Diagnostic device, diagnostic method, program, and processing system
The diagnostic device uses frequency analysis and bandpass filters to detect machining tool abnormalities, improving accuracy and preventing defective product production, thereby enhancing machining center productivity.
Patent Information
- Application Number
- JP2020181869
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2020-10-29
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2040-10-29
AI Technical Summary
Existing diagnostic technologies struggle to accurately detect abnormalities in machining tools, leading to sudden production of defective products due to issues like chipping or runout, which reduces productivity and increases costs.
A diagnostic device that uses a rotating shaft, receiving unit, frequency analyzer, and bandpass filters to analyze rotation and tool information, extracting feature information through frequency analysis to determine the machining state, enabling precise detection of abnormalities.
The device accurately detects machining abnormalities, preventing defective product production and enhancing productivity by automating machining centers with improved accuracy.
Smart Images

Figure 0007589505000002 
Figure 0007589505000003 
Figure 0007589505000004
Abstract
Description
[Technical field]
[0001] The present invention relates to a diagnostic device, a diagnostic method, a program, and a processing system. [Background technology]
[0002] There is a diagnostic device that diagnoses whether the machining state of a machining device, such as a machining center that automatically changes tools required for various types of machining and performs machining continuously, is normal or abnormal.The diagnostic device interrupts machining or issues an alert depending on the diagnosis result of the machining device, and is used to suppress the generation of a large number of defective products due to abnormal machining of the machining device and to prevent the outflow of defective products.
[0003] Patent Document 1 discloses a diagnostic technology that acquires context information from a machine tool in advance, generates a trained model that is trained using physical quantity data acquired by a sensor for each operating state of the machine tool based on the context information, and, when diagnosing the machine tool, determines whether the machining of the machine tool is normal or abnormal by comparing a score obtained by applying the context information acquired from the machine tool and the physical quantity data acquired by the sensor to the trained model with a preset threshold value for determining whether the machining is normal or abnormal.
[0004] Patent document 2 relates in particular to a method and device for monitoring the condition of tools in a milling machine, and discloses a technology in which, in the case of normal cutting blades, the load corresponding to the cutting blades is generated evenly, whereas if even one cutting blade is defective (worn or broken), the load on the cutting blade becomes uneven. In order to identify the defective condition from the corresponding change in vibration, the technology discloses a technology in which the signal from an acceleration sensor arranged on the frame of the machine is passed through multiple adjustable bandpass filters, and the signal obtained is compared with a reference level, and an alarm signal is output if the deviation is unacceptable.
[0005] Patent Document 3 relates to a method and device for detecting abnormalities in a rotating tool during cutting processing using an NC (Numerical Control) machine tool equipped with an automatic tool changer, and discloses technology that detects cutting force, performs frequency analysis of the cutting force signal, calculates the ratio between the voltage level of a frequency equivalent to the spindle rotation speed x number of teeth and the voltage level of a frequency equivalent to the spindle rotation speed, detects the magnitude of runout of the rotating tool, and issues an abnormality signal when a predetermined threshold is exceeded. Summary of the Invention [Problem to be solved by the invention]
[0006] However, if machining is performed with damage such as chipping or breakage on the cutting edge of the tool, defective products will be produced because the machining dimensions are outside the tolerances. However, breakage occurs suddenly or accidentally, so it is very difficult to prevent it in advance. Also, when changing tools, if cutting chips get mixed in between the taper part of the tool holder and the spindle and are chucked, the runout of the cutting edge will increase. Even if machining is performed in this state, defective products will be produced. Although there is technology that can detect the runout of the cutting edge before machining and immediately after changing tools to prevent the generation of defective products, it is inevitable that productivity will decrease because a new process will be added.
[0007] Therefore, in order to deal with these problems, rather than preventing sudden or accidental machining defects, it may be possible to reduce overall production costs by preventing the outflow of defective products that have already occurred or by preventing the continuous occurrence of defective products. Also, although machining centers can automate production and increase productivity by performing various machining operations using multiple tools, in order to diagnose abnormalities in the machining state that result in the above-mentioned defective products, it is necessary to improve the accuracy of abnormality judgment according to each machining state.
[0008] The present invention has been made in consideration of the above, and aims to provide a diagnostic device, a diagnostic method, a program, and a machining system that make it possible to detect and monitor with high accuracy the occurrence of abnormalities in the machining state of a machining device for various machining operations performed in a machining center. [Means for solving the problem]
[0009] In order to solve the above problems and achieve the object, the present invention provides a rotary shaft constituting a processing device. Rotation information and the rotation axis of the tool attached to Tool information and A receiver receives context information that specifies an operation and detection information of a time-varying physical quantity generated by the tool while the tool is executing a machining operation on a workpiece, a frequency analyzer performs frequency analysis on the detection information, a range setting unit that sets a frequency range, a bandwidth setting unit that sets a bandwidth of a frequency band of interest within the frequency range, and sets a band pass filter (BPF) using a plurality of center frequencies that are set using the rotation information, the tool information, and the frequency range, and the bandwidth. A fundamental rotation frequency calculated using the rotation information and a frequency that is an integer multiple of the fundamental rotation frequency are set as the plurality of center frequencies. a characteristic information extraction unit that uses the BPF to extract characteristic information of the detection information from a frequency analysis result of the detection information; and a determination unit that uses the characteristic information to determine a processing state of the processing device. a frequency shift estimation unit that calculates an autocorrelation function as a result of frequency analysis of the detection information, obtains a delay value at which the autocorrelation function is maximum within a range in which a delay value of the autocorrelation function exceeds the fundamental frequency of rotation, and estimates a number of teeth of a tool using the delay value, and the BPF setting unit sets the BPF using the plurality of center frequencies that are set using the estimated number of teeth instead of the tool information. . Effect of the Invention
[0010] According to the present invention, it is possible to detect with high accuracy the occurrence of an abnormality in the machining state of a machining device for various machining operations carried out in a machining center. [Brief description of the drawings]
[0011] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a machining system to which a diagnostic device according to the present embodiment is applied. [Diagram 2] FIG. 2 is a block diagram showing an example of a hardware configuration of the processing device according to the present embodiment. [Diagram 3] FIG. 3 is a block diagram showing an example of a hardware configuration of the diagnostic device according to the present embodiment. [Figure 4] FIG. 4 is a block diagram showing an example of a functional configuration of the diagnostic device according to the present embodiment. [Diagram 5]FIG. 5 is a diagram showing an example of the correspondence between the context information and the learning model stored in the diagnostic device according to the present embodiment. [Figure 6] FIG. 6 is a diagram showing an example of an average spectrum obtained by frequency analysis by the frequency analysis unit of the diagnostic device according to this embodiment. [Figure 7] FIG. 7 is a diagram showing an example of an average spectrum obtained by frequency analysis by the frequency analysis unit of the diagnostic device according to this embodiment. [Figure 8] FIG. 8 is a diagram showing an example of an average spectrum obtained by frequency analysis by the frequency analysis unit of the diagnostic device according to this embodiment. [Figure 9] FIG. 9 is a diagram showing an example of an average spectrum obtained by frequency analysis by the frequency analysis unit of the diagnostic device according to this embodiment. [Figure 10] FIG. 10 is a flowchart showing an example of the flow of diagnostic processing in the diagnostic device according to the present embodiment. [Figure 11] FIG. 11 is a flowchart showing an example of the flow of the model generation process performed by the diagnosis device according to the present embodiment. [Figure 12] FIG. 12 is a flowchart showing an example of the flow of the process of extracting characteristic information according to the BPF by the diagnosis device according to the present embodiment. [Figure 13] FIG. 13 is a diagram for explaining an example of a method for selecting a BPF by the diagnosis device according to the present embodiment. [Figure 14] FIG. 14 is an enlarged view of the vicinity of the BPF of the average spectrum calculated in the diagnostic device according to this embodiment. [Figure 15] FIG. 15 is a diagram showing an example of an average spectrum obtained by the diagnostic device according to the present embodiment. [Figure 16] FIG. 16 is a flowchart showing another example of the flow of the feature information extraction process according to the BPF by the diagnosis device according to the present embodiment. [Figure 17] FIG. 17 is a diagram showing an example of an autocorrelation function obtained by the diagnostic device according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0012] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of a diagnostic device, a diagnostic method, a program, and a machining system will be described in detail with reference to the accompanying drawings.
[0013] 1 is a block diagram showing a configuration example of a processing system to which a diagnostic device according to the present embodiment is applied. As shown in FIG. 1, the processing system according to the present embodiment includes a processing device 200 and a diagnostic device 100.
[0014] The processing device 200 and the diagnostic device 100 may be connected in any form. For example, the processing device 200 and the diagnostic device 100 are connected by a dedicated connection line, a wired network such as a wired LAN (local area network), a wireless network, or the like.
[0015] The machining device 200 includes a machine control unit 201, a tool exchange device 202, a display unit 203, a memory unit 204, a communication control unit 205, a numerical control unit 206, a tool information input unit 207, an alarm unit 208, an input / output unit 209, a machine tool 220, etc.
[0016] 1 and equipped with a drive unit. The Z-axis stage 226 is equipped with a spindle 221 which is an example of a rotation axis constituting the processing apparatus 200. A tool holder 222 which holds a tool 223 is attached to the spindle 221. The machine tool 220 is equipped with an XY-axis stage 225 which is equipped with a drive unit and which is movable in two axial directions in a plane perpendicular to the Z-axis stage 226, below the spindle 221. The XY-axis stage 225 holds a workpiece 224.
[0017] The numerical control unit 206 executes machining by the machining device 200 through numerical control. For example, the numerical control unit 206 reads a machining program from the input / output unit 209, and generates and outputs numerical control data for controlling the spindle rotation and the position of each axis stage. Also, the machining program describes the storage number of the tool exchange device 202, and the numerical control unit 206 performs tool exchange according to the description.
[0018] The numerical control unit 206 outputs the context information to the communication control unit 205. The context information is information that specifies the operation of the tool 223 of the processing device 200, and a plurality of pieces of context information are defined for each type of operation of the tool 223. In this embodiment, the context information includes, for example, tool information that identifies the tool 223, rotation information of the rotating spindle 221 (for example, the spindle rotation speed that is the rotation speed of the rotating spindle 221), movement information of the Z-axis stage 226 and the XY-axis stage 225 (movement speed, information during movement), and the like.
[0019] The tool information includes at least the tool type (e.g. drill, reamer, end mill, etc.), the number of cutting edges, etc. This tool information is input by an operator from the tool information input unit 207 according to the information displayed on the display unit 203. Alternatively, the tool information can be obtained by reading a list file of the tool information from the input / output unit 209, or by inputting the information from an external computer (not shown) via the communication control unit 205. Moreover, the tool information can be stored in the storage unit 204 and can be referred to from the machining program.
[0020] The numerical control unit 206 transmits, for example, context information that specifies the current operation of the tool 223 to the diagnosis device 100 via the communication control unit 205. When machining the workpiece 224 according to the machining program, the numerical control unit 206 controls the type of the tool 223, the positions of the Z-axis stage 226 and the XY-axis stage 225, the rotation speed of the rotating spindle 221, and the like, depending on the machining process. The numerical control unit 206 transmits, from the context information, context information corresponding to a predetermined operation to the diagnosis device 100 via the communication control unit 205. Here, the predetermined operation is a preset operation among the operations of the tool 223. In this embodiment, the numerical control unit 206 sequentially transmits context information corresponding to the changed type of operation to the diagnosis device 100 via the communication control unit 205 every time the type of operation of the tool 223 is changed.
[0021] The communication control unit 205 (an example of a transmission unit) controls communication with an external device such as the diagnostic device 100. For example, the communication control unit 205 transmits context information corresponding to the current operation of the tool 223 to the diagnostic device 100.
[0022] The physical quantity information detection unit 227 has a sensor that detects a time-varying physical quantity that is generated while the tool 223 is performing a machining operation on the workpiece 224 as an analog signal. The physical quantity information detection unit 227 also has a function of appropriately amplifying the analog signal detected by the sensor, cutting an arbitrary frequency range, and converting the analog signal into a digital signal. The physical quantity information detection unit 227 also functions as an example of a transmission unit that transmits the digital signal to the diagnosis device 100 as detection information. The type of sensor that the physical quantity information detection unit 227 has and the physical quantity that it detects may be any type. For example, the sensor that the physical quantity information detection unit 227 has may be a microphone, an acceleration sensor, an AE (acoustic emission) sensor, or the like, and each of the sensors outputs acoustic data, acceleration data, or data indicating an AE wave as detection information. The number of physical quantity information detection units 227 that the diagnosis device 100 has may be any number and may be more than one. For example, the diagnosis device 100 may include multiple sensors that detect different physical quantities.
[0023] In FIG. 1, the physical quantity information detection unit 227 has a sensor attached to a side of a structure that holds the rotating spindle 221 and a sensor attached to a side of the XY-axis stage 225. An acceleration sensor is built into the sensor of the physical quantity information detection unit 227. When the processing by the processing device 200 is started, the physical quantity information detection unit 227 detects the acceleration of vibration generated by the rotation of the rotating spindle 221. In the processing device 200, when the tool 223 and the workpiece 224 come into contact with each other and actual cutting is started, a cutting force is generated, which becomes a vibration force and vibrates the tool 223 and the workpiece 224, and the vibration propagates between them. The physical quantity information detection unit 227 transmits the acceleration of the vibration and the like to the diagnosis device 100 as detection information.
[0024] For example, when the cutting blade of the tool 223 breaks or chips during processing, the cutting force that was uniform for each cutting blade during normal processing becomes uneven, and the generated vibration changes. Alternatively, when chips get between the tool holder 222 and the rotating spindle 221 during replacement of the tool 223, the cutting blade at the tip of the tool 223 swings (vibrates) more against the rotating shaft. This causes the cutting amount per cutting blade of the tool 223 to become uneven, and a change in vibration occurs due to the uneven cutting force, as when the cutting blade is damaged.
[0025] The diagnostic device 100 receives the detection information of the vibration by the communication control unit 101. In addition, the communication control unit 101 controls communication with the processing device 200 to receive context information from the processing device 200. The determination unit 102 determines whether the processing state of the processing device 200 is normal or not by referring to the context information and the detection information. Furthermore, when the diagnostic device 100 diagnoses that the processing state of the processing device 200 is abnormal, it transmits alert information to the processing device via the communication control unit 101. When the processing device 200 receives the alert information by the communication control unit 205, the processing device 200 displays the alert information on the display unit 203 and operates the alarm unit 208. The alarm unit 208 is a patrol lamp, a buzzer, a speaker, or the like. Furthermore, the machine control unit 201 can interrupt the operation of the processing device 200 according to the processing program and stop the processing of the processing device 200.
[0026] Fig. 2 is a block diagram showing an example of a hardware configuration of the processing device according to the present embodiment. As shown in Fig. 2, the processing device 200 according to the present embodiment includes a CPU (Central Processing Unit) 251, a ROM (Read Only Memory) 252, a RAM (Random Access Memory) 253, a communication I / F (Interface) 254, a drive control circuit 255, a motor 256, an input / output I / F 257, an input device 258, and a display 259, which are connected by a bus 260.
[0027] The CPU 251 controls the entire processing apparatus 200. The CPU 251 controls the operation of the entire processing apparatus 200 and realizes various functions of the processing apparatus 200 by, for example, executing a program stored in the ROM 252 or the like using the RAM 253 as a work area.
[0028] The communication I / F 254 is an interface for communicating with an external device such as the diagnostic device 100. The drive control circuit 255 is a circuit for controlling the drive of the motor 256. Each of the rotating spindle 221, the Z-axis stage 226, and the XY-axis stage 225 is provided with a drive unit such as the motor 256. The sensor 270 is attached to the processing device 200 and converts a physical quantity that changes according to the operation of the processing device 200 into an electric signal. The signal conversion circuit 271 amplifies the electric signal output from the sensor 270 to a desired magnitude, cuts out noise components contained in the electric signal, and converts it into a digital signal. Then, the signal conversion circuit 271 outputs the digital signal to the diagnostic device 100 as detection information. That is, the sensor 270 and the signal conversion circuit 271 correspond to, for example, the physical quantity information detection unit 227 shown in FIG. 1.
[0029] The numerical control unit 206 and the communication control unit 205 shown in FIG. 1 may be realized by the CPU 251 executing a program stored in the ROM 252, that is, by software, or may be realized by hardware such as an IC (Integrated Circuit), or may be realized by a combination of software and hardware.
[0030] Fig. 3 is a block diagram showing an example of a hardware configuration of a diagnostic device according to the present embodiment. As shown in Fig. 3, diagnostic device 100 according to the present embodiment includes CPU 151, ROM 152, RAM 153, communication I / F 154, auxiliary storage device 155, and input / output I / F 157, which are connected via bus 160.
[0031] The CPU 151 controls the entire diagnostic device 100. For example, the CPU 151 executes a program stored in the ROM 152 or the like using the RAM 153 as a work area, thereby controlling the operation of the entire diagnostic device 100 and realizing a diagnostic function of the processing device 200.
[0032] The communication I / F 154 is an interface for communicating with an external device such as the processing device 200. The auxiliary storage device 155 stores various information such as setting information of the diagnosis device 100, context information received from the processing device 200, and detection information output from the physical quantity information detection unit 227. The auxiliary storage device 155 also stores various calculation results used to determine whether the processing state of the processing device 200 is normal or not. The auxiliary storage device 155 is composed of a non-volatile storage means such as an HDD (Hard Disk Drive), an EEPROM (Electrically Erasable Programmable Read-Only Memory), or an SSD (Solid State Drive).
[0033] The input / output I / F 157 sequentially displays the detection information input from the physical quantity information detection unit 227 on the display 159, and displays the determination result by the determination unit 102. The input / output I / F 157 also accepts settings required for diagnosing the processing apparatus 200 that are input by the user while watching the display 159 via an input device 158 such as a keyboard or a mouse.
[0034] 4 is a block diagram showing an example of a functional configuration of a diagnostic device according to the present embodiment. The diagnostic device 100 according to the present embodiment includes a storage unit 103, a generation unit 104, a display control unit 105, a display unit 106, an input unit 107, a reception unit 120, and a feature extraction unit 110 in addition to the above-mentioned communication control unit 101 and determination unit 102.
[0035] The storage unit 103 stores various information required for the diagnosis function of the diagnosis device 100. The storage unit 103 is realized by, for example, the RAM 153 and the auxiliary storage device 155 shown in FIG. 3. For example, the storage unit 103 stores one or more models (hereinafter referred to as learning models) used to determine an abnormality in the machining state of the machining device 200. Here, the learning model is generated by learning, for example, using detection information output from the physical quantity information detection unit 227 when the machining state of the machining device 200 is normal. The learning method of the learning model and the format of the learning model may be any method and format. For example, the learning model and the learning method of the learning model may be a learning model such as a GMM (Gaussian mixture model) or an HMM (hidden Markov model), and a model learning method corresponding to the learning model.
[0036] Furthermore, the storage unit 103 may set normal and abnormal machining states of the machining device 200 as rules and store them as learning models. For example, the rule stored in the storage unit 103 as a learning model may be that the first 10 machining operations after a new tool 223 is attached and machining is started are a learning period for determining diagnostic rules. The rule stored in the storage unit 103 as a learning model may be determined in advance separately from the actual machining, and the determined rule may be stored in the storage unit 103 as a learning model.
[0037] In this embodiment, the learning model stored in the storage unit 103 is generated for each piece of context information. The storage unit 103 stores, for example, the context information and the learning model corresponding to the context information in association with each other.
[0038] 5 is a diagram showing an example of the correspondence between the context information and the learning model stored in the diagnosis device according to the present embodiment. As shown in FIG. 5, machining process 1, machining process 2, and machining process 3 use the same end mill A and have the same rotation speed. On the other hand, machining process 4, machining process 5, and machining process 6 use different tools 223, and the tools 223 rotate at different rotation speeds. In this embodiment, if the rotation speed and tool type of the tool 223 are different, the diagnosis device 100 generates a learning model for each of them and stores it in the storage unit 103.
[0039] Furthermore, machining process 1, machining process 2, and machining process 3 are consecutive machining processes of the same part using the same end mill A. However, machining processes 1 to 3 have different machining conditions, and therefore different vibration intensities. Therefore, even when machining is performed by rotating the same tool 223 at the same rotation speed, the diagnosis device 100 generates a separate learning model for each machining process and judges whether the machining state of the machining device 200 is normal or not.
[0040] Returning to Fig. 4, the communication control unit 101 includes a receiving unit 101a and a transmitting unit 101b. The receiving unit 101a receives various information transmitted from the processing device 200 or an external device. For example, the receiving unit 101a receives context information corresponding to the current operation of the tool 233 and detection information output from the physical quantity information detection unit 227. The transmitting unit 101b transmits various information to the processing device 200.
[0041] The feature extraction unit 110 generates a learning model and extracts feature information (feature amount) used in the judgment by the judgment unit 102 from the detection information. Here, the feature information may be information indicating the feature of the detection information. For example, when the detection information is acoustic data collected by a microphone, the feature extraction unit 110 extracts feature amounts such as energy, frequency spectrum, and MFCC (Mel-Frequency Cepstrum Coefficients) from the detection information. In this embodiment, the feature extraction unit 110 has a BPF (Band Pass Filter) setting unit 111, a frequency shift estimation unit 114, a frequency analysis unit 115, and a processed waveform extraction unit 116. Furthermore, the BPF setting unit 111 has a bandwidth setting unit 112, a range setting unit 117, a band selection unit 113, and a natural frequency excluding unit 118.
[0042] The generating unit 104 generates a learning model for judging the normal processing state of the processing device 200 by learning using feature information extracted from the detection information when the processing device 200 is in a normal processing state. However, when the learning model is generated by an external device, the diagnostic device 100 does not need to include the generating unit 104. Specifically, the learning model may be generated by an external device, and the learning model generated by the external device may be received by the receiving unit 101a and stored in the storage unit 103. When context information for which a learning model is not defined and detection information corresponding to the context information are input, the generating unit 104 may generate a learning model corresponding to the context information by using feature information extracted from the detection information.
[0043] The determination unit 102 determines the processing state of the processing device 200 by using the feature information extracted from the detection information. In the present embodiment, the determination unit 102 determines the processing state of the processing device 200 by using the feature information and a learning model corresponding to the context information. For example, the determination unit 102 requests the feature extraction unit 110 to extract the feature information from the detection information. The determination unit 102 calculates a likelihood indicating the likelihood that the feature information extracted from the detection information is normal by using the corresponding learning model. The determination unit 102 compares the likelihood with a predetermined threshold. Then, when the likelihood is equal to or greater than the threshold, the determination unit 102 determines that the processing state of the processing device 200 is normal. Also, when the likelihood is less than the threshold, the determination unit 102 determines that the processing state of the processing device 200 is abnormal.
[0044] The method of determining the processing state of the processing device 200 is not limited to this, and any method may be used as long as the processing state of the processing device 200 can be determined using the feature information and the model. For example, the determination unit 102 may determine whether the processing state of the processing device 200 is normal or not by comparing a value indicating the fluctuation of the likelihood with a threshold value instead of directly comparing the likelihood with the threshold value. Alternatively, the determination unit 102 calculates a score that is a positive numerical value that is equal to or greater than zero by taking the logarithm of the likelihood and inverting the sign. The score is close to zero if the processing state of the processing device 200 is normal, and increases as the degree of abnormality of the processing state of the processing device 200 increases. Therefore, the determination unit 102 determines that the processing state of the processing device 200 is normal if the score is equal to or less than a predetermined threshold value, and determines that the processing state of the processing device 200 is abnormal if the score is equal to or greater than the threshold value. That is, the determination unit 102 determines the processing state of the processing device 200 by comparing at least one of the likelihood or a value calculated using the likelihood with a threshold value.
[0045] Each unit shown in FIG. 4 (communication control unit 101, judgment unit 102, reception unit 120, feature extraction unit 110, generation unit 104) may be realized by CPU 151 shown in FIG. 3 executing a program, that is, by software, or may be realized by hardware such as an IC (Integrated Circuit), or may be realized by a combination of software and hardware.
[0046] The diagnostic device 100 according to this embodiment is characterized by a feature extraction unit 110 and a reception unit 120. In this embodiment, the feature extraction unit 110 has a BPF (Band Pass Filter) setting unit 111, a frequency shift estimation unit 114, a frequency analysis unit 115, and a machining waveform extraction unit 116. Furthermore, the BPF setting unit 111 has a bandwidth setting unit 112, a range setting unit 117, a band selection unit 113, and a natural frequency excluding unit 118. In this embodiment, the diagnostic device 100 requires context information such as a spindle speed and tool information during or before and after the operation of the machining device 200, and therefore the reception unit 120 has a spindle speed reception unit 122, a tool information reception unit 121, and a machining process reception unit 123.
[0047] Next, the operation of the diagnostic device 100 according to this embodiment will be described in detail with reference to FIG.
[0048] In this embodiment, the processing device 200 has the physical quantity information detection unit 227 installed near the rotating main shaft 221, and an acceleration sensor is used as the sensor 270 of the physical quantity information detection unit 227. The physical quantity information detection unit 227 amplifies an analog signal detected by the sensor 270 with a preamplifier of the sensor 270, samples the analog signal at a predetermined time interval, and converts the sampled analog signal into a digital signal with an analog / digital (A / D) converter (signal conversion circuit 271). The diagnosis device 100 receives the digital signal output from the physical quantity information detection unit 227 as detection information with the receiving unit 101a. The digital signal output from the physical quantity information detection unit 227 is converted into a unit of acceleration by a calibration value of the sensor 270 as necessary, but these processes are omitted here and the description is given in a state that does not depend on the sensitivity of the sensor 270 or the specifications of the A / D converter (signal conversion circuit 271). Therefore, the receiving unit 101a receives, as detection information, a time domain waveform of an observation value proportional to the acceleration detected by the sensor 270 of the physical quantity information detecting unit 227.
[0049] The processing device 200 drilled a pilot hole of 5.0 mm in depth in an aluminum alloy plate with a drill of 8.2 mm in diameter, and then performed contouring by rotating an end mill with a diameter of 8.0 mm and four blades at 7500 rpm. The contouring process was divided into three steps, and the cutting depths in the radial direction of the tool 223 were set to 100.0 um (microns), 200.0 um, and 32.0 um, respectively. The XY axis stage 225 rotates so as to widen the diameter of the pilot hole at this cutting depth. At this time, the processing device 200 rotated the XY axis stage 225 at 90.0 mm / min in the same direction as the rotational direction of the rotating spindle 221.
[0050] The reception unit 120 of the diagnostic device 100 requests the machining device 200 to transmit the respective context information from the spindle rotation speed reception unit 122, the tool information reception unit 121, and the machining process reception unit 123, and transmits and receives the context information via the respective communication control units 205 and 101. Here, the context information includes rotation information, machining process information, tool information, etc.
[0051] The rotation information may be either the spindle speed set from the machining program read by the machining device 200 or the spindle speed measured by a tachometer in the machining device 200. The rotation information is, for example, 7500 rpm set from the machining program. The machining process information includes a number for identifying the machining process described in the machining program and information regarding the start and end of the operation of the rotating spindle 221 and the stage (here, the XY-axis stage 225 and the Z-axis stage 226). The machining process information is, for example, information regarding the start and end of rotation of the XY-axis stage 225. The tool information includes the type of tool, the diameter, the number of teeth, etc. However, the tool information is not limited to the context information from the machining device 200, and may be context information input from the input unit 158 to the diagnosis device 100, context information stored in the auxiliary storage device 155, context information received by the receiving unit 101a from an external device other than the machining device 200, etc. The tool information is, for example, the number of teeth of the tool 223 (for example, 4 teeth).
[0052] The in-machining waveform extraction unit 116 of the feature extraction unit 110 extracts in-machining waveform data, which is acceleration waveform data of time intervals corresponding to each of three machining steps, namely, the rotational motion of the XY-axis stage 225, the start of the rotational motion, and the end of the rotational motion, from the detection information (for example, acceleration waveform data, which is acceleration waveform data) input from the physical quantity information detection unit 227, for each cutting depth of the tool 223. The frequency analysis unit 115 is an example of a frequency analysis unit that performs frequency analysis on the detection information. The frequency analysis unit 115 performs Fourier transform on a predetermined number of consecutive samples of in-machining waveform data among the extracted in-machining waveform data, for example, using an FFT (Fast Fourier Transform) algorithm. The data string of the in-machining waveform data to be Fourier transformed may use the extracted in-machining waveform data for all of it, or may be one in which a part of it is replaced with zero. However, the processing waveform extraction unit 116 determines the frequency resolution that can be analyzed by the Fourier transform based on the sampling time interval of the detection information (acceleration waveform data) before A / D conversion by the signal conversion circuit 271 and the data length of the detection information (the number of data in the data string). Here, an embodiment will be described on the assumption that the combination of the time interval and the data length is set to obtain a frequency resolution of approximately 5.8 Hz.
[0053] 6 to 9 are diagrams showing an example of an average spectrum obtained by frequency analysis by the frequency analysis unit of the diagnosis device according to the present embodiment. The average spectrum shown in Figs. 6 to 9 is an average spectrum obtained by extracting a data string shifted by an arbitrary time from the beginning of the waveform data during machining, performing Fourier transform to obtain the power of amplitude, and performing averaging processing on the power. The average spectrum shown in Fig. 6 is an average spectrum when the cutting depth of the tool 223 is 200.0 um, and the average spectrum shown in Fig. 7 is an average spectrum when the cutting depth of the tool 223 is 32.0 um. The average spectrum shown in Fig. 8 is an average spectrum when the spindle rotation speed of the rotating spindle 221 is changed. The average spectrum shown in Fig. 9 is an average spectrum when the tool 223 is changed. In Figs. 6 to 9, only average spectra with frequencies of 1600.0 Hz or less are shown. In Figs. 6 and 7, the solid lines represent average spectra when the blade runout of the tool 223 is measured with a dial gauge and the maximum and minimum runout widths are adjusted to 2.0 um. 6 and 7, the dotted lines represent the average spectrum when the runout of the cutting edge of the tool 223 is adjusted to about 15.0 um to 20.0 um.
[0054] When the rotation speed of the rotating spindle 221 is 7500 rpm, the 7500 rpm is converted into a frequency of 125.0 Hz. In addition, since the number of blades of the end mill, which is an example of the tool 223, is four, the processing device 200 repeats intermittent cutting at 125.0 Hz x 4 = 500.0 Hz, generating a cutting force, which generates and propagates vibrations in the processing device 200 and the workpiece 224. Therefore, the diagnosis device 100 receives acceleration waveform data corresponding to this as detection information. The frequency of the acceleration waveform data is called the tool passing frequency (TPF). Therefore, the average spectrum obtained by the frequency analysis unit 115 ideally has a spectrum structure having sharp peaks at the TPF and its multiple harmonic components.
[0055] The solid lines shown in Figs. 6 and 7 are average spectra in which the runout of the cutting edge of the tool 223 is suppressed to 2.0 um or less. The peaks marked with ▽ in the average spectrum indicate the TPF, the second harmonic of the TPF, and the third harmonic of the TPF, and show large power. However, in reality, it is not possible to reduce the runout of the cutting edge of the tool 223 to zero, so other peaks are observed. On the other hand, the dotted lines shown in Figs. 6 and 7 are average spectra (large runout) in which the runout of the cutting edge of the tool 223 is adjusted to be between 15.0 um and 20.0 um. In the average spectra shown in Figs. 6 and 7, it is observed that the power is large in peaks other than the TPF and the harmonic components of the TPF. The peak marked with □ in the average spectrum shown in Figs. 6 and 7 is 125.0 Hz, which corresponds to the rotation speed of the rotating spindle 221, and this frequency is called the fundamental frequency of rotation. The peaks marked with circles in the average spectrum shown in Figs. 6 and 7 are components that increase or decrease by the fundamental frequency of rotation from the TPF and its harmonic components, i.e., are modulated components, and are called sideband waves. When the waveform of the TPF component is viewed in time series, the amplitude of the waveform of the sideband wave becomes uneven due to the vibration of the cutting edge, and appears as a characteristic of the spectrum. In addition, when damage to the cutting edge occurs, the cutting force becomes uneven, similar to the vibration of the cutting edge, and the generated vibration also becomes uneven, and a similar increase in sideband waves is observed. In this way, compared to the sideband waves in the normal machining state of the processing device 200, the sideband waves in the abnormal machining state increase, showing a positive correlation. On the other hand, when focusing on the TPF and its harmonic components, it is observed that the power of the average spectrum decreases as the vibration of the cutting edge increases. Therefore, the power of the average spectrum of the TPF and its harmonic components shows a negative correlation when the machining state changes from normal to abnormal.
[0056] Such characteristics are observed even when the rotation speed of the rotating spindle 221 and the number of blades of the tool 223 are changed. The average spectrum shown in FIG. 8 is an average spectrum of a machining process using the same end mill with four cutting blades, with the rotation speed of the rotating spindle 221 at 4000 rpm and the cutting depth at 200.0 um. The average spectrum shown in FIG. 9 is an average spectrum of a machining process using an end mill with a diameter of 8.0 mm and three cutting blades, with the rotation speed of the rotating spindle 221 at 4000 rpm and the cutting depth at 200.0 um. In both the average spectra of FIG. 8 and FIG. 9, the above-mentioned characteristics are observed for the TPF, its harmonic components, and most of the respective sidebands. Furthermore, in the rotation fundamental frequency of the average spectra shown in FIG. 8 and FIG. 9, the power increases as the runout of the cutting edge increases. This is a part of the component generated by runout caused by chucking error caused by chips getting mixed between the rotating spindle 221 and the tapered part of the tool holder, as shown in Patent Document 3.
[0057] In the average spectra shown in Figures 6 and 7, it can be seen that the shape of the envelope of the entire average spectrum and the power at the same frequency are different due to the difference in the cutting depth of the tool 223. This shows the need to distinguish models for different machining processes even when the same tool 223 is used at the same rotation speed, and it is useful to generate a learning model based on context information for each machining process, as shown in Figure 5.
[0058] 10 is a flowchart showing an example of the flow of diagnosis processing in the diagnosis device according to the present embodiment. The numerical control unit 206 of the processing device 200 sequentially transmits context information indicating the current operation of the tool 223 to the diagnosis device 100. The receiving unit 101a receives the context information transmitted from the processing device 200 (step S101). Next, the feature extraction unit 110 reads a learning model corresponding to the received context information from the storage unit 103 (step S102). The timing of step S102 may be immediately before step S106, as long as it is before step S106 in which feature information according to the BPF described later is extracted.
[0059] The physical quantity information detection unit 227 of the processing device 200 sequentially outputs detection information during processing by the processing device 200. The receiving unit 101a receives detection information (sensor data) transmitted from the processing device 200 (step S103). The processing waveform extraction unit 116 of the feature extraction unit 110 extracts processing waveform data from the received detection information and context information related to the processing operation (step S104). Next, the frequency analysis unit 115 performs frequency analysis of the processing waveform data using an FFT algorithm or the like (step S105). This frequency analysis is performed while shifting a preset number of data samples and a start position of a data string from the processing waveform data. The result of the frequency analysis of the processing waveform data is three-dimensional data in which multiple spectra are arranged in a time series. The feature extraction unit 110 extracts feature information from multiple spectra or an average spectrum obtained by averaging multiple spectra over an arbitrary time range (step S106). In this embodiment, since the BPF is recorded in the learning model read in step S102, the feature extraction unit 110 uses the BPF to extract feature information from the average spectrum.
[0060] The determination unit 102 determines the machining state of the machining device 200 using the feature information extracted by the feature extraction unit 110 and the learning model corresponding to the received context information (step S107). This allows the machining state of the machining device 200 to be determined using the feature information and learning model extracted according to the tool 223 and the type of machining, making it possible to detect and monitor the occurrence of an abnormality in the machining state of the machining device 200 with high accuracy for various machining operations performed in the machining center. The determination unit 102 outputs the determination result to the display unit 106 via the display control unit 105 (step S108). Alternatively, the determination unit 102 transmits alert information to the machining device 200 or an external device via the transmission unit 101b (step S108).
[0061] Next, an example of the model generation process by the diagnostic device 100 according to the present embodiment will be described with reference to FIG. 11. FIG. 11 is a flowchart showing an example of the flow of the model generation process by the diagnostic device according to the present embodiment. In the present embodiment, the generation unit 104 executes the model generation process in advance, for example, before the diagnostic process of the processing device 200. Alternatively, as described above, the generation unit 104 may execute the model generation process when context information for which a learning model is not defined is input. Also, when the learning model is generated externally as described above, the model generation process does not need to be executed in the diagnostic device 100.
[0062] The receiving unit 101a receives the context information transmitted from the processing device 200 (step S201). In addition, the receiving unit 101a receives the detection information (sensor data) transmitted from the processing device 200 (step S202).
[0063] The context information and detection information received in this manner are used to generate a learning model. In this embodiment, since the generation unit 104 generates a learning model for each piece of context information, the detection information needs to be associated with the corresponding context information. For this reason, for example, the receiving unit 101a associates the received detection information with the context information received at the same timing and temporarily stores it in the storage unit 103 or the like. Then, the generation unit 104 confirms that the detection information stored in the storage unit 103 is information in a normal state, and generates a learning model using only the detection information in a normal state. That is, the generation unit 104 generates a learning model using detection information labeled as normal.
[0064] The confirmation (labeling) of whether the detection information is normal or not may be performed at any timing after the detection information is stored in the storage unit 103 or the like, or may be performed in real time while the processing device 200 is operating. Alternatively, the generation unit 104 may generate a learning model by assuming that the detection information is normal without performing labeling of the detection information. If the detection information assumed to be normal is actually abnormal detection information, the generated learning model will not correctly execute the determination process of whether the processing state of the processing device 200 is normal or not. Therefore, it is possible to determine whether the learning model was generated using abnormal detection information based on the frequency at which the processing state of the processing device 200 is determined to be abnormal, and measures such as deleting the erroneously generated learning model can be taken. Alternatively, the learning model generated using abnormal detection information may be used as a learning model for determining abnormality.
[0065] The in-process waveform extraction unit 116 of the feature extraction unit 110 extracts in-process waveform data based on the received detection information and the context information during the processing operation (step S203). Next, the frequency analysis unit 115 performs frequency analysis on the extracted in-process waveform data using an FFT algorithm or the like (step S204). The frequency analysis unit 115 performs frequency analysis on the in-process waveform data while shifting a preset number of data samples and the start position of the data string. The obtained frequency analysis result is three-dimensional data in which multiple spectra are arranged in time series. The feature extraction unit 110 extracts feature information from multiple spectra or a spectrum obtained by averaging multiple spectra over an arbitrary time range according to the BPF (step S205). This method will be described later.
[0066] The generating unit 104 generates a learning model corresponding to the same context information by using feature information extracted from the detection information associated with the same context information (step S206). The generating unit 104 stores the generated learning model in the storage unit 103 (step S207).
[0067] Next, an example of the flow of the feature information extraction process according to the BPF by the feature extraction unit 110 according to the present embodiment will be described with reference to Fig. 12. Fig. 12 is a flowchart showing an example of the flow of the feature information extraction process according to the BPF by the diagnosis device according to the present embodiment.
[0068] The BPF setting unit 111 receives tool information such as the number of cutting edges (number of teeth) t and rotation information such as the spindle rotation speed r from the tool information receiving unit 121 and the spindle rotation speed receiving unit 122 of the receiving unit 120 (step S301). Then, the BPF setting unit 111 calculates the fundamental rotation frequency and the TPF using the formulas (1) and (2) (step S302). Here, the spindle rotation speed r is the rotation speed of the rotating spindle 221 set by the machining program. Fundamental rotation frequency [Hz] = r [rpm] / 60 (1) TPF = fundamental frequency [Hz] × t [sheets] (2) That is, the BPF setting unit 111 calculates a fundamental rotational frequency using the rotation information, and calculates a TPF using the fundamental rotational frequency and the number of teeth included in the tool information.
[0069] Next, the BPF setting unit 111 of the feature extraction unit 110 sets (calculates) a BPF center frequency (an example of a center frequency) using the rotation information, the tool information, and the frequency range. In this embodiment, the BPF setting unit 111 uses the formula (3) to set a BPF center frequency in a frequency range (frequency range) set by the range setting unit 117 of the BPF setting unit 111. Here, the range setting unit 117 sets a frequency range of interest. For example, the range setting unit 117 sets a lower limit frequency and an upper limit frequency, and calculates a natural number n such that the BPF center frequency of the formula (3) is in this range. Alternatively, the natural number n may be any number that can specify the lower limit and upper limit of the frequency, such as the order of the harmonic of the rotation fundamental frequency [Hz] or the order of the overtone of the TPF. That is, the BPF setting unit 111 sets the rotation fundamental frequency and an integer multiple frequency of the rotation fundamental frequency as the BPF center frequency. Alternatively, the BPF setting unit 111 may set the TPF and sidebands of frequencies that are integer multiples of the TPF as the BPF center frequency.
[0070] Furthermore, the BPF setting unit 111 calculates the BPF from the BPF center frequency and the bandwidth [Hz] set by the bandwidth setting unit 112 (step S303). Here, the bandwidth setting unit 112 sets the bandwidth of a frequency band of interest within the frequency range set by the range setting unit 117. BPF center frequency = fundamental frequency of rotation [Hz] × n (3)
[0071] When the bandwidth [Hz] is set to b by the bandwidth setting unit 112, the BPF setting unit 111 calculates BPFs for the number of BPF center frequencies so as to satisfy the equation (4). BPF center frequency -b / 2≦BPF(n)≦BPF center frequency+b / 2...(4)
[0072] Next, band selection unit 113 (an example of a BPF selection unit) selects a BPF to be used for extracting feature information from among the multiple BPFs (step S304). Band selection unit 113 selects a BPF by one of the following selection methods (a) to (e). (a) Fullselect (b) Overtones of the TPF (TPF, 2×TPF, …) and their sidebands (c) Only the sidebands of the harmonics of the TPF (d) The fundamental frequency of rotation and (b) or (c) (e) Interactive selection
[0073] 13 is a diagram for explaining an example of a method for selecting a BPF by the diagnosis device according to the present embodiment. For example, when selecting a BPF by the selection method (e), the display control unit 105 displays a BPF selection screen 500 on the display unit 106. The BPF selection screen 500 includes a context information display unit 310 displaying the fundamental rotation frequency and the number of teeth, a range display unit 320 displaying the range set by the range setting unit 117, a bandwidth display unit 330 displaying the bandwidth set by the bandwidth setting unit 112, a band display unit 340 displaying the band selected by the band selection unit 113, a data display unit 350, and the like.
[0074] The range display unit 320 displays the upper limit frequency of the range set by the range setting unit 117. Specifically, the range display unit 320 includes a frequency input text box 323 in which the upper limit frequency can be input, and an order input text box 324 in which the upper limit frequency is input in the harmonic order of the TPF. The user of the diagnostic device 100 can input the upper limit frequency exclusively into the frequency input text box 323 and the order input text box 324. The frequency radio button 321 and the TPF order radio button 322 can exclusively set which of the frequency input text box 323 and the order input text box 324 the upper limit frequency is input into. In the range display unit 320 shown in FIG. 13, the TPF harmonic order is input into the order input text box 324, and the upper limit frequency is input as the second order TPF (2×TPF).
[0075] The bandwidth display section 330 has a bandwidth input text box 331. In the bandwidth input text box 331 shown in Fig. 13, 40.0 Hz is set as the bandwidth. In the data display section 350, an average spectrum labeled as normal is displayed, which is an average spectrum obtained by averaging a plurality of spectra obtained by frequency analysis of the in-process waveform data used for learning the model generation process shown in Fig. 11. The average spectrum displayed in the data display section 350 may be an average spectrum obtained from test in-process waveform data obtained by processing with a pseudo blade vibration and labeled as abnormal, or may be a plurality of average spectra labeled as both normal and abnormal (average spectra shown in Figs. 6 to 9).
[0076] 12. The band display unit 340 displays a TPF selection toggle button 341 for selecting use of a TPF for extracting feature information, a sideband selection toggle button 342 for selecting use of a sideband for extracting feature information, and an other harmonics toggle button 343, so as to correspond to the BPF displayed in the BPF display unit 351. Of the TPF selection toggle button 341, the sideband selection toggle button 342, and the other harmonics toggle button 343, those used for extracting feature information in step S306, which will be described later, are turned on, and those not used for extracting feature information are turned off.
[0077] Next, the natural frequency excluding unit 118 excludes BPFs close to the natural frequency of the machining device 200 and the tool 223 from the BPFs set by the BPF setting unit 111 (step S305). The tool 223, the holder, the rotating spindle 221, etc. have natural frequencies due to their shapes, dimensions, and weights. The frequency components of the natural frequencies tend to be greater than the power of other frequency components in machining caused by damage or whirling of the cutting tool, whether the machining state of the machining device 200 is in an abnormal state or in a normal state. Therefore, if the frequency components of the natural frequencies are included in the characteristic information, the accuracy of determining the machining state of the machining device 200 decreases. Therefore, the natural frequency is input in advance from the input unit 107 and stored in the memory unit 103, and the natural frequency excluding unit 118 calls up the natural frequency from the memory unit 103, and if there is a BPF including the natural frequency among the BPFs calculated using the formula (4), the BPF is excluded.
[0078] Fig. 14 is an enlarged view of the vicinity of the BPF of the average spectrum calculated by the diagnostic device according to this embodiment. In Fig. 14, the solid line average spectrum is the average spectrum near the BPF (indicated by reference number 361) labeled as normal, and the dashed line average spectrum is the average spectrum near the BPF (indicated by reference number 361) labeled as abnormal.
[0079] 14, the peak indicated by reference numeral 362 is the natural frequency. When selecting a BPF by the above-mentioned selection method (e), the user turns off the sideband selection toggle button 360 on the BPF selection screen 500 to exclude selection methods (b) and (c). Alternatively, in the case of the above-mentioned selection methods (a) and (e), the BPFs including the natural frequency are automatically excluded from the BPFs.
[0080] The feature extraction unit 110 extracts only the power whose center frequency is within the range of the BPF from the power of the average spectrum obtained by the Fourier transform as feature information of the average spectrum (step S306). That is, the feature extraction unit 110 extracts feature information using the BPF selected by the band selection unit 113. For example, the feature extraction unit 110 may set the bandwidth to zero, select the center frequency of the Fourier transform closest to the BPF center frequency of equation (3), and extract the power corresponding to the center frequency in the average spectrum as feature information. The feature extraction unit 110 converts the amplitude and power extracted as feature information from the average spectrum into optimal values such as a linear scale or a log scale (dB) depending on the machining method and the type of tool.
[0081] In step S105 of Fig. 10, the frequency analysis unit 115 performs frequency analysis of the waveform data being processed using an FFT algorithm or the like, while shifting a preset number of data samples and the start position of the data string from among the waveform data being processed, similarly to step S204 of Fig. 11. This results in a three-dimensional data group in which multiple spectra SPj(f) are arranged in time series. Here, j (=1 to J) is the number of spectra, and corresponds to the number of spectra for which frequency analysis was performed while shifting the start position of the data string.
[0082] Next, a first judgment method of the processing state of the processing device 200 will be described. In the first judgment method, first, the feature extraction unit 110 calculates an average spectrum SP(f) of a plurality of spectra SPj(f). Next, the feature extraction unit 110 extracts the power or amplitude closest to the BPF center frequency from the average spectrum SP(f) as feature information. When the extracted feature information is the TPF and its overtones, the judgment unit 102 compares the TPF and its overtones with thresholds preset for the TPF and its overtones, respectively, and judges that the processing state of the processing device 200 is abnormal when the feature information is less than the threshold. In addition, when the extracted feature information is a sideband wave and other harmonics of the rotation fundamental frequency, the judgment unit 102 compares the sideband wave and the harmonics with thresholds preset for the sideband wave and the harmonics, respectively, and judges that the processing state of the processing device 200 is abnormal when the feature information exceeds the threshold. Alternatively, when the extracted characteristic information is a sideband wave or other harmonics of the fundamental rotation frequency, the judgment unit 102 may calculate the establishment rate at which the characteristic information exceeds a threshold, compare the establishment rate with a threshold value previously set for the establishment rate, and judge that the processing state of the processing device 200 is abnormal if the establishment rate exceeds the threshold.
[0083] Next, a second judgment method of the processing state of the processing device 200 will be described. In the second judgment method, the feature extraction unit 110 extracts feature information in the same manner as the first judgment method. Next, the judgment unit 102 performs one-class SVM learning using the multidimensional feature information, and judges whether the processing state of the processing device 200 is abnormal by outlier detection.
[0084] Next, a third judgment method of the processing state of the processing device 200 will be described. In the third judgment method, the judgment unit 102 reads a learning model (a learning model when the processing state of the processing device 200 is normal) generated by the model generation process shown in FIG. 11 from the storage unit 103. The actual state of this learning model may be a probability density function P(X) such as a GMM (Gaussian mixture model). Here, X(={x1, x2, . . . xn}) is an n-dimensional feature extracted according to the BPF when the learning model is learned. The BPF is stored in the storage unit 103 together with this learning model, and in step S106 of FIG. 11, the feature values of each of the multiple spectra SPj(f) are extracted using this BPF.
[0085] 11, if the likelihood obtained by inputting the feature into the probability density function P(X) is equal to or greater than a preset threshold, the determination unit 102 determines that the processing state of the processing device 200 is normal, and if the likelihood is less than the threshold, the determination unit 102 determines that the processing state of the processing device 200 is abnormal. Alternatively, as shown in the following formula (5), the determination unit 102 defines a value with the sign of the log likelihood reversed as the anomaly score aj, sets the anomaly score as an index value such that the anomaly score increases when the abnormal state of the processing device 200 is strong, and obtains the anomaly score aj for the number j=1 to J of spectra. aj = -log(P(Xj)) (5)
[0086] The judgment unit 102 selects, as the total score of the abnormality scores aj, for example, the maximum value of the abnormality scores aj, the average of the abnormality scores aj, or a value suitable for the tool or processing method, as shown in equation (6). A = (Σaj) / J (6)
[0087] Then, the judgment unit 102 compares the abnormality score aj with a preset threshold value, and if the abnormality score aj is greater than or equal to the threshold value, judges that the processing state of the processing device 200 is abnormal, and if the abnormality score aj is less than the threshold value, judges that the processing state of the processing device 200 is normal.
[0088] Next, we will explain the operation of the diagnostic device 100 when contouring is performed by adjusting the runout of the cutting edge of the end mill used in the above-mentioned contouring process to 2.0 um or less, increasing the rotation speed of the XY-axis stage 225 to 520.0 mm / min, and leaving the other processing conditions unchanged.
[0089] The feature extraction unit 110 extracts data of the machining section with a cutting depth of 200.0 um from the detection information received from the machining device 200, and obtains an average spectrum by the frequency analysis shown in step S204 of FIG.
[0090] Fig. 15 is a diagram showing an example of an average spectrum obtained by the diagnosis device according to this embodiment. In Fig. 15, the solid line represents the average spectrum when processing is performed by the processing device 200 under the above processing conditions, and the dashed line represents the average spectrum labeled as normal. As described above, the only processing condition that differs is the rotation speed (90.0 mm / min) of the XY axis stage 225.
[0091] As the rotation speed of the XY-axis stage 225 increases, the harmonics of the fundamental rotation frequency (and TPF) shift to the lower frequency side. The shift of the fundamental rotation frequency (125.0 Hz) was approximately 3.0 Hz, and the shift of the second harmonic of the TPF (=1000.0 Hz) was approximately 25.0 Hz. This shift in the fundamental rotation frequency and TPF is caused by the cutting blade changing the contact position with the arc surface of the pilot hole in the workpiece due to the rotational motion of the XY-axis stage 225.
[0092] 10 and step S205 in FIG. 11, with a shifted BPF (a BPF that does not correspond to the shift of the rotation fundamental frequency), correct feature information cannot be extracted, and the judgment result of the processing state of the processing device 200 becomes doubtful. Therefore, if the shifted rotation fundamental frequency F is obtained, the rotation fundamental frequency in the formulas (2) to (4) can be replaced with the shifted rotation fundamental frequency F, and the BPF can be calculated. In this embodiment, the frequency shift estimation unit 114 corrects the rotation fundamental frequency using the frequency analysis result (e.g., average spectrum) of the detection information by the frequency analysis unit 115 and the rotation information, and calculates the shifted rotation fundamental frequency F.
[0093] Fig. 16 is a flowchart showing another example of the flow of the characteristic information extraction process according to the BPF by the diagnosis device according to this embodiment. In the flow of the characteristic information extraction process shown in Fig. 16, step S302 shown in Fig. 12 is replaced by the process shown in step S307. Here, the BPF setting unit 111 calculates the shifted rotation fundamental frequency F and the TPF (step S307).
[0094] Next, an example of the correction process of the rotation fundamental frequency in the diagnosis device 100 according to the present embodiment will be described. As shown in Fig. 15, large peaks are observed in the average spectrum at the TPF and its harmonic components. The frequency shift estimation unit 114 searches for a frequency at which the average spectrum shows a maximum in the vicinity of the TPF obtained in step S302 shown in Fig. 12 and an integer multiple frequency of the TPF as a reference, and corrects the rotation fundamental frequency based on the searched frequency. Alternatively, the frequency shift estimation unit 114 may appropriately mix the rotation fundamental frequency obtained in step S302 and the integer multiple frequency of the TPF.
[0095] For example, the frequency shift estimation unit 114 may divide the found frequencies by their respective reference integer values, and calculate the average value as the corrected fundamental rotation frequency. Alternatively, the frequency shift estimation unit 114 may plot the integer values on the horizontal axis and the found frequencies on the vertical axis, and calculate the slope by the least squares method.
[0096] Next, another example of the correction process of the fundamental rotation frequency in the diagnosis device 100 according to this embodiment will be described. If the fundamental rotation frequency calculated from the spindle speed included in the context information according to equation (1) is f, the fundamental rotation frequency f is 125.0 Hz in the average spectrum 401 shown in Fig. 15. The frequency shift estimation unit 114 calculates the autocorrelation function of the average spectrum 401 shown in Fig. 15 by using the following equation (7). Fig. 17 is a diagram showing an example of the autocorrelation function calculated by the diagnosis device according to this embodiment.
number
[0097] Since the autocorrelation of the spectrum has been calculated, the harmonics can be identified. h is selected as the index with an upper frequency limit of 7f = 875.0 Hz. Therefore, n is the index with an upper frequency limit of 14f = 1750.0 Hz. As shown in Figure 15, it can be seen that the TPF and its harmonic components have a large power in the average spectrum. Therefore, the autocorrelation function shown in Figure 17 is also maximum at h = TPF. If h at this time is H, then equation (8) holds. F = H / t (8)
[0098] Furthermore, the frequency shift estimation unit 114 can also estimate the number of teeth t based on the fundamental rotational frequency f calculated from the context information, as shown in equation (9). t = round(H / f) (9) Here, round() is a function that rounds its argument to the nearest integer.
[0099] That is, the frequency shift estimation unit 114 calculates an autocorrelation function of the frequency analysis result of the detection information, and obtains a delay value at which the autocorrelation function is maximum within a range where the delay value of the autocorrelation function exceeds the corrected fundamental rotation frequency. Next, the frequency shift estimation unit 114 estimates the number of teeth of the tool 223 using the obtained delay value. Then, the BPF setting unit 111 sets the BPF using multiple BPF center frequencies that are set (calculated) using the estimated number of teeth instead of the tool information.
[0100] In this way, according to the machining system of this embodiment, the machining state of the machining device 200 can be determined using feature information extracted according to the tool 223 and the type of machining, so that it is possible to detect and monitor with high accuracy the occurrence of abnormalities in the machining state of the machining device 200 for various machining operations performed in a machining center.
[0101] The program executed by the diagnostic device 100 of this embodiment is provided in advance in the ROM 152 or the like. The program executed by the diagnostic device 100 of this embodiment may be provided as a computer program product by being recorded in a computer-readable recording medium such as a CD-ROM, a flexible disk (FD), a CD-R, or a digital versatile disk (DVD) in the form of an installable or executable file.
[0102] Furthermore, the program executed by the diagnostic device 100 of this embodiment may be configured to be stored on a computer connected to a network such as the Internet and provided by being downloaded via the network. Also, the program executed by the diagnostic device 100 of this embodiment may be configured to be provided or distributed via a network such as the Internet.
[0103] The program executed by the diagnostic device 100 of this embodiment has a modular configuration including the above-mentioned units (communication control unit 101, judgment unit 102, generation unit 104, display control unit 105, feature extraction unit 110, reception unit 120, etc.), and as actual hardware, the CPU 151 (an example of a processor) reads out the program from the ROM 152 and executes it, thereby loading the above-mentioned units onto a main storage device and generating the units on the main storage device. Also, the hardware of the diagnostic device 100 of this embodiment may be incorporated into the processing device 200, which is configured to execute the program, to form a processing device with a diagnostic function. [Explanation of symbols]
[0104] 100 Diagnostic Equipment 101 Communication control unit 101a Receiver 101b Transmitter 102 Judgment section 103 Storage section 104 Generation part 105 Display control unit 106 Display section 107 Input section 110 Feature Extraction Unit 111 BPF setting section 112 Bandwidth setting section 113 Band Selection Section 114 Frequency shift estimator 115 Frequency Analysis Section 116 Processing waveform extraction section 117 Range setting section 118 Natural frequency exclusion section 120 Reception 121 Tool Information Reception Department 122 Spindle speed reception unit 123 Processing Process Reception Department 200 Processing equipment 205 Communication control section 221 Rotating shaft 223 Tools 227 Physical quantity information detection unit [Prior art documents]
Patent Document
[0105]
Patent Document 1
Patent Document 2
Patent Document 3
Claims
1. a receiving unit that receives context information including rotation information of a rotary shaft constituting a processing device and tool information of a tool attached to the rotary shaft and that specifies an operation of the tool, and detection information of a time-varying physical quantity generated by the tool while the tool is executing a processing operation on a workpiece; A frequency analysis unit that performs frequency analysis on the detection information; A range setting unit that sets a frequency range; a bandwidth setting unit that sets a bandwidth of a frequency band of interest in the frequency range; a band pass filter (BPF) setting unit that sets a plurality of center frequencies that are set using the rotation information, the tool information, and the frequency range, and the bandwidth, and sets a rotation fundamental frequency calculated using the rotation information and a frequency that is an integer multiple of the rotation fundamental frequency as the plurality of center frequencies; a feature information extraction unit that extracts feature information of the detection information from a frequency analysis result of the detection information by using the BPF; a determination unit that determines a processing state of the processing device using the characteristic information; a frequency shift estimation unit that calculates an autocorrelation function of a result of frequency analysis of the detection information, obtains a delay value at which the autocorrelation function is maximum within a range in which a delay value of the autocorrelation function exceeds the fundamental frequency of rotation, and estimates the number of teeth of the tool using the delay value; The BPF setting unit sets the BPF using the plurality of center frequencies that are set using the estimated number of teeth instead of the tool information.
2. A generation unit that generates a model using the feature information, The diagnostic device according to claim 1 , wherein the determining unit determines the machining state by using the model.
3. a BPF selection unit that selects the BPF to be used for extracting the feature information from the BPFs set by the BPF setting unit, The diagnostic device according to claim 1 , wherein the feature information extraction unit extracts the feature information by using the BPF selected by the BPF selection unit.
4. The diagnostic device according to claim 1 , further comprising a natural frequency exclusion unit that excludes the BPF close to a natural frequency of the processing device and the tool from the BPF.
5. the plurality of center frequencies include a fundamental rotation frequency calculated using the rotation information and a frequency that is an integer multiple of the fundamental rotation frequency, The diagnostic device according to claim 1 , wherein the BPF setting unit corrects the fundamental rotation frequency by using a frequency analysis result of the detection information and the rotation information.
6. receiving context information including rotation information of a rotary shaft constituting a processing device and tool information of a tool attached to the rotary shaft and defining an operation of the tool, and detection information of a time-varying physical quantity generated by the tool while the tool is executing a processing operation on a workpiece; a frequency analysis step of performing a frequency analysis on the detection information; a range setting step for setting a frequency range; a bandwidth setting step of setting a bandwidth of a frequency band of interest in the frequency range; a BPF setting step of setting a band pass filter (BPF) using a plurality of center frequency information sets that are set using the rotation information, the tool information, and the frequency range, and the bandwidth, and setting a rotation fundamental frequency calculated using the rotation information and a frequency that is an integer multiple of the rotation fundamental frequency as the plurality of center frequencies; a feature extraction step of extracting feature information of the detection information by using a frequency analysis result of the detection information and the BPF; a determination step of determining a processing state of the processing device using the characteristic information; an estimation step of calculating an autocorrelation function of a result of frequency analysis of the detection information, determining a delay value at which the autocorrelation function is maximum within a range in which a delay value of the autocorrelation function exceeds the fundamental frequency of rotation, and estimating the number of teeth of the tool using the delay value, The BPF setting step sets the BPF using the plurality of center frequencies that are set by using the estimated number of teeth instead of the tool information.
7. A generating step of generating a model using the feature information, The diagnostic method according to claim 6 , wherein the determining step determines the machining state by using the model.
8. 8. The diagnostic method according to claim 6, wherein the BPF setting step sets, as the plurality of center frequencies, a cutting edge passing frequency (TPF) calculated using the fundamental rotation frequency calculated using the rotation information and the number of teeth of the tool information, and sidebands of each frequency of an integer multiple of the TPF.
9. a BPF selection step of selecting the BPF to be used for extracting the feature information from the BPFs set in the BPF setting step, 9. The diagnostic method according to claim 6, wherein the feature extraction step extracts the feature information by using the BPF selected in the BPF selection step.
10. 8. The diagnostic method according to claim 7, wherein the determining step determines the machining state by calculating a likelihood that the feature information is normal using the model, and comparing at least one of the likelihood or a value calculated using the likelihood with a preset threshold value.
11. On the computer, A program for executing the diagnostic method according to any one of claims 6 to 10.
12. A processing system including the diagnostic device according to any one of claims 1 to 5 and a processing device to be diagnosed by the diagnostic device, The processing system includes a transmission unit that transmits the context information and the detection information to the diagnostic device.
13. A processing system including a diagnostic device that executes the diagnostic method according to any one of claims 6 to 10 and a processing device to be diagnosed by the diagnostic device, The processing system includes a transmission unit that transmits the context information and the detection information to the diagnostic device.
14. A processing system including a diagnostic device including the computer that executes the program according to claim 11 and a processing device to be diagnosed by the diagnostic device, The processing system includes a transmission unit that transmits the context information and the detection information to the diagnostic device.
Citation Information
Patent Citations
A condition monitoring system
EP3527961A1
Method and device for monitoring the status of machine tool machining cycle
JP1983500605A
Fully automatic transfer system of automatic private branch exchange
JP1986056566A
Abnormality detection method and device for rotating tool
JP1997174383A
Inspection method, inspection apparatus and diagnostic apparatus for facility
JP2005121639A
