KR20210065581A - System and method for monitoring of machine tool abnormality - Google Patents

Abstract

The present invention provides a system and a method for monitoring machine tool abnormality which can measure the parameters of a machine tool, extract the characteristic data of the measured parameters, compare the determination reference range set for the parameters of the machine tool, and monitor the abnormal state of the machine tool. A system and a method for monitoring machine tool abnormality according to an embodiment of the present invention includes an edge computing device that collects data measured by a machine tool and extracts feature data from the collected data; and a server for receiving data from the edge computing device through mobile communication, and setting a determination reference range based on the received data. The edge computing device receives the determination reference range from the server through mobile communication, and monitors the abnormal state of the machine tool by comparing the characteristic data with the determination reference range.

Description

Machine tool abnormality monitoring system and method {System and method for monitoring of machine tool abnormality}

The present invention relates to a machine tool abnormal state monitoring system and method, and more particularly, by measuring the parameters of the machine tool, extracting characteristic data of the measured parameters, and comparing the judgment reference range set for the parameters of the machine tool. It relates to a machine tool abnormal state monitoring system and method for monitoring the abnormal state of

A machine tool is a machine that processes a material to make a desired shape. A machine tool consists of a rotation system part that can move at a desired speed in order to process a material accurately, and a feed system part that can be transported to an accurate position.

When using a machine tool, tool wear or breakage, bearing wear or damage, and spindle runout abnormality occur frequently. Solutions are being commercialized to solve problems such as wear or breakage of tools, wear or breakage of bearings, and abnormal spindle runout that occur when using machine tools.

These solutions judge machine tool abnormalities by judging by rule base or by judging how far from the average based on sampling values. However, such a solution has a problem in that it is not reliable in the actual field by performing monitoring based on a small number of samples.

The present invention is to solve the problem described above, by measuring the parameters of the machine tool, extracting characteristic data of the measured parameters, and comparing the judgment reference range set for the parameters of the machine tool to monitor the abnormal state of the machine tool. An object of the present invention is to provide a machine tool abnormal state monitoring system and method.

In addition to the technical problems of the present invention mentioned above, other features and advantages of the present invention will be described below, or will be clearly understood by those skilled in the art from such description and description.

A machine tool abnormal state monitoring system for monitoring an abnormal state of a machine tool according to an embodiment of the present invention for achieving the above-described object collects data measured in a machine tool, and an edge for extracting feature data from the collected data Comprising a computing device and a server for receiving data from the edge computing device through mobile communication, and setting a determination reference range based on the received data, the edge computing device receives the determination reference range from the server through mobile communication, , it is possible to monitor the abnormal state of the machine tool by comparing the characteristic data and the criterion range.

In addition, the machine tool abnormal state monitoring method for monitoring the abnormal state of the machine tool according to an embodiment of the present invention for achieving the above-described object collects data measured in the machine tool in the edge computing device, and from the collected data The steps of extracting the feature data, receiving data from the edge computing device through mobile communication in the server, and setting a judgment reference range based on the received data, and the judgment criterion from the server through mobile communication in the edge computing device It may include the step of receiving the range, and comparing the characteristic data and the judgment reference range to monitor the abnormal state of the machine tool.

In the present invention according to an embodiment of the present invention, the machine tool abnormal state monitoring system measures the parameters of the machine tool, extracts the characteristic data of the measured parameters, and compares the judgment reference range set for the parameters of the machine tool. Anomalies can be monitored.

In addition, by monitoring the abnormal state of the machine tool by using an edge computing device rather than a server, it is possible to determine the abnormal state of the machine tool without data delay.

In addition, by monitoring the abnormal state of the machine tool using an edge computing device rather than the server, the amount of data received by the server can be reduced.

In addition, since the server and the edge computing device communicate using mobile communication, data can be transmitted at a fast response speed, and data loss can be reduced in the data transmission/reception process.

In addition, other features and advantages of the present invention may be newly recognized through embodiments of the present invention.

1 is a diagram showing the configuration of a machine tool abnormal state monitoring system according to an embodiment of the present invention.
2 is a diagram showing the configuration of an edge computing device according to an embodiment of the present invention.
3 is a diagram showing the configuration of a server according to an embodiment of the present invention.
4 is an exemplary view for determining wear or breakage of a tool according to an embodiment of the present invention.
5 is an exemplary diagram for determining whether chatter occurs according to an embodiment of the present invention.
Figure 6 is an exemplary view for determining whether or not the expansion hole in accordance with an embodiment of the present invention.
7 is a view showing a machine tool abnormal state monitoring method according to an embodiment of the present invention.

In order to clearly explain the present invention, parts irrelevant to the description are omitted, and the same reference numerals are given to the same or similar elements throughout the specification.

Although not defined otherwise, all terms including technical and scientific terms used herein have the same meaning as commonly understood by those of ordinary skill in the art to which the present invention belongs. Commonly used terms defined in the dictionary are additionally interpreted as having a meaning consistent with the related technical literature and the presently disclosed content, and unless defined, they are not interpreted in an ideal or very formal meaning.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art can easily implement them. However, the present invention may be embodied in various different forms and is not limited to the embodiments described herein.

1 is a diagram showing the configuration of a machine tool abnormal state monitoring system according to an embodiment of the present invention.

Referring to FIG. 1 , a machine tool abnormal state monitoring system 1000 according to an embodiment of the present invention may include a machine tool 100 , an edge computing device 200 , a server 300 , and an analysis device 400 . have.

The edge computing device 200 may collect data measured by the machine tool 100 and monitor an abnormal state of the machine tool 100 using the collected data. The edge computing device 200 may determine the abnormal state of the machine tool 100 by extracting feature data from the collected data, and determining whether the extracted feature data exists within a judgment reference range. Here, the feature data extracted from the collected data may be valid data among data within a certain range.

As a result of determining the abnormal state of the machine tool 100 , the edge computing device 200 transmits an alarm notifying that an abnormality has occurred to the machine tool 100 when it is determined that there is an abnormality in the machine tool 100 , or It is possible to transmit a control signal capable of responding to an abnormal state occurring in (100).

Here, the control signal that can correspond to the abnormal state occurring in the machine tool 100 may be a signal for calling another tool in the group. For example, a plurality of the same tool may be grouped, and when it is determined that the current tool is in an abnormal state, it may be exchanged with another tool in the group by a control signal received from the edge computing device 200 . Accordingly, the operation of the machine tool 100 may not be stopped, and the productivity of the workpiece may be increased by not stopping the operation of the machine tool 100 .

In addition, the control signal that can respond to the abnormal state generated in the machine tool 100 may be a signal for operating an air blow (Air blow) to remove the chip. For example, in the case of inspecting hole processing, the diameter of the hole may be increased because the amount of processing is greater than a predetermined dimension, and a chip may be pinched between the spindle and the tool to rotate larger than the predetermined radius. When it is determined that the machine tool 100 is in an abnormal state, an air blow is operated by a control signal received from the edge computing device 200, and a strong wind is applied between the spindle and the tool by the operated air blow, so that the chip is can be removed.

The server 300 may receive the characteristic data and the result of determining the abnormal state of the machine tool 100 from the edge computing device 200 , and set the determination reference range based on the received characteristic data and the determination result. For example, the server 300 may generate a learning model by learning the characteristic data and the determination result received from the edge computing device 200 , and may set a determination reference range based on an output value of the learning model.

The determination reference range may be different depending on the type of data measured by the machine tool 100 . The server 300 may transmit the set determination reference range to the edge computing device 200 . Accordingly, the edge computing device 200 may determine whether the feature data exists within the determination reference range.

The server 300 may adjust the range of the set judgment reference range according to the user's manipulation. For example, when the emphasis is placed on production rather than the quality of the processed product generated by the machine tool 100, the user can adjust the range of the judgment standard range widely. Accordingly, the criteria for determining the abnormal state of the machine tool 100 can be widened.

On the other hand, when the emphasis is placed on quality rather than the production of the processed product generated by the machine tool 100 , the user can narrow the range of the judgment reference range. Accordingly, the criterion for determining the abnormal state of the machine tool 100 may be narrowed.

The analysis device 400 may analyze the quality of the workpiece generated by the machine tool 100 . A QR code or RFID chip may be attached to the workpiece generated by the machine tool 100, and the QR code or RFID chip may include a material ID, processing information, and process information of the workpiece. Accordingly, the analysis device 400 may read the material ID, processing information, and process information of the processed product by recognizing the QR code or RFID chip attached to the processed product.

The analysis device 400 may analyze the quality of the processed product, and transmit the quality analysis result of the processed product, the material ID of the processed product, processing information, and process information to the server 300 . The server 300 may match and store the quality analysis result of the processed product received from the analysis device 400 with the material ID, processing information, and process information of the processed product, and the stored quality analysis result of the processed product, the material ID of the processed product, and processing information , process information can be used to set the criteria for judgment. For this reason, the server 300 can more accurately set the determination reference range, and by analyzing the abnormal state of the machine tool 100 using the determination reference range set in the edge computing device 200, the Anomalies can also be analyzed more accurately. Due to this, the reliability of the machine tool abnormal state monitoring system 1000 may increase.

2 is a diagram showing the configuration of an edge computing device according to an embodiment of the present invention.

Referring to FIG. 2 , the edge computing device 200 according to an embodiment of the present invention includes a sensor 210 , a data collection unit 220 , a data processing unit 230 , a monitoring unit 240 , and a communication unit 250 . can do.

The sensor 210 may be attached to the machine tool 100 to measure parameters generated when the machine tool 100 operates. The parameter measured by the sensor 210 may include at least one of vibration, volume, temperature, flow rate, pneumatic/hydraulic pressure, and displacement. In addition, the sensor 210 is a vibration sensor for measuring the vibration of the machine tool 100, an acoustic sensor for measuring the sound of the machine tool 100, a temperature sensor for measuring the temperature of the machine tool 100, the machine tool 100 ) may include a flow sensor for measuring the flow rate, a pneumatic/hydraulic sensor for measuring the pneumatic/hydraulic pressure of the machine tool 100 , and a displacement sensor for measuring the displacement of the machine tool 100 . Data for each parameter measured by the sensor 210 may be transmitted to the data collection unit 220 .

The data collection unit 220 may collect parameters transmitted from the sensor 210 , and for example, the data collection unit 220 may include a data acquisition system (DAQ). In addition, the data collection unit 220 may receive the process information of the machine tool 100 , processing conditions, and input/output data of a PLC (Programmable Logic Controller) contact point from the machine tool 100 . Here, the machine tool 100 may include a separate communication module, and may transmit process information of the machine tool 100, processing conditions, and input/output data of PLC contacts to the data collection unit 220 through the communication module. . The machine tool 100 may transmit data to the data collection unit 220 through a network through Ethernet or the like.

The data processing unit 230 may process the data collected by the data collection unit 220 and extract feature data from the processed data. The data processing unit 230 may convert data into digital signals and extract effective feature data according to the type of data. For example, the data processing unit 230 may extract valid characteristic data for each of data transmitted from the vibration sensor and data transmitted from the temperature sensor.

The corresponding characteristic data are RMS (Root Mean Square), acceleration, displacement, FFT conversion result, moving average line graph, threshold value, axis temperature data, maximum, minimum, average, peak-to-peak, asymmetry ( skewness, kurtosis, crest factor, shape factor, impulse factor, rmse, rotating frequency multiple components, moving average, BPFI (Ball Passing Frequency Inner) Race), a Ball Passing Frequency Outer Race (BPFO), a Fundamental Train Frequency (FTF), a Ball Spin Frequency (BSF), and the like.

The monitoring unit 240 may compare the feature data extracted by the data processing unit 230 with the determination reference range received from the server 300 . The monitoring unit 240 may compare the determination criterion range for each characteristic data and the characteristic data. For example, in the case of the characteristic data for vibration and temperature, the monitoring unit 240 may compare it with a reference range for vibration and temperature.

The monitoring unit 240 may determine that the machine tool 100 is normal when the feature data is within the judgment reference range, and when the feature data is not within the judgment reference range, the machine tool 100 is abnormal. can judge

When it is determined that an abnormality has occurred in the machine tool 100 , the monitoring unit 240 may transmit an alarm for informing that the abnormality has occurred to the machine tool 100 through the communication unit 250 . In addition, the monitoring unit 240 may transmit a control signal capable of responding to an abnormal state occurring in the machine tool 100 to the machine tool 100 through the communication unit 250 .

The communication unit 250 may receive data from the machine tool 100 and the sensor 210 through a network such as Ethernet or a digital input/output module, and may transmit an alarm or a control signal to the machine tool 100 .

In addition, as a result of analyzing the abnormal state of the machine tool 100 in the monitoring unit 240 , the communication unit 250 transmits NC data and characteristic data including process information and processing conditions of the machine tool 100 to the server 300 . can be sent to Here, the communication unit 250 analyzes the abnormal state of the machine tool 100 with the server 300 using mobile communication such as LTE and 5G, and as a result, NC including process information and processing conditions of the machine tool 100 It is possible to transmit data and characteristic data, and receive a determination reference range from the server 300 . The communication unit 250 may transmit data to the server 300 or receive data from the server 300 using mobile communication to transmit data at a fast response speed, and loss of data in the process of transmitting and receiving data may occur. can decrease.

The machine tool abnormal state monitoring system 1000 according to an embodiment of the present invention can determine the abnormal state of the machine tool 100 without data delay by using the edge computing device 200 . That is, the data obtained by measuring the parameters generated when the machine tool 100 is operated is not transmitted to the server 300 for storage and analysis, but the data is analyzed by the edge computing device 200 so that the server 300 is received. It can reduce the amount of data. That is, the edge computing device 200 processes data, and transmits only some data to the server 300 , so the amount of data received by the server 300 can be reduced, which causes the server 300 to increase the load. can be reduced.

3 is a diagram showing the configuration of a server according to an embodiment of the present invention.

Referring to FIG. 3 , the server 300 according to an embodiment of the present invention may include a communication unit 310 , a database 320 , a data learning unit 330 , and a determination reference range setting unit 340 .

The communication unit 310 transmits the judgment reference range to the edge computing device 200 using mobile communication such as LTE, 5G, and analyzes the abnormal state of the machine tool 100 from the edge computing device 200. Process information of the machine 100, NC data including processing conditions, and feature data may be received.

In addition, the communication unit 310 may receive the quality inspection result of the processed product and the material ID, processing information, and process information of the processed product from the analysis device 400 .

The database 320 may store data received from the communication unit 310 . For example, the database 320 is a result of analyzing the abnormal state of the machine tool 100 received by the communication unit 310 from the edge computing device 200 , NC data including process information and processing conditions of the machine tool 100 . and feature data.

In addition, the database 320 may store the quality inspection result of the processed product received by the communication unit 310 from the analysis device 400 and the material ID, processing information, and process information of the processed product. Here, the database 320 may match and store the quality analysis result of the processed product with the material ID, processing information, and process information of the processed product. The data stored in the database 320 may be used by the determination reference range setting unit 340 to set the determination reference range.

The data learning unit 330 may generate a learning model by learning the data stored in the database 320 . For example, the data learning unit 330 learns data through deep learning using a Convolutional Neural Network (CNN), a Deep Neural Network (DNN), a Recurrent Neural Network (RNN), and a Generative Adversarial Network (GAN), etc., and a learning model can create

Here, the data learning unit 330 may learn data for each type of abnormal state of the machine tool 100 . For example, the type of abnormal state may be wear or breakage of bearings and tools, spindle runout, thermal displacement, chatter generation, and the like.

The determination criterion range setting unit 340 may select a learning model that outputs the most accurate output value among the learning models generated by the data learning unit 330 . The determination reference range setting unit 340 may set the determination reference range based on the output value of the selected learning model. For example, the determination reference range setting unit 340 may set a range from more than a predetermined value of the output value of the selected learning model to less than or equal to a predetermined value of the output value of the selected learning model as the determination reference range.

Also, the determination reference range setting unit 340 may set the determination reference range using values of data received through the communication unit 310 . For example, the determination reference range setting unit 340 may calculate a maximum value and a minimum value of the parameter, and may calculate an average value of the maximum and minimum values. The determination reference range setting unit 340 may set a value equal to or less than a predetermined size of the calculated average value as the determination reference range.

The determination reference range setting unit 340 may set the determination reference range and transmit the set determination reference range to the edge computing device 200 through the communication unit 310 . Accordingly, the edge computing device 200 may determine the abnormal state of the machine tool 100 by using the determination reference range transmitted from the determination reference range setting unit 340 .

The server 300 according to an embodiment of the present invention performs deep learning based on the values of the parameters measured in the machine tool 100, and then the values of the parameters measured in the machine tool 100 are also deep learning results. can be reflected in Accordingly. By reflecting the latest measured parameters in the machine tool 100, the edge computing device 200 may determine an abnormal state with respect to the current state of the machine tool 100. That is, the edge computing device 200 may determine the abnormal state of the machine tool 100 in the current state.

4 is an exemplary view for determining wear or breakage of a tool according to an embodiment of the present invention.

Referring to FIG. 4 , a solid line 30 indicates characteristic data, a gray line 10 indicates a reference value for a sampled value, and a green line 20 indicates a judgment reference range. .

The data collection unit 220 may collect parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions. For example, in order to determine wear or breakage of the tool, the data collection unit 220 may collect parameter information about revolutions per minute (RPM), feed, doc, vibration, and temperature generated in the machine tool 100 .

The data processing unit 230 may process the data collected by the data collection unit 220 . For example, the data processing unit 230 may convert the data collected by the data collection unit 220 into a digital signal, and perform Fast Fourier Transform (FFT) conversion on the digital signal. The data processing unit 230 may extract a signal within a predetermined range from the converted signal. Here, the predetermined range may be 1 Hz to 10000 Hz. The data processing unit 230 may extract a valid signal suitable for the machining tool from the extracted signal within a predetermined range, and extract feature data based on the amplitude value of the extracted valid signal. Here, the feature data may be in the form of a graph.

Here, the data processing unit 230 may calculate a maximum value and a minimum value for each predetermined section of the data. For example, the data processing unit 230 may calculate a maximum value and a minimum value for every 0.25 second interval for the vibration signal. The data processing unit 230 may calculate an average value of the calculated maximum value and minimum value, and arrange and graph the maximum value, minimum value, and average value in an array.

The database 320 may store parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions.

The data learning unit 330 may learn information stored in the database 320 . The data learning unit 330 may generate a learning model by learning information stored in the database 320 . The learning model may be configured in plurality, and may be configured for each type of abnormal state of the machine tool 100 .

The determination reference range setting unit 340 may convert the data collected in the database 320 into a digital signal, and FFT (Fast Fourier Transform) transform the signal converted into the digital signal. The determination reference range setting unit 340 may extract a signal within a predetermined range from the converted signal. Here, the predetermined range may be 1 Hz to 10000 Hz. The determination reference range setting unit 340 may extract a valid signal suitable for the machining tool from the extracted signal within a predetermined range, and extract feature data based on the amplitude value of the extracted valid signal. Here, the feature data may be in the form of a graph.

Here, the determination reference range setting unit 340 may calculate a maximum value and a minimum value for each predetermined section of the data. For example, the data processing unit 230 may calculate a maximum value and a minimum value for every 0.25 second interval for the vibration signal. The determination reference range setting unit 340 may calculate an average value of the calculated maximum value and minimum value, and arrange and graph the maximum value, minimum value, and average value in an array.

The determination reference range setting unit 340 may calculate a maximum value and a minimum value for valid signals every predetermined time, and may calculate an average value of the maximum value and the minimum value. The determination reference range setting unit 340 may determine the difference between the maximum value and the minimum value as a ratio, and set the determined ratio as the determination reference range.

Also, the determination criterion range setting unit 340 may select a learning model that outputs the most accurate output value among the learning models generated by the data learning unit 330 . The determination reference range setting unit 340 may set the determination reference range based on the output value of the selected learning model. The determination reference range setting unit 340 may transmit the determination reference range to the edge computing device 200 through the communication unit 310 .

The communication unit 210 of the edge computing device 200 may receive the determination reference range transmitted from the server 300 .

The monitoring unit 240 may compare the feature data 30 extracted by the data processing unit 230 with the determination reference range 20 received from the server 300 . The monitoring unit 240 may determine that the machine tool 100 is normal when the characteristic data 30 exists within the determination reference range 20, and the characteristic data 30 does not exist within the determination reference range 20. If not (32), it can be determined that wear or breakage of the tool has occurred.

The monitoring unit 240 calculates the maximum value and the minimum value for a certain range, and reflects the ratio of the difference between the maximum value and the minimum value to the average value and compares it with the characteristic data 30 based on the determination reference range 20 set have. At this time, the monitoring unit 240 may determine whether an abnormal state according to the current state of the machine tool 100 by reflecting the judgment reference range set through deep learning auxiliary. For example, as a result of comparing the monitoring unit 240 with the characteristic data 30 based on the determination criterion range 20 set by reflecting the ratio of the difference between the maximum value and the minimum value to the average value, the characteristic data 30 is the determination criterion may not be within range 20 . At this time, when the feature data 30 exists within the determination reference range set through deep learning, the monitoring unit 240 may determine that an abnormality does not occur in the machine tool 100 . This may vary depending on the user's settings or circumstances, and the feature data 30 does not exist within the judgment reference range 20 set by reflecting the ratio of the difference between the maximum value and the minimum value to the average value, and through deep learning Even if the characteristic data 30 exists within the set determination reference range, the monitoring unit 240 may determine that an abnormality has occurred in the machine tool 100 .

5 is an exemplary diagram for determining whether chatter occurs according to an embodiment of the present invention.

Referring to FIG. 5 , the data collection unit 220 may collect parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions. For example, in order to determine whether chatter occurs, the data collection unit 220 may collect parameter information on vibrations generated in the machine tool 100 .

The data processing unit 230 may process the data collected by the data collection unit 220 . For example, the data processing unit 230 may convert the data collected by the data collection unit 220 into a digital signal, and perform Fast Fourier Transform (FFT) conversion on the digital signal. The data processing unit 230 may extract a signal within a predetermined range from the converted signal.

The data processing unit 230 may extract a valid signal suitable for the machining tool from the extracted signal within a predetermined range, and extract feature data based on the amplitude value of the extracted valid signal. Here, the feature data may be in the form of a graph.

The database 320 may store parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions.

The data learning unit 330 may learn information stored in the database 320 . The data learning unit 330 may generate a learning model by learning information stored in the database 320 . The learning model may be configured in plurality, and may be configured for each type of abnormal state of the machine tool 100 .

The determination criterion range setting unit 340 may select a learning model that outputs the most accurate output value among the learning models generated by the data learning unit 330 . The determination reference range setting unit 340 may set the determination reference range based on the output value of the selected learning model. The determination reference range setting unit 340 may transmit the determination reference range to the edge computing device 200 through the communication unit 310 .

The communication unit 210 of the edge computing device 200 may receive the determination reference range transmitted from the server 300 .

The monitoring unit 240 may compare the feature data 30 extracted by the data processing unit 230 with the determination reference range 20 received from the server 300 . The monitoring unit 240 may determine that the machine tool 100 is normal when the characteristic data 30 exists within the determination reference range 20, and the characteristic data 30 does not exist within the determination reference range 20. If not (32), it can be determined that wear or breakage of the tool has occurred. Here, the monitoring unit 240 determines that the machine tool 100 is normal when the specific data 30 exists in the stable area, and determines that chatter has occurred in the machine tool 100 when it exists in the unstable area. can

As another embodiment, when the vibration level among the parameter information exceeds a threshold value, the determination reference range setting unit 340 may divide the vibration signal into a predetermined range and perform FFT conversion. For example, the predetermined range may be an interval of 10 Hz. Also, the determination reference range setting unit 340 may determine whether a frequency peak exists in the converted signal. As a result of determining whether a frequency peak exists in the converted signal, if a frequency peak exists, it is possible to determine whether a frequency peak exists by further subdividing a predetermined range. For example, the determination reference range setting unit 340 may determine whether a frequency peak exists by dividing the converted signal by an interval of 1 Hz.

The determination reference range setting unit 340 may sequentially search for frequency peaks when it is determined that a frequency peak exists. For example, the determination reference range setting unit 340 may determine whether there is a multiple component of the frequency and, if there is a multiple component, may delete the multiple component. Also, the determination reference range setting unit 340 searches for the next peak, and when the next peak to be searched for exists, determines whether there is a multiple component of the frequency and deletes the multiple component. The determination reference range setting unit 340 may repeat the above process until the next peak does not exist.

The monitoring unit 240 may determine the forced vibration when abnormal signals other than the blade frequency and the main shaft frequency exist in the signal except for the multiple component of the frequency. On the other hand, the monitoring unit 240 may determine that chatter has occurred when there is no abnormal signal other than the blade frequency and the main shaft frequency with respect to the signal processed by the data processing unit 230 .

The monitoring unit 240 can determine the abnormal state of the machine tool 100 according to the presence of abnormal signals other than the blade frequency and the spindle frequency. At this time, the monitoring unit 240 may determine whether an abnormal state according to the current state of the machine tool 100 by reflecting the judgment reference range set through deep learning auxiliary. For example, the monitoring unit 240 may determine that an abnormality has occurred in the machine tool 100 because there is an abnormal signal other than the blade frequency and the spindle frequency. At this time, when the feature data 30 exists within the determination reference range set through deep learning, the monitoring unit 240 may determine that an abnormality does not occur in the machine tool 100 . This may vary depending on the user's settings or circumstances, and even if there is an abnormal signal other than the blade frequency and the main axis frequency, and the characteristic data 30 exists within the judgment reference range set through deep learning, the monitoring unit 240 performs the work. It may be determined that an abnormality has occurred in the machine 100 .

Figure 6 is an exemplary view for determining whether or not the expansion hole according to an embodiment of the present invention.

Referring to FIG. 6 , the data collection unit 220 may collect parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions. For example, the data collection unit 220 may collect parameter information about vibrations occurring in the machine tool 100 in order to determine whether the drilling occurs. Here, the sensor 210 may measure the X-axis and Y-axis vibration by installing the vibration sensor at a right angle to the spindle of the tool.

The data processing unit 230 may process the data collected by the data collection unit 220 . For example, the data processing unit 230 may generate concentric circles by convolutionally synthesizing the data collected by the data collection unit 220 .

The database 320 may store parameter information generated when the machine tool 100 is driven, process information of the machine tool 100, and information on processing conditions.

The data learning unit 330 may learn information stored in the database 320 . The data learning unit 330 may generate a learning model by learning information stored in the database 320 . The learning model may be configured in plurality, and may be configured for each type of abnormal state of the machine tool 100 .

The determination criterion range setting unit 340 may select a learning model that outputs the most accurate output value among the learning models generated by the data learning unit 330 . The determination reference range setting unit 340 may set the determination reference range based on the output value of the selected learning model. The determination reference range setting unit 340 may transmit the determination reference range to the edge computing device 200 through the communication unit 310 . Here, the criterion range may be in the form of concentric circles.

The communication unit 210 of the edge computing device 200 may receive the determination reference range transmitted from the server 300 .

The monitoring unit 240 may compare the feature data 30 extracted by the data processing unit 230 with the determination reference range 20 received from the server 300 . The monitoring unit 240 may determine that the machine tool 100 is normal when the characteristic data 30 exists within the determination reference range 20, and the characteristic data 30 does not exist within the determination reference range 20. If not (32), it can be determined that the tool has a hole in it.

7 is a view showing a machine tool abnormal state monitoring method according to an embodiment of the present invention.

Referring to FIG. 7 , parameters generated during operation of the machine tool 100 may be measured and collected ( S100 ). When the sensor 210 measures a parameter generated when the machine tool 100 operates and transmits it to the data collection unit 220 , the data collection unit 220 collects information about the parameter received from the sensor 210 . can Here, the parameters may include vibration, volume, temperature, flow rate, pneumatic/hydraulic pressure, and displacement.

The data processing unit 230 may extract feature data from the data collected by the data collection unit 220 (S200). The data processing unit 230 may convert the data collected by the data collection unit 220 into digital signals and extract effective feature data according to the type of data. For example, the data processing unit 230 may extract valid characteristic data for each of data transmitted from the vibration sensor and data transmitted from the temperature sensor.

The monitoring unit 240 may compare the characteristic data with the determination reference range transmitted from the server 300 to determine whether the characteristic data exists within the determination reference range (S300).

As a result of comparing the characteristic data and the determination reference range, the monitoring unit 240 may determine that no abnormality has occurred in the machine tool 100 when the characteristic data is within the determination reference range (S400).

On the other hand, as a result of comparing the characteristic data and the determination reference range, the monitoring unit 240 may determine that an abnormality has occurred in the machine tool 100 when the characteristic data does not exist within the determination reference range (S500).

When it is determined that an abnormality has occurred in the machine tool 100 , an alarm may be transmitted to the machine tool 100 or a control signal may be transmitted ( S600 ). The monitoring unit 240 may transmit an alarm indicating that an abnormality has occurred to the machine tool 100 through the communication unit 250 . In addition, the monitoring unit 240 may transmit a control signal capable of responding to an abnormal state occurring in the machine tool 100 through the communication unit 250 .

As described above, according to an embodiment of the present invention, the machine tool measures the parameter, extracts the characteristic data of the measured parameter, and compares the judgment reference range set for the machine tool parameter to monitor the abnormal state of the machine tool. A machine tool abnormal state monitoring system and method can be realized.

Those skilled in the art to which the present invention pertains should understand that the present invention can be embodied in other specific forms without changing the technical spirit or essential characteristics thereof, so the embodiments described above are illustrative in all respects and not restrictive. only do The scope of the present invention is indicated by the following claims rather than the detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be construed as being included in the scope of the present invention. .

100: machine tool 200: edge computing device
300: server 400: analysis device
210: sensor 220: data collection unit
230: data processing unit 240: monitoring unit
250: communication unit 310: communication unit
320: database 330: data learning unit
340: judgment standard range setting unit

Claims (18)

In the machine tool abnormal state monitoring system for monitoring the abnormal state of the machine tool,
an edge computing device for collecting data measured by the machine tool and extracting feature data from the collected data; and
A server for receiving data from the edge computing device through mobile communication and setting a judgment reference range based on the received data,
The edge computing device receives the determination reference range from the server through mobile communication, and compares the characteristic data with the determination reference range to monitor the abnormal state of the machine tool.
According to claim 1,
Further comprising an analysis device for analyzing the quality of the workpiece generated by the machine tool,
The analysis device recognizes at least one of a QR code or an RFID chip attached to the workpiece to read the material ID, processing information, and process information of the workpiece.
3. The method of claim 2,
The analysis device transmits the result of analyzing the material ID, processing information, process information and the quality of the processed product to the server,
The server is a machine tool abnormal state monitoring system for setting a judgment reference range using the information received from the analysis device.
According to claim 1,
The edge computing device measures at least one parameter of vibration, volume, temperature, flow rate, pneumatic/hydraulic pressure, and displacement of the machine tool,
A machine tool abnormality monitoring system that converts measured data into digital signals and extracts valid characteristic data according to the type of data.
According to claim 1,
The edge computing device determines that the machine tool is in a normal state when the feature data is within the determination reference range, and determines that the machine tool is in an abnormal state when the feature data is not within the determination reference range machine tool abnormal condition monitoring system.
According to claim 1,
The server stores the data received from the edge computing device, a machine tool abnormal state monitoring system for learning the stored data to generate a learning model.
7. The method of claim 6,
The server selects a learning model that outputs the most accurate output value among the generated learning models, and sets the range from a value above a certain value to a value below a certain value of the output value of the selected learning model as a judgment reference range. monitoring system.
According to claim 1,
The edge computing device is a machine tool abnormal condition monitoring system for transmitting an alarm to the machine tool when it is determined that the machine tool is in an abnormal state.
According to claim 1,
The server adjusts the range of the judgment reference range set according to the user's operation, and the machine tool abnormal state monitoring system in which the quality of the workpiece generated by the machine tool is determined according to the range of the judgment reference range.
In the machine tool abnormal state monitoring method for monitoring the abnormal state of the machine tool,
collecting data measured by the machine tool in an edge computing device, and extracting feature data from the collected data;
receiving data from the edge computing device through mobile communication in a server, and setting a determination reference range based on the received data; and
Receive the determination reference range from the server through mobile communication in the edge computing device, and compare the characteristic data with the determination reference range to monitor the abnormal state of the machine tool. .
11. The method of claim 10,
The analysis device further comprises the step of analyzing the quality of the workpiece produced by the machine tool,
The step of analyzing the quality of the processed product,
A machine tool abnormal state monitoring method for recognizing at least one of a QR code or an RFID chip attached to the workpiece to read the material ID, processing information, and process information of the workpiece.
12. The method of claim 11,
The step of analyzing the quality of the processed product,
The material ID of the processed product, processing information, process information, and the result of analyzing the quality of the processed product are transmitted to the server,
The step of setting the criterion range is,
A method for monitoring an abnormal state of a machine tool for setting a judgment criterion range using the received material ID, processing information, process information, and the result of analyzing the quality of the processed product.
11. The method of claim 10,
The step of extracting the feature data,
Measuring at least one parameter of vibration, volume, temperature, flow rate, pneumatic/hydraulic pressure, and displacement of the machine tool,
A machine tool abnormal state monitoring method that converts measured data into digital signals and extracts valid characteristic data according to the type of data.
11. The method of claim 10,
The step of monitoring the abnormal state of the machine tool,
When the characteristic data is within the determination reference range, it is determined that the machine tool is in a normal state, and when the characteristic data is not within the determination reference range, the machine tool abnormal state is determined to be in an abnormal state. monitoring method.
11. The method of claim 10,
The step of setting the criterion range is,
A machine tool abnormal state monitoring method for storing data received from the edge computing device, and learning the stored data to generate a learning model.
16. The method of claim 15,
The step of setting the criterion range is,
A machine tool abnormal state monitoring method that selects a learning model that outputs the most accurate output value among the generated learning models, and sets the range from a value above a certain value to a value below a certain value of the output value of the selected learning model as a judgment reference range.
11. The method of claim 10,
The step of monitoring the abnormal state of the machine tool,
A machine tool abnormal state monitoring method for transmitting an alarm to the machine tool when it is determined that the machine tool is in an abnormal state.
11. The method of claim 10,
The step of setting the criterion range is,
A machine tool abnormal state monitoring method in which a range of a judgment reference range set according to a user's operation is adjusted, and the quality of a workpiece generated by the machine tool is determined according to the range of the judgment reference range.

Images