layer, and an output gate; the input of the input gate is the labeled training set; the forget gate generates a forget factor through the Sigmoid function to judge the retention and forgetting of the input information of the input gate. The weight matrix in the forget gate is continuously optimized during model training to maintain the balance between information retention and forgetting; the candidate state layer generates a candidate state vector through the tanh activation function according to the current input data; the state update layer updates the current state through element-wise multiplication and addition based on the information filtered by the input gate and the candidate state vector generated by the candidate state layer; the output gate calculates the output vector of the LSTM layer at the current moment through the sigmoid function and the tanh function according to the current input and the updated state; the SVM classifier includes an input layer, a kernel function layer, and a decision layer; the input layer receives the output vector of the LSTM layer, including the key information left after the analysis and processing of the machine tool operation data by the LSTM layer, such as the feature vector representation after the LSTM layer captures the vibration and temperature change trends of the machine tool, as the direct input for SVM classification; the kernel function layer uses the Gaussian kernel function to perform a non-linear transformation on the data received by the input layer, mapping the data that is linearly inseparable in the low-dimensional space to a high-dimensional space to enhance the separability of the data; the decision layer, based on the principle of maximizing the margin, determines the decision boundary according to the trained vector and outputs the classification result, including whether the machine tool operation data is normal or abnormal. When the output indicates that the machine tool operation data is abnormal, the corresponding machine tool fault type is also output;

[0075] Model training and validation: After the model is built, set the model parameters. Use the divided training set as the input to train the model. After each round of training, use the divided validation set to validate the trained model.

[0076] Model evaluation: When overfitting occurs, stop the model training and validation operations, and use the divided test set for model evaluation.

[0077] Model optimization: Adjust the model parameters according to the model evaluation results and retrain until the optimal combination of model parameters is obtained, and determine the corresponding model version.

[0078] Step S40: The cloud server uses big data analysis algorithms to analyze the trend of abnormal machine tool operations, find the correlations between different machine tool abnormalities, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device makes optimization adjustments according to the feedback suggestions.

[0079] Specifically, the steps for the cloud server to analyze the trend of abnormal machine tool operations using big data analysis algorithms in step S40 include:

[0080] Data aggregation and integration: The cloud server regularly collects machine tool operation abnormal data information from each edge computing device, including the model number, serial number, abnormal occurrence time, abnormal type, and abnormal data value of the machine tool.

[0081] Time series analysis: According to the collected machine tool abnormal data information, construct a time series model, and use the sliding window technique to select abnormal data windows at different time scales for analysis, including the past 1 hour, the past 1 day, and the past 1 week. For example, for a certain model of machine tool, observe the occurrence frequency of spindle overheating abnormalities within the past 1 day. By counting the number of abnormalities in different time periods, draw a curve of abnormal frequency changing with time to visually display the development trend of the spindle overheating problem of this machine tool.

[0082] Association analysis: In the machine tool operation abnormal data information regularly collected by the cloud server from each edge computing device, use the Apriori algorithm to analyze the combinations of simultaneously occurring machine tool operation abnormalities. For example, when it is found that multiple machine tools in the workshop simultaneously show slight vibration abnormalities and the hardness of the processed material is relatively high, subsequent spindle overheating or tool breakage and other abnormalities may occur, prompting the workshop management personnel to adjust the processing technology in a timely manner or perform preventive maintenance on the machine tools.

[0083] Data visualization: Convert the machine tool operation abnormal trends obtained from time series analysis and the machine tool abnormal associations obtained from association analysis into visual charts. Use bar charts to display the occurrence frequencies of different machine tool abnormal types, line charts to display the trends of machine tool operation abnormalities changing with time, and network charts to display the association status between machine tool operation abnormalities.

[0084] In addition, the present invention also provides an edge computing system for abnormal operation data of machine tools. Please refer to Figure 2 , and the edge computing system for abnormal operation data of machine tools includes:

[0085] Machine tool operation data acquisition module: Sensors are deployed at key parts of the machine tool to collect machine tool operation data and transmit it to an edge computing device with computing capabilities and a local database;

[0086] Machine tool operation data transmission module: During the data transmission process, the SDN network technology is used to dynamically optimize the data transmission path, and the optimal transmission path is allocated for different abnormal operation data of the machine tool according to the real-time state of the machine tool operation and the data priority;

[0087] Edge computing module: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and the detection results in the local database, and uploads them to the cloud server regularly;

[0088] Collaboration and optimization module: The cloud server uses big data analysis algorithms to analyze the trend of abnormal machine tool operation, find the correlations between different machine tool abnormalities, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device makes optimization adjustments according to the feedback suggestions;

[0089] Among them, the sensors deployed at key parts of the machine tool in the machine tool operation data acquisition module include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time; and the installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocols.

[0090] The edge computing system for abnormal operation data of machine tools provided by the present application adopts the edge computing method for abnormal operation data of machine tools in the above embodiment, and can solve the technical problem of low efficiency of the traditional edge computing method for abnormal operation data of machine tools. Compared with the prior art, the beneficial effects of the edge computing system for abnormal operation data of machine tools provided by the present application are the same as those of the edge computing method for abnormal operation data of machine tools provided by the above embodiment, and the other technical features in the edge computing system for abnormal operation data of machine tools are the same as the features disclosed in the method of the above embodiment, and will not be elaborated here.

[0091] The present application provides an edge computing device for abnormal operation data of a machine tool. The edge computing device for abnormal operation data of a machine tool includes: at least one processor; and a memory communicatively connected to the at least one processor. Wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute an edge computing method for abnormal operation data of a machine tool in Embodiment 1 above.

[0092] Reference is made below to Figure 3 , which shows a schematic structural diagram of an edge computing device for abnormal operation data of a machine tool suitable for implementing the embodiments of the present application. An edge computing device for abnormal operation data of a machine tool in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Descriptions: tablet computers), PMPs (Portable Media Players), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The edge computing device for abnormal operation data of a machine tool shown is only an example and should not impose any limitation on the functions and scope of use of the embodiments of the present application.

[0093] Figure 3An edge computing device for abnormal operation data of a machine tool as shown may include a processing system 1001 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1002 or a program loaded from a storage system 1003 into a random access memory (RAM) 1004. In the RAM 1004, various programs and data required for the operation of an edge computing device for abnormal operation data of a machine tool are also stored. The processing system 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input system 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output system 1008 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage system 1003 including, for example, a magnetic tape, a hard disk, etc.; and a communication system 1009. The communication system 1009 may allow an edge computing device for abnormal operation data of a machine tool to communicate with other devices wirelessly or wiredly to exchange data. Although an edge computing device for abnormal operation data of a machine tool with various systems is shown in the figure, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be alternatively implemented or had.

[0094] Particularly, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program may be downloaded and installed from a network through the communication system, or installed from the storage system 1003, or installed from the ROM 1002. When the computer program is executed by the processing system 1001, the above functions defined in the methods of the embodiments disclosed in the present application are executed.

[0095] An edge computing device for abnormal operation data of machine tools provided by this application, adopting an edge computing method for abnormal operation data of machine tools in the above-mentioned embodiment, can solve the technical problem of low efficiency of traditional edge computing methods for abnormal operation data of machine tools. Compared with the prior art, the beneficial effects of the edge computing device for abnormal operation data of machine tools provided by this application are the same as those of the edge computing method for abnormal operation data of machine tools provided by the above-mentioned embodiment, and other technical features in this edge computing device for abnormal operation data of machine tools are the same as those disclosed in the method of the previous embodiment, which will not be elaborated here.

[0096] Each part disclosed in this application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0097] This application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, it implements the steps of an edge computing method for abnormal operation data of machine tools as described above.

[0098] The computer program product provided by this application can solve the technical problem of low efficiency of traditional edge computing methods for abnormal operation data of machine tools. Compared with the prior art, the beneficial effects of the computer program product provided by this application are the same as those of the edge computing method for abnormal operation data of machine tools provided by the above-mentioned embodiment, which will not be elaborated here.

[0099] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalent technologies, the present invention is also intended to include these changes and modifications.

Claims

1. An edge computing method for abnormal operation data of a machine tool, characterized in that, The method includes the following steps: Step S10: Deploy sensors at key parts of the machine tool to collect the operation data of the machine tool and transmit it to an edge computing device with computing capabilities and a local database; Step S20: During the data transmission process, use the SDN network technology to dynamically optimize the data transmission path, and allocate the optimal transmission path for different machine tool abnormal operation data according to the real-time state of the machine tool operation and the data priority; Step S30: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and the detection results in the local database, and uploads them to the cloud server regularly; Step S40: The cloud server uses big data analysis algorithms to analyze the trend of machine tool operation anomalies, find the correlations between different machine tool anomalies, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device makes optimization adjustments according to the feedback suggestions; The sensors deployed at key parts of the machine tool in step S10 include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time. The installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocols.

2. The edge computing method for abnormal operation data of a machine tool according to claim 1, characterized in that The steps of using the SDN network technology to dynamically optimize the data transmission path and allocate the optimal transmission path for different machine tool abnormal operation data according to the real-time state of the machine tool operation and the data priority in step S20 include: Construct the workshop edge computing network architecture: Deploy network devices supporting SDN technology inside the workshop where the machine tool is located, including switches and routers, to form a communication network covering the entire workshop; Set data priority classification: Divide the data according to the importance and urgency of various operation data generated by the machine tool, and preset the priority division rules in the edge computing device; Monitor the machine tool operation state in real time: Monitor the operation state of the machine tool in real time through the installed sensors, and analyze the operation state of the machine tool and the change trend of the data in each operation state according to the machine tool operation data collected by the sensors; Formulate a dynamic path allocation strategy: According to the set data priority and the analyzed machine tool operation state, formulate a dynamic data transmission path allocation strategy; Network resource allocation and optimization: During the transmission process of the machine tool operation data, continuously monitor the occupancy of the network bandwidth and the transmission delay. When congestion occurs in some links, the SDN controller reallocates the network resources.

3. A method for edge computing of abnormal operation data of a machine tool according to claim 1, characterized in that, The process of the edge computing device performing data preprocessing and abnormal data detection after receiving the machine tool operation data in step S30 includes: The edge computing device receives the sensor data in real time and performs data preprocessing; Set the abnormal data threshold for preliminary abnormal judgment. When the machine tool operation data is greater than the set abnormal data threshold, it is initially determined as abnormal data, and a first-level abnormal warning signal is generated, and the relevant data and warning information are stored in the local database; When no abnormality is found in the preliminary abnormality judgment, the machine tool operation data is further analyzed through the machine tool abnormality data detection model based on the machine learning algorithm. When the model determines that the machine tool operation data is abnormal, a secondary abnormality warning signal is generated, and the relevant data and warning information are stored in the local database; The local database sorts and stores the abnormal operation data of the machine tool according to the time sequence and the abnormality level, which is convenient for subsequent query and analysis.

4. A method for edge computing of abnormal operation data of a machine tool according to claim 3, characterized in that, The steps for the edge computing device to receive sensor data in real time and perform data preprocessing include: Noise removal: According to the machine tool operation environment and the characteristics of the sensor itself, analyze the noise types and characteristics of various data, and use appropriate filters to remove the noise respectively; Set the standard format and convert: Convert the machine tool operation data into a unified standard data format. Set the standard data format of the machine tool operation data as the decimal encoding format, the data length is 4 bytes, the high-order bits are in the front and the low-order bits are in the back, and the floating-point precision is reserved to two decimal places. For the original data formats of different sensors, formulate mapping rules and convert them into the standard encoding format.

5. The edge computing method for abnormal operation data of a machine tool according to claim 3, characterized in that, The steps for constructing and training the machine tool abnormality data detection model based on the machine learning algorithm include: Dataset preparation: Obtain the data samples of normal and abnormal operations in the historical operation data of the machine tool, including temperature data, vibration frequency data and current data. After data preprocessing, they are divided into training set, validation set and test set according to the ratio of 70%:15%:15% respectively, and the data samples are labeled in the training set data. Normal data is labeled as 0, and abnormal data is labeled with a unique identifier according to the type; Determine the loss function and optimizer: Use the multi-class cross-entropy as the loss function, and the Adam optimization algorithm is used as the optimizer to adaptively adjust the learning rate and dynamically update the model parameters. The initial learning rate is set to 0.0015 and is adjusted according to the number of training epochs; Model construction: An SVM (Support Vector Machine) combined with an LSTM (Long Short-Term Memory) network is used to construct the model. The entire model includes an LSTM layer and an SVM classifier. The LSTM layer includes an input gate, a forget gate, a candidate state layer, a state update layer, and an output gate. The input of the input gate is the labeled training set. The forget gate generates a forgetting factor through the Sigmoid function to determine the retention and forgetting of the input information of the input gate. The weight matrix in the forget gate is continuously optimized during model training to maintain the balance between information retention and forgetting. The candidate state layer generates a candidate state vector through the tanh activation function based on the current input data. The state update layer updates the current state through element-wise multiplication and addition based on the information filtered by the input gate and the candidate state vector generated by the candidate state layer. The output gate calculates the output vector of the LSTM layer at the current moment through the sigmoid function and the tanh function based on the current input and the updated state. The SVM classifier includes an input layer, a kernel function layer, and a decision layer. The input layer receives the output vector of the LSTM layer. The kernel function layer performs a non-linear transformation on the data received by the input layer using the Gaussian kernel function. The decision layer determines the decision boundary based on the principle of maximizing the margin and outputs the classification result, including whether the machine tool operation data is normal or abnormal. When the output indicates that the machine tool operation data is abnormal, the corresponding machine tool fault type is also output. Model training and verification: After the model is constructed, set the model parameters, and use the divided training set as the input to train the model. After each round of training, use the divided verification set to verify the trained model. Model evaluation: Stop the model training and verification operations when overfitting occurs, and use the divided test set for model evaluation. Model optimization: Adjust the model parameters according to the model evaluation results and retrain until the optimal combination of model parameters is obtained, and determine the corresponding model version.

6. A method for edge computing of abnormal operation data of a machine tool according to claim 1, characterized in that The steps in which the cloud server analyzes the trend of abnormal machine tool operation using big data analysis algorithms in step S40 include: Data aggregation and integration: The cloud server regularly collects machine tool operation abnormal data information from each edge computing device, including the model, number, abnormal occurrence time, abnormal type, and abnormal data value of the machine tool. Time series analysis: According to the collected machine tool abnormal data information, construct a time series model, and use the sliding window technique to select abnormal data windows of different time scales for analysis, including the past 1 hour, the past 1 day, and the past 1 week. Association analysis: In the machine tool operation abnormal data information regularly collected by the cloud server from each edge computing device, use the Apriori algorithm to analyze the combinations of machine tool operation abnormalities that occur simultaneously. Data visualization: Convert the machine tool operation abnormal trend obtained from time series analysis and the machine tool abnormal association status obtained from association analysis into visual charts. Use bar charts to show the frequencies of different machine tool abnormal types, line charts to show the trend of machine tool operation abnormalities over time, and network charts to show the association status between machine tool operation abnormalities.

7. An edge computing system for abnormal operation data of a machine tool, characterized in that, The described edge computing system for abnormal machine tool operation data includes: Machine tool operation data acquisition module: Sensors are deployed at key parts of the machine tool to collect machine tool operation data and transmit it to an edge computing device with computing capabilities and a local database; Machine tool operation data transmission module: During the data transmission process, the SDN network technology is used to dynamically optimize the data transmission path, and the optimal transmission path is allocated for different machine tool abnormal operation data according to the real-time state of the machine tool operation and the data priority; Edge computing module: After receiving the machine tool operation data, the edge computing device performs data preprocessing and abnormal data detection, stores the machine tool operation data and the detection results in the local database, and regularly uploads them to the cloud server; Collaboration and optimization module: The cloud server uses big data analysis algorithms to analyze the trend of machine tool operation abnormalities, find the correlations between different machine tool abnormalities, and feedback suggestions to the edge computing device according to the analysis results. The edge computing device makes optimization adjustments according to the feedback suggestions; The sensors deployed at key parts of the machine tool in the machine tool operation data acquisition module include installing temperature sensors at the bearing parts of the machine tool spindle to collect the temperature data of the bearings in real time; installing vibration sensors on the motor housing of the machine tool to collect the vibration frequency data during motor operation in real time; installing current sensors in the electrical control cabinet of the machine tool to collect the current data of each circuit in real time. The installed sensors are connected to the edge computing device through industrial Ethernet or wireless communication protocols.

8. An edge computing device for abnormal operation data of a machine tool, characterized in that, The edge computing device for abnormal operation data of a machine tool includes: A memory, a processor, and a machine tool abnormal operation data edge computing program stored on the memory and executable on the processor. When the machine tool abnormal operation data edge computing program is executed by the processor, it implements a machine tool abnormal operation data edge computing method according to any one of claims 1 to 6.

9. A computer program product, characterized in that, The computer program product includes a machine tool abnormal operation data edge computing program. When the machine tool abnormal operation data edge computing program is executed by the processor, it implements a machine tool abnormal operation data edge computing method according to any one of claims 1 to 6.

Images