Processing diagnosis device, learning device, inference device, processing diagnosis method, processing diagnosis program, and machine tool system

Abstract

The machining diagnosis device (1) includes a data acquisition unit (11) that acquires a physical quantity related to the operating state of a machining machine (90) that machines a workpiece, and the machining conditions when the physical quantity is obtained, a feature quantity calculation unit (12) that calculates, for each machining condition, a statistical feature quantity of the physical quantity obtained during machining under the same machining conditions based on the physical quantity and the machining conditions, and an inference unit (16) that performs machining diagnosis based on the statistical feature quantity for each machining condition.

Description

Technical Field

[0001] The present disclosure relates to a machining diagnosis apparatus, a learning apparatus, an inference apparatus, a machining diagnosis method, a machining diagnosis program, and a machine tool system that estimate the results of machining performed by a numerically controlled (NC) machine tool and the like.

Background Art

[0002] When the tool used by a numerically controlled machine tool wears during machining, a good machining result cannot be obtained. Therefore, techniques for determining whether a predetermined machining result can be obtained, techniques for determining the necessity of tool change, and the like have been proposed so far.

[0003] For example, Patent Document 1 describes a technique of collecting the drive current, vibration, temperature, etc. of a spindle motor as state data indicating the operating state of a numerically controlled machine tool, and predicting machining quality and tool replacement timing based on the collected state data. In the technique described in Patent Document 1, when performing machining that repeatedly creates machining result objects of the same shape, for example, assuming that state data can be repeatedly obtained under the same machining conditions, a regression equation is generated as a prediction model used for predicting machining quality.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in the technology described in Patent Document 1, machining conditions such as the rotational speed of the spindle, the feed rate of the tool during machining, and the depth of cut with respect to the workpiece are not considered. For this reason, in machining for multi-variety and small-batch production where the same machining is not repeatedly performed, such as machining for mold making and machining of custom products, it is difficult to repeatedly collect state data under the same machining conditions, so there is a problem that it is impossible to predict machining quality and determine the tool replacement timing.

[0006] In machining to obtain one machined product, usually, machining is carried out while changing machining conditions such as the depth of cut of the tool with respect to the workpiece and the feed rate. If machining to obtain machined products of the same shape is repeated, in each machining carried out repeatedly, the machining conditions change at the same timing, and the state data during machining changes in the same way at the same timing every time. Therefore, even if the machining conditions are unknown, it is possible to predict machining quality and determine the tool replacement timing from the state data obtained in each machining that is repeatedly executed. On the other hand, in machining for multi-variety and small-batch production where the timing at which the machining conditions change is different for each machining, it is difficult to predict machining quality and determine the tool replacement timing based only on the state data.

[0007] The present disclosure has been made in view of the above, and an object thereof is to obtain a machining diagnosis apparatus capable of predicting machining quality and determining the tool replacement timing even when machining for multi-variety and small-batch production is performed.

Means for Solving the Problems

[0008] In order to solve the above-described problems and achieve the object, a machining diagnosis apparatus according to the present disclosure includes a physical quantity related to the operating state of a machining machine that machines a workpiece, and the machining conditions when the physical quantity is obtained, a data acquisition unit that acquires them, a feature quantity calculation unit that calculates statistical feature quantities of the physical quantities obtained during machining under the same machining conditions for each machining condition based on the physical quantity and the machining conditions, and an estimation unit that performs machining diagnosis based on the statistical feature quantities for each machining condition. Independent of the target shape And a feature quantity calculation unit that calculates statistical feature quantities of the physical quantities obtained during machining under the same machining conditions for each machining condition based on the physical quantity and the machining conditions, and an estimation unit that performs machining diagnosis based on the statistical feature quantities for each machining condition. Unaffected by the required time until the end of processing It is characterized by comprising , the processing machine is provided in a numerically controlled machine tool having a machining program generation unit that automatically generates a machining program based on the material of the workpiece input and the tool used in the machining and a database of machining parameters, performs machining using the machining program generated by the machining program generation unit, the data acquisition unit acquires machining conditions from the machining program generation unit, and the feature quantity calculation unit calculates statistical feature quantities using the physical quantities newly acquired by the data acquisition unit and the physical quantities acquired in the past by the data acquisition unit. this.

Advantages of the Invention

[0009] The machining diagnosis device according to the present disclosure has an effect that even when performing machining for multi-variety and small-batch production, it is possible to predict the machining quality and determine the tool replacement timing.

Brief Description of the Drawings

[0010]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Embodiments for Carrying Out the Invention

[0011] Hereinafter, a machining diagnosis apparatus, a learning apparatus, an inference apparatus, a machining diagnosis method, a machining diagnosis program, and a machine tool system according to embodiments of the present disclosure will be described in detail with reference to the drawings.

[0012] Embodiment FIG. 1 is a diagram showing an example of a machine tool system 100 realized by applying a machining diagnosis apparatus 1 according to an embodiment. In FIG. 1, the configuration of a numerically controlled machine tool 7 that is the object of diagnosis by the machining diagnosis apparatus 1 is also shown.

[0013] First, the configuration of the numerically controlled machine tool 7 will be described. The numerically controlled machine tool 7 includes a numerical control device 70 that processes numerical control data (NC data) and outputs a control signal, and a processing machine 90 that applies machining to a workpiece (work) WA according to the control signal from the numerical control device 70.

[0014] The processing machine 90 includes a spindle 91, a tool 92 attached to the spindle 91, a spindle motor 93 that rotationally drives the spindle 91, a table 94 that fixes the workpiece WA, a moving mechanism 95 that moves the table 94, two servo motors 96x, 96y that align the moving mechanism 95 in the x-axis and y-axis directions, and a servo motor 96z that aligns the spindle motor 93 in the z-axis direction. A drive current is supplied to the spindle motor 93 from a spindle amplifier 85, and drive currents are supplied to the servo motors 96x, 96y, 96z from a servo amplifier 86. The positions of the tool 92 with respect to the workpiece WA can be independently controlled in the x-axis, y-axis, and z-axis directions by the servo motors 96x, 96y, 96z.

[0015] A current sensor S1, a vibration sensor S2, and a temperature sensor S3 are arranged on the spindle motor 93. In addition, devices 98 such as contactors, solenoids, and lamps, and limit switches, sensors, and other switches 97 that detect the states of respective parts inside the processing machine 90 and output detection signals are arranged on the processing machine 90.

[0016] The numerical control device 70 includes a processor 71 that processes information, a storage unit 72 that stores data, an external connection interface (hereinafter referred to as I / F) 73 connected to the machining diagnosis device 1, an operation panel I / F 74 connected to the NC operation panel 82, a spindle control I / F 75 connected to the spindle amplifier 85, and a servo control I / F 76 connected to the servo amplifier 86.

[0017] The processor 71 supplies a drive signal to the spindle amplifier 85 via the spindle control I / F 75 according to the machining program and numerical information stored in the storage unit 72 to drive the spindle motor 93, and supplies a drive signal to the servo amplifier 86 via the servo control I / F 76 to operate the servo motors 96x, 96y, 96z. Thereby, the spindle 91 is rotationally driven, the tool 92 is positioned, and the workpiece WA is cut. Note that the machining program stored in the storage unit 72 is also referred to as a numerical control program.

[0018] Also, the digital input unit 77 inputs data from the current sensor S1, vibration sensor S2, temperature sensor S3, limit switch, and sensors and other switches 97 and supplies it to the processor 71. The processor 71 stores these data in the storage unit 72. Further, the processor 71 transmits data from the current sensor S1, vibration sensor S2, and temperature sensor S3 to the machining diagnosis device 1 via the external connection I / F 73.

[0019] The external connection I / F 73 performs data communication with an external device. In particular, in the present embodiment, the external connection I / F 73 adds time data and type data to the data output by the current sensor S1, vibration sensor S2, and temperature sensor S3 under the control of the processor 71 and transmits it to the machining diagnosis device 1.

[0020] The operation panel I / F 74 is connected to the NC operation panel 82 including a display unit, a key operation unit, etc. (not shown), transmits display data to the NC operation panel 82, receives a key operation signal from the NC operation panel 82, and notifies the processor 71.

[0021] Further, the numerical control device 70 includes a digital input unit 77 that receives a digital signal from the processing machine 90, a digital output unit 78 that transmits a digital signal to the processing machine 90, a machining program generation unit 79 that generates a machining program stored in the storage unit 72, and an internal timer IT built into the processor 71.

[0022] Here, the machining program generation unit 79 will be described. The machining program generation unit 79 is based on various information indicating machining details such as the target shape, which is the final shape of the workpiece WA obtained when machining of the workpiece WA is performed normally, the shape of the workpiece WA before machining, the workpiece material which is the material of the workpiece WA, the cutting tool used in the machining of the workpiece WA, the machining pattern, the peripheral speed, the feed rate, the radial depth of cut, and the axial depth of cut. It automatically generates a machining program.

[0023] Among the various information required for the machining program generation unit 79 to generate a machining program, for the target shape, the shape before machining, the workpiece material, the cutting tool used, and the machining pattern, it is acquired from outside the numerical control machine tool 7 (for example, from an operator, etc.). The machining program generation unit 79 acquires the target shape, the shape before machining, the workpiece material, the cutting tool used, and the machining pattern via, for example, the NC operation panel 82 and the operation panel I / F 74.

[0024] Among the information acquired by the machining program generation unit 79 from outside the numerical control machine tool 7, for the cutting tool used and the machining pattern, they are individually set for each of the plurality of machining sections that form the path (hereinafter referred to as the tool path) along which the tool moves from the start to the end of the machining to obtain the target shape. The designation of the machining section and the setting of the cutting tool used and the machining pattern in each machining section are performed by, for example, an operator.

[0025] Among the various pieces of information necessary for generating a machining program, the peripheral speed, feed rate, radial depth of cut, and axial depth of cut are determined as values that can be set according to the cutting tool used and the machining pattern. If inappropriate values are set, normal machining cannot be performed. Therefore, the peripheral speed, feed rate, radial depth of cut, and axial depth of cut are determined in advance for each combination of each of the multiple types of cutting tools that can be used in machining and each of the multiple machining patterns that can be specified, and are assumed to be stored in a database. The database for the peripheral speed, feed rate, radial depth of cut, and axial depth of cut is constructed, for example, in the storage unit 72. The determination of the peripheral speed, feed rate, radial depth of cut, and axial depth of cut is performed by, for example, the manufacturer of each cutting tool that can be selected as the cutting tool used, the manufacturer of the numerical control machine tool 7, and the like.

[0026] A specific example of the machining program generation unit 79 generating a machining program will be described. For example, when creating a machining program for turning the workpiece 203 with the tool 201 using the cutting edge 202 as shown in FIG. 2, the machining program generation unit 79 generates a machining program based on the machining parameters illustrated in FIG. 3. Note that FIG. 2 is a diagram showing an example of machining performed by the numerical control machine tool 7 of the machine tool system 100 according to the embodiment, and FIG. 3 is a diagram showing an example of machining parameters necessary for creating a machining program. Among the machining parameters shown in FIG. 3, the machining program generation unit 79 acquires "unit", "machining part", "finish allowance - Z", "tool", "starting point - X", "starting point - Z", "ending point - X", and "ending point - Z" from the outside via the NC operation panel 82 and the operation panel I / F 74. Further, among the machining parameters shown in FIG. 3, for "radial depth of cut", "peripheral speed", and "feed rate", the machining program generation unit 79 searches a database in which "radial depth of cut", "peripheral speed", and "feed rate" are registered, using the "tool", "unit", and "machining part" acquired from the outside as keys. A configuration example of the database in which "radial depth of cut", "peripheral speed", and "feed rate" are registered is shown in FIG. 4. FIG. 4 is a diagram showing a configuration example of the database referred to when the machining program generation unit 79 of the numerical control device 70 generates a machining program. In the database shown in FIG. 4, "work material", "tool number", "machining pattern", "peripheral speed", "feed rate", "radial depth of cut", "axial depth of cut", etc. are registered.

[0027] In the turning process shown in Fig. 2, the cutting edge 202 of the tool 201 is brought into contact with the end face of the workpiece 203 being rotated, and from this state, the tool 201 is moved in the X-axis direction, thereby cutting (engaging) the workpiece 203 in the Z-axis direction with a cutting depth according to the "radial cutting depth". When the cutting depth required to achieve the target shape is greater than the cutting depth per cut, the positions of the tool 201 and the cutting edge 202 in the Z-axis direction are changed and the same cutting is repeated. For example, when the processing parameters for performing the turning process shown in Fig. 2 are as shown in Fig. 3, since "start - Z" is "4", "end - Z" is "0", and "finish allowance - Z" is "0.15", the cutting depth required in the Z-axis direction is the value obtained by subtracting the finish allowance (=0.15) from the movement amount of the tool 201 and the cutting edge 202 in the Z-axis direction (=4), which is (=3.85). The cutting depth in the Z-axis direction per cut is indicated by the "radial cutting depth" of the processing parameters, and in Fig. 3 it is 1.5. Since the cutting depth in the Z-axis direction per cut (=1.5) is smaller than the cutting depth required in the Z-axis direction (=3.85), it is necessary to repeat the cutting a plurality of times. Specifically, it is necessary to repeat the cutting 3 times. Also, the difference between the total cutting depth in the Z-axis direction (=3) at the time when cutting with a cutting depth of 1.5 is performed twice and the cutting depth required in the Z-axis direction (=3.85) (=0.85) becomes the cutting depth in the third cut. The machining program generation unit 79 creates a machining program for performing such machining based on the machining parameters shown in Fig. 3. In this embodiment, when the same machining is repeatedly executed for one section, the section where cutting is performed in one machining may be referred to as a pass. For example, when cutting is executed 3 times for one section to machine to the target shape, it is 3 passes.

[0028] The machining program generation unit 79 also acquires information (machining parameters) indicating the machining content for the machining programs of other machining sections required to machine the workpiece to the target shape and creates them in the same way.

[0029] Note that the machining program stored in the storage unit 72 is not limited to the machining program automatically generated by the machining program generation unit 79. The storage unit 72 may store a machining program manually created by an operator or the like, and the numerical control machine tool 7 may machine the workpiece WA according to this machining program.

[0030] When manually creating the machining program stored in the storage unit 72, the program creator refers to the cutting conditions such as the depth of cut, peripheral speed, and feed described in the catalog etc. by the tool manufacturer as recommended cutting values, and also considers the material of the workpiece (work material), the tool to be used, the machining pattern, etc., and sets values such as the peripheral speed, feed rate, radial depth of cut, and axial depth of cut to values that seem appropriate. However, in order to set values such as the peripheral speed, feed rate, radial depth of cut, and axial depth of cut to appropriate values, a lot of knowledge and experience are required, and it is difficult for inexperienced creators to create the program. If the settings such as the peripheral speed, feed rate, radial depth of cut, and axial depth of cut are not appropriate, there is a high possibility of problems such as the desired machining quality not being achievable, the machining machine malfunctioning, and the tool being damaged. In order to prevent the occurrence of these problems, it is necessary to repeatedly perform trial machining etc. to adjust the program, which takes time. On the other hand, the machining program generation unit 79 automatically sets values such as the peripheral speed, feed rate, radial depth of cut, and axial depth of cut by using a database in which values such as the peripheral speed, feed rate, radial depth of cut, and axial depth of cut are registered in advance for each combination of work material, tool, and machining pattern. Therefore, the creator only needs to set the work material, the tool to be used, the machining pattern, and the target shape, and as described above, can automatically generate a program that repeatedly performs machining according to the peripheral speed, feed rate, radial depth of cut, axial depth of cut, etc. retrieved from the settings until the target shape is reached, without the need for experience and furthermore, without the need for program creation technology, and can efficiently create a machining program in a short time.

[0031] Next, the processing diagnosis device 1 will be described. The processing diagnosis device 1 is, for example, a PC (Personal Computer), an IPC (Industrial Personal Computer), a PLC (Programmable Logic Controller), or other FA (Factory Automation) device. Also, the numerical control device 70 may function as part or all of the processing diagnosis device 1. FIG. 5 is a diagram showing the hardware configuration of the processing diagnosis device 1 according to the embodiment. As shown in FIG. 5, the processing diagnosis device 1 includes a processor 51, a main storage unit 52, an auxiliary storage unit 53, an input unit 54, an output unit 55, and a communication unit 56 as its hardware configuration. The main storage unit 52, the auxiliary storage unit 53, the input unit 54, the output unit 55, and the communication unit 56 are all connected to the processor 51 via an internal bus 57.

[0032] The processor 51 includes an integrated circuit CPU (Central Processing Unit), MPU (Micro Processing Unit), arithmetic GPU (Graphics Processing Unit), or FPU (Floating Point Unit). The processor 51 realizes various functions of the processing diagnosis device 1 and executes the following-described processing by executing the program P1 stored in the auxiliary storage unit 53.

[0033] The main storage unit 52 includes a RAM (Random Access Memory). The program P1 is loaded from the auxiliary storage unit 53 into the main storage unit 52. Then, the main storage unit 52 is used as the working area of the processor 51.

[0034] The auxiliary storage unit 53 includes a non-volatile memory typified by an EEPROM (Electrically Erasable Programmable Read-Only Memory). In addition to the program P1, the auxiliary storage unit 53 stores various data used in the processing by the processor 51. The auxiliary storage unit 53 supplies data used by the processor 51 to the processor 51 in accordance with an instruction from the processor 51, and stores the data supplied from the processor 51.

[0035] The input unit 54 includes input devices typified by input keys, buttons, switches, keyboards, pointing devices, and DI (Digital Input) contacts (photo-coupler inputs). The input unit 54 acquires information input by the user of the machining diagnosis device 1 and other information provided from the outside, and notifies the acquired information to the processor 51.

[0036] The output unit 55 includes output devices typified by LEDs (Light Emitting Diodes), LCDs (Liquid Crystal Displays), DO (Digital Output) contacts (photo-coupler outputs), and speakers. The output unit 55 presents various information to the user or outputs it to the outside in accordance with an instruction from the processor 51.

[0037] The communication unit 56 has a network interface circuit and an analog signal circuit for communicating with an external device. The external device is, for example, the numerical control device 70 and the external sensor 80 described above. The external sensor 80 is a sensor provided outside the numerically controlled machine tool 7, and includes a temperature sensor, a vibration sensor, and the like. The communication unit 56 receives a signal from the outside and outputs the information indicated by this signal to the processor 51. Further, the communication unit 56 transmits a signal indicating the information output from the processor 51 to an external device or outputs an analog signal.

[0038] By the cooperation of the above-described hardware configuration, the machining diagnosis apparatus 1 performs learning of a model for diagnosing the numerical control machine tool 7 using the data collected from the numerical control machine tool 7. Further, when the learning of the model is completed, the machining diagnosis apparatus 1 diagnoses the machining by the processing machine 90 provided in the numerical control machine tool 7, that is, diagnoses the result of the machining performed by the processing machine 90, based on the learned model and the data collected when the numerical control machine tool 7 is machining the workpiece WA. Details of the machining diagnosis will be described later.

[0039] FIG. 6 is a diagram showing a functional block configuration example of the machining diagnosis apparatus 1 according to the embodiment. The machining diagnosis apparatus 1 includes a data acquisition unit 11, a feature quantity calculation unit 12, a storage unit 13, a learning target specification unit 14, a learning unit 15, an inference unit 16, and a notification unit 17.

[0040] The data acquisition unit 11 and the notification unit 17 are realized by the communication unit 56 shown in FIG. 5. Further, the storage unit 13 is realized by the main storage unit 52 and the auxiliary storage unit 53 shown in FIG. 5. The feature quantity calculation unit 12, the learning target specification unit 14, the learning unit 15, and the inference unit 16 are realized by the processor 51, the main storage unit 52, and the auxiliary storage unit 53 shown in FIG. 5. That is, the feature quantity calculation unit 12, the learning target specification unit 14, the learning unit 15, and the inference unit 16 are realized by the processor 51 executing a program P1 for operating as each of these units. Note that the program P1 for operating as the feature quantity calculation unit 12, the learning target specification unit 14, the learning unit 15, and the inference unit 16 is assumed to be stored in advance in the auxiliary storage unit 53, but is not limited thereto. The program P1 may be supplied to the user of the machining diagnosis apparatus 1 in a state of being written in a recording medium such as a CD (Compact Disc)-ROM or a DVD (Digital Versatile Disc)-ROM, and the user may install the above-described program P1 in the auxiliary storage unit 53. Further, the above-described program P1 may be in a form downloaded by the communication unit 56 from a server via a network.

[0041] Each unit of the machining diagnosis apparatus 1 shown in FIG. 6 will be described.

[0042] The data acquisition unit 11 acquires, from the numerical control device 70, physical quantities and the like related to the machining operation of the machine tool 90 that were collected while the machine tool 90 was machining the workpiece WA, and also acquires analog data from the external sensor 80. As physical quantities collected during machining, for example, the data acquisition unit 11 acquires detection values from a current sensor S1, a vibration sensor S2, and a temperature sensor S3 arranged on the spindle motor 93 of the machine tool 90. The data acquisition unit 11 may acquire the rotational speed, moving amount, torque, force, acceleration, temperature, etc. of each motor provided in the machine tool 90. Note that the numerical control device 70 periodically collects the above physical quantities at predetermined regular time intervals, and controls the machine tool 90 based on the collected physical quantities.

[0043] In addition, the data acquisition unit 11 acquires information that the machining program generation unit 79 of the numerical control device 70 referred to when generating the machining program, specifically, workpiece material, cutting tool used, peripheral speed, feed rate, radial depth of cut, axial depth of cut, etc. The data acquisition unit 11 also acquires the machining program name generated by the machining program generation unit 79 of the numerical control device 70, the number of times the cutting tool is used counted for each cutting tool used in the numerical control device 70, the cumulative value of the tool use time, information on the steps during the execution of the machining program, etc. In the following description, the information (such as information on workpiece material, cutting tool used, peripheral speed, feed rate, radial depth of cut, axial depth of cut, etc.) that the data acquisition unit 11 acquires and that the machining program generation unit 79 referred to when generating the machining program may be collectively referred to as machining conditions. The machining conditions shall include at least workpiece material, cutting tool used, peripheral speed, feed rate, radial depth of cut, and axial depth of cut.

[0044] The feature quantity calculation unit 12 calculates the feature quantity of the physical quantity acquired by the data acquisition unit 11. The feature quantity calculation unit 12 calculates statistical feature quantities, such as the average value, maximum value, minimum value, median value, standard deviation, kurtosis, skewness, etc. over a predetermined period, based on a plurality of physical quantities collected during the predetermined period.

[0045] The memory unit 13 stores various data necessary for the operation of the machining diagnosis device 1.

[0046] The learning target specifying unit 14 extracts, as search conditions, the machining condition information collected from the storage unit 72 of the numerical control device 70 and the machining program generation unit 79, the feature amounts to be used as learning targets when the learning unit 15 learns a model for diagnosing the numerical control machine tool 7, from the feature amounts calculated by the feature amount calculation unit 12.

[0047] The learning unit 15 performs machine learning using the learning data based on the feature amounts extracted by the learning target specifying unit 14, and generates a learned model for diagnosing the numerical control machine tool 7. The learning unit 15 generates a learned model for each machining condition. That is, the learning unit 15 learns the relationship between the feature amounts of the physical quantities collected when machining is performed under the same machining conditions and the quality information of the workpiece machined by the numerical control machine tool 7.

[0048] The inference unit 16 performs machining diagnosis of the numerical control machine tool 7 based on the learned models for each machining condition generated by the learning unit 15 and the feature amounts calculated by the feature amount calculation unit 12. Specifically, the inference unit 16 determines whether the machining result by the machining unit 90 of the numerical control machine tool 7 is normal, and determines the state such as the wear amount and wear tendency of the tool. In the determination of whether the machining result is normal, the inference unit 16 performs, for example, one or both of the determination of whether the difference between the dimension of the machined section where machining has ended and the target dimension is within a specified range, and the determination of whether the difference between the shape of the workpiece after machining has ended and the target shape is within a specified range. Further, the inference unit 16 estimates the dimension, curvature, straightness, and surface roughness of the workpiece after machining by the machining unit 90 has ended. Note that the inference unit 16 may execute all of the determination of the machining result, the state determination, and the estimation of the workpiece after machining has ended (estimation of dimension, curvature, straightness, and surface roughness) in the machining diagnosis, or may execute a part of them in the machining diagnosis.

[0049] The notification unit 17 notifies the inference result by the inference unit 16, that is, the diagnosis result of the numerical control machine tool 7, to the numerical control device 70 or the like.

[0050] Next, a learning phase in which the machining diagnosis device 1 generates a learned model for diagnosing the numerically controlled machine tool 7 and a utilization phase in which the machining diagnosis device 1 diagnoses the numerically controlled machine tool 7 using the learned model will be described.

[0051] <Learning phase> FIG. 7 is a flowchart showing an example of an operation in which the machining diagnosis device 1 generates a learned model for diagnosing the numerically controlled machine tool 7.

[0052] First, the machining diagnosis device 1 acquires data necessary for generating a model (step S11). Specifically, the data acquisition unit 11 acquires, from the numerical control device 70, physical quantities related to the machining operation collected while the processing machine 90 is machining the workpiece WA and the machining conditions. The data acquisition unit 11 may acquire the above data while the numerically controlled machine tool 7 is machining the workpiece WA with the processing machine 90, or may acquire, after the machining is completed, the data collected when the numerically controlled machine tool 7 machined the workpiece WA with the processing machine 90 and stored in the numerical control device 70. The machining conditions at the time when the physical quantities are collected and acquired by the data acquisition unit 11 are the same machining conditions as those used when the machining program generation unit 79 generates the machining program used when the numerical control device 70 controls the processing machine 90 to machine the workpiece WA. Note that the workpiece material may be acquired once during the machining of the same workpiece WA.

[0053] Next, the machining diagnosis device 1 calculates the feature amount of the physical quantity acquired in step S11 (step S12). Specifically, the feature amount calculation unit 12 calculates the feature amount of the physical quantity. Here, even for machining performed under the same machining conditions, if the machining range (the length of the machining section) is different, the required time until machining is completed is also different. For this reason, the feature amount calculation unit 12 calculates a statistical feature amount that is not affected by the required time until machining is completed, such as an integrated value of the physical quantity, but a statistical feature amount that is not affected by the required time until machining is completed. Specifically, the feature amount calculation unit 12 calculates, as statistical feature amounts, the average value, maximum value, minimum value, median value, standard deviation, kurtosis, skewness, etc. in a defined period. The feature amount calculation unit 12 calculates, for example, the average value, maximum value, minimum value, median value, etc. of the physical quantity collected from the start of machining under certain machining conditions until the machining conditions are changed. The feature amount calculation unit 12 may divide the section where machining is performed under the same machining conditions into subsections of a certain width and calculate the feature amount for each subsection. The feature amount calculation unit 12 may calculate the feature amount using the latest data acquired and the data acquired in the past each time the data acquisition unit 11 acquires the above data in step S11, or may calculate the feature amount at a defined timing using the data acquired in the past. The defined timing may be the timing when the number of acquired data reaches a certain number, or the timing when the machining diagnosis device 1 receives a predetermined operation by the user, for example, the timing when it receives an operation instructing the generation of a learned model. The data acquired in the past may include data obtained in machining using different machining programs, that is, data obtained when machining with different target shapes is performed on each of different workpieces and are mixed. Further, the feature amount calculation unit 12 may calculate the feature amount using the data acquired by the data acquisition unit 11 in a certain period each time a certain period of a defined length has elapsed.

[0054] Next, the machining diagnosis device 1 creates learning data for each machining condition based on the feature amount calculated in step S12 (step S13). Specifically, the learning target specifying unit 14 groups the feature amounts calculated by the feature amount calculating unit 12 for each feature amount with the same machining condition to create learning data for each machining condition. At this time, when the same machining is repeatedly executed on the same section until the target shape (target dimension) is machined, the learning target specifying unit 14 specifies the approach section, which is the section from when the machining condition is set and the tool starts approaching the workpiece until the tool contacts the workpiece and cutting starts with the cutting depth specified by the machining condition, and excludes the feature amounts of the physical quantities collected in the approach section from the learning data. The approach section is provided to prevent the tool from colliding violently with the workpiece. Also, the learning target specifying unit 14 excludes the feature amounts of the physical quantities collected in the target dimension adjustment section, which will be described later, from the learning data. That is, as shown in FIG. 8, the learning target specifying unit 14 acquires the feature amounts in the repeated path (step S21) and specifies the feature amounts to be the learning target (step S22). The feature amounts in the repeated path are the feature amounts of the physical quantities collected when the same machining is repeatedly executed on the same section.

[0055] In step S21, the learning target specifying unit 14 extracts the feature amounts in the repetitive path from the feature amounts calculated by the feature amount calculating unit 12. In step S22, the learning target specifying unit 14 sets, as the learning target, the remaining feature amounts among the feature amounts in the repetitive path, excluding the feature amounts in the approach section and the feature amounts in the target dimension alignment section. FIG. 9 is a diagram for explaining the approach section and the target dimension alignment section. In FIG. 9, the horizontal axis represents the cutting time (seconds), which is the elapsed time since the start of the cutting operation, and the vertical axis represents the feature amount. In the example shown in FIG. 9, after the start of the cutting operation, the tool contacts the workpiece around the time when the cutting time exceeds 40 seconds and cutting starts, and the feature amount increases for each pass. The magnitude of the feature amount is correlated with the depth of cut, and when the depth of cut becomes constant (prescribed depth of cut), the magnitude of the feature amount for each pass becomes substantially constant. However, when the machining target surface of the workpiece is not uniform and has irregularities or the like, the depth of cut does not become the prescribed depth of cut indicated by the machining conditions immediately after the tool contacts the workpiece and actual cutting starts, and the physical quantity and the feature amount may gradually increase. Considering such characteristics, the learning target specifying unit 14 specifies the approach section based on the change amount of the feature amount for each pass. For example, the learning target specifying unit 14 sets, as the approach section, the section from the start of the cutting operation until the increase amount of the feature amount for each pass becomes 10% or less after the increase of the feature amount for each pass starts. Note that the change amount of 10% used for specifying the approach section is an example. The learning target specifying unit 14 may determine, as the approach section, another value, for example, the section until the change amount becomes 5% or less. Also, when the depth of cut required to achieve the target shape is not an integer multiple of the depth of cut per cutting operation, the depth of cut in the last pass (hereinafter referred to as the final pass) of the repetitive path becomes less than the depth of cut in the immediately preceding pass, and the feature amount also decreases. Therefore, when the feature amount in the final pass decreases significantly more than the feature amount in the previous pass, for example, when the decrease amount (change amount) exceeds 10%, the learning target specifying unit 14 determines the final pass as the target dimension alignment section and excludes it from the learning target. Note that when the depth of cut required to achieve the target shape is an integer multiple of the depth of cut per cutting operation, the final pass does not become the target dimension alignment section.In this way, the learning target specifying unit 14 specifies the feature amount of the physical quantity collected in the section that does not belong to either the approach section or the target dimension alignment section among the feature amounts in the repetitive path, and sets it as the feature amount of the learning target.

[0056] When the learning target specifying unit 14 specifies the feature amount of the learning target from the feature amounts in the repetitive path, the feature amount in the path where the change amount of the feature amount between adjacent paths is within a determined range, centered on the path with the largest feature amount such as the average value or median value of the physical quantity, may be determined as the learning target. For example, the feature amount in the path where the change amount of the feature amount is within ±10% may be determined as the learning target.

[0057] The learning target specifying unit 14 may limit the feature amount of the learning target to the feature amount of the physical quantity collected when repeatedly performing the same processing on the same section until the target dimension is processed. In this case, in the machine learning for the learning unit 15 to generate a learned model, it becomes possible to learn the trend of the feature amount when repeatedly performing the same processing.

[0058] Returning to the description of FIG. 7, the machining diagnosis apparatus 1 then generates a learned model for each machining condition (step S14). Specifically, the learning unit 15 performs machine learning using the learning data for each machining condition created by the learning target specifying unit 14, and generates a plurality of learned models corresponding to each of the machining conditions. The learning generates a model from the feature amounts that are the learning data for each machining condition, calculates a principal component distance value, an abnormality degree of a neural network, etc. from the learning data used for model generation or other learning target data, and extracts the peak value in the learning target data as the normal / abnormal determination threshold value. Alternatively, a quality prediction model may be generated by machine learning using the learning data for each machining condition with the quality value as the target variable. That is, the learning unit 15 generates a learned model for inferring the machining result based on the feature amounts of the physical quantities collected while the machining machine 90 of the numerically controlled machine tool 7 is machining the workpiece WA. In the machining after the learning is completed, if the learned model does not exceed the normal / abnormal determination threshold value determined by the principal component distance value or the abnormality degree of the neural network based on the feature amounts calculated after the machining, it is regarded as normal machining, or when predicting the quality value with the learned model, if the error between the predicted values such as dimensions (hereinafter sometimes referred to as actual dimensions), curvature, straightness, surface roughness, etc. and the dimensions of the target shape (hereinafter sometimes referred to as target dimensions) is within the specified range, it is regarded as normal. The learning unit 15 uses machine learning methods such as principal component analysis, multiple regression analysis, support vector machine, decision tree, gradient boosting decision tree, neural network, etc. in the machine learning.

[0059] The learning unit 15 may use the error between the actual dimension and the target dimension as the machining result and learn the relationship between the error and the feature amounts that are the learning data. In this case, the learning unit 15 generates a learned model for inferring the machining quality based on the feature amounts of the physical quantities collected while the machining machine 90 of the numerically controlled machine tool 7 is machining the workpiece WA, specifically, a learned model for inferring the error between the target dimension and the actual dimension.

[0060] The learning unit 15 may learn the relationship between the wear state of the tool used for machining and the feature amounts that are the learning data. In this case, the learning unit 15 generates a learned model for inferring the wear state of the tool based on the feature amounts of the physical quantities collected while the processing machine 90 of the numerically controlled machine tool 7 is machining the workpiece WA. The wear state may be information on whether the cutting edge of the tool is worn to the extent that replacement is necessary, or may be the actual wear amount from the initial state (new state) of the cutting edge of the tool, or may be a numerical value indicating the degree of wear, for example, with the new state being 0% and the state where replacement is necessary being 100%, indicating what percentage of wear has occurred. When the processing machine 90 repeatedly executes the same processing for one section and performs cutting to a target dimension, the learning unit 15 may learn the relationship between the feature amounts and the wear state of the tool in each of the repeated paths, or may learn the relationship between the trend of the feature amounts for each repeated path and the wear state of the tool. The difference in the feature amounts may be taken before and after the repeated path, and the relationship between the trend of the difference and the wear state of the tool may be learned. When the tool wears, a change occurs in the feature amounts of the physical quantities collected during machining. That is, there is a correlation between the wear state of the tool and the feature amount or the difference in the feature amounts. Therefore, by learning the trend of the feature amounts for each repeated path, a learned model capable of inferring the wear state of the tool with high accuracy can be generated.

[0061] As described above, in learning the model for diagnosing the numerically controlled machine tool 7, the machining diagnosis apparatus 1 acquires from the numerical control device 70 the physical quantities collected during machining and the machining conditions when the physical quantities were collected, creates learning data for each machining condition based on the feature amounts of the physical quantities and the machining conditions, and performs learning using the learning data for each machining condition to generate a learned model for each machining condition. Further, when generating the learning data for each machining condition, the machining diagnosis apparatus 1 specifies the feature amounts calculated based on the physical quantities actually collected while machining the workpiece under the acquired machining conditions among the feature amounts of the physical quantities collected during machining, and creates the learning data.

[0062] The processing diagnosis device 1 may generate multiple types of learned models for each processing condition. For example, the processing diagnosis device 1 may create a learned model for inferring whether the processing performed by the processing machine 90 is normal or abnormal and a learned model for inferring the wear state of the tool for each processing condition. For example, the processing diagnosis device 1 may generate a learned model for inferring the error between the target dimension and the actual dimension and a learned model for inferring the wear state of the tool for each processing condition.

[0063] <Utilization phase> FIG. 10 is a flowchart showing an example of the operation of the processing diagnosis device 1 according to the embodiment for diagnosing the processing result by the numerical control machine tool 7 using the learned model.

[0064] First, the processing diagnosis device 1 acquires data necessary for the diagnosis process of the numerical control machine tool 7 (step S31). Specifically, the data acquisition unit 11 periodically acquires the physical quantities related to the processing operation collected from the processing machine 90 during the processing of the workpiece WA and the processing conditions from the numerical control device 70. The data acquired by the data acquisition unit 11 in this step S31 is the same type of data as the data collected in step S11 of the above-described learning phase.

[0065] Next, the processing diagnosis device 1 calculates the feature amounts of the physical quantities included in the data acquired in step S31 (step S32). Specifically, the feature amount calculation unit 12 calculates the same feature amounts as those calculated in step S12 of the above-described learning phase at a predetermined timing. For example, the processing diagnosis device 1 calculates the feature amounts at a period that is an integer multiple of the period at which the data acquisition unit 11 acquires data in step S31. When the processing conditions included in the data acquired by the data acquisition unit 11 in step S31 change, that is, when the processing conditions included in the latest data acquired by the data acquisition unit 11 are different from the processing conditions included in the previously acquired data, the processing diagnosis device 1 may calculate the feature amounts of the physical quantities collected during the processing under the processing conditions before the change. The feature amount calculation unit 12 outputs the calculated feature amounts and the processing conditions when the physical quantities used for the calculation of the feature amounts were collected to the inference unit 16.

[0066] Next, the machining diagnosis device 1 estimates the machining result by the machining unit 90 of the numerically controlled machine tool 7 based on the feature amount calculated in step S32 and the learned model generated in the above-described learning phase (step S33). Specifically, the inference unit 16, which is an estimation unit, performs a machining diagnosis for estimating the machining result based on the feature amount input from the feature amount calculation unit 12 and the learned model corresponding to the machining conditions input from the feature amount calculation unit 12. The inference unit 16 inputs the feature amount into the learned model, and obtains the estimated machining result output from the learned model accordingly.

[0067] Next, the machining diagnosis device 1 outputs the estimation result in step S33 (step S34). Specifically, the notification unit 17 outputs the machining result estimated by the inference unit 16 in step S33 to the numerical control device 70.

[0068] Note that the machining diagnosis device 1 may estimate the wear state of the tool in step S33, and if it is estimated that the wear amount has reached the wear amount at which tool replacement is necessary, in step S34, notify the numerical control device 70 of the information on the tool that needs to be replaced and instruct tool replacement.

[0069] As also described in the learning phase, even for machining repeatedly executed under the same machining conditions, in the approach section, machining with the cutting amount indicated by the machining conditions is not performed. That is, the physical quantities collected in the section from the start of machining to the end of the approach section are significantly different from the physical quantities collected in other sections where machining is performed under the same machining conditions. Therefore, in the approach section, it is not possible to perform a correct diagnosis using the learned model, and if a diagnosis is performed, the diagnosis result will be abnormal machining. For this reason, the machining diagnosis device 1 waits until the approach section ends and the diagnosis result becomes normal machining for the first time after starting repeated machining, and starts the actual diagnosis. Specifically, the machining diagnosis device 1 treats the diagnosis results during the period from the start of repeated machining until the diagnosis result becomes normal machining for the first time as invalid. FIG. 11 is a flowchart showing an example of the diagnosis operation by the machining diagnosis device 1 according to the embodiment.

[0070] The machining diagnosis device 1 first executes the same processing as steps S31 to S33 shown in FIG. 10 to estimate the machining result (step S41), and checks whether the estimation result is normal machining (step S42). If the estimation result is abnormal machining (step S42: No), the machining diagnosis device 1 determines that the estimation result is invalid, returns to step S41, and estimates the machining result again.

[0071] If the estimation result is normal machining (step S42: Yes), the machining diagnosis device 1 starts the diagnosis (step S43) and estimates the machining result (step S44). In this step S44, the same processing as steps S31 to S34 shown in FIG. 10 is executed to estimate the machining result and output the estimated machining result.

[0072] After the machining diagnosis device 1 estimates the machining result in step S44, it checks whether the estimation result is abnormal machining (step S45). If the estimation result is normal machining (step S45: No), the machining diagnosis device 1 returns to step S44 and estimates the machining result and outputs the estimated machining result again. On the other hand, if the estimation result is abnormal machining (step S45: Yes), the machining diagnosis device 1 checks whether it is the final pass, that is, whether the estimation result is abnormal machining in the final pass (step S46). If the estimation result is abnormal machining in the final pass (step S46: Yes), since it is caused by machining with a cutting amount different from that in other passes, the machining diagnosis device 1 simply ends the operation. On the other hand, if the estimation result is abnormal machining outside the final pass (step S46: No), the change amount of the feature amount tends to be different from that in normal machining due to breakage or loss of the tool used, and the estimation result may be abnormal machining. Therefore, the machining diagnosis device 1 notifies the numerical control device 70 of the numerical control machine tool 7 that the tool is abnormal (step S47) and ends the diagnosis operation.

[0073] In this way, the machining diagnosis apparatus 1 acquires, from the numerical control device 70, the physical quantities related to the machining operation collected from the machine tool 90 during the machining of the workpiece WA and the machining conditions at the time when the physical quantities are collected, and uses the feature quantities of the acquired physical quantities and the learned model corresponding to the acquired machining conditions to perform machining diagnosis of the numerically controlled machine tool 7, specifically, to estimate the machining result by the machine tool 90 provided in the numerically controlled machine tool 7.

[0074] Note that, in the present embodiment, the numerical control device 70 is provided with a machining program generation unit 79, and the machining program generation unit 79 is configured to generate a machining program. However, the present invention is not limited to such a configuration. A configuration may be adopted in which a CAM (Computer Aided Manufacturing) system creates a machining program based on the drawing data of the target shape of the workpiece WA created by a CAD (Computer Aided Design) system.

[0075] When the CAM system creates a machining program, the operator refers to the drawing of the target shape created in the CAD system and past experience of similar machining (cutting conditions for cutting with the same tool in the past), etc., and sets the cutting amount in the CAM system, and also registers the coordinate information indicating the passing points of the tool in the CAM system. When the necessary settings and registrations are made by the operator and further instructed to create a machining program, the CAM system automatically generates a machining program based on the registered coordinate information and the set machining conditions (work material, tool used, peripheral speed, feed rate, radial depth of cut, axial depth of cut). The CAM system holds information on what machining conditions to use for each line of the machining program (hereinafter referred to as information on machining conditions per line). Therefore, the machining diagnosis device 1 acquires the information on machining conditions per line from the CAM system in advance together with the machining program name and holds it in the storage unit 13. When machining is carried out on the numerically controlled machine tool 7, the machining diagnosis device 1 acquires from the numerical control device 70 the name of the machining program being executed, the physical quantities collected by the numerical control device 70, and the information on the steps during the execution of the machining program (hereinafter referred to as information on steps during execution), and by comparing the acquired information on steps during execution with the information on machining conditions per line held in the storage unit 13, identifies the machining conditions when the acquired physical quantities are obtained. Thereby, the machining diagnosis device 1 can create learning data for each machining condition, perform machine learning, and generate a learned model for each machining condition. Further, the machining diagnosis device 1 can diagnose the machining result by the numerically controlled machine tool 7 using the learned model for each machining condition.

[0076] As described above, the machining diagnosis apparatus 1 according to the present embodiment acquires, from the numerically controlled machine tool 7 including the numerical control device 70 and the machining machine 90, the physical quantities collected when the machining machine 90 is machining the workpiece WA, and the machining conditions when the physical quantities are collected, performs machine learning using the acquired physical quantities and machining conditions, and generates a learned model for diagnosing the numerically controlled machine tool 7 for each machining condition. Further, after generating the learned model, the machining diagnosis apparatus 1 acquires the physical quantities collected when the machining machine 90 is machining the workpiece WA, and the machining conditions when the physical quantities are collected, and performs a machining diagnosis of the numerically controlled machine tool 7 based on the learned model corresponding to the acquired machining conditions and the feature amounts of the acquired physical quantities. Since the machining diagnosis apparatus 1 according to the present embodiment performs machining diagnosis in consideration of the machining conditions (workpiece material, cutting tool used, peripheral speed, feed rate, radial depth of cut, axial depth of cut) when the physical quantities are collected, even when the machining machine 90 of the numerically controlled machine tool 7 to be diagnosed performs machining for small-lot production of multiple varieties, the machining diagnosis of the numerically controlled machine tool 7 can be performed.

[0077] In the machining diagnosis operation of the numerically controlled machine tool 7, the machining diagnosis apparatus 1 may perform diagnoses with different contents for each machining section. For example, when the machining machine 90 repeatedly executes the same machining for one machining section until the target dimension is cut, the machining diagnosis apparatus 1 diagnoses the wear state of the cutting tool and the machining result, and when the machining machine 90 does not repeatedly execute the same machining for one machining section (performs machining only once), the machining diagnosis apparatus 1 may diagnose only the machining result. Further, the machining diagnosis apparatus 1 may diagnose either one or both of the wear state of the cutting tool and the machining result only when the machining machine 90 repeatedly executes the same machining for one machining section until the target dimension is cut.

[0078] The configurations shown in the above embodiments are merely examples, and it is possible to combine them with other known technologies, and it is also possible to omit or change a part of the configurations without departing from the gist.

[0079] For example, in the embodiment, the configuration in which the learning unit 15 and the inference unit 16 are provided inside the machining diagnosis device 1 has been described. However, one or both of the learning unit 15 and the inference unit 16 may be provided outside the machining diagnosis device 1. For example, the learning unit 15 may be provided in an external learning device, and the inference unit 16 may be provided in an external inference device, and these learning device and inference device may be connected to the machining diagnosis device 1 via a communication network or the like. Also, the machining diagnosis device 1 may be configured to be included in the numerically controlled machine tool 7.

Explanation of Reference Numerals

[0080] 1 Machining diagnosis device, 7 Numerically controlled machine tool, 11 Data acquisition unit, 12 Feature amount calculation unit, 13,72 Storage unit, 14 Learning target identification unit, 15 Learning unit, 16 Inference unit, 17 Notification unit, 51,71 Processor, 52 Main memory unit, 53 Auxiliary memory unit, 54 Input unit, 55 Output unit, 56 Communication unit, 70 Numerical control device, 73 External connection I / F, 74 Operation panel I / F, 75 Spindle control I / F, 76 Servo control I / F, 77 Digital input unit, 78 Digital output unit, 79 Machining program generation unit, 80 External sensor, 82 NC operation panel, 85 Spindle amplifier, 86 Servo amplifier, 90 Machine tool, 91 Spindle, 92,201 Tool, 93 Spindle motor, 94 Table, 95 Moving mechanism, 96x,96y,96z Servo motor, 100 Machine tool system, 202 Cutting edge, 203 Workpiece.

Claims

1. A data acquisition unit that acquires a physical quantity related to the operating state of a processing machine for processing a workpiece, and processing conditions that do not depend on the target shape when the physical quantity is obtained; A feature quantity calculation unit that calculates, for each of the processing conditions, a statistical feature quantity that is not affected by the time required until the end of processing of the physical quantity obtained during processing under the same processing conditions, based on the physical quantity and the processing conditions; An estimation unit that performs processing diagnosis based on the statistical feature quantity for each of the processing conditions; comprising; The processing machine is provided in a numerically controlled machine tool having a processing program generation unit that automatically generates a processing program based on the input material of the workpiece and the tool used in the processing, and a database of processing parameters, and performs the processing using the processing program generated by the processing program generation unit, The data acquisition unit acquires the processing conditions from the processing program generation unit, The feature quantity calculation unit calculates the statistical feature quantity using the physical quantity newly acquired by the data acquisition unit and the physical quantity previously acquired by the data acquisition unit, A processing diagnosis device characterized by the above.

2. A data acquisition unit that acquires a physical quantity related to the operating state of a processing machine for processing a workpiece, and the processing conditions when the physical quantity is obtained; A feature quantity calculation unit that calculates, for each of the processing conditions, a statistical feature quantity of the physical quantity obtained during processing under the same processing conditions, based on the physical quantity and the processing conditions; An estimation unit that performs processing diagnosis based on the statistical feature quantity for each of the processing conditions; comprising; The estimation unit performs the processing diagnosis based on the statistical feature quantity of the physical quantity obtained when the processing machine repeatedly executes processing under the same processing conditions for the same section, A processing diagnosis device characterized by the above.

3. The processing conditions are the material of the workpiece, the tool used in the processing, the peripheral speed, the feed speed, the radial depth of cut, and the axial depth of cut when performing the processing, The processing diagnosis device according to claim 1 or 2, characterized by the above.

4. The processing machine is provided in a numerically controlled machine tool having a processing program generation unit that automatically generates a processing program based on the input material of the workpiece and the tool used in the processing, and a database of processing parameters, and performs the processing using the processing program generated by the processing program generation unit, The data acquisition unit acquires the processing conditions from the processing program generation unit, The processing diagnosis apparatus according to claim 2, characterized in that...

5. The processing machine performs the processing using a processing program generated by the CAM system based on the target shape of the workpiece, The data acquisition unit acquires the processing conditions from the CAM system, The processing diagnosis apparatus according to claim 1 or 2, characterized in that...

6. The estimation unit estimates whether the processing performed by the processing machine is normal or abnormal based on the statistical feature amount, The processing diagnosis apparatus according to claim 1 or 2, characterized in that...

7. The estimation unit estimates the quality of the processing performed by the processing machine based on the statistical feature amount, The processing diagnosis apparatus according to claim 1 or 2, characterized in that...

8. The processing diagnosis result by the estimation unit includes the wear state of the tool used by the processing machine in the processing when the physical quantity was obtained, When the wear state indicates that the wear amount has reached the amount that requires tool replacement, a notification unit that instructs tool replacement, The processing diagnosis apparatus according to claim 1 or 2, characterized by comprising...

9. The estimation unit performs the processing diagnosis based on the statistical feature amount of the physical quantity obtained when the processing machine repeatedly executes processing under the same processing conditions for the same section, The processing diagnosis apparatus according to claim 1, characterized in that...

10. The estimation unit performs the processing diagnosis using the remaining statistical feature amount after excluding the statistical feature amount of the physical quantity obtained when the cutting amount of the tool with respect to the workpiece does not reach the specified cutting amount indicated by the processing conditions, The processing diagnosis apparatus according to claim 2 or 9, characterized in that...

11. The estimation unit repeatedly performs the processing diagnosis based on the feature amount of the physical quantity obtained in each of the repeated processings under the same processing conditions for the same section by the processing machine, and invalidates the result of the processing diagnosis performed before the cutting amount of the tool with respect to the workpiece reaches the specified cutting amount indicated by the processing conditions, and sets the result of the processing diagnosis performed after the cutting amount of the tool with respect to the workpiece reaches the specified cutting amount indicated by the processing conditions as a valid processing diagnosis result, The processing diagnosis apparatus according to claim 10, characterized in that...

12. After the depth of cut of the tool with respect to the workpiece reaches the specified depth of cut indicated by the machining conditions, if the result of the machining diagnosis based on the characteristic quantities of the physical quantities obtained in the machining of passes other than the final pass in the repeated machining under the same machining conditions is abnormal machining, the estimation unit estimates that the tool used in the machining has broken or chipped. The machining diagnosis apparatus according to claim 10, characterized in that.

13. The characteristic quantity calculation unit calculates the statistical characteristic quantity based on the physical quantity obtained in the machining using different machining programs. The machining diagnosis apparatus according to claim 1 or 2, characterized in that.

14. A learning unit that performs machine learning using learning data created based on the statistical characteristic quantities for each machining condition and generates a learned model for performing the machining diagnosis based on the statistical characteristic quantities for each machining condition. Comprising When the data acquisition unit acquires the physical quantity and the machining condition, the estimation unit performs the machining diagnosis using the characteristic quantity calculated by the characteristic quantity calculation unit based on the acquired physical quantity and the learned model corresponding to the acquired machining condition. The machining diagnosis apparatus according to claim 1 or 2, characterized in that.

15. A learning target specifying unit that determines the statistical characteristic quantity to be used in the machine learning as the remaining statistical characteristic quantity after excluding the statistical characteristic quantity of the physical quantity obtained when the depth of cut of the tool with respect to the workpiece does not reach the specified depth of cut indicated by the machining conditions among the statistical characteristic quantities of the physical quantities obtained when the machining machine repeatedly executes machining under the same machining conditions for the same section. The machining diagnosis apparatus according to claim 14, characterized by comprising.

16. A data acquisition unit that acquires a physical quantity related to the operating state of a machining machine that machines a workpiece and a machining condition that does not depend on the target shape when the physical quantity is obtained. A characteristic quantity calculation unit that calculates, for each machining condition, a statistical characteristic quantity that is not affected by the required time until the end of machining of the physical quantity obtained during machining under the same machining conditions based on the physical quantity and the machining condition. A learning unit that performs machine learning using learning data created based on the statistical characteristic quantities for each machining condition and generates a learned model for performing a machining diagnosis based on the statistical characteristic quantities for each machining condition. Comprising The processing machine is provided in a numerically controlled machine tool having a machining program generation unit that automatically generates a machining program based on the material of the workpiece input and the tool used in the machining, and the machining is performed using the machining program generated by the machining program generation unit. The data acquisition unit acquires the machining conditions from the machining program generation unit. The feature amount calculation unit calculates the statistical feature amount using the physical quantity newly acquired by the data acquisition unit and the physical quantity already acquired by the data acquisition unit in the past. A learning device characterized by this.

17. A data acquisition unit that acquires a physical quantity related to the operating state of a processing machine for processing a workpiece and a processing condition independent of the target shape when the physical quantity is obtained, A feature amount calculation unit that calculates, for each processing condition, a statistical feature amount that is not affected by the time required until the end of processing of the physical quantity obtained during processing under the same processing condition based on the physical quantity and the processing condition. A learned model for performing machining diagnosis based on the statistical feature amount, generated by a learning device that performs machine learning using learning data created based on the statistical feature amount for each processing condition, and the statistical feature amount for each processing condition calculated by the feature amount calculation unit. An estimation unit for performing machining diagnosis based on the above. Comprising The processing machine is provided in a numerically controlled machine tool having a machining program generation unit that automatically generates a machining program based on the material of the workpiece input and the tool used in the machining, and the machining is performed using the machining program generated by the machining program generation unit. The data acquisition unit acquires the machining conditions from the machining program generation unit. The feature amount calculation unit calculates the statistical feature amount using the physical quantity newly acquired by the data acquisition unit and the physical quantity already acquired by the data acquisition unit in the past. An inference device characterized by this.

18. A machining diagnosis method in which a machining diagnosis device estimates the machining result of a workpiece by a machining machine, A data acquisition step of acquiring a physical quantity related to the operating state of the processing machine and a processing condition independent of the target shape when the physical quantity is obtained. A feature quantity calculation step of calculating, for each processing condition, a statistical feature quantity that is not affected by the time required until the end of processing of the physical quantity obtained during processing under the same processing condition, based on the physical quantity and the processing condition; A diagnosis step of diagnosing processing based on the statistical feature quantity for each processing condition; comprising; The processing machine is provided in a numerically controlled machine tool having a processing program generation unit that automatically generates a processing program based on the material of the workpiece to be processed and the tool used in processing, and the input database of processing parameters. The processing is performed using the processing program generated by the processing program generation unit. In the data acquisition step, the processing conditions are acquired from the processing program generation unit. In the feature quantity calculation step, the statistical feature quantity is calculated using the physical quantity newly acquired in the data acquisition step and the physical quantity acquired in the past in the data acquisition step. A processing diagnosis method characterized by the above.

19. A data acquisition step of acquiring a physical quantity related to the operating state of a processing machine for processing a workpiece and processing conditions independent of the target shape when the physical quantity is obtained; A feature quantity calculation step of calculating, for each processing condition, a statistical feature quantity that is not affected by the time required until the end of processing of the physical quantity obtained during processing under the same processing condition, based on the physical quantity and the processing condition; A diagnosis step of diagnosing processing based on the statistical feature quantity for each processing condition; executed by a computer, The processing machine is provided in a numerically controlled machine tool having a processing program generation unit that automatically generates a processing program based on the material of the workpiece to be processed and the tool used in processing, and the input database of processing parameters. The processing is performed using the processing program generated by the processing program generation unit. In the data acquisition step, the processing conditions are acquired from the processing program generation unit. In the feature quantity calculation step, the statistical feature quantity is calculated using the physical quantity newly acquired in the data acquisition step and the physical quantity acquired in the past in the data acquisition step. A processing diagnosis program characterized by the above.

20. A processing machine for processing a workpiece; A numerical control device for controlling the processing machine according to a processing program; A processing diagnosis device for diagnosing the processing by the processing machine, comprising: The processing diagnosis device is A data acquisition unit that acquires a physical quantity related to the operating state of a processing machine for processing a workpiece and a processing condition independent of the target shape when the physical quantity is obtained; A feature quantity calculation unit that calculates, for each of the processing conditions, a statistical feature quantity that is not affected by the time required until the end of processing of the physical quantity obtained during processing under the same processing conditions based on the physical quantity and the processing condition; An estimation unit that performs processing diagnosis based on the statistical feature quantity for each of the processing conditions; Comprising; The processing machine is provided in a numerically controlled machine tool having a processing program generation unit that automatically generates a processing program based on the material of the input workpiece, the tool used in the processing, and a database of processing parameters, and performs the processing using the processing program generated by the processing program generation unit, The data acquisition unit acquires the processing condition from the processing program generation unit, The feature quantity calculation unit calculates the statistical feature quantity using the physical quantity newly acquired by the data acquisition unit and the physical quantity already acquired by the data acquisition unit in the past, A machine tool system characterized by this.

21. The numerical control device, Based on the material of the input workpiece, the tool used in the processing, and a database of processing parameters, determines the processing condition and generates the processing program, The machine tool system according to claim 20, characterized by this.

Images