JP2023106206A - Machine tool diagnosis system, diagnosis method, and diagnosis program - Google Patents

Abstract

To enhance diagnostic precision and minimize implementation of a diagnostic motion.SOLUTION: A diagnosis system 20 includes a first diagnosis unit 21 that uses operation data, which is acquired during normal operation of a machine tool 1, to diagnose whether the machine tool 1 falls under preliminarily classified normalcy or any of plural abnormalities signifying abnormal regions, a second diagnosis unit 22 that, when an abnormal region is not identified even though the first diagnosis unit 21 diagnoses the machine tool as not being normal, causes the machine tool 1 to implement a predetermined diagnostic motion and identifies the abnormal region using acquired diagnostic motion data, and a diagnostic processing update unit 23 that updates the classification of normalcy and/or abnormality signifying an abnormal region, which is made by the first diagnosis unit 21, on the basis of a result of diagnosis conducted by the second diagnosis unit 22.SELECTED DRAWING: Figure 1

Description

The present disclosure relates to a diagnostic device, diagnostic method, and diagnostic program for diagnosing an abnormality in a machine tool.

Unexpected failures of machinery and equipment cause unplanned losses, so it is required to detect signs before failure occurs from the operating status of machinery and equipment. In addition, when detecting a sign of trouble, it is also necessary to specify an abnormal location that causes the trouble in order to eliminate the trouble.
In Patent Document 1, a normal model is constructed using the state quantities acquired at the time when the plant is known to be normal, and whether or not the state quantities acquired from the plant at the time of diagnosis are classified into the normal model. is determined, and if it is classified as a normal model, it is displayed as normal, and if it is not classified as a normal model, it is displayed as an unknown state.
In addition, in Patent Document 2, pattern matching is performed using actual measurement data of normal or abnormal results and simulated abnormal data that simulates abnormalities in facility equipment generated in consideration of hypothetical abnormal conditions. A method for diagnosing the presence or absence of is disclosed.

WO2012/73289 JP 2018-190245 A

Methods of acquiring data for diagnosis can be broadly divided into methods that use operations during normal operation performed in the production process, such as during operation of a machining program, and methods that use specific diagnostic operation patterns.
In the method using operations during normal operation, the normal range is often set and diagnosed using data shortly after the product is manufactured. Therefore, it is difficult to determine whether the change is an abnormal state due to the deterioration of parts or a change that should be regarded as normal, such as a change in the environment or usage conditions, when the state is out of the normal state. . In addition, when the normal state is reset due to the absence of manufacturing defects in a state where the number of years of use has passed, there is a possibility of overlooking the failure sign state that the parts are actually deteriorated although there are no manufacturing defects.
On the other hand, the method of performing diagnosis using operation patterns for diagnosis can be performed by operations set by the machine manufacturer side. It also makes it easier to collect data on unusual abnormal conditions. Therefore, there is an advantage that the diagnostic accuracy of normality and abnormality is high and the location of abnormality can be specified. However, since it is necessary to perform an operation that is not a normal operation, there is a problem that time is lost for diagnosis if performed periodically. This loss of time becomes conspicuous particularly in machine tools that manufacture a wide variety of products, leading to a decrease in production efficiency.

Therefore, an object of the present disclosure is to provide a machine tool diagnostic apparatus, diagnostic method, and diagnostic program capable of improving diagnostic accuracy while minimizing the execution of diagnostic operations and suppressing the occurrence of time loss. do.

In order to achieve the above object, a first configuration of the present disclosure is a diagnostic device for diagnosing an abnormality in a machine tool, comprising:
a first diagnosing unit that diagnoses whether the machine tool is normal or one of a plurality of abnormal types classified in advance using operation data acquired during normal operation of the machine tool;
When the type of the abnormality cannot be specified even if the first diagnosis unit diagnoses that the abnormality is not normal, the machine tool is caused to perform a predetermined diagnostic operation, and the acquired operation data for diagnosis is used to identify the abnormality. a second diagnostic unit that identifies the type;
a diagnostic processing updating unit that updates the classification of the normal and/or the abnormal type in the first diagnostic unit based on the diagnostic result of the second diagnostic unit.
According to another aspect of the first configuration, in the above configuration, the first diagnosis unit, when neither the diagnosis of normality nor the identification of the type of abnormality can be executed, the diagnosis result and the type of each abnormality Calculate the rank of relevance to
The second diagnosis unit is characterized in that the type of abnormality is identified based on the calculated order.
In addition, in the present disclosure, the “abnormality type” includes at least an abnormal site.
In another aspect of the first configuration, in the configuration described above, when the second diagnostic unit identifies the type of abnormality, the diagnostic processing updating unit performs It is characterized by updating the setting range of the type of abnormality.
In another aspect of the first configuration, in the configuration described above, the diagnostic processing update unit causes the normal state in the first diagnostic unit when the type of the abnormality is not specified by the second diagnostic unit. is characterized by updating the setting range of
In order to achieve the above object, a second configuration of the present disclosure is a diagnostic method for diagnosing an abnormality in a machine tool, comprising:
a first diagnosis step of diagnosing whether the machine tool is normal or one of a plurality of abnormal types classified in advance using operation data acquired during normal operation of the machine tool;
If the type of the abnormality cannot be specified even if the first diagnosis step diagnoses that the machine tool is not normal, the machine tool is caused to perform a predetermined diagnostic operation, and the acquired operation data for diagnosis is used to identify the abnormality. a second diagnostic step of type identification;
and a diagnostic processing update step of updating the classification of the normal and/or the abnormal type in the first diagnostic step based on the diagnostic result of the second diagnostic step.
Another aspect of the second configuration is that in the above configuration, in the first diagnosis step, if neither the diagnosis of normality nor the identification of the type of abnormality can be executed, the diagnosis result and the type of each abnormality Calculate the rank of relevance to
In the second diagnosis step, the type of abnormality is specified based on the calculated ranking.
Another aspect of the second configuration is that, in the above configuration, in the diagnostic process update step, when the type of abnormality is specified in the second diagnostic step, the specified type of abnormality in the first diagnostic step is It is characterized by updating the setting range of the type of abnormality.
According to another aspect of the second configuration, in the above configuration, in the diagnostic processing update step, if the type of abnormality is not specified in the second diagnostic step, the normal state in the first diagnostic step is determined. It is characterized by updating the setting range.
In order to achieve the above object, a third configuration of the present disclosure is a diagnostic program for a machine tool, in which a machine tool control device executes the machine tool diagnostic method according to any one of the second configurations. It is characterized by

According to the present disclosure, diagnosis using operation data and diagnosis using operation data for diagnosis are used together, and operation data for diagnosis is used only when the type of abnormality cannot be specified in the diagnosis result based on the operation data. run the diagnostics Therefore, it is possible to improve diagnostic accuracy while suppressing the occurrence of time loss by minimizing the execution of diagnostic operations. Therefore, it is particularly useful in machinery and equipment for manufacturing a wide variety of products.

1 is a block configuration diagram of a diagnostic device for a machine tool; FIG. FIG. 10 is an explanatory diagram showing an example of diagnostic setting by a diagnostic processing updating unit; 4 is a flow chart of a diagnostic method for a machine tool;

Embodiments of the present disclosure will be described below with reference to the drawings.
FIG. 1 is a block configuration diagram showing an example of a diagnostic device for a machine tool according to a first configuration. The machine tool 1 is controlled by an NC device 10 . The NC device 10 has a program interpretation section 11 and a machine operation command section 12 . The program interpreter 11 interprets a program input by an operator through input means (not shown) into a spindle rotation command, a feed axis operation command, and a peripheral device operation command. The machine operation command unit 12 controls the spindle motor, each feed shaft, and peripheral devices based on commands sent from the program interpretation unit 11 .
In the diagnostic device 20, based on the diagnostic program according to the third configuration input to the NC device 10, information used for controlling the spindle motor and each feed axis from the machine operation command unit 12, and information provided for separate diagnosis. It acquires the information from the sensor and diagnoses the presence or absence of anomalies. The diagnostic device 20 includes a first diagnostic section 21 , a second diagnostic section 22 and a diagnostic processing update section 23 .

The first diagnosis unit 21 performs diagnosis using operation data during normal operation. Diagnosis processing is performed by a diagnosis model created by extracting state feature values using normal operation data in the early stage of manufacture of the machine tool and abnormal operation data in accumulated failure cases. In addition, operation data during normal operation is divided into machining operation data and non-machining operation data, each of which has a diagnostic model. The operation data from the operation during machining will differ greatly if the machining details such as the tool used for machining, the material of the workpiece, and the machining conditions are different. During machining, there are many similar operations that do not depend on the machine, such as rapid traverse of the feed axis and acceleration/deceleration of the spindle, making it easy to operate a diagnostic model across machines and models.

The second diagnosis unit 22 diagnoses by performing a specific diagnostic operation that is effective for identifying the location of the abnormality. For example, in order to diagnose bearing damage of the feed shaft, a full stroke operation is performed at a specific feed speed, and in order to diagnose bearing damage of the main shaft, an idle rotation operation is performed at a specific feed speed.
The diagnostic process updating unit 23 receives the diagnostic result of the second diagnostic unit 22 and updates and re-learns the diagnostic model used for diagnosis in the first diagnostic unit 21 .
FIG. 2 shows an example of diagnostic setting by the diagnostic processing updating unit 23. As shown in FIG. In the first diagnosis unit 21, as shown in FIG. 2A, the types of normal and abnormal parts (here, two categories of abnormality A and abnormality B) are classified, and classified into any frame. Suppose that a new diagnostic point is detected. At that time, diagnosis is performed in the second diagnosis unit 22 . When the second diagnostic unit 22 diagnoses the abnormality A, as shown in FIG. to update. If the second diagnosis unit 22 does not diagnose any abnormality, the first diagnosis unit 21 classifies the new diagnosis point as normal, as shown in FIG. 2(C). Update diagnostic setting range.
In addition, when multiple machines are connected via a network, a diagnostic model can be shared.

FIG. 3 shows a flow chart of a diagnostic method according to the second configuration executed by the diagnostic device 20 .
First, the first diagnostic unit 21 determines whether or not the system is in normal operation (S0). Here, it is determined that the program is operating normally, such as when the machining program is running, and diagnosis is not performed when the operator is manually operating, such as when the override is being changed.
If it is in normal operation, it is next determined whether or not it is in machining operation (S1). Here, an operation during which the spindle is not rotating is determined as a non-machining operation, and an operation during which the spindle is rotating is determined as a machining operation.
If it is determined in S1 that machining is in progress, a diagnostic determination process for machining is performed in S2, and it is determined in S3 whether or not it is normal. If it is judged to be normal here, the fact that it is normal is displayed on the monitor or the like of the NC unit 10 in S4. However, it is not necessary to display anything in particular as a normal state.
On the other hand, if it is determined that there is an abnormality in S3, it is determined in S5 whether or not an abnormal portion has been identified. If the abnormal part is identified as one of the types classified in advance as shown in FIG. do.

Even if it is determined to be abnormal in the determination of S3, if the abnormal site is not specified in the determination of S5, the relationship with the abnormal site is calculated in S7. For example, as shown in FIG. 2A, if a new diagnosis point that does not correspond to either abnormality A or B is detected, the first diagnosis unit 21 determines the centroid position of each cluster of normal and abnormal site types. Calculate the order of relevance to the abnormal site based on the distance from. Further, if the diagnosis result of the first diagnosis unit 21 indicates that the nearby abnormal portion is related to the feed shaft, the second diagnosis unit 22 does not perform the diagnosis of the abnormal portion related to the parts of the spindle. The relevance may be determined based on the abnormal site in unit units, such as by setting as follows. The processing from S1 to S7 excluding S4 and S6 is the first diagnostic step of the present disclosure.

When the relationship with the abnormal site is calculated in S7, the diagnostic operation is performed in the second diagnostic unit 22 in S8. Since this diagnosis is performed with a diagnostic operation suitable for the abnormal site to be identified, it is performed in descending order based on the order of relevance to the abnormal site calculated by the first diagnosis unit 21. Then it is efficient. The second diagnostic unit 22 interrupts the operation for diagnosis at the end of the machining program. The diagnostic operation is performed by, for example, issuing an instruction to perform the
After performing the diagnostic operation in S8, it is determined in S9 whether or not an abnormal portion has been identified. If an abnormal portion is identified here, in S10, the monitor or the like of the NC unit 10 displays that an abnormality has occurred in the identified abnormal portion. Update the diagnosis judgment process in That is, as shown in FIG. 2B, the setting range of diagnosis is updated so that the new diagnosis point is included in the abnormality category to which the abnormality part belongs.

On the other hand, if the abnormal part could not be specified in S9, the monitor of the NC unit 10 or the like could not identify the abnormal part in S12. is displayed, and in S11, the diagnosis determination process during machining in the first diagnosis unit 21 is updated. That is, as shown in FIG. 2C, the diagnostic setting range is updated so that the new diagnostic point is included in the normal category. The processing from S8 to S12 excluding S11 is the second diagnostic step of the present disclosure, and S11 is the diagnostic processing update step of the present disclosure. .
Even if it is determined in S1 that processing is not being performed, normal and abnormal portions can be specified by executing the processing of S22 to S32 in the same flow as the diagnosis during processing. Also in this case, the processing from S22 to S27 excluding S24 and S26 is the first diagnostic step of the present disclosure, and the processing from S28 to S32 excluding S31 is the second diagnostic step of the present disclosure. Then, S31 is the diagnostic processing update step of the present disclosure.

In this manner, the diagnostic device 20 of the above embodiment uses the operation data acquired during the normal operation of the machine tool 1 to classify the machine tool 1 into pre-classified normals and a plurality of types of abnormalities including abnormal parts. a first diagnosis unit 21 for diagnosing whether the In addition, if the first diagnostic unit 21 diagnoses that the machine tool 1 is not normal but the abnormal part cannot be specified, the diagnostic device 20 causes the machine tool 1 to perform a predetermined diagnostic operation, and stores the acquired diagnostic operation data. A second diagnostic unit 22 is provided for identifying an abnormal site using the diagnosing unit. The diagnostic apparatus 20 includes a diagnostic processing updating section 23 that updates the classification of the normal and/or abnormal parts in the first diagnostic section 21 based on the diagnostic result of the second diagnostic section 22 . The diagnostic device 20 executes the diagnostic method shown in FIG. 3 based on the diagnostic program.
According to this configuration, the diagnosis using the operation data and the diagnosis using the operation data for diagnosis are used in combination, and the operation data for diagnosis is used only when the abnormal part cannot be specified by the diagnosis result based on the operation data. Will run diagnostics. Therefore, it is possible to improve diagnostic accuracy while suppressing the occurrence of time loss by minimizing the execution of diagnostic operations. Therefore, it is particularly useful in machinery and equipment for manufacturing a wide variety of products.

It should be noted that the types of abnormality are not limited to the two categories described above. You may set three or more categories.
In the above embodiment, the abnormal part is identified as the type of abnormality, but in addition to this, the degree of abnormality, the cause of abnormality, etc. can also be identified.
The present disclosure is agnostic of the type of machine tool. Also, this disclosure is not directed to only one machine tool. As described above, when a plurality of machine tools are connected via a network, it is possible to diagnose a plurality of machine tools by sharing a diagnosis model. Accordingly, the diagnostic device may be installed at a location separate from the one or more machine tools.

DESCRIPTION OF SYMBOLS 1... Machine tool, 10... NC unit, 11... Program interpretation part, 12... Machine operation command part, 20... Diagnosis device, 21... First diagnosis part, 22... Second diagnosis part , 23 .

Claims (9)

A diagnostic device for diagnosing an abnormality in a machine tool,
a first diagnosing unit that diagnoses whether the machine tool is normal or one of a plurality of abnormal types classified in advance using operation data acquired during normal operation of the machine tool;
When the type of the abnormality cannot be specified even if the first diagnosis unit diagnoses that the abnormality is not normal, the machine tool is caused to perform a predetermined diagnostic operation, and the acquired operation data for diagnosis is used to identify the abnormality. a second diagnostic unit that identifies the type;
a diagnostic processing updating unit that updates the classification of the normal and/or the abnormal type in the first diagnostic unit based on the diagnostic result of the second diagnostic unit;
A diagnostic device for a machine tool, comprising: When neither the diagnosis of normality nor the identification of the type of abnormality can be performed, the first diagnosis unit calculates the ranking of the relevance between the diagnosis result and each type of abnormality,
2. The machine tool diagnostic apparatus according to claim 1, wherein the second diagnostic unit identifies the type of the abnormality based on the calculated order. The diagnostic process updating unit updates a set range of the specified type of abnormality in the first diagnostic unit when the type of abnormality is identified by the second diagnostic unit. The machine tool diagnosis device according to claim 1 or 2. 2. The diagnostic process updating unit updates the normal set range in the first diagnostic unit when the type of the abnormality is not specified by the second diagnostic unit. 4. The machine tool diagnosis device according to any one of 1 to 3. A diagnostic method for diagnosing an abnormality in a machine tool, comprising:
a first diagnosis step of diagnosing whether the machine tool is normal or one of a plurality of abnormal types classified in advance using operation data acquired during normal operation of the machine tool;
If the type of the abnormality cannot be specified even if the first diagnosis step diagnoses that the machine tool is not normal, the machine tool is caused to perform a predetermined diagnostic operation, and the acquired operation data for diagnosis is used to identify the abnormality. a second diagnostic step of type identification;
a diagnostic processing update step of updating the classification of the normal and/or abnormal type in the first diagnostic step based on the diagnostic result of the second diagnostic step;
A diagnostic method for a machine tool, characterized by: In the first diagnosis step, if neither the diagnosis of normality nor the identification of the type of abnormality can be executed, the ranking of the relevance between the diagnosis result and each type of abnormality is calculated;
6. The method of diagnosing a machine tool according to claim 5, wherein, in said second diagnosis step, the type of said abnormality is specified based on said calculated order. In the diagnostic processing update step, when the type of abnormality is specified in the second diagnostic step, the setting range of the specified type of abnormality in the first diagnostic step is updated. The machine tool diagnosis method according to claim 5 or 6. 5. The diagnostic process update step updates the normal set range in the first diagnostic step when the type of abnormality is not specified in the second diagnostic step. 8. The method for diagnosing a machine tool according to any one of 7. A machine tool diagnostic program for causing a machine tool control device to execute the machine tool diagnostic method according to any one of claims 5 to 8.

Images