Diagnostic device, diagnostic system, diagnostic method, and program

Abstract

To make it possible to diagnose a state of a machine tool with high precision.SOLUTION: A diagnostic device comprises: a time-series data acquisition unit that acquires time-series data for indicating an active state of a machine tool 200 of a diagnostic target; a work-piece state data acquiring unit that acquires work-piece state data for indicating a state of a work-piece 1 before machining, the work-piece being machined by the machine tool 200; a feature amount calculating unit that calculates a feature amount from the time-series data acquired by the time-series data acquiring unit; and a diagnostic unit that determines whether the machine tool 200 is normal or abnormal by determining whether the feature amount calculated by the feature amount calculating unit falls outside a normal range of the feature amount corresponding to the work-piece state data.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a diagnostic device, a diagnostic system, a diagnostic method, and a program. [Background technology]

[0002] Techniques for diagnosing the condition of machine tools are known. For example, Patent Document 1 discloses a diagnostic device that includes an acquisition unit that acquires time-series information such as a current value and frequency characteristics output to a motor that drives the axis of the machine tool, and a storage unit that stores a normal range of the time-series information corresponding to a physical quantity that indicates the environment of the machine tool measured by a sensor, and that determines that the machine tool is abnormal if the condition information acquired by the acquisition unit is not within the normal range corresponding to the physical quantity measured by the sensor. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent Publication No. 2021-47617 Summary of the Invention [Problem to be solved by the invention]

[0004] Generally, time-series information fluctuates depending on the condition of the workpiece, and the normal range of this time-series information may change depending on the condition of the workpiece before machining. In contrast, the diagnostic device of Patent Document 1 does not take into account the condition of the workpiece before machining. For this reason, there is room for improvement in terms of accurately determining whether or not an abnormality has occurred in the machine tool.

[0005] The present disclosure has been made in consideration of the above-described circumstances, and aims to provide a diagnostic device, a diagnostic system, a diagnostic method, and a program that are capable of accurately diagnosing the condition of a machine tool. [Means for solving the problem]

[0006] In order to achieve the above object, the diagnostic device according to the present disclosure includes a time series data acquisition unit that acquires time series data indicating the operating state of the machine tool to be diagnosed; a workpiece state data acquisition unit that acquires workpiece state data that indicates the state of the workpiece before it is machined by the machine tool; a feature calculation unit that calculates feature amounts from the time series data acquired by the time series data acquisition unit; and a diagnostic unit that determines whether the feature amounts calculated by the feature calculation unit are outside the normal range of the feature amounts corresponding to the workpiece state data, thereby determining whether the machine tool is abnormal or normal. [Effects of the Invention]

[0007] According to the present disclosure, whether the machine tool is normal or abnormal is determined by determining whether the feature quantities of the time-series data of the machine tool fall outside the normal range of the feature quantities corresponding to the workpiece condition data, thereby enabling the condition of the machine tool to be diagnosed with high accuracy. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram showing a functional configuration of a diagnostic system according to a first embodiment. [Figure 2] 2A is a diagram showing an example of the rotation speed of a spindle motor, (B) is a diagram showing an example of a processing feed signal, and (C) is a diagram showing an example of a current value flowing through a spindle motor, all of which are acquired by the processing data acquisition device shown in FIG. 1. [Figure 3] FIG. 1 is a block diagram showing an example of a physical configuration of an information processing device. [Figure 4] 1 is a flowchart showing a learning model generation process performed by the diagnostic system according to the first embodiment. [Figure 5] FIG. 2 is a diagram showing an example of learning data generated by the signal processing device shown in FIG. 1; [Figure 6] 1 is a flowchart showing a diagnostic process performed by the diagnostic system according to the first embodiment. [Figure 7] FIG. 1 is a block diagram showing a functional configuration of a diagnostic system according to a first embodiment. [Figure 8] FIG. 8 is a diagram showing an example of a normal range DB shown in FIG. 7 . [Figure 9] 10 is a flowchart showing a diagnostic process performed by a diagnostic system according to a second embodiment. [Figure 10] FIG. 8 is a diagram showing an example of a normal range set by the normal range DB shown in FIG. 7; DETAILED DESCRIPTION OF THE INVENTION

[0009] Hereinafter, a diagnostic device, a diagnostic system, a diagnostic method, and a program according to embodiments of the present disclosure will be described with reference to the drawings. In the drawings, the same or corresponding parts are designated by the same reference numerals.

[0010] (Embodiment 1) A diagnostic device according to an embodiment of the present disclosure will be described using an example in which it is applied to diagnostic system 100 shown in FIG. 1. Diagnostic system 100 acquires workpiece condition data indicating the pre-machining condition of a workpiece, which is an object to be machined by machine tool 200, and collects time-series data indicating the operating condition during machining from machine tool 200. The workpiece condition data includes information such as the temperature, hardness, appearance image, surface texture, lot number, and manufacturer of the workpiece before machining. Diagnostic system 100 calculates feature amounts from the collected time-series data. Diagnostic system 100 applies the calculated feature amounts and the acquired workpiece condition data to a machine learning model to determine whether or not there is an abnormality in machine tool 200.

[0011] As shown in the figure, the diagnostic system 100 of this embodiment includes a machining data acquisition device 2 that acquires machining data including time series data from the machine tool 200 to be diagnosed, a workpiece condition data acquisition device 3 that acquires workpiece condition data of the workpiece 1, a signal processing device 4 that processes the time series data, a memory device 5 that stores various information generated by the signal processing device 4, a diagnostic device 6 that determines whether or not there is an abnormality in the machine tool 200, a learning model generation device 7 that generates a learning model to be used by the diagnostic device 6, a status display 9 that displays the diagnosis results by the diagnostic device 6, and a control device 8 that controls the machine tool 200 based on the determination results by the diagnostic device 6.

[0012] The machine tool 200 is a machine for performing processes such as cutting, cutting, and polishing on the workpiece 1, and is, for example, a machining center, lathe, drill press, milling machine, polishing machine, turning center, electric discharge machine, laser processing machine, ultrasonic processing machine, etc.

[0013] Machine tool 200 is equipped with sensors such as a current sensor, a vibration sensor, a microphone, an AE (Acoustic Emission) meter, an acceleration sensor, a torque sensor, a rotation sensor, and a temperature sensor. The sensors detect physical quantities such as the value of current flowing through motors that drive tools such as drills, end mills, face mills, long drills, bit tips, and grinding wheels, the motor rotation speed, rotational torque, and vibration magnitude, and output information on the detected physical quantities. These sensors are attached, for example, to the outer wall of machine tool 200, cables in a control panel, the inside of a machining chamber, etc.

[0014] The machining data acquisition device 2 acquires machining data from the machine tool 200, including time-series data indicating the operating state during machining, temporarily stores the data, compiles a certain amount of machining data, and provides the data to the signal processing device 4. The machining data acquired by the machining data acquisition device 2 includes time-series data indicating the operating state of the machine tool 200, such as information on physical quantities detected by sensors provided in the machine tool 200 and instruction signals output from the control device 8 that controls the operation of the machine tool 200, as well as machining-related information that identifies the machining process, such as the machining program number, tool number, power consumption, coordinate information of the machining axis, hydraulic / pneumatic pressure, battery voltage, and number of days of equipment use. As illustrated in FIG. 2, the time-series data includes the rotation speed of the spindle motor provided in the machine tool 200, the machining feed signal output from the control device 8, and the current value flowing through the spindle motor. The machining data acquisition device 2 is an example of a time-series data acquisition unit.

[0015] The workpiece state data acquisition device 3 acquires workpiece state data indicating the state of the workpiece 1. The workpiece state data acquisition device 3 includes a hardness sensor, a surface roughness sensor, a camera, etc., and acquires physical quantities measured by various sensors or appearance images captured by a camera as workpiece state data. The workpiece state data acquisition device 3 also acquires information such as the lot number and manufacturer of the workpiece from a storage device 5 that stores workpiece management data in advance. The workpiece state data acquisition device 3 is an example of a workpiece state data acquisition unit.

[0016] The signal processing device 4 processes time-series data indicating the operating state of the machine tool 200 transmitted from the machining data acquisition device 2, and generates input data to be input to the learning model. Specifically, the signal processing device 4 performs signal processing such as amplification of the analog signal acquired from the sensor by the machining data acquisition device 2, filtering using a high-pass filter or a low-pass filter, and A / D (Analog to Digital) conversion based on a sampling frequency and resolution preset by the user. The signal processing device 4 also performs cleansing processing such as removing noise and outliers and filling in missing values.

[0017] After performing signal processing, the signal processing device 4 calculates feature quantities from the processed time-series data. The signal processing device 4, for example, extracts time-series data for an actual machining interval from the start to the end of actual machining, and calculates feature quantities. Specifically, in the example of FIG. 2, the signal processing device 4 extracts the current value of the spindle motor during an interval in which the machining feed signal is at a high level, and calculates feature quantities. The feature quantities include, for example, the maximum value, minimum value, average value, standard deviation, variance, kurtosis, skewness, etc. of the current value. Note that, in addition to the current value, feature quantities may also be calculated from the frequency, magnitude, temperature, etc. of vibration, sound, or AE. When calculating feature quantities from vibration, sound, or AE, frequency analysis may be performed to calculate feature quantities from the magnitude of a specific frequency.

[0018] The signal processing device 4 calculates a feature amount for each cycle, which is the period from the start to the end of machining of one workpiece. The signal processing device 4 associates the calculated feature amount with workpiece state data to generate input data. The signal processing device 4 associates the generated input data with machining-related information including the machining program number, tool number, etc. output from the machining data acquisition device 2 and stores the associated data in the storage device 5. Note that, when multiple machining processes are performed in one cycle, a feature amount may be calculated for each machining process, and input data may be generated for each machining process. Note that the signal processing device 4 is an example of a feature amount calculation unit.

[0019] Returning to FIG. 1, the storage device 5 stores input data including the feature quantities and workpiece state data transmitted from the signal processing device 4 in association with processing-related information including the processing program number, tool number, etc.

[0020] Diagnostic device 6 diagnoses machine tool 200 using input data generated by signal processing device 4 and stored in storage device 5 and a learning model generated by learning model generation device 7. Specifically, diagnostic device 6 applies input data generated based on workpiece state data of workpiece 1 acquired by workpiece state data acquisition device 3 before machining and time-series data acquired sequentially during machining by machine tool 200 to the learning model, and outputs a diagnostic result determining whether machine tool 200 is normal or not. Diagnostic device 6 transmits the diagnostic result to status display device 9 and control device 8. Note that diagnostic device 6 is an example of a diagnostic unit.

[0021] The learning model generation device 7 generates a learning model that has learned the normal range of feature quantities corresponding to the workpiece state data. Specifically, the learning model generation device 7 generates the learning model by performing machine learning on the learning model by providing the learning model with learning data including input data generated by the signal processing device 4 and label information indicating whether each input data is normal data or abnormal data. The user determines whether the machining process of each cycle executed by actual machining or simulation is normal or not, and creates learning data by assigning label information indicating normality or abnormality to the combination of feature quantities and workpiece state data associated with each cycle.

[0022] The learning model is a model based on a machine learning algorithm such as a decision tree, k-nearest neighbor method, or support vector machine. It receives a combination of feature values ​​calculated from time-series data and workpiece condition data as input, and outputs determination result information indicating whether the feature value corresponding to the workpiece condition data is normal. The input data included in the learning data does not have to be actual data acquired from the machine tool 200. For example, simulated data generated by computer simulation based on the type of machine tool 200, the processing content, the state of the workpiece before processing, and the like may be used. The label information included in the learning data is not limited to two patterns, normal and abnormal. The abnormal label may be further subdivided into dimensional abnormality, poor surface texture, and the like, resulting in three or more label patterns. The learning model generation device 7 is an example of a learning model generation unit.

[0023] Control device 8 controls the operation of machine tool 200. Control device 8 is, for example, a numerical control device that controls machine tool 200 by commands based on an NC (Numerical Control) program. When control device 8 receives diagnosis result information from diagnosing device 6 indicating that it has been determined that an abnormality has occurred in machine tool 200, control device 8 stops machine tool 200. Note that control device 8 may not stop machine tool 200 completely, but may limit some of the functions of machine tool 200 depending on the degree of the abnormality occurring in machine tool 200.

[0024] The status display 9 displays the diagnosis result by the diagnostic device 6. The status display 9 is, for example, an operation panel installed on the machine tool 200. The status display 9 notifies the user of the occurrence of an abnormality by displaying information indicating that an abnormality has been diagnosed on a screen. The status display 9 may also include a lamp, a speaker, etc., and may turn on a lamp and emit an alarm sound from the speaker when an abnormality is diagnosed. Furthermore, the status display 9 may have a function to automatically send an email indicating that an abnormality has been diagnosed to a pre-set email address. The status display 9 is an example of a display.

[0025] In the diagnostic system 100 having the functional configuration described above, the information processing device 10 in which the processing data acquisition device 2, the workpiece state data acquisition device 3, the signal processing device 4, the diagnostic device 6, the learning model generation device 7, and the control device 8 are realized physically comprises, as shown in Figure 3, a CPU (Central Processing Unit) 11 that executes processing according to a program, a RAM (Random Access Memory) 12 which is a volatile memory, a ROM (Read Only Memory) 13 which is a non-volatile memory, a memory unit 14 that stores data, an input unit 15 that accepts input of information, and a communication unit 16 that transmits and receives information, which are connected via an internal bus 99.

[0026] The CPU 11 executes various processes by reading out the programs stored in the storage unit 14 into the RAM 12 and executing them. As main functions provided by the programs, the CPU 11 executes a process of acquiring machining data including time-series data by the machining data acquisition device 2 and providing it to the signal processing device 4, a process of acquiring workpiece state data by the workpiece state data acquisition device 3 and providing it to the signal processing device 4, a process of processing the time-series data transmitted from the machining data acquisition device 2 by the signal processing device 4 to calculate feature amounts, a diagnostic process by the diagnostic device 6 to determine whether the machine tool 200 is normal, and a process of generating a learning model by the learning model generation device 7, etc.

[0027] The RAM 12 is used as a work area for the CPU 11. The ROM 13 stores the control program executed by the CPU 11, a BIOS (Basic Input Output System), and the like.

[0028] The storage unit 14 includes a hard disk drive, stores the programs executed by the CPU 11, and stores various data used when the programs are executed. The storage unit 14 stores the machining data acquired from the machine tool 200 by the machining data acquisition device 2, the workpiece state data acquired by the workpiece state data acquisition device 3, the input data generated by the signal processing device 4, the learning model generated by the learning model generation device 7, and the learning data used for learning the learning model.

[0029] The input unit 15 is a user interface including a keyboard, a mouse, a communication device, etc. The communication unit 16 is a network termination device or a wireless communication device that connects to a network, and a serial interface or a LAN (Local Area Network) interface that connects to them.

[0030] Next, a description will be given of the operation of diagnostic system 100. In the following, an example will be given in which machine tool 200 to be diagnosed by diagnostic system 100 is a cutting machine, and the presence or absence of an abnormality in machine tool 200 is determined based on feature amounts calculated from the current flowing through a spindle motor having a drill attached to its tip.

[0031] (Learning model generation process) Diagnostic system 100 inputs input data, including feature amounts calculated from time-series data of machine tool 200 and workpiece state data, into a learning model to determine whether or not machine tool 200 has an abnormality. A learning model generation process for generating a learning model will now be described with reference to FIG. 4. The learning model generation process is executed before the first diagnostic process is performed, when the learning model needs to be updated due to a change in machine tool 200 or tool, or when it is desired to improve the accuracy of determination using more data. The following describes a case in which, in a machining process in which only cutting is performed, one cycle is defined as the period from the start to the end of machining of one workpiece, and a learning model is generated based on feature amounts calculated from current data for each cycle and workpiece state data, including the temperature and lot number of the workpiece being cut.

[0032] When the user operates the input unit 15 of the learning model generation device 7 to request the start of the learning model generation process, the learning model generation device 7 starts the process.

[0033] The workpiece state data acquisition device 3 acquires workpiece state data of the workpiece (step S11). The workpiece state data acquisition device 3 acquires the temperature of the workpiece measured by a temperature sensor, for example, when the workpiece is placed on the processing table and immediately before cutting processing begins. The workpiece state data acquisition device 3 also accesses the management data stored in the storage device 5 to acquire the lot number of the workpiece. The workpiece state data acquisition device 3 outputs the acquired temperature and lot number to the signal processing device 4.

[0034] Next, when cutting processing by the machine tool 200 is started, the processing data acquisition device 2 acquires processing data including time-series data indicating the current value flowing through the spindle motor, the motor rotation speed, and the cutting feed signal, as well as processing-related information such as the processing program number and tool number (step S12). The processing data acquisition device 2 lumps together the time-series data for each cycle, which is the period from the start to the end of processing one workpiece, and outputs the lumped data together with the processing-related information to the signal processing device 4.

[0035] Next, the signal processing device 4 performs a cleansing process on the time-series data output from the processing data acquisition device 2 (step S13). Specifically, the signal processing device 4 performs signal processing on the received time-series data and extracts time-series data for the actual processing interval from the processed time-series data. Specifically, the signal processing device 4 extracts current data for the interval from when the cutting feed signal changes from low level to high level to when the cutting feed signal changes from high level to low level. The signal processing device 4 also performs processing to remove noise and abnormal values ​​from the current data and to complement missing values. Note that, when the signal processing device 4 uses an appearance image of the workpiece as the workpiece state data, the cleansing process may include trimming the image captured by the camera. Furthermore, when processing data for a processing step when a workpiece different from the target workpiece is processed is acquired by the processing data acquisition device 2, the cleansing process may include removing this processing data.

[0036] Next, the signal processing device 4 calculates feature amounts from the machining data cleansed in step S13 (step S14). Specifically, the signal processing device 4 calculates feature amounts by finding the average value, maximum value, minimum value, standard deviation, variance, kurtosis, skewness, etc. of the current values ​​in the cleansed actual machining section. The signal processing device 4 calculates feature amounts for each cycle from the start to the end of cutting processing of one workpiece.

[0037] Note that after calculating the features, a cleansing process may be performed again. This is to prevent, for example, when a workpiece other than the workpiece for which anomaly detection is targeted enters the mass production line, input data including time-series data and workpiece status data for a workpiece other than the target workpiece from being used to train the learning model. Specifically, features are calculated from time-series data and workpiece status data in a normal state, and outliers are removed or smoothed from the set of features. Note that outliers can be identified by calculating the standard deviation of all the features and identifying those that deviate from the standard deviation as outliers, or by calculating a moving average from features calculated earlier to smooth out the outliers. Alternatively, instead of using a moving average, the median value for a certain time period may be calculated and smoothed.

[0038] Next, the learning model generation device 7 generates learning data (step S15). Specifically, based on input from the user, the learning model generation device 7 generates learning data by assigning labels to the feature amounts for each cycle and the workpiece state data corresponding to the feature amounts stored in the storage device 5 in step S14, indicating whether the data is data obtained when the machine tool 200 is operating normally or data obtained when an abnormality occurs.

[0039] 5, the learning data includes "machining-related information" including a tool number and a machining program number for identifying the machining process, "workpiece status data" including the lot number and temperature of the workpiece acquired in step S11, "features" indicating the features calculated by the signal processing device 4 in step S14, and "status labels" indicating labels assigned to each feature. The learning model generation device 7 stores the generated learning data in the storage device 5. Note that, in step S14, multiple feature amounts may be calculated for each cycle, and in that case, the learning data may store a data set of multiple feature amounts for each cycle.

[0040] Returning to FIG. 4, next, the learning model generation device 7 generates a learning model (step S16). Specifically, the learning model is a model based on a machine learning algorithm such as a decision tree, k-nearest neighbor method, or support vector machine, and receives as input feature values ​​calculated from the current data and workpiece state data, and outputs determination result information that determines whether the feature values ​​are normal or not. The learning model generation device 7 generates the learning model by performing machine learning using learning data that includes feature values ​​contained in the learning data generated in step S15, workpiece state data including the lot number and temperature of the workpiece, and label information indicating whether the machine tool 200 is normal or abnormal. The learning model generation device 7 stores the generated learning model in the storage device 5. In this manner, the learning model is completed.

[0041] (Diagnosis processing) Next, the operation of the diagnostic process for determining whether machine tool 200 is normal or not using a learning model in an actual machining process will be described with reference to FIG.

[0042] The storage device 5 stores the learning model generated by the above-mentioned learning model generation process, and the learning model stored in the storage device 5 is pre-installed in the diagnostic device 6 by the user.

[0043] Steps S21 to S24 of the diagnostic process are similar to steps S11 to S14 of the learning model generation process shown in Figure 4, in which workpiece state data and machining data are acquired sequentially, and feature values ​​are calculated for each cycle of the cutting process from the current data contained in the machining data.

[0044] Returning to Fig. 6, next, diagnostic device 6 determines whether the target machine tool 200 is normal or abnormal (step S25). Specifically, diagnostic device 6 inputs input data including the workpiece state data acquired in step S21 and the feature values ​​calculated in step S24 into the learning model, and obtains determination result information indicating whether the feature values ​​are normal or not. If diagnostic device 6 obtains a determination result that the feature values ​​are normal, it determines that the target machine tool 200 is normal, and if it obtains a determination result that the feature values ​​are outside the normal range, it determines that the target machine tool 200 is abnormal.

[0045] When the diagnostic device 6 determines that the target machine tool 200 is normal (step S26; Yes), it determines whether the machining process has ended, and if it determines that it has ended (step S27; Yes), it ends the diagnostic processing. When the diagnostic device 6 determines that the machining process has not ended (step S27; No), it returns to step S21 and acquires workpiece state data for the next cycle (step S21).

[0046] Returning to step S26, if diagnostic device 6 determines that the target machine tool 200 is abnormal (step S26; No), it executes a preset operation of notifying status display device 9 and control device 8 that machine tool 200 has been determined to be abnormal (step S28), and ends the processing. Thereafter, status display device 9 notifies the user of the occurrence of an abnormality by displaying information indicating that the machine tool has been diagnosed as abnormal on the screen. Control device 8 stops machine tool 200.

[0047] As described above, diagnosis system 100 determines whether or not there is an abnormality in machine tool 200 based on workpiece state data indicating the state of the workpiece before machining and feature amounts calculated from time-series data indicating the operating state of machine tool 200. Therefore, it is possible to determine whether or not there is an abnormality in machine tool 200 by taking into account the effect on the time-series data due to differences in the state of the workpiece, making it possible to accurately determine whether or not there is an abnormality.

[0048] (Embodiment 2) In the above embodiment, diagnostic system 100 diagnosed whether machine tool 200 was normal by applying input data including workpiece state data and feature quantities of time-series data to a learning model. In contrast, diagnostic system 100a according to embodiment 2 determines whether a feature quantity calculated from collected time-series data is outside a normal range, based on a normal range database (DB) that defines a normal range for feature quantities corresponding to workpiece state data, thereby determining whether or not there is an abnormality in the machine tool.

[0049] 7 includes a machining data acquisition device 2, a workpiece state data acquisition device 3, a signal processing device 4, a state display device 9 that displays the diagnosis result by the diagnostic device 6a, and a control device 8 that controls the machine tool 200 based on the determination result by the diagnostic device 6a, all of which are included in the diagnostic system 100. Instead of the storage device 5 and the diagnostic device 6 that are included in the diagnostic system 100, the diagnostic system 100a includes a storage device 5a that stores a normal range DB 51, and a diagnostic device 6a that references the normal range DB 51 to determine whether or not there is an abnormality in the machine tool 200.

[0050] The storage device 5a stores a normal range DB 51 that defines normal ranges of feature quantities corresponding to workpiece condition data. As illustrated in FIG. 8, the normal range DB 51 includes the following items: "machining-related information" including the machining program number and tool number; "workpiece condition data" including the lot number and temperature of the workpiece; and "feature quantity" including upper and lower limit values ​​of feature quantities. The normal range DB 51 illustrated defines the upper and lower limit values ​​of the normal ranges of feature quantities when machining a workpiece with lot number "10001" and a temperature of 25°C to 28°C. Note that the normal range DB 51 is not limited to the illustrated example. Normal ranges may be defined for different machining processes identified by the machining-related information, or for different types of workpiece condition data or different combinations of workpiece condition data.

[0051] Diagnosis device 6a diagnoses machine tool 200 using input data generated by signal processing device 4 and stored in storage device 5a and normal range DB 51. Specifically, diagnosis device 6a determines whether machine tool 200 is normal by determining whether a feature calculated based on time-series data acquired sequentially during machining by machine tool 200 deviates from a normal range of a feature corresponding to workpiece state data of workpiece 1 acquired by workpiece state data acquisition device 3 before machining.

[0052] Next, the operation of the diagnostic process by the diagnostic system 100a will be described with reference to Fig. 9. Note that Fig. 9 includes steps in common with the flowchart shown in Fig. 6, so differences will be mainly described.

[0053] A normal range DB 51, an example of which is shown in FIG. 8, created by a user is stored in advance in the storage device 5a of the diagnostic system 100a.

[0054] When processing similar to steps S21 to S24 in Fig. 6 is executed, diagnostic device 6a determines whether machine tool 200 is normal or abnormal by referring to normal range DB 51 illustrated in Fig. 8 (step S31). Specifically, if diagnostic device 6a determines that the feature amount calculated in step S24 deviates from normal range 52 defined by normal range DB 51 illustrated in Fig. 10, it determines that machine tool 200 is abnormal (step S32; No), and if it determines that the feature amount does not deviate from normal range 52, it determines that machine tool 200 is normal (step S32; Yes). For example, if lot number "l0001" and temperature "25 degrees" are acquired as workpiece data in step S21 and feature amount "12" is calculated in step S24, diagnostic device 6a determines that machine tool 200 is normal because feature amount "12" is within normal range 52. Also, for example, if the lot number "l0001" and the temperature "25 degrees" are acquired as workpiece data in step S21, and the feature value "18" is calculated in step S24, the feature value "18" deviates from the normal range 52, and therefore the diagnostic device 6a determines that the machine tool 200 is abnormal.

[0055] If the diagnostic device 6a determines that the machine tool 200 is normal (step S32; Yes), it determines whether the machining process has ended by processing similar to step S27 in Figure 6, and if it determines that the machining process has ended (step S27; Yes), it terminates the diagnostic processing, and if it determines that the machining process has not ended (step S27; No), it returns to step S21 and obtains workpiece state data for the next cycle (step S21).

[0056] On the other hand, if the diagnostic device 6a determines that the machine tool 200 is abnormal (step S32; No), it executes a preset operation, such as notifying the status display device 9 and the control device 8 that the machine tool 200 has been determined to be abnormal, by processing similar to step S28 in Figure 6, and then terminates the processing.

[0057] (Variation) In the above embodiment, the functions of the diagnostic system 100 have been described as being executed by individual devices, namely, the processing data acquisition device 2, the workpiece state data acquisition device 3, the signal processing device 4, the storage device 5, the diagnostic device 6, the learning model generation device 7, the status display device 9, and the control device 8. However, this is not limited to this, and the functions of each device may be executed by an information processing device that is a single computer. Furthermore, each process of the diagnostic system 100 may be executed by multiple computers that execute the functions of several devices, such as by executing the functions of the processing data acquisition device 2, the workpiece state data acquisition device 3, the signal processing device 4, the storage device 5, the diagnostic device 6, and the learning model generation device 7 by a single computer, and by executing the functions of the status display device 9 and the control device 8 by a single computer.

[0058] In the above embodiment, the signal processing device 4 is described as extracting time-series data of an actual machining interval from the start to the end of actual machining and calculating feature quantities, but this is not limited to this. A calculation interval in which an abnormality is likely to occur in each machining process may be specified in advance, and a rule for extracting time-series data of the calculation interval may be defined for each machining process. In this case, in the cleansing process of step S13 and step S23, an extraction rule for the machining process specified by the machining-related information included in the machining data acquired in step S12 or step S22 may be read, and time-series data may be extracted based on the extraction rule, such as extracting current data for 5 seconds from the time a cutting signal is output, and feature quantities may be calculated.

[0059] Furthermore, in the above embodiment, diagnostic system 100 diagnoses the presence or absence of an abnormality in machine tool 200 based on feature amounts calculated from time-series data of machine tool 200 during machining and workpiece state data indicating the state of the workpiece before machining. However, environmental data indicating the environment of machine tool 200 may also be used for diagnosis. Specifically, physical quantities such as the ambient temperature and humidity around machine tool 200 during machining are acquired as environmental data from a temperature sensor, humidity sensor, etc. equipped in machine tool 200, and a learning model may be generated by providing learning data obtained by adding label information to input data including the workpiece state data, environmental data, and feature amounts. Then, in the diagnostic processing, the input data including the workpiece state data, environmental data, and feature amounts for each machining cycle is applied to the learning model to determine the presence or absence of an abnormality in machine tool 200.

[0060] Furthermore, the diagnostic process may be performed by referring to the normal range DB 51, which takes into account the feature amounts calculated from the time-series data, the workpiece state data, and the above-mentioned environmental data. Specifically, the normal range DB 51, which defines the normal range of the feature amount for each value of the workpiece data and for each value of the environmental data, is stored in the storage device 5a, and the diagnostic device 6a may refer to this normal range DB 51 to determine whether or not there is an abnormality in the machine tool 200.

[0061] Furthermore, in the above embodiment, diagnostic device 6 has been described as determining that machine tool 200 is abnormal when it is determined that the feature amount is outside the normal range, but this is not limited to this. For example, it may be determined that machine tool 200 is abnormal based on a preset abnormality detection pattern, such as determining that machine tool 200 is abnormal when a determination result that the feature amount is outside the normal range is acquired a set number of times in succession.

[0062] The status display 9 may not only display the diagnostic results but also the processed data stored in the storage device 5. Furthermore, since the feature amounts calculated from the time-series data are sequentially stored in the storage device 5 during the diagnostic process, the feature amounts may be displayed in real time using trend graphs, scatter diagrams, etc.

[0063] Diagnostic device 6 may not only determine whether or not an abnormality is currently occurring in machine tool 200, but may also predict when an abnormality will occur. Specifically, diagnostic device 6 creates a trend graph of the feature quantities, performs linear regression analysis or nonlinear regression analysis, and fits an approximate curve to predict future feature quantity values ​​and calculate the time until an abnormality occurs. When the time to error shortens and exceeds a preset threshold, status indicator 9 may sound an alarm to an operator or display a message urging maintenance. Furthermore, status indicator 9 may display the threshold together with the trend graph.

[0064] Furthermore, in the above embodiment, a case where one machining process is performed in one cycle has been described, but the present invention is not limited to this. For example, a plurality of machining processes, such as cutting, cutting, and polishing, may be performed in one cycle. In this case, feature amounts may be calculated from the time-series data of each actual machining section to generate a learning model. Then, the diagnostic device 6 inputs each feature amount into each learning model to determine whether or not there is an abnormality in the machine tool 200.

[0065] Furthermore, the information stored in the storage device 5 may be collectively managed by a cloud server present on the network, and the workpiece state data acquisition device 3, the signal processing device 4, the diagnostic device 6, and the learning model generation device 7 may access the cloud server as necessary to read and write information. In this case, the diagnostic system 100 does not need to include the storage device 5. Furthermore, the signal processing of data by the signal processing device 4 and the learning model generation processing by the learning model generation device 7 may be performed on the cloud using information stored in the cloud server.

[0066] Furthermore, the processed data acquisition device 2, the signal processing device 4, the storage device 5, the diagnostic device 6, and the learning model generation device 7 can be realized using a normal computer system, rather than using dedicated devices. For example, a program for realizing each function may be stored and distributed on a computer-readable recording medium such as a CD-ROM (Compact Disc Read Only Memory) or a DVD-ROM (Digital Versatile Disc Read Only Memory), and this program may be installed on a computer to configure a computer capable of realizing each of the above-mentioned functions.

[0067] Furthermore, when each function is realized by sharing the functions between an OS (Operating System) and an application, or by cooperation between the OS and the application, only the application may be stored on the recording medium.

[0068] The present disclosure allows various embodiments and modifications without departing from the broad spirit and scope of the present disclosure. Furthermore, the above-described embodiments are intended to explain the present disclosure and do not limit the scope of the present disclosure. In other words, the scope of the present disclosure is defined by the claims, not the embodiments. Various modifications made within the scope of the claims and the meaning of equivalent disclosures are considered to be within the scope of the present disclosure.

[0069] Various aspects of the present disclosure are summarized below as appendices.

[0070] (Appendix 1) a time series data acquisition unit that acquires time series data indicating the operating state of the machine tool to be diagnosed; a workpiece state data acquisition unit that acquires workpiece state data indicating a state of a workpiece before it is machined by the machine tool; a feature amount calculation unit that calculates feature amounts from the time series data acquired by the time series data acquisition unit; a diagnosis unit that determines whether the machine tool is normal or abnormal by determining whether the feature calculated by the feature calculation unit falls outside a normal range of the feature corresponding to the workpiece state data; A diagnostic device comprising:

[0071] (Appendix 2) the diagnosing unit determines whether the machine tool to be diagnosed is normal or abnormal by inputting the workpiece state data acquired by the workpiece state data acquiring unit and the feature amount calculated by the feature amount calculating unit into a preset machine learning model and determining whether the feature amount falls outside the normal range based on the output obtained. 10. The diagnostic device of claim 1.

[0072] (Appendix 3) a learning model generation unit that generates the machine learning model based on learning data including the workpiece state data acquired by the workpiece state data acquisition unit, combinations of the feature amounts calculated by the feature amount calculation unit, and label information assigned to each combination; 3. The diagnostic device of claim 2, further comprising:

[0073] (Appendix 4) a normal range database that defines the normal range of the feature corresponding to the workpiece state data; the diagnosing unit determines that the machine tool is abnormal when it determines that the feature amount calculated by the feature amount calculating unit is outside the normal range of the feature amount corresponding to the workpiece state data acquired by the workpiece state data acquiring unit. 10. The diagnostic device of claim 1.

[0074] (Appendix 5) The workpiece state data includes at least one of the temperature, hardness, appearance image, surface texture, lot number, and manufacturer of the workpiece. 5. The diagnostic device of any one of appendices 1 to 4.

[0075] (Appendix 6) the learning data includes simulated learning data generated by computer simulation, the simulated learning data including combinations of the workpiece state data and the feature amounts, and the label information assigned to each combination; 4. The diagnostic device of claim 3.

[0076] (Appendix 7) A diagnostic device according to any one of appendices 1 to 6; a display that displays a diagnosis result by the diagnostic device; a control device that stops operation of a part or all of the machine tool when a determination result indicating that the machine tool to be diagnosed is abnormal is received from the diagnosing device; A diagnostic system comprising:

[0077] (Appendix 8) The computer acquiring time-series data indicating an operating state of the machine tool to be diagnosed; acquiring workpiece state data indicating a state of a workpiece before it is machined by the machine tool; calculating a feature amount from the acquired time-series data; determining whether the calculated feature amount falls outside a normal range of the feature amount corresponding to the workpiece state data, thereby determining whether the machine tool is abnormal or normal; A diagnostic method to perform.

[0078] (Appendix 9) On the computer, A process of acquiring time-series data indicating the operating state of the machine tool to be diagnosed; A process of acquiring workpiece state data indicating a state of a workpiece before being machined by the machine tool; A process of calculating feature amounts from the acquired time-series data; a process of determining whether the calculated feature amount falls outside a normal range of the feature amount corresponding to the workpiece state data, thereby determining whether the machine tool is abnormal or normal; A program that executes the following. [Explanation of symbols]

[0079] 100,100a Diagnostic system, 200 Machine tool, 1 Workpiece, 2 Machining data acquisition device, 3 Workpiece status data acquisition device, 4 Signal processing device, 5,5a Storage device, 51 Normal range DB, 52 Normal range, 6,6a Diagnostic device, 7 Learning model generation device, 8 Control device, 9 Status display, 10 Information processing device, 11 CPU, 12 RAM, 13 ROM, 14 Memory unit, 15 Input unit, 16 Communication unit, 99 Internal bus.

Claims

1. a time series data acquisition unit that acquires time series data indicating the operating state of the machine tool to be diagnosed; a workpiece state data acquisition unit that acquires workpiece state data indicating a state of a workpiece before it is machined by the machine tool; a feature amount calculation unit that calculates feature amounts from the time series data acquired by the time series data acquisition unit; a diagnosis unit that determines whether the machine tool is normal or abnormal by determining whether the feature calculated by the feature calculation unit falls outside a normal range of the feature corresponding to the workpiece state data; A diagnostic device comprising:

2. the diagnosing unit determines whether the machine tool to be diagnosed is normal or abnormal by inputting the workpiece state data acquired by the workpiece state data acquiring unit and the feature amount calculated by the feature amount calculating unit into a preset machine learning model and determining whether the feature amount falls outside the normal range based on the output obtained. The diagnostic device of claim 1 .

3. a learning model generation unit that generates the machine learning model based on learning data including the workpiece state data acquired by the workpiece state data acquisition unit, combinations of the feature amounts calculated by the feature amount calculation unit, and label information assigned to each combination; The diagnostic device of claim 2 further comprising:

4. a normal range database that defines the normal range of the feature corresponding to the workpiece state data; the diagnosing unit determines that the machine tool is abnormal when it determines that the feature amount calculated by the feature amount calculating unit is outside the normal range of the feature amount corresponding to the workpiece state data acquired by the workpiece state data acquiring unit. The diagnostic device of claim 1 .

5. The workpiece state data includes at least one of the temperature, hardness, appearance image, surface texture, lot number, and manufacturer of the workpiece. The diagnostic device according to claim 1 or 2.

6. the learning data includes simulated learning data generated by computer simulation, the simulated learning data including combinations of the workpiece state data and the feature amounts, and the label information assigned to each combination; The diagnostic device of claim 3 .

7. The diagnostic device according to claim 1 or 2; a display that displays a diagnosis result by the diagnostic device; a control device that stops operation of a part or all of the machine tool when a determination result indicating that the machine tool to be diagnosed is abnormal is received from the diagnosing device; A diagnostic system comprising:

8. The computer acquiring time-series data indicating an operating state of the machine tool to be diagnosed; acquiring workpiece state data indicating a state of a workpiece before it is machined by the machine tool; calculating a feature amount from the acquired time-series data; determining whether the calculated feature amount falls outside a normal range of the feature amount corresponding to the workpiece state data, thereby determining whether the machine tool is abnormal or normal; A diagnostic method to perform.

9. On the computer, A process of acquiring time-series data indicating the operating state of the machine tool to be diagnosed; A process of acquiring workpiece state data indicating a state of a workpiece before being machined by the machine tool; A process of calculating feature amounts from the acquired time-series data; a process of determining whether the calculated feature amount falls outside a normal range of the feature amount corresponding to the workpiece state data, thereby determining whether the machine tool is abnormal or normal; A program that executes the following.

Images