Machine tool system, tool type estimation method and feature extraction method

Abstract

To provide a machine tool system capable of easily estimating the type of a tool, a tool type estimation method, and a feature amount extraction method.SOLUTION: A machine tool system 1 comprises a machine tool 10, an acquisition unit 22, and an estimation unit 23. The machine tool 10 processes a workpiece W using a tool 15. The acquisition unit 22 acquires processing data including coordinate data of the tool 15 in processing of the workpiece W. The estimation unit 23 estimates the type of the tool 15 on the basis of the processing data.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present disclosure relates to a machine tool system, a tool type estimation method, and a feature extraction method. [Background technology]

[0002] BACKGROUND ART Conventionally, computer numerically controlled machine tools (hereinafter referred to as "machine tools") that machine a workpiece into a desired shape according to an NC (Numerical Control) program have been known.

[0003] Patent Document 1 discloses a method for identifying the type of tool based on the shape of the tool recognized from image data of the tool attached to a machine tool. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2020-032475 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the technique described in Patent Document 1 requires the installation of an imaging device, a lighting device, a controller, and wiring for these devices in order to capture an image of the tool.

[0006] An object of the present disclosure is to provide a machine tool system, a tool type estimation method, and a feature extraction method that can easily estimate a tool type. [Means for solving the problem]

[0007] A machine tool system according to one aspect of the present disclosure includes a machine tool, an acquisition unit, and an estimation unit. The machine tool machines a workpiece using a tool. The acquisition unit acquires machining data including coordinate data of the tool used in machining the workpiece. The estimation unit estimates the tool type of the tool based on the machining data. [Effects of the Invention]

[0008] According to the present disclosure, it is possible to provide a machine tool system capable of easily estimating a tool type, a tool type estimation method, and a learning model generation method. [Brief explanation of the drawings]

[0009] [Figure 1] A block diagram showing the configuration of a machine tool system according to an embodiment. [Figure 2] Flowchart showing the flow of the tool type estimation method DETAILED DESCRIPTION OF THE INVENTION

[0010] (Configuration of machine tool system 1) The configuration of a machine tool system 1 according to this embodiment will be described with reference to the drawings. Figure 1 is a block diagram showing the configuration of the machine tool system 1.

[0011] The machine tool system 1 includes a machine tool 10 and a tool type estimation device 20. There is no particular limitation on the number of machine tools 10 in the machine tool system 1, as long as it is one or more.

[0012] [Machine tool 10] The machine tool 10 has a machine body 11 and a CNC (Computer Numerical Control) control unit 12.

[0013] The machine body 11 processes a workpiece (so-called workpiece) W into a desired shape. Cutting is a typical type of processing performed by the machine body 11. Cutting includes drilling, milling, turning, and the like.

[0014] The machine body 11 has a table 14 , a tool 15 , and a spindle 16 .

[0015] A workpiece W is placed on the table 14. A tool 15 is used to machine the workpiece W. The tool 15 is attached to a spindle 16 and driven to rotate. The tool 15 moves relative to the workpiece W placed on the table 14. The tool 15 moves relative to the workpiece W as the table 14 or the spindle 16 moves.

[0016] The type of tool 15 (hereinafter abbreviated as "tool type") is appropriately selected depending on the machining content to be performed on the workpiece W. Therefore, in the machine tool 10, a plurality of types of tools 15 are usually used.

[0017] Examples of tool types include, but are not limited to, milling cutters, end mills, drills, reamers, taps, boring tools, and cutters.

[0018] The CNC control unit 12 controls the machine body 11 in accordance with an NC (Numerical Control) program. The NC program includes machining conditions for the workpiece W. The machining conditions include the feed rate of the tool 15 (so-called F code, hereinafter referred to as "feed rate") and the rotation speed of the spindle 16 (hereinafter referred to as "spindle rotation speed").

[0019] The CNC control unit 12 generates machining data upon completing machining of the workpiece W. The machining data includes coordinate data of the tool 15 used in machining the workpiece W and machining condition data for the workpiece W.

[0020] The coordinate data represents changes over time in the coordinates indicating the position of the cutting edge of tool 15. The coordinate data is a set of coordinates (x, y, z) of the cutting edge of tool 15 in the machine coordinate system (X, Y, Z) of machine tool 10.

[0021] The machining condition data includes the feed rate of the tool 15 and the spindle rotation speed of the spindle 16. The feed rate and the spindle rotation speed are written in the NC program. However, if the operator changes the feed rate and the spindle rotation speed, the changed values ​​are used.

[0022] The CNC control unit 12 transmits the generated machining data to the tool type estimation device 20.

[0023] [Tool type estimation device 20] The tool type estimation device 20 is capable of mutual communication with the machine tool 10 and the terminal device 30 via a network. The functions of the tool type estimation device 20 can be achieved by a server. The server may be a cloud server.

[0024] The tool type estimation device 20 includes a storage unit 21, an acquisition unit 22, and an estimation unit 23.

[0025] The storage unit 21 stores a plurality of index feature amounts and a plurality of tool types in association with each other. The plurality of index feature amounts are used to estimate the tool type based on machining data. Each of the plurality of index feature amounts indicates a different tool type. Each index feature amount is represented by a vector of two or more dimensions.

[0026] As an example of a method for extracting index features, the extraction of index features indicating a milling cutter will be described. First, machining data obtained when a workpiece is machined according to an NC program for the milling cutter is acquired. The machining data includes coordinate data of the milling cutter, the feed rate of the milling cutter, and the spindle rotation speed of the milling cutter. Next, index features indicating the milling cutter are extracted from the machining data. Artificial intelligence technology such as an autoencoder is used to extract the index features. The index features are represented by vectors of two or more dimensions. Then, the milling cutter as a tool type and the index features indicating the milling cutter are associated with each other and stored in the storage unit 21.

[0027] In this way, index feature amounts are obtained in advance not only for milling cutters but also for various tools such as end mills, drills, reamers, taps, boring bars, and cutters, and the tool types and index feature amounts are stored in the storage unit 21 in association with each other.

[0028] The acquisition unit 22 acquires machining data from the machine tool 10. The acquisition unit 22 stores the acquired machining data in the storage unit 21. Every time the acquisition unit 22 acquires machining data, the acquisition unit 22 stores the newly acquired machining data in the storage unit 21.

[0029] The estimation unit 23 estimates the tool type of the tool 15 based on the machining data. As shown in Fig. 1, the estimation unit 23 includes an extraction unit 23a and a selection unit 23b.

[0030] The extraction unit 23a acquires machining data from the storage unit 21. The extraction unit 23a extracts an extracted feature quantity indicating the type of tool 15 from the acquired machining data. An artificial intelligence technique such as an autoencoder is used to extract the extracted feature quantity. The extracted feature quantity is represented by a vector of two or more dimensions.

[0031] The selection unit 23b acquires a plurality of index feature amounts and the tool types associated with each index feature amount from the storage unit 21. The selection unit 23b acquires the extracted feature amount from the extraction unit 23a.

[0032] The selection unit 23b calculates the similarity between each of the plurality of index features and the extracted feature. The similarity can be calculated by using the inner product of each of the plurality of index features and the extracted feature. The closer the inner product value is to 1 (i.e., the larger the inner product value is), the higher the similarity between the index feature and the extracted feature.

[0033] Based on the calculated similarity, the selection unit 23b selects, from the plurality of tool types, a candidate tool type for the tool 15. The selection unit 23b selects, as a candidate tool type for the tool 15, a tool type associated with an index feature having a high similarity to the extracted feature among the plurality of index features.

[0034] When there are two or more index features with high similarity, the selection unit 23b may select two or more candidates for the tool type of the tool 15. When the selection unit 23b selects two or more candidates for the tool type of the tool 15, it is preferable that the selection unit 23b assigns a priority to each of the two or more tool types based on the similarity. The selection unit 23b can assign a higher priority to the higher the similarity.

[0035] (Tool type estimation method) Next, we will explain the method for estimating the tool type in the machine tool system 1. Fig. 2 is a flowchart showing the flow of the tool type estimation method.

[0036] In step S1, a plurality of tool types and a plurality of index feature amounts are associated with each other and stored in the storage unit 21. As described above, each index feature amount is extracted from machining data for each tool type.

[0037] In step S2, the acquisition unit 22 acquires machining data from the machine tool 10.

[0038] In step S3, the extraction unit 23a extracts an extraction feature amount indicating the type of the tool 15 from the machining data.

[0039] In step S4, the selection unit 23b selects a candidate tool type for the tool 15 from among the plurality of tool types based on the similarity between each of the plurality of index features and the extracted feature.

[0040] (Features) The machine tool system 1 includes a machine tool 10, an acquisition unit 22, and an estimation unit 23. The acquisition unit 22 acquires machining data for machining a workpiece W. The estimation unit 23 estimates the tool type of the tool 15 based on the machining data. In this way, the tool type of the tool 15 can be easily estimated based on the machining data, so there is no need for an operator to manually input the tool type of the tool 15 or to identify the tool type from image data of the tool 15. As a result, machining data can be compiled for each tool type and then compared, which makes it easy to optimize machining conditions.

[0041] The machining data includes coordinate data of the tool 15 used in machining the workpiece W, and machining condition data for the workpiece W. The machining condition data includes the feed rate of the tool 15 and the number of revolutions of the spindle 16. In this way, by using the machining condition data in addition to the coordinate data of the tool 15, the tool type of the tool 15 can be more accurately estimated.

[0042] The estimation unit 23 has an extraction unit 23a that extracts extracted feature amounts indicating the tool type of the tool 15 from the machining data, and a selection unit 23b that selects tool type candidates based on the similarity between each of a plurality of index feature amounts and the extracted feature amounts. In this way, the tool type of the tool 15 can be automatically estimated by selecting tool type candidates using the extracted feature amounts extracted from the machining data.

[0043] When the selection unit 23b selects two or more tool types as candidates for the tool type of the tool 15, the selection unit 23b assigns a priority to each of the two or more tool types based on the similarity. This makes it easier to identify the tool type of the tool 15 from the two or more selected tool types.

[0044] (Modification of the embodiment) The present disclosure is not limited to the above-described embodiments, and various modifications and alterations are possible without departing from the scope of the present disclosure.

[0045] (Variation 1) In the above embodiment, the functions of the tool type estimation device 20 can be achieved by a server, but this is not limited to this. At least some of the functions of the tool type estimation device 20 can be achieved by a terminal device. For example, a terminal device capable of communicating with the machine tool 10 may include the storage unit 21, the acquisition unit 22, and the estimation unit 23.

[0046] (Variation 2) In the above embodiment, the machining data includes coordinate data of the tool 15 used in machining the workpiece W and machining condition data for the workpiece W, but it may also include only coordinate data.

[0047] In the above embodiment, the candidate tool types are selected using the extracted feature amounts extracted from the machining data, but the candidate tool types may be selected directly from the machining data.

[0048] For example, when machining using a milling cutter, the x- and z-coordinates change but the y-coordinate does not, so by looking at the coordinate data included in the machining data, it can be inferred that the tool 15 is a milling cutter. Therefore, even if the machining data includes only coordinate data, the tool type can be inferred directly from the machining data.

[0049] Furthermore, when machining using a lathe tool, the x and y coordinates change but the z coordinate does not, so by looking at the coordinate data included in the machining data, it can be assumed that the tool 15 is a lathe tool. Therefore, even if the machining data includes only coordinate data, the tool type can be directly estimated from the machining data.

[0050] Furthermore, although the coordinate data for drills and taps is basically the same, the thread pitch of a tap is usually between 0.6 and 6.0, so if the feed rate is between [0.6 × spindle rotation speed] and [6.0 × spindle rotation speed], the tool 15 can be estimated to be a tap, and if the feed rate is outside this range, the tool 15 can be estimated to be a drill. Therefore, if the machining data includes coordinate data and machining condition data, the tool type can be estimated directly from the machining data. [Explanation of symbols]

[0051] 1 Machine tool system 10 Machine tools 11 Machine body 12 CNC control unit 14 tables 15 Tools 16 spindle 20 Tool type estimation device 21 Memory section 22 Acquisition Department 23 Estimation part 23a Extraction part 23b Selection section

Claims

1. a machine tool that processes a workpiece using a tool; an acquisition unit that acquires machining data including coordinate data of the tool used in machining the workpiece; an estimation unit that estimates a tool type of the tool based on the machining data; a storage unit that stores a plurality of index feature amounts that indicate a plurality of tool types; Equipped with the estimation unit includes an extraction unit that extracts an extracted feature amount indicating a tool type of the tool from the machining data, and a selection unit that selects a candidate tool type of the tool from the plurality of tool types based on a similarity between each of the plurality of index feature amounts and the extracted feature amount, the plurality of index features and the extracted feature are each represented by a vector of two or more dimensions; the similarity is indicated by a value of an inner product of each of the plurality of index features and the extracted feature. Machine tool systems.

2. the processing data includes processing condition data for the workpiece, The machining condition data includes a feed rate of the tool and a rotation speed of a spindle to which the tool is attached. The machine tool system according to claim 1 .

3. When the selection unit selects two or more tool types from the plurality of tool types as candidates for the tool type of the tool, the selection unit assigns a priority to each of the two or more tool types based on the similarity. The machine tool system according to claim 1 .

4. acquiring machining data including coordinate data of a tool used in machining a workpiece; estimating a tool type of the tool based on the machining data; Equipped with The step of estimating the tool type includes: extracting an extracted feature amount indicating a tool type of the tool from the machining data; selecting a candidate tool type of the tool from among the plurality of tool types based on a similarity indicated by a value of an inner product of each of a plurality of index feature amounts indicating a plurality of tool types and the extracted feature amount; Including, the plurality of index features and the extracted features are each represented by a vector of two or more dimensions; Tool type estimation method.

Images