Method for detecting the mounting status of a tool holder and machine tool to which the tool holder is mounted.

Abstract

To enable reliable detection of a fact that a tool holder is attached to a spindle of a machine tool in an appropriate state when a tool is automatically replaced by using an automatic tool changer (ATC).SOLUTION: A machine tool includes an automatic tool changer. In the machine tool, a displacement sensor that detects a distance between a tool holder replaceably attached to a spindle and the spindle is provided in the vicinity of the spindle. The machine tool includes: an operation part that performs conversion and creates a histogram for each axial direction of the spindle from data of a change in a distance with an outer peripheral part of the tool holder detected by the displacement sensor; a preprocessing part that creates a histogram group including a histogram of a normal attachment state and at least one abnormal attachment state with respect to the spindle and the tool holder and stores the histogram group as master data; and a determination part that determines normal or abnormal of an attachment state indicated by the histogram created by the operation part on the basis of the master data stored in the preprocessing part.SELECTED DRAWING: Figure 3

Description

Technical Field

[0001] The present invention relates to a method for detecting the mounting state of a tool holder in an NC machine tool or the like and a machine tool on which the tool holder is mounted, and more particularly to a method for detecting the mounting state of a tool holder suitable for a machine tool having an automatic tool changer (ATC) and a machine tool on which the tool holder is mounted.

Background Art

[0002] In an NC machine tool or a machining center equipped with an automatic tool changer (ATC), the tool is automatically exchanged according to the machining content for the workpiece. This tool exchange is programmed in advance in the NC machine tool or the machining center, and a tool holder with a tool attached at a predetermined position in the ATC is mounted on the spindle of the machine tool according to the program. Since these are automatically executed without human intervention, detection means for detecting whether or not the tool holder is properly mounted on the spindle is provided near the spindle of the machine tool.

[0003] For example, in the machine tool described in Patent Document 1, when there is foreign matter such as chips sandwiched between the spindle and the tool holder, the inclination of the tool holder can be detected with high precision using an inexpensive sensor. Specifically, an eddy current sensor is arranged on the side surface of the flange portion of the tool holder so as to be obliquely opposed to the side surface of the eddy current sensor. The spindle is rotated, and the inclination of the tool holder is detected based on the change in the output of the eddy current sensor caused by the change in the relative position between the flange portion of the tool holder and the head portion of the eddy current sensor.

[0004] Another conventional method for detecting the mounting state of a tool holder is described in Patent Document 2. The machine tool described in this publication is configured to automatically measure the eccentricity of the tool holder when clamped and to detect clamping abnormalities. Specifically, the headstock of the machine tool rotatably supports the spindle, and the tool holder is mounted on the spindle by an automatic tool changer. The tool holder has a flange, and an eccentricity detection device for the tool holder is positioned opposite the flange. The eccentricity detection device has an eddy current sensor head and measures the distance between the tool holder and the flange non-contact. If the change in the measured distance is large, it is determined that the tool holder is clamped eccentrically, and a clamping abnormality is detected.

[0005] Another conventional example of a method for detecting the mounting state of a tool holder is described in Patent Document 3. In the machine tool described in this publication, a tool holder with a tool attached is mounted on the spindle, and the spindle is rotated to machine a workpiece. At that time, a sensor measures the displacement of the outer surface of the flange of the tool holder, and a data processing device performs Fourier analysis on the measured data and determines whether there is an abnormality in the mounting state of the tool holder on the spindle from the power spectrum shape of the analysis result.

[0006] Furthermore, since general-purpose tool holders have detachable notches formed at symmetrical positions in the circumferential direction, the measurement results will be interrupted at these notches regardless of which method is used. To resolve this problem, Patent Document 4 describes determining invalid data portions from the measurement results and interpolating those invalid portions. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2007-260885 [Patent Document 2] Japanese Patent Publication No. 2002-331442 [Patent Document 3] Japanese Patent Publication No. 2004-276145 [Patent Document 4] Japanese Patent Publication No. 2008-93750 [Overview of the project] [Problems that the invention aims to solve]

[0008] As described in the above patent documents, when changing tools using an ATC without human intervention, there is a risk that chips remaining in the environment where the machine tool is placed may get caught between the tool holder and the spindle, causing the tool holder to be tilted and / or eccentrically mounted on the spindle. To resolve this problem, it is necessary to install some kind of sensor near the spindle to confirm that the tool holder is properly mounted on the spindle of the machine tool.

[0009] In the machine tool described in Patent Document 1, the eddy current sensor must be positioned diagonally to the spindle or tool holder, and therefore must be installed in a way that does not interfere with the tool change operation by the ATC, which limits the position of the eddy current sensor. Furthermore, in order to increase the sensitivity of the eddy current sensor, it is necessary to place the sensor as close as possible to the tool holder, but since tool holders generally have notches, the closer the sensor is to the tool holder, the more the influence of the notch will be included in the measurement data, and it is necessary to eliminate this influence.

[0010] The machine tools described in Patent Documents 2 and 3 differ from those in Patent Document 1 in that the eddy current sensor is positioned almost horizontally opposite the side of the tool holder. However, the other configurations are similar, and it seems necessary to eliminate the influence of the notch formed in the tool holder, but Patent Document 2 does not disclose how to eliminate the influence of the notch.

[0011] Furthermore, Patent Document 3 compares the measurement data with basic data measured during normal installation, so even if there is a notch, the influence of the notch on the measurement data can be eliminated. However, the machine tool described in this publication removes the data for the notch, resulting in data loss, which is a point that needs improvement.

[0012] Patent Document 4, which takes into account the handling of the notch portion of the tool holder, describes a data processing device that identifies the notch portion and interpolates the data for that portion to cover the data loss. This allows for more accurate detection of abnormal tool holder mounting. In the machine tool described in Patent Document 4, in order to use the interpolated data as data for determining mounting abnormalities, registered data representing normal mounting is obtained and compared with the measured data. In this comparison, it is necessary to align the circumferential position (phase) in the registered data and the measured data, but in rare cases, a correct judgment may not be made due to differences in circumferential position.

[0013] This invention has been made in view of the shortcomings of the prior art described above, and its purpose is to reliably detect that the tool holder is properly mounted on the spindle of a machine tool when automatically changing tools using an ATC. In addition to this purpose, it also aims to detect abnormal mounting of the tool holder to the spindle in the shortest possible time. [Means for solving the problem]

[0014] A feature of the present invention that achieves the above objective is a method for detecting the mounting state of a tool holder when automatically mounting the tool holder to the spindle of a machine tool using a displacement sensor provided near the spindle, wherein the displacement sensor includes the steps of: measuring the distance between the displacement sensor and the spindle for one or more turns of the spindle; converting the measured distance data into a histogram for each axial position of the spindle; and determining whether the tool holder is correctly mounted on the spindle within an acceptable range using the HBOS (Histogram Outlier Detection) method based on the obtained histogram and a plurality of pre-stored histograms.

[0015] Furthermore, in this feature, the tool holder has notches at substantially symmetrical positions in the circumferential direction, and the histogram may be created based on the change in the output of the displacement sensor at the circumferential position of the tool holder excluding the notches. It is also desirable to include a step of obtaining a histogram when the tool holder is correctly mounted on the spindle within the tolerance range and at least one histogram when the tool holder is mounted on the spindle outside the tolerance range, before automatically mounting the tool holder on the spindle. It is also preferable to include a step of adding the histogram when the mounting abnormality is determined in the determination step to the stored histogram group.

[0016] Another feature of the present invention that achieves the above objective is a machine tool equipped with an automatic tool changer, wherein a displacement sensor is provided near the spindle to detect the distance between a tool holder, which is replaceably attached to the spindle, and the spindle, and the machine tool is equipped with an automatic tool changer, wherein the machine tool is equipped with a displacement sensor that detects the distance between the spindle and a tool holder, and the spindle, and the displacement sensor is provided with a calculation unit that converts data on the change in distance between the outer circumference of the tool holder and the spindle detected by the displacement sensor into a histogram for each axial direction of the spindle, a preprocessing unit that creates a group of histograms having histograms for a normal mounting state and at least one abnormal mounting state relating to the spindle and the tool holder, and stores them as master data, and a determination unit that determines whether the mounting state shown in the histogram created by the calculation unit is normal or abnormal based on the master data stored in the preprocessing unit.

[0017] In this feature, it is preferable that the master data stores histograms of normal mounting states corresponding to the number of tool holders to be replaced, and histograms of abnormal mounting states equal to or greater than the number of tool holders. It is also preferable that the determination unit uses the HBOS (Histogram-Based Outlier Detection) method to determine the mounting abnormality between the spindle and the tool holder. Furthermore, it is preferable that the machine tool is an NC machine tool or a machining center. [Effects of the Invention]

[0018] According to the present invention, from the data of the distance between the displacement sensor detected by the displacement sensor and the tool holder, the appearance frequency of the displacement at the axially divided positions is obtained, and the appearance frequency is compared with the appearance frequency in the reference data. Therefore, the influence of the circumferential direction (phase) can be removed, and it can be surely detected that the tool holder is properly mounted on the spindle of the machine tool. Further, since phase alignment is not required only by converting the measurement data into a frequency distribution, the detection of abnormal mounting of the tool holder on the spindle can be performed in as short a time as possible.

Brief Description of the Drawings

[0019] [Figure 1] It is a front view and a bottom view of a displacement sensor mounting portion in an embodiment of the present invention. [Figure 2] It is a graph showing an output example of the displacement sensor shown in FIG. 1. [Figure 3] It is a diagram for explaining a method of creating a histogram from displacement data. [Figure 4] It is a diagram showing an example of comparing displacement data with master data. [Figure 5] It is a diagram showing an example of displacement data used for data processing. [Figure 6] It is a block diagram of an embodiment of the data arithmetic processing unit shown in FIG. 1.

Modes for Carrying Out the Invention

[0020] Hereinafter, a method for detecting the presence or absence of an abnormality when mounting a tool holder on a spindle of a machine tool according to the present invention will be described with reference to the drawings. In the following embodiments, the machine tool is described by taking an NC machine tool or a machining center equipped with an automatic tool changer (ATC) as an example, but the machine tool is not limited thereto and includes all those that automatically exchange tools.

[0021] Figure 1 shows only the mounting portion of the tool holder 110 of the machine tool 100. It shows the tool holder 110, to which a drill tool 112 has been attached using an ATC (not shown), mounted on the spindle 120 of the machine tool 100. Figure 1(a) is a front view of the spindle 120, and Figure 1(b) is a view from the bottom. Figure 1 also shows a mounting condition detection device 200 that detects whether or not there is an abnormality in the mounting condition.

[0022] In the machine tool 100, a tool holder 110 to which a tool 112 is attached is interchangeably mounted on the spindle 120. In this mounting, a conical fitting portion 114 provided on the upper end of the tool holder 110 is fitted with a complementary conical fitting portion 116 formed on the spindle 120 to fit the fitting portion 114 onto the spindle 120. A mounting state detection device 200 is provided to monitor for mounting abnormalities caused by chips 102 or the like adhering to the fitting portion during this fitting process.

[0023] The mounting state detection device 200 in this embodiment automatically detects mounting abnormalities of the tool holder 110 mounted on the spindle 120, particularly using an ATC, and as shown in Figure 1(a), it mainly consists of a sensor 210 and a processing unit 220. The sensor 210 is a displacement sensor that detects the displacement of the cylindrical outer surface 110a of the flange portion of the tool holder 110 without contact. Therefore, the sensor 210 is attached to the spindle 120 via a bracket 202 at the lower end of the spindle 120, facing the outer surface 110a of the tool holder 110, and is positioned at a slight distance from the tool holder 110. The distance from the outer surface 110a of the tool holder 110 detected by the sensor 210 is input to the processing unit 220 as a displacement signal Sg.

[0024] The arithmetic processing unit 220 detects an abnormal mounting of the tool holder 110 based on the displacement signal Sg detected by the sensor 210. The arithmetic processing unit 220 includes an A / D converter 222 that converts the detected displacement signal Sg into a digital signal, an arithmetic processing unit 224 that performs various calculations on the digitally converted displacement signal Sg (details to be described later), a storage unit 226 that stores master data, etc., and an input / output unit 228 that manages input / output processing such as outputting calculation results to the control device 150 of the machine tool 100.

[0025] Referring to Figure 1(b), two notches 110c are formed on the outer circumferential surface 110a of the flange portion 110b of the tool holder 110, at positions approximately symmetrical in the circumferential direction, for attaching and detaching the tool holder 110 to the spindle 120. As a result, the displacement signal detected by the sensor 210 shows abrupt changes at two locations in the circumferential direction (see Figure 2).

[0026] Figure 2 shows an example of a displacement signal output from the sensor 210 when the sensor 210 is positioned as shown in the embodiment of Figure 1. The horizontal axis of Figure 2 shows the circumferential position of the tool holder 110, expressed as a percentage of the entire circumference. The vertical axis of Figure 2 is the output value of the sensor 210, which can be set arbitrarily. If the output range that guarantees the linearity of the sensor 210 is, for example, ±1V, then Figure 2 shows 0 to -500mV. Since the sensor 210 is positioned opposite the outer circumferential surface 110a of the tool holder 110, it can be seen that the output signal of the sensor 210 drops sharply at approximately 15% and 65% of the entire circumference. As mentioned above, this corresponds to the notch 110c of the tool holder 110 shown in Figure 1(b).

[0027] Conventionally, the displacement signal Sg detected by the sensor 210 for more than one full rotation of the tool holder 110 is compared with a master, which is a reference displacement signal, and a judgment of abnormality / normality is made according to the degree of deviation from the master. In this process, many conventional methods use techniques to exclude the influence of the notch 110c included in the tool holder 110. However, since the data needed to determine abnormality / normality is not the data of the notch 110c but the data of the cylindrical outer surface 110a, it is not a good idea to spend time processing the notch 110c.

[0028] Another method involves acknowledging the presence of the notches 110c and treating the entire data, including the data of the notches 110c, as a displacement signal Sg for more than one full turn around the tool holder 110, and comparing it with the displacement signal stored as a master in the storage unit 226 for normal mounting. In this case, a misalignment of a pair of notches 110c significantly affects the determination of whether the condition is abnormal or normal. For example, it is necessary to distinguish between an abnormal state where chips 102 are contained in the mating portion between the tool holder 110 and the spindle 120, causing the tool holder 110 to be mounted eccentrically or at an angle, and a state where the mounting condition is normal, but the alignment between the master data used to detect the abnormality and the displacement signal Sg data detected by the sensor 210 is not good. A typical example of the latter is when the positions of two nearly symmetrically formed notches 110c are determined to be in opposite positions by the master and the detected displacement signal.

[0029] To correctly implement this distinction, it is necessary to align the master and data, taking into account that the circumferential data of the tool holder 110 detected by the displacement sensor 210 may fluctuate within the assembly tolerance. Performing this alignment requires either installing additional equipment such as a camera for imaging, or establishing a new reference point on the tool holder, which makes it less versatile and difficult to automate.

[0030] In this invention, an outlier detection method using the histogram shown below is employed to determine whether the mounting state of the spindle 120 and the tool holder 110 is abnormal or normal. Figure 3 shows a diagram illustrating the method for creating the histogram. Figure 3(a) is a diagram showing the displacement signal Sg shown in Figure 2 with the vertical and horizontal axes swapped, and Figure 3(b) is a histogram obtained as a result of processing the displacement signal Sg in Figure 3(a) with the arithmetic processing unit 224.

[0031] First, in Figure 3(a), count the number of intersections 16 and 18 between the straight lines 12 and 14, which extend vertically with constant displacement, and the displacement data 10 detected by the displacement sensor 210. Here, the displacement data 10 includes data for the entire circumference of the tool holder 110. In the case of the straight line 12 that intersects with the data of the notch 110c, the number of intersections 16 is four, corresponding to both edges of the notch 110c, since the outer surface of the notch 110c is a flat surface 110d rather than a cylindrical surface (see Figure 1(b)). If the straight line with constant displacement is moved to the right in Figure 3(a), the number of points intersecting with the displacement data 10 from the edge of the notch 110c increases. In the ideal state where the outer circumference of the tool holder 110 is a perfect circle except for the notch 110c, and it is mounted perfectly concentrically and straight on the spindle 120, the number of intersections between the displacement data 10 and the straight line with constant displacement becomes infinite. However, in actual installation, the tool holder 110 is installed with variations within the assembly tolerance, and the outer surface 110a of the tool holder 110 cannot be a perfect circle when viewed microscopically. Therefore, a finite number of intersections 18 are obtained, as in the case of the straight line 14 and the displacement data 10. Figure 3(b) shows the number of intersections obtained in this way, with the output value of the displacement data 10 on the horizontal axis.

[0032] As is clear from the displacement data in Figure 3(a), the number of intersections 16 is constant in the notch 110c section, and the number of intersections does not change in this section even if the tool holder 110 is mounted at an angle or abnormally eccentrically (section B). On the other hand, near the position where the output of the displacement sensor 210 is at its maximum value, i.e., where the outer circumference of the displacement sensor 210 and the tool holder 110 are closest together (section A), the number of intersections 18 increases significantly compared to other sections. Therefore, it can be seen that in subsequent decisions, it is sufficient to consider only this section A.

[0033] Figure 4 shows a magnified view of area A. Figure 4(a) is an example of a tool holder 110 in a previously detected normal mounting state, which will hereafter be referred to as the master 240 or master data 240. The master 240 is stored in the storage unit 226 of the arithmetic processing unit 220 shown in Figure 1. When creating the master, the vertices of the histogram shown in Figure 3(b) are connected by polylines to make it a convenient format for subsequent processing. The sharp increase near zero in the output value is as explained in the description of Figure 3.

[0034] Figure 4(b) is a diagram comparing the master data 240 shown in Figure 4(a) with the displacement data 250 obtained by the sensor 210 when the tool holder 110 was actually attached to the spindle 120. In Figure 4(b), the solid line graph connects the vertices of the histogram of the displacement data 250 detected by the sensor 210, and the dashed line graph connects the vertices of the histogram of the master 240. The storage unit 226 stores the displacement data when the tool holder 110 is attached to the spindle 120 in an appropriate or normal state, and, if possible, the displacement data when an abnormal state is present.

[0035] Since the tool holder 110 is replaced by the ATC according to the workpiece machining content, a number of tool holders 110 corresponding to the workpiece machining content or the type of workpiece are used in the same machine tool 100. For this reason, the storage unit 226 stores master data 240 for at least the number of tool holders 110 that are replaced and used using the ATC. If only the tool 112 is changed while using the same tool holder 110, then it is sufficient to have a number of master data 240 corresponding to the number of tool holders 110, not the number of tools 112. However, as will be described later, if multiple master data for the same tool holder 110 are stored in the storage unit 226, mounting abnormalities can be determined with even greater accuracy.

[0036] The abnormal installation of the tool holder 110 is detected below using the HBOS (Histogram-Based Outlier Detection) method. In Figure 4(b), there are discrepancies between the line graphs of the two data sets 240 and 250. The hatched area 20 is the common part of the line graphs of the two data sets 240 and 250, the white area 22 is a part that exists only in the master 240, and the cross-hatched area 24 is a part that exists only in the detected displacement data 250. However, the overall shape can be considered almost identical, and the changes in output values ​​can also be considered almost identical. Whether the two data sets 240 and 250 can be considered identical, that is, whether the installation can be considered normal, depends on whether these differing areas 22 and 24 are within an acceptable range.

[0037] For determining these differences, for example, supervised machine learning with normal / abnormal labels can be used. Specifically, for example, when introducing a new tool holder, the operation of attaching the tool holder 110 to the spindle 120 is repeated, and a displacement signal Sg is detected for each instance. Then, multiple patterns of histograms similar to those in Figure 4(a) are obtained as line graphs. The operator then determines whether each of these is normal or abnormal based on the results obtained by visual inspection or simulated machining. The obtained large number of data are classified into normal and abnormal groups, the classified data is trained on a binary classification machine learning model, and the data and binary classification machine learning model are stored in the storage unit 226. Here, the normal group contains at least one data for each tool holder 110, and the abnormal group contains one or more data for each tool holder 110. Thereafter, the binary classification machine learning model is used to determine whether it is normal or abnormal. Preferably, the results of the determination are added as belonging to either the normal or abnormal group, and the binary classification machine learning model is retrained. This allows for the acquisition of displacement data for various mounting conditions. As the amount of data increases, the number of stored mounting conditions that can be compared increases, making it possible to obtain a more accurate mounting condition and thus enable highly accurate determination of the mounting condition. Furthermore, the data used by the binary classification machine learning model to train on normal and abnormal groups can be not only data actually acquired from the device, but also data generated through simulation.

[0038] Figure 5 shows an example of comparing displacement data 250 obtained when the same tool holder 110 is mounted on the same spindle 120 multiple times, against a single master data 240. In Figure 5, as in Figure 4(a), a line graph is created by connecting the peaks of the histogram. The master data 240 is located in the center (Figure 5(a)), and the detected displacement data 250 are located around it (Figures 5(b) to (g)). In terms of calculation, these two types of data 240 and 250 are superimposed and their differences are judged, but for the sake of clarity in the figure, each data is shown individually.

[0039] The multiple detected displacement data are classified into a first group G1, shown in Figures 5(b) to (d), and a second group G2, shown in Figures 5(e) to (g). Each data point belonging to the first group G1 has multiple peaks near the center and also has a peak on the negative side of the horizontal axis. This characteristic is similar to that of the master data 240. On the other hand, each data point belonging to the second group G2 has fewer peaks on the positive side of the horizontal axis, or no peaks on the negative side of the horizontal axis, and its overall gain (vertical axis scale) is low. Since the differences can be visually determined in this way, this is automatically determined using the labeled supervised machine learning described above.

[0040] Each of the above processes is executed by the arithmetic processing unit 224 of the arithmetic processing unit 220 of the mounting state detection device 200. The details of the arithmetic processing unit 224 are explained in the block diagram shown in Figure 6. The arithmetic processing unit 224 is broadly divided into a data preprocessing unit 310, a learning model generation unit 330, and a data determination unit 320. The data preprocessing unit 310 is responsible for the process up to obtaining the image (HBOS image) shown in Figure 4(b).

[0041] First, distance (displacement) data between the tool holder 110 and the sensor 210 is acquired before machining when each tool holder 110 intended for use is properly attached to the spindle 120, and the data registration unit 312 registers the acquired data as master data in the storage unit 226. Next, the learning data generation unit 314 generates multiple displacement data through simulation, either randomly or according to a certain criterion. The created multiple displacement data are defined as normal data and abnormal data respectively based on the master data, and then grouped into normal data and abnormal data groups and registered in the storage unit 226.

[0042] Furthermore, the groups for normal data and abnormal data may include displacement data detected by the sensor 210 when machining was performed using the same spindle 120 and the same tool holder 110 in previous or subsequent machining operations. In this case as well, data detected by the sensor 210 when the mounting of the tool holder 110 to the spindle 120 was abnormal should be grouped into the abnormal data group, and data detected when the mounting was normal should be grouped into the normal data group.

[0043] Next, the preprocessing unit (HBOS image generation unit) 316 histograms (binarizes) the master data and the data within the normal group stored in the storage unit 226. Then, it overlays the histogram of the master data with the histogram of one data point within the normal group to generate an HBOS image as shown in Figures 4(b) and 5. This process is repeated for each normal group data point contained within the normal group. Similarly, the abnormal data within the abnormal group stored in the storage unit 226 is histogrammed, and the histogram of the master data with one data point within the abnormal group is overlaid to generate an HBOS image as shown in Figures 4(b) and 5. After the above processing is complete, the storage unit 226 may store only the HBOS image and the master histogram.

[0044] The above describes the preparation before attaching the workpiece to the machining device. After the workpiece is attached to the machining device, the sensor 210 detects the mounting status of the tool holder 110 to the spindle 120. The detected displacement signal (Sg) 10 is converted into a digital signal via the A / D converter 222 and transmitted to the data acquisition unit 322 of the data determination unit 320. The displacement signal 10 acquired by the data acquisition unit 322 is histogrammed in the preprocessing unit (HBOS image generation unit) 316, and superimposed on the master histogram to generate an HBOS image as shown in Figures 4(b) and 5.

[0045] The learning model generation unit 330 learns the HBOS images created by the data preprocessing unit 310 using supervised learning with "normal / abnormal" labels. As a result of the learning, it generates a binary classification machine learning model 332. This determines the criteria for determining abnormal / normal. The judgment unit 326, by referring to the generated binary classification machine learning model 332 and learning through this process, makes a judgment of abnormal / normal for the HBOS images of the displacement signals 10 that were newly acquired. The judged HBOS images are added to the histogram of the normal fitting group or the abnormal fitting group.

[0046] Furthermore, other conventionally known methods can be used to determine whether an HBOS image is abnormal or normal after it has been generated. For example, outlier detection methods such as the 3σ method or the interquartile method can be used.

[0047] As described above, according to the embodiment of the present invention, when detecting the mounting state of the tool holder on the spindle, it is possible to determine whether the mounting is abnormal or normal without considering the relative circumferential position between the tool holder and the spindle. Therefore, especially when changing tool holders with an ATC in accordance with machining in unmanned NC machine tools and machining centers, abnormal mounting of the tool holder on the spindle can be reliably detected without taking time. As a result, the reliability of machining in machine tools is improved, and machining throughput is also improved. [Explanation of Symbols]

[0048] 10...Displacement signal, 12, 14...Straight line (constant output value), 16, 18...Intersection, 20...Hatched area (common part), 22, 24...Difference area, 100...Machine tool, 102...Chip, 110...Tool holder, 110a...Outer surface, 110b...Flange part, 110c...Notch, 110d...Plane, 112...Tool, 114, 116...Matching part, 120...Spindle, 150...Control device, 200...Mounting state detection device, 202...Bracket, 210...(Displacement ) Sensor, 220... Processing Unit, 222... A / D Converter, 224... Processing Unit, 226... Storage Unit, 228... Input / Output Unit, 240... Master (Data), 250... Displacement Data, 310... Data Preprocessing Unit, 312... Data Registration Unit, 314... Training Data Generation Unit, 316... Preprocessing Unit (HBOS Image Generation Unit), 320... Data Judgment Unit, 322... Data Acquisition Unit, 326... Judgment Unit, 330... Training Model Generation Unit, 332... Binary Classification Machine Learning Model

Claims

1. It includes a determination unit that determines the mounting status of the tool holder attached to the spindle, The determination unit includes a machine learning model that determines the mounting state based on a histogram representing the frequency of occurrence of the output value, which is generated from data consisting of the circumferential position of the tool holder measured by a displacement sensor and the output value of the displacement sensor at the circumferential position.

2. The machine tool according to claim 1, wherein the machine learning model is generated by supervised learning based on data on which the mounting state is labeled to a group of histograms acquired in advance.

3. Furthermore, it is equipped with a data registration unit, The machine tool according to claim 2, wherein the data registration unit adds the histogram when the determination unit determines that there is an abnormal installation to the histogram group.

Images