Apparatus and method for inspecting cutting tools
The cutting tool inspection device addresses the inefficiency of manual waveform estimation by automatically estimating and aligning time ranges for tool abnormalities detection, ensuring accurate analysis without timing signals.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- AZBIL CORP
- Filing Date
- 2022-07-27
- Publication Date
- 2026-06-22
AI Technical Summary
Conventional methods for detecting abnormalities in machine tool tools fail to extract power or current waveforms for specific processes when timing signals are not supplied, necessitating manual estimation which is inefficient.
A cutting tool inspection device that estimates a time range for specific processes using waveform characteristics and extracts training and inspection data, aligning time positions for accurate comparison.
Enables automated detection of tool abnormalities even without timing signals, correcting for timing discrepancies within machine tool cycles.
Smart Images
Figure 0007877109000001 
Figure 0007877109000002 
Figure 0007877109000003
Abstract
Description
Technical Field
[0001] The present invention relates to a tool inspection apparatus and method for detecting abnormalities in tools of a machine tool.
Background Art
[0002] Conventionally, as technologies for detecting abnormalities in tools of a machine tool, there are technologies disclosed in Patent Document 1 and Patent Document 2. In the technologies disclosed in Patent Document 1 and Patent Document 2, it is determined whether an abnormality has occurred in the tool by comparing the power waveform or current waveform of a motor that drives the tool with the waveform during normal operation.
[0003] By the way, a machine tool has many tools in one machine tool, and different tools are applied according to the cutting process. Therefore, in order to automatically determine the abnormalities of different tools for each process, it is necessary to compare with the waveform during normal operation for each process.
[0004] However, when a timing signal of, for example, a finishing process is not supplied from the machine tool, it is necessary to estimate the section of the finishing process on the diagnostic device side and extract the power waveform or current waveform of the finishing process. However, in the conventional technology, such a waveform extraction function has not been realized.
Prior Art Documents
Patent Documents
[0005] [[ID=Z30]]
Patent Document 1
Patent Document 2
Summary of the Invention
Problems to be Solved by the Invention
[0006] The present invention has been made to solve the above problems, and aims to provide a cutting tool inspection device and method that can extract training data and inspection data for a specific process even when a timing signal for a specific process is not supplied from the machine tool. [Means for solving the problem]
[0007] The present invention provides a cutting tool inspection device comprising: a data acquisition unit configured to acquire one batch of waveform data of the current supplied to a motor that drives a cutting tool of a machine tool; a training data generation unit configured to estimate a time range corresponding to a specific process of the machine tool based on the characteristics of the waveform data when generating training data for comparison with inspection data of the cutting tool, and to extract waveform data within this time range as the training data; and an inspection data generation unit configured to extract waveform data at the same time position as the time range estimated by the training data generation unit from the waveform data acquired by the data acquisition unit during the inspection of the cutting tool as the inspection data.
[0008] Furthermore, in one example configuration of the cutting tool inspection device of the present invention, the inspection data generation unit divides the training data into blocks along the time direction and aligns the time positions of the training data and the inspection data for each block.
[0009] Furthermore, in one configuration example of the cutting tool inspection apparatus of the present invention, the training data generation unit includes a first level determination unit configured to determine a first threshold level for extracting waveform data within the range of the finishing process of the machine tool from one batch of waveform data acquired by the data acquisition unit when the training data is generated; a first extraction processing unit configured to extract, from the waveform data of one batch acquired by the data acquisition unit when the training data is generated, a range of values below the first threshold level immediately before the end of the batch as waveform data within the range of the finishing process; a second level determination unit configured to determine a second threshold level for detecting the rising edge of the waveform within the range of the finishing process; and within the range of the finishing process, the current value first reaches the second threshold level. The present invention comprises: an extraction length determination unit configured to determine the extraction length of training data by adding a predetermined time to the length of time from the point in time exceeding a threshold to the end of the batch; a second extraction processing unit configured to extract data from one batch of waveform data acquired by the data acquisition unit when the training data is generated, where the current value is below the first threshold level and the length is within the range of the extraction length, just before the end of the batch, as the training data; and a third extraction processing unit configured to extract data from one batch of waveform data acquired by the data acquisition unit when the cutting tool is inspected, where the current value is below the first threshold level and the length is within the range of the extraction length, just before the end of the batch, as the inspection data.
[0010] Furthermore, in one configuration example of the cutting tool inspection apparatus of the present invention, the first level determination unit, the first cutting processing unit, the second level determination unit, and the cutting length determination unit each process the waveform data for one batch acquired multiple times by the data acquisition unit, the second cutting processing unit uses the average value of the first threshold level and the average value of the cutting length obtained in the multiple processing times to extract data from the waveform data for one batch acquired by the data acquisition unit when generating the training data, just before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the cutting length, as the training data, and the third cutting processing unit extracts data from the waveform data for one batch acquired by the data acquisition unit when inspecting the cutting tool, just before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the cutting length, as the inspection data.
[0011] Furthermore, in one configuration example of the cutting tool inspection apparatus of the present invention, the training data generation unit includes: a first level determination unit configured to determine a first threshold level for extracting waveform data within the range of the finishing process of the machine tool from one batch of waveform data acquired by the data acquisition unit when the training data is generated; a first extraction processing unit configured to extract, from the waveform data of one batch acquired by the data acquisition unit when the training data is generated, a range of values below the first threshold level immediately before the end of the batch as waveform data within the range of the finishing process; a second level determination unit configured to determine a second threshold level for detecting the rising edge of the waveform within the range of the finishing process; an extraction length determination unit configured to determine the length of the training data extraction length as the length of time from the time when the current value first exceeds the second threshold level within the range of the finishing process to the end of the batch plus a predetermined time; and, from the waveform data of one batch acquired by the data acquisition unit when the training data is generated, the current value below the first threshold level immediately before the end of the batch The inspection data generation unit includes a second cutting processing unit configured to cut out data with a current value below the threshold level and a length within the range of the cutting length as training data, the inspection data generation unit includes a third cutting processing unit configured to cut out data with a current value below the first threshold level and a length within the range of the cutting length, immediately before the end of a batch, from the waveform data for one batch acquired by the data acquisition unit during the inspection of the cutting tool, as inspection data, the data immediately before the end of the batch, the data with a current value below the first threshold level and a length within the range of the cutting length, the inspection data generation unit includes a third level determination unit configured to determine a third threshold level for detecting the start timing of position correction for the inspection data based on the training data, a block extraction unit configured to define the start timing of position correction as the point in time when the current value exceeds the third threshold level in the training data, and the range between the start timings of position correction in the training data as a position correction block, and a correlation coefficient calculation unit configured to calculate the correlation coefficient between the position correction block of the training data and the inspection data in a block section at the same time position as the position correction block for each position correction block.The system is characterized by comprising a correction unit configured to move the time position of the inspection data in the block section for each position correction block so as to maximize the correlation coefficient.
[0012] Furthermore, in one configuration example of the cutting tool inspection apparatus of the present invention, the first level determination unit, the first cutting processing unit, the second level determination unit, and the cutting length determination unit each process the waveform data for one batch acquired multiple times by the data acquisition unit, and the second cutting processing unit uses the average value of the first threshold level and the average value of the cutting length obtained in the multiple processing to select from the waveform data for one batch acquired by the data acquisition unit when generating the training data the current value is less than or equal to the average value of the first threshold level and the length is the cutting length immediately before the end of the batch. The third extraction processing unit extracts data within the range of average values as training data, extracts data from one batch of waveform data acquired by the data acquisition unit during the inspection of the cutting tool, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the extraction length, just before the end of the batch, as inspection data, the third level determination unit determines the third threshold level based on the average data of the plurality of training data, and the block extraction unit extracts the position correction blocks for the average data of the plurality of training data.
[0013] Furthermore, the cutting tool inspection method of the present invention is characterized by including: a first step of acquiring one batch of waveform data of the current supplied to a motor that drives a cutting tool of a machine tool; a second step of estimating a time range corresponding to a specific process of the machine tool based on the characteristics of the waveform data when generating training data for comparison with inspection data of the cutting tool, and extracting waveform data of this time range as the training data; and a third step of extracting waveform data at the same time position as the time range estimated by the second step from the waveform data acquired in the first step when inspecting the cutting tool as the inspection data.
[0014] Furthermore, in one example configuration of the cutting tool inspection method of the present invention, the third step is characterized by including the step of dividing the training data into block units along the time direction and aligning the time positions of the training data and the inspection data for each block. [Effects of the Invention]
[0015] According to the present invention, by providing a training data generation unit and an inspection data generation unit, even if timing signals for specific processes such as finishing processes are not supplied from the machine tool, the tool inspection device can estimate the interval of a specific process and extract training data and inspection data for that specific process.
[0016] Furthermore, in this invention, the inspection data generation unit divides the training data into blocks along the time direction, and aligns the time positions of the training data and the inspection data for each block. This makes it possible to correct for time discrepancies that occur in the machining timing of the machine tool in units of the control timing cycle inside the machine tool. [Brief explanation of the drawing]
[0017] [Figure 1] Figure 1 is a block diagram showing the configuration of a cutting tool inspection system according to an embodiment of the present invention. [Figure 2] Figure 2 is a block diagram showing the configuration of a cutting tool inspection device according to an embodiment of the present invention. [Figure 3] Figure 3 is a flowchart illustrating the training data generation process according to an embodiment of the present invention. [Figure 4] Figure 4 is a flowchart illustrating the training data generation process according to an embodiment of the present invention. [Figure 5] Figure 5 shows an example of the current waveform for one batch supplied to the motor. [Figure 6] Figure 6 shows an example of a current waveform within the finishing process. [Figure 7]FIG. 7 is a diagram showing an example of a current waveform for a cut length within the range of the finishing process. [Figure 8] FIG. 8 is a flowchart for explaining inspection data generation processing according to an embodiment of the present invention. [Figure 9] FIG. 9 is a diagram showing an example of an average waveform of teacher data for the number of learning times. [Figure 10] FIG. 10 is a diagram showing a position correction block of teacher data, a waveform cut out from the position correction block, and a waveform cut out from inspection data. [Figure 11] FIG. 11 is a block diagram showing a configuration example of a computer that realizes a tool inspection apparatus according to an embodiment of the present invention.
Embodiments of the Invention
[0018] Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a tool inspection system according to an embodiment of the present invention. The tool inspection system includes a tool inspection apparatus 1, a CT (Current Transformer) 2 that is a current transformer for converting a current supplied to a motor that drives a tool of a machine tool 5 into a current of a magnitude that can be handled by the tool inspection apparatus 1, a display 3 for displaying inspection results and the like of the tool by the tool inspection apparatus 1, and an external storage device 4 that stores data obtained by the tool inspection apparatus 1.
[0019] The machine tool 5 includes a tool (not shown) for machining a workpiece that is a workpiece to be machined, a motor 50 that drives the tool, a control unit 51 that controls the motor 50, and a PLC (Programmable Logic Controller) 52 that controls the entire machine tool.
[0020] Figure 2 is a block diagram showing the configuration of the cutting tool inspection device 1. The cutting tool inspection device 1 consists of a data acquisition unit 10 that acquires waveform data of the current supplied to the motor 50 in one batch via the CT2, a training data generation unit 11 that estimates a time range corresponding to the finishing process of the machine tool 5 based on the characteristics of the waveform data when generating training data for comparison with inspection data of the cutting tool, and extracts waveform data within this time range as training data, a training data storage unit 12 that stores the training data, an inspection data generation unit 13 that extracts waveform data at the same time position as the time range estimated by the training data generation unit 11 from the waveform data acquired by the data acquisition unit 10 during the inspection of the cutting tool as inspection data, an inspection data storage unit 14 that stores the inspection data, a determination unit 15 that compares the training data and the inspection data to determine whether an abnormality has occurred in the cutting tool, and a management unit 16 that receives input from the operator and sets data acquisition conditions.
[0021] The training data generation unit 11 includes a level determination unit 110 that determines a first threshold level for extracting waveform data for the finishing process range of the machine tool 5 from one batch of waveform data acquired by the data acquisition unit 10 when generating training data, an extraction processing unit 111 that extracts from the waveform data for one batch acquired by the data acquisition unit 10 when generating training data the range of values below the first threshold level immediately before the end of the batch as waveform data for the finishing process range, and a second threshold for detecting the rising edge of the waveform within the finishing process range. The system consists of a level determination unit 112 that determines the threshold level, a cut-out length determination unit 113 that determines the length of the cut-out section of the training data as the length obtained by adding a predetermined time to the length of time from the moment the current value first exceeds the second threshold level within the range of the finishing process until the end of the batch, a cut-out processing unit 114 that cuts out data as training data from the waveform data for one batch acquired by the data acquisition unit 10 when the training data is generated, where the current value is below the first threshold level and the length is within the range of the cut-out section, just before the end of the batch, and a storage unit 115.
[0022] The inspection data generation unit 13 consists of: an extraction processing unit 131 that extracts data from one batch of waveform data acquired by the data acquisition unit 10 during tool inspection, where the current value is below a first threshold level and the length is within the range of the extraction length, just before the end of the batch, as inspection data; a level determination unit 132 that determines a third threshold level for detecting the start timing of position correction for the inspection data based on the training data; a block extraction unit 133 that defines the point in time when the current value exceeds the third threshold level in the training data as the start timing of position correction, and defines the range between the start timings of position correction in the training data as a position correction block; a correlation coefficient calculation unit 134 that calculates the correlation coefficient between the position correction block of the training data and the inspection data in the block section at the same time position as the position correction block for each position correction block; a correction unit 135 that moves the time position of the inspection data in the block section for each position correction block so as to maximize the correlation coefficient; and a storage unit 136.
[0023] First, we will explain how to generate training data for tool inspection. Figures 3 and 4 are flowcharts illustrating the training data generation process. The operator of the cutting tool inspection device 1 operates the terminal 6 when the cutting tool of the machine tool 5 is in a normal state, and instructs the cutting tool inspection device 1 to generate training data.
[0024] When the cutting tool inspection device 1 receives instructions from the operator via the control unit 16, it performs a predetermined number of parameter determination processes to extract current waveform data for the workpiece finishing process from the current waveform data supplied to the motor 50 of the machine tool 5 as training data (Figure 3, steps S100, S101).
[0025] In the cutting parameter determination process, the data acquisition unit 10 of the cutting tool inspection device 1 acquires current waveform data for one batch from the start of the rough cutting process to the end of the finishing process of the workpiece via the CT2 (Figure 4, step S200). The data acquired by the data acquisition unit 10 is stored in the storage unit 115 of the training data generation unit 11. A timing signal roughly indicating one batch can be obtained from the PLC 52. An example of the current waveform for one batch is shown in Figure 5(A). Figure 5(B) shows the waveform of the portion of the current waveform in Figure 5(A) that is extracted as training data or inspection data.
[0026] Next, the training data generation unit 11 of the cutting tool inspection device 1 determines a threshold level (hereinafter referred to as THLEVEL) for extracting current waveform data within the finishing process range from the current waveform data for one batch acquired by the data acquisition unit 10 (Figure 4, step S201). Specifically, the level determination unit 110 within the training data generation unit 11 sets the THLEVEL to a current value that is half the current peak value in the current waveform data for one batch. Alternatively, a histogram of the current for one batch may be created and the median value may be set as the THLEVEL.
[0027] Then, the extraction processing unit 111 within the training data generation unit 11 selects the last point in time when the current value changes from below THLEVEL (Low) to above THLEVEL (High) in the current waveform data for one batch acquired by the data acquisition unit 10 as the trigger point TP (end of batch), and extracts the range of current values below THLEVEL immediately before the trigger point TP as the current waveform data for the finishing process range SH (Figure 4, step S202). The finishing process range SH is shown in Figure 6.
[0028] As shown in Figure 6, the reason why the finishing process range SH can be extracted is that the data for one batch has the following characteristics: in the first half, the current waveform fluctuates greatly due to rough cutting and other operations, in the second half, the finishing process consists of waveforms with small peaks, and at the end of one batch, the waveform rises sharply once due to the back electromotive force when the motor stops, thus ending the process.
[0029] Next, the level determination unit 112 within the training data generation unit 11 determines a threshold level (hereinafter referred to as UPLEVEL) for detecting the rising edge of the current waveform within the finishing process range SH (Figure 4, step S203). Specifically, the level determination unit 112 classifies the current waveform data of the finishing process range SH into two clusters, for example, using the K-means method. Then, the level determination unit 112 sets the current value at the midpoint of the centroids of the two clusters as UPLEVEL.
[0030] Next, the cut-out length determination unit 113 within the training data generation unit 11 determines the length of the training data cut-out (hereinafter referred to as THLENGTH) as the length obtained by adding the length T1 of the time from the point in the finishing process range SH when the current value first exceeds UPLEVEL to the trigger point TP mentioned above, to the length T2 of the predetermined header range (Figure 4, step S204). With this, the two cut-out parameters, THLEVEL and THLENGTH, can be determined for one batch of current waveform data, and the cut-out parameter determination process is completed. An example of the current waveform for THLENGTH within the finishing process range SH is shown in Figure 7.
[0031] The data acquisition unit 10 and the training data generation unit 11 perform the extraction parameter determination process in step S101 (S200~S204) each time they acquire one batch of current waveform data.
[0032] After performing the extraction parameter determination process a predetermined number of times, the extraction processing unit 114 in the training data generation unit 11 sets the average value of multiple THLEVELs obtained from the current waveform data for the number of parameter determinations as THLEVELLa, and the average value of multiple THLENGTHs obtained from the current waveform data for the number of parameter determinations as THLENGTHa (Figure 3, step S102).
[0033] Next, the extraction processing unit 114 performs the process of extracting training data from one batch of current waveform data stored in the storage unit 115 a predetermined number of times (Figure 3, steps S103, S104). Specifically, the extraction processing unit 114 extracts data from one batch of current waveform data stored in the storage unit 115 as training data, where the current value is less than or equal to THLEVELa and the length is in the range of THLENGTHa, immediately before the trigger point TP.
[0034] The extraction processing unit 114 stores the extracted training data in the training data storage unit 12 (Figure 3, step S105). Similarly, the trigger point TP (end of batch) is the last point in time in the current waveform data for one batch stored in the storage unit 115 when the current value changes from below THLEVELa (Low) to above THLEVELa (High).
[0035] The training data generation process ends when the training data has been extracted and saved for a predetermined number of training cycles (NO in step S103 of Figure 3). Since the data acquisition unit 10 acquires current waveform data for one batch the number of times parameter determination is performed during the parameter determination process, the number of learning iterations should be less than or equal to the number of parameter determination iterations.
[0036] Next, the operator of the cutting tool inspection device 1 operates the terminal 6 to instruct the cutting tool inspection device 1 to perform an inspection of the cutting tools of the machine tool 5. When the cutting tool inspection device 1 receives instructions from the operator via the control unit 16, it performs inspection data generation processing. Figure 8 is a flowchart illustrating the inspection data generation processing.
[0037] The data acquisition unit 10 of the cutting tool inspection device 1 acquires current waveform data for one batch via the CT2 (Figure 8, step S300). The data acquired by the data acquisition unit 10 is stored in the storage unit 136 within the inspection data generation unit 13.
[0038] The extraction processing unit 131 within the test data generation unit 13 extracts data from one batch of current waveform data stored in the storage unit 136, specifically the data immediately preceding the trigger point TP, where the current value is less than or equal to THLEVELa and the length is within the range of THLENGTHa, as test data (Figure 8, step S301). The test data is stored in the storage unit 136 within the test data generation unit 13. Similarly, the trigger point TP (end of batch) is the last point in the current waveform data for one batch stored in the storage unit 136 where the current value changes from less than or equal to THLEVELa (Low) to more than THLEVELa (High).
[0039] Next, the level determination unit 132 within the inspection data generation unit 13 determines a threshold level (hereinafter referred to as UPLEVEL2) for detecting the start timing of position correction for the inspection data, based on the average data of the training data for the number of training iterations stored in the training data storage unit 12 (Figure 8, step S302). Specifically, the level determination unit 132 classifies the average data of the training data for the number of training iterations into two clusters, for example, using the K-means method. Then, the level determination unit 132 sets the current value at the midpoint of the centroids of the two clusters as UPLEVEL2.
[0040] Next, the block extraction unit 133 within the inspection data generation unit 13 defines the point at which the current value exceeds UPLEVEL2 in the average data of the training data for the number of training iterations as the start timing TC for position correction for the inspection data, and defines the range from the start timing TC of the average data of the training data to the next start timing TC as the position correction block CB (Figure 8, step S303). An example of the average waveform of the training data for the number of training iterations is shown in Figure 9.
[0041] The correlation coefficient calculation unit 134 in the inspection data generation unit 13 extracts data for a window width before and after centered on the position correction start timing TC on the start side of the position correction block CB from the position correction block CB for which the correlation coefficient is to be calculated, and also extracts data for a window width before and after centered on the position correction start timing TC from the inspection data in the block section at the same time position as the position correction block CB. Then, it calculates the correlation coefficient between the data extracted from the position correction block CB and the data extracted from the inspection data (Fig. 8, step S305). The block section of the inspection data refers to the range at the same time position as the position correction block CB when the start of the average data of the teacher data is set to time 0 and the start of the inspection data is set to time 0.
[0042] Fig. 10(A) shows the position correction block CB of the teacher data, Fig. 10(B) shows the waveform extracted from the position correction block CB, and Fig. 10(C) shows the waveform extracted from the inspection data in the block section.
[0043] Next, the correlation coefficient calculation unit 134 moves the time position of the inspection data by a predetermined time t (t < window) (Fig. 8, step S306), extracts data for a window width before and after centered on the position correction start timing TC on the start side of the position correction block CB from the position correction block CB for which the correlation coefficient is to be calculated, and also extracts data for a window width before and after centered on the position correction start timing TC from the inspection data in the block section at the same time position as the position correction block CB after moving the time position. Then, it calculates the correlation coefficient between the data extracted from the position correction block CB and the data extracted from the inspection data again (step S305).
[0044] The correlation coefficient calculation unit 134 repeatedly calculates the correlation coefficient while moving the time position of the inspection data within a predetermined time range (for example, TC - window to TC + window) centered on the position correction start timing TC on the start side of the position correction block CB for which the correlation coefficient is to be calculated.
[0045] After calculating all correlation coefficients within a predetermined time range (YES in step S307 of Figure 8), the correction unit 135 in the inspection data generation unit 13 uses the position correction start timing TC on the starting side of the position correction block CB as the reference position, determines the amount of deviation of the time position where the correlation coefficient was maximized relative to the reference position, and updates the inspection data stored in the storage unit 136 by the above amount of deviation, only for the inspection data of the block section for which the correlation coefficient was calculated among the inspection data stored in the storage unit 136 (step S308 of Figure 8).
[0046] In this way, the time position of the position correction block CB and the time position of the inspection data for the corresponding block section can be aligned. Note that in step S306, the time position of the entire inspection data is moved, but the inspection data stored in the storage unit 136 is not updated. On the other hand, in step S308, only the time position of the inspection data for the block section is changed, and the inspection data stored in the storage unit 136 is updated.
[0047] Furthermore, the correction unit 135 only moves the time position of the inspection data for the block section for which the correlation coefficient is calculated. For example, if it is moved in the direction of time delay, the data at the trailing end of this block section will be deleted by the amount of the above-mentioned shift. In addition, since a blank area without data is created at the leading end of the block section, the correction unit 135 interpolates the data by adding data at the leading end of the block section equal to the amount of the above-mentioned shift from the trailing end of the previous block section.
[0048] Furthermore, if the correction unit 135 moves the inspection data of the block section for which the correlation coefficient is calculated in a direction that advances time, it will delete the data at the leading edge of the block section by the amount of the above-mentioned shift. In addition, since a blank area without data will be created at the trailing edge of the block section, the correction unit 135 interpolates the data by adding data at the trailing edge of the block section equal to the amount of the above-mentioned shift at the leading edge of the immediately following block section.
[0049] The inspection data generation unit 13 performs the processing in steps S305 to S308 for each position correction block CB of the average data of the training data. When processing is completed for all position correction blocks CB, the correction unit 135 saves the inspection data stored in the storage unit 136 to the inspection data storage unit 14 (Figure 8, step S309).
[0050] The determination unit 15 of the cutting tool inspection device 1 compares the average data of the training data stored in the training data storage unit 12 with the inspection data stored in the inspection data storage unit 14 to determine whether an abnormality has occurred in the cutting tool of the machine tool 5. The determination method at this time is a well-known technique, as disclosed in, for example, Patent Document 1 and Patent Document 2. Furthermore, the present invention is not limited to a specific determination method.
[0051] The determination unit 15 displays, for example, the determination result, the waveform of the average data of the training data, and the waveform of the test data on the display unit 3. The determination unit 15 also stores the determination result, the average data of the training data, and the test data in the external storage device 4.
[0052] As described above, in this embodiment, even if the machine tool 5 does not supply a timing signal for the finishing process, the tool inspection device 1 can estimate the finishing process interval and extract the training data and inspection data for the finishing process. Furthermore, although a time difference may occur in the machining timing of the machine tool 5 in units of the internal control timing cycle of the machine tool, in this embodiment, the time positions of the training data and inspection data can be aligned in block units, and the time difference can be corrected.
[0053] The cutting tool inspection device 1 described in this embodiment can be realized by a computer equipped with a CPU (Central Processing Unit), a storage device, and an interface, and a program that controls these hardware resources. An example of the configuration of this computer is shown in Figure 11.
[0054] The computer comprises a CPU 200, a storage device 201, and an interface device (I / F) 202. A display 3, an external storage device 4, a machine tool 5, a terminal 6, etc., are connected to the I / F 202. In such a computer, the program for implementing the cutting tool inspection method of the present invention is stored in the storage device 201. The CPU 200 executes the processing described in this embodiment according to the program stored in the storage device 201. Furthermore, at least a part of the cutting tool inspection device 1 may be implemented in hardware. [Industrial applicability]
[0055] This invention can be applied to techniques for detecting abnormalities in cutting tools of machine tools. [Explanation of symbols]
[0056] 1...Cutting tool inspection device, 2...CT, 3...Display unit, 4...External storage device, 5...Machine tool, 6...Terminal, 10...Data acquisition unit, 11...Training data generation unit, 12...Training data storage unit, 13...Inspection data generation unit, 14...Inspection data storage unit, 15...Determination unit, 16...Management unit, 50...Motor, 51...Control unit, 52...PLC, 110, 112, 132...Level determination unit, 111, 114, 131...Cutting processing unit, 113...Cutting length determination unit, 115, 136...Storage unit, 133...Block extraction unit, 134...Correlation coefficient calculation unit, 135...Correlation unit.
Claims
1. A data acquisition unit configured to capture one batch of waveform data of the current supplied to the motor that drives the cutting tool of a machine tool, A training data generation unit is configured to generate training data for comparison with inspection data of the cutting tool, estimate a time range corresponding to a specific process of the machine tool based on the characteristics of the waveform data, and extract waveform data within this time range as the training data. The system includes an inspection data generation unit configured to extract waveform data at the same time position as the time range estimated by the training data generation unit from the waveform data acquired by the data acquisition unit during the inspection of the cutting tool, as the inspection data, The cutting tool inspection device is characterized in that the inspection data generation unit divides the training data into blocks along the time direction and aligns the time positions of the training data and the inspection data for each block.
2. A data acquisition unit configured to acquire one batch of waveform data of the current supplied to a motor that drives a cutting tool of a machine tool, A training data generation unit is configured to generate training data for comparison with inspection data of the cutting tool, estimate a time range corresponding to a specific process of the machine tool based on the characteristics of the waveform data, and extract waveform data within this time range as the training data. The system includes an inspection data generation unit configured to extract waveform data at the same time position as the time range estimated by the training data generation unit from the waveform data acquired by the data acquisition unit during the inspection of the cutting tool, as the inspection data, The aforementioned training data generation unit, A first level determination unit is configured to determine a first threshold level for extracting waveform data within the range of the finishing process of the machine tool from one batch of waveform data acquired by the data acquisition unit when generating the training data, A first extraction processing unit is configured to extract, from the waveform data for one batch acquired by the data acquisition unit when generating the training data, the range of values below the first threshold level immediately before the end of the batch as waveform data for the finishing process, A second level determination unit configured to determine a second threshold level for detecting the rising edge of the waveform within the range of the finishing process, A cutting length determination unit is configured to determine the cutting length of the training data as the length obtained by adding a predetermined time to the length of time from the moment the current value first exceeds the second threshold level within the range of the finishing process until the end of the batch, The system includes a second extraction processing unit configured to extract, as the training data, data from one batch of waveform data acquired by the data acquisition unit during the generation of the training data, where the current value is below the first threshold level and the length is within the range of the extraction length, immediately before the end of the batch. The aforementioned inspection data generation unit is: A cutting tool inspection device characterized by comprising a third cutting processing unit configured to extract, as inspection data, data immediately before the end of a batch, where the current value is below the first threshold level and the length is within the range of the cutting length, from the waveform data of one batch acquired by the data acquisition unit during the inspection of the cutting tool.
3. In the cutting tool inspection device according to claim 2, The first level determination unit, the first extraction processing unit, the second level determination unit, and the extraction length determination unit each process the waveform data for one batch that the data acquisition unit has acquired multiple times. The second extraction processing unit uses the average value of the first threshold level and the average value of the extraction length obtained in the multiple processing steps to extract data from one batch of waveform data acquired by the data acquisition unit during the generation of the training data, specifically data immediately before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the extraction length, as the training data. The cutting tool inspection device is characterized in that the third cutting processing unit extracts data from one batch of waveform data acquired by the data acquisition unit during the inspection of the cutting tool, specifically data immediately before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the cutting length, as the inspection data.
4. In the cutting tool inspection device according to claim 1, The aforementioned training data generation unit, A first level determination unit is configured to determine a first threshold level for extracting waveform data within the range of the finishing process of the machine tool from one batch of waveform data acquired by the data acquisition unit when generating the training data, A first extraction processing unit is configured to extract, from the waveform data for one batch acquired by the data acquisition unit when generating the training data, the range of values below the first threshold level immediately before the end of the batch as waveform data for the finishing process, A second level determination unit configured to determine a second threshold level for detecting the rising edge of the waveform within the range of the finishing process, A cutting length determination unit is configured to determine the cutting length of the training data as the length obtained by adding a predetermined time to the length of time from the moment the current value first exceeds the second threshold level within the range of the finishing process until the end of the batch, The system includes a second extraction processing unit configured to extract, as the training data, data from one batch of waveform data acquired by the data acquisition unit during the generation of the training data, where the current value is below the first threshold level and the length is within the range of the extraction length, immediately before the end of the batch. The aforementioned inspection data generation unit is: A third cutting processing unit is configured to extract, as inspection data, data from one batch of waveform data acquired by the data acquisition unit during inspection of the cutting tool, where the current value is below the first threshold level and the length is within the range of the cutting length, just before the end of the batch, and the data is extracted as inspection data. A third level determination unit is configured to determine a third threshold level for detecting the start timing of position correction for the inspection data based on the aforementioned training data, A block extraction unit is configured such that, in the aforementioned training data, the point in time when the current value exceeds the third threshold level is defined as the position correction start timing, and the range between the position correction start timings in the training data is defined as a position correction block. A correlation coefficient calculation unit is configured to calculate the correlation coefficient between the position correction block of the training data and the inspection data of a block section located at the same time position as the position correction block, for each position correction block. A cutting tool inspection device comprising a correction unit configured to move the time position of the inspection data in the block section for each position correction block so as to maximize the correlation coefficient.
5. In the cutting tool inspection device according to claim 4, The first level determination unit, the first extraction processing unit, the second level determination unit, and the extraction length determination unit each process the waveform data for one batch that the data acquisition unit has acquired multiple times. The second extraction processing unit uses the average value of the first threshold level and the average value of the extraction length obtained in the multiple processing steps to extract data from one batch of waveform data acquired by the data acquisition unit during the generation of the training data, specifically data immediately before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the extraction length, as the training data. The third cutting processing unit extracts, from the waveform data for one batch acquired by the data acquisition unit during the inspection of the cutting tool, data immediately before the end of the batch, where the current value is less than or equal to the average value of the first threshold level and the length is within the range of the average value of the cutting length, as the inspection data. The third level determination unit determines the third threshold level based on the average data of the plurality of training data, The cutting tool inspection device is characterized in that the block extraction unit extracts the position correction blocks from the average data of a plurality of training data.
6. The first step is to acquire one batch of waveform data of the current supplied to the motor that drives the cutting tool of the machine tool, In generating training data for comparison with inspection data of the cutting tool, a second step is to estimate a time range corresponding to a specific process of the machine tool based on the characteristics of the waveform data, and to extract waveform data within this time range as the training data. The third step includes extracting waveform data from the waveform data acquired in the first step during the inspection of the cutting tool, specifically waveform data at the same time position as the time range estimated in the second step, as the inspection data. The third step is a cutting tool inspection method characterized by including the step of dividing the training data into blocks along the time direction and aligning the time positions of the training data and the inspection data for each block.