JP3780299B1 - Diagnostic method for target equipment, computer program, and apparatus for diagnosing target equipment - Google Patents

Abstract

A diagnostic method of a target facility, a computer program, and an apparatus for diagnosing the target facility are provided that can more accurately diagnose the target facility.
A step of dividing the acquired waveform data into T pieces of divided waveform data, a step of Fourier transforming each of the T pieces of divided waveform data to obtain T frequency spectra, and the T pieces of waveform data And calculating the principal component score based on the intensity of the frequency spectrum obtained for each of the divided frequency bands. And calculating in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal as the eigenvector used when determining the principal component score. Using a reference eigenvector that has been placed.
[Selection] Figure 1

Description

  The present invention relates to a diagnosis method for a target facility, a computer program, and an apparatus for diagnosing the target facility.

Various methods are known for acquiring waveform data (vibration data, etc.) of a target facility and diagnosing the target facility based on the acquired waveform data.
As an example of such a method, there is a method of determining whether the equipment is normal by observing disturbance of the acquired waveform data. In this method, for example, a threshold is set for the amplitude of the waveform data, and when the amplitude of the acquired waveform data exceeds the threshold, it is determined that the equipment has shifted to an abnormal state.
However, this method includes the following problems. In other words, depending on the target equipment, even if the equipment shifts to an abnormal state, the waveform data may not be noticeably disturbed, and the measures using the thresholds described above properly perform equipment diagnosis. It may not be done.

As a method for solving such a problem, Japanese Patent No. 3382240 discloses a method by principal component analysis for obtaining a principal component score based on acquired waveform data. In this method, when the operation of the facility is normal, for example, the eigenvector and the principal component score used for obtaining the principal component score are obtained based on the waveform data acquired at the initial operation of the facility. When calculating the principal component score based on the waveform data acquired during the subsequent operation of the facility, for example, during the middle or late operation of the facility, based on the waveform data acquired during the initial operation of the facility. The obtained eigenvector is used. Then, by comparing the principal component score obtained based on the waveform data obtained during the subsequent operation of the equipment with the principal component score obtained based on the waveform data obtained at the initial operation time, Even if there is no significant disturbance in the data, it can be determined whether or not the equipment has shifted to an abnormal state.
Japanese Patent No. 3382240

  By the way, as the characteristics of the target equipment, as long as the operation of the equipment is normal, the equipment having the characteristics that the waveform data does not change so much from the initial operation to the late operation, and the operation of the equipment is normal. Even so, there is equipment having a characteristic that the waveform data can be remarkably changed from the initial operation time to the late operation time. For example, the former can be a pump system in which the impeller of the pump keeps rotating at a constant speed, and the latter can be a press working system for pressing a material such as metal.

  Here, to supplement the latter example, in the press working system, the material is molded so that the material has a desired shape using a mold composed of a fixed mold part and a movable mold part. Even if molding is performed normally (for example, even if no cracks or cracks occur in the material), from the initial stage of molding (at the beginning of movement of the movable mold) to the late stage of molding (at the latter stage of movement of the movable mold) ) Until the vibration of the mold, in other words, the waveform data (vibration data) acquired by the sensor provided in the mold changes significantly.

  And when adopting the method disclosed in Japanese Patent No. 3382240 for diagnosing equipment having the characteristics according to the former, the main component score obtained based on the waveform data acquired at the initial stage of operation, In comparison, whether or not the equipment is in an abnormal state or a normal state can be determined depending on whether or not the principal component score obtained based on the waveform data acquired during the middle stage of operation or the late stage of operation is significantly different (that is, If there is a significant difference, it is determined that the equipment is in an abnormal state, and if there is no difference, it is determined that the equipment is in a normal state), so that the equipment diagnosis is appropriately executed.

  However, when adopting the method disclosed in Japanese Patent No. 3382240 for diagnosing equipment having the characteristics related to the latter, even if the equipment is operating normally, the operation is performed from the initial operation to the late operation. Compared with the principal component score obtained based on the waveform data acquired at the initial stage of operation, even though the operation of the equipment is normal at the middle stage or late stage of operation, since the waveform data can change significantly. There is a possibility that the principal component scores obtained based on the waveform data acquired at the middle stage of the operation or at the latter stage of the operation are significantly different. Therefore, when the method is used for diagnosing equipment having the characteristics related to the latter, even if the equipment is diagnosed, its accuracy is insufficient.

  The present invention has been made in view of such a problem, and the object of the present invention is to diagnose a target facility diagnosis method, a computer program, and a target facility in which the target facility is diagnosed more accurately. It is in realizing the apparatus for.

  The main aspect of the present invention is a step of acquiring waveform data of a target facility in a predetermined period of at least a part of an operation period from the start of operation of the facility to the end of operation, and the acquired in the predetermined period Dividing the waveform data into T pieces of divided waveform data; Fourier transforming each of the T pieces of divided waveform data to obtain T frequency spectra; and each of the T frequency spectra. A principal component score is obtained for each of the T frequency spectra based on the step of obtaining the strength for each of the divided frequency bands divided into P pieces and the intensity of the frequency spectrum obtained for each of the divided frequency bands. A wave acquired during the predetermined period when the operation of the facility is normal as an eigenvector used in obtaining the principal component score. Data, a method of diagnosing a subject facility and having the steps of: using the reference eigenvectors, obtained in advance based on.

  Other features of the present invention will become apparent from the description of the present specification and the accompanying drawings.

  At least the following will be made clear by the description of the present specification and the accompanying drawings.

The waveform data of the target equipment is acquired in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation, and the waveform data acquired in the predetermined period of time, Dividing into T pieces of divided waveform data; obtaining a T frequency spectrum by Fourier-transforming each of the T pieces of divided waveform data; and calculating the intensity of each of the T frequency spectra to P pieces. Determining for each divided frequency band, and determining a principal component score for each of the T frequency spectra based on the strength of the frequency spectrum determined for each of the divided frequency bands, Based on the waveform data acquired during the predetermined period when the operation of the facility is normal as the eigenvector used when obtaining the principal component score. Method of diagnosing a subject amenities and a step of using the reference eigenvectors, obtained in advance, the Te.
According to the method for diagnosing a target facility, the target facility can be diagnosed more accurately.

A step of comparing the obtained principal component score with a reference principal component score obtained in advance based on the waveform data obtained during the predetermined period when the operation of the equipment is normal; It is good also as having.
In such a case, the main component score is compared with the reference main component score, so that the diagnosis of the target facility is performed more appropriately.

Further, in the step of acquiring the waveform data of the target equipment in the predetermined period, in the step of acquiring a plurality of the waveform data in a plurality of the predetermined periods and obtaining the principal component score, a plurality of The principal component score may be obtained based on each of the waveform data.
In such a case, it is possible to reliably grasp whether an abnormality has occurred in the target equipment.

Next, obtaining waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation, and the waveform acquired in the predetermined period Dividing the data into T pieces of divided waveform data; Fourier transforming each of the T pieces of divided waveform data to obtain T frequency spectra; and the strength of each of the T frequency spectra. For each of the divided frequency bands divided into P, and the frequency spectrum strength corresponding to one of the frequency spectrum strengths obtained for each of the divided frequency bands and the other divided frequencies A step of obtaining a correlation coefficient with the intensity of the frequency spectrum corresponding to the band, and the obtained correlation coefficient when the operation of the facility is normal. Method of diagnosing a subject facility and having a reference correlation coefficient obtained in advance based period the acquired waveform data, in the step of comparing, the.
According to the method for diagnosing a target facility, the target facility can be diagnosed more accurately.

Further, in the step of acquiring the waveform data of the target equipment in the predetermined period, in the step of acquiring a plurality of the waveform data in a plurality of the predetermined periods and obtaining the correlation coefficient, a plurality of The correlation coefficient may be obtained based on each of the waveform data.
In such a case, it is possible to reliably grasp whether an abnormality has occurred in the target equipment.

In addition, a plurality of the predetermined periods may be selected such that the end of one of the plurality of predetermined periods coincides with the start of the other predetermined period.
In such a case, the time required for diagnosis is shortened.

In addition, a plurality of the predetermined periods may be selected so that a part of one of the plurality of predetermined periods overlaps a part of the other predetermined period.
In such a case, it is possible to prevent inconvenience that the diagnosis is not properly performed when an abnormality occurs in the target equipment at the boundary of each predetermined period.

Next, as the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and the second Obtaining waveform data in a second predetermined period which is a part of the second predetermined period of the operation period, wherein a start time of the second predetermined period coincides with an end of the first predetermined period; Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data; the T first divided waveform data and the T pieces of divided waveform data; Fourier transforming each of the second divided waveform data to obtain T first frequency spectra and T second frequency spectra, and setting the intensity of each of the T first frequency spectra to P Divided division frequency And calculating the intensity of each of the T second frequency spectra for each of the divided frequency bands divided into P pieces, and the first frequency spectrum obtained for each of the divided frequency bands. The step of obtaining eigenvectors used when obtaining the principal component score based on the strength and the T principal component scores based on the strength of the second frequency spectrum obtained for each of the divided frequency bands. Using the eigenvector obtained based on the strength of the first frequency spectrum as the eigenvector used when obtaining the principal component score. A diagnostic method for target equipment characterized by
According to the method for diagnosing a target facility, the target facility can be diagnosed more accurately.

Further, in the step of obtaining the eigenvector, a principal component score is obtained for each of the T first frequency spectra together with the eigenvector based on the strength of the first frequency spectrum, and the strength of the second frequency spectrum is obtained. The principal component score obtained in the step of obtaining the principal component score based on the step may be compared with the principal component score obtained based on the strength of the first frequency spectrum.
In such a case, the principal component score obtained on the basis of the strength of the second frequency spectrum is compared with the principal component score obtained on the basis of the strength of the first frequency spectrum. Is more appropriately performed.

Next, as the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and the second Obtaining waveform data in a second predetermined period which is a part of the second predetermined period of the operation period, wherein a start time of the second predetermined period coincides with an end of the first predetermined period; Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data; the T first divided waveform data and the T pieces of divided waveform data; Fourier transforming each of the second divided waveform data to obtain T first frequency spectra and T second frequency spectra, and setting the intensity of each of the T first frequency spectra to P Divided division frequency And calculating the strength of each of the T second frequency spectra for each of the divided frequency bands divided into P pieces, and the strength of the first frequency spectrum obtained for each of the divided frequency bands. Determining a first correlation coefficient between the strength of the first frequency spectrum corresponding to one of the divided frequency bands and the strength of the first frequency spectrum corresponding to the other divided frequency band, and for each of the divided frequency bands The second correlation coefficient between the strength of the second frequency spectrum corresponding to one divided frequency band and the strength of the second frequency spectrum corresponding to another divided frequency band among the strengths of the second frequency spectrum obtained in And a method of comparing the second correlation coefficient with the first correlation coefficient.
According to the method for diagnosing a target facility, the target facility can be diagnosed more accurately.

Further, as the waveform data of the target equipment, the first waveform data is acquired in a first predetermined period of a part of the operation period from the start of operation of the equipment to the end of operation, and the second waveform Instead of acquiring data in a second predetermined period of a part of the operating period, wherein the second predetermined period coincides with the end of the first predetermined period. As the waveform data of the target equipment, the first waveform data is acquired in a first predetermined period of a part of the operation period from the start of operation of the equipment to the end of operation, and the second waveform The data is a second predetermined period that comes after the first predetermined period of a part of the operation period, and a part of the second predetermined period overlaps a part of the first predetermined period. (Ii) acquiring in a predetermined period; It is also possible to be.
In such a case, the target facility is diagnosed more accurately.

Further, the target facility may be a processing system for processing an object.
In such a case, the effect described above, that is, the effect that the diagnosis of the target facility is performed more accurately is more effectively exhibited.

  The waveform data may be vibration data.

Next, obtaining waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation, and the waveform acquired in the predetermined period Dividing the data into T pieces of divided waveform data; Fourier transforming each of the T pieces of divided waveform data to obtain T frequency spectra; and the strength of each of the T frequency spectra. For each of the divided frequency bands divided into P, and a step of obtaining a principal component score for each of the T frequency spectra based on the strength of the frequency spectrum obtained for each of the divided frequency bands. As eigenvectors used when obtaining the principal component score, waveform data acquired during the predetermined period when the operation of the facility is normal, A step of using a reference eigenvector obtained in advance, and the obtained principal component score is obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal. And a step of comparing with the reference principal component score, and in the step of acquiring the waveform data of the target equipment in the predetermined period, a plurality of the waveform data are acquired in a plurality of the predetermined periods. In the step of obtaining the principal component score, a principal component score is obtained based on each of the plurality of waveform data, and the end of one of the plurality of predetermined periods is the other A plurality of the predetermined periods are selected to coincide with the start of the predetermined period, and the target equipment is a processing system for processing an object, and the wave Data diagnostic methods of the subject equipment, which is a vibration data.
In this way, the object of the present invention can be achieved more effectively because most of the effects described above can be achieved.

Next, as the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and the second Obtaining waveform data in a second predetermined period which is a part of the second predetermined period of the operation period, wherein a start time of the second predetermined period coincides with an end of the first predetermined period; Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data; the T first divided waveform data and the T pieces of divided waveform data; Fourier transforming each of the second divided waveform data to obtain T first frequency spectra and T second frequency spectra, and setting the intensity of each of the T first frequency spectra to P Divided division frequency And calculating the intensity of each of the T second frequency spectra for each of the divided frequency bands divided into P pieces, and the first frequency spectrum obtained for each of the divided frequency bands. The step of obtaining eigenvectors used when obtaining the principal component score based on the strength and the T principal component scores based on the strength of the second frequency spectrum obtained for each of the divided frequency bands. Using each eigenvector obtained based on the strength of the first frequency spectrum as an eigenvector used when obtaining the principal component score. In the step of obtaining the eigenvector, the T principal component scores are set together with the eigenvector based on the strength of the first frequency spectrum. The principal component obtained for each first frequency spectrum and the principal component score obtained in the step of obtaining the principal component score based on the strength of the second frequency spectrum is determined based on the strength of the first frequency spectrum. A method for diagnosing the target facility, wherein the target facility is a processing system for processing an object, and the waveform data is vibration data.
In this way, the object of the present invention can be achieved more effectively because most of the effects described above can be achieved.

A computer program for realizing the above-described target facility diagnosis method can also be realized.
According to such a computer program, the diagnosis of the target equipment can be performed more accurately.

Further, it is possible to realize an apparatus for diagnosing a target facility, which includes a sensor for acquiring waveform data of a target facility and a computer including the above-described computer program.
According to the apparatus for diagnosing the target facility, the target facility is diagnosed more accurately.

=== First Embodiment According to the Diagnosis Method of Target Equipment ===
First, a first embodiment according to a diagnosis method for target equipment will be described with reference to FIGS. 1 to 8B. FIG. 1 is a flowchart showing a first embodiment according to a method for diagnosing a target facility. FIG. 2 is an explanatory diagram for explaining a predetermined period related to acquisition of vibration data. FIG. 3 is an explanatory diagram for explaining the division of vibration data into divided vibration data. 4 to 8B will be described later.

  In this section, as a target facility, a processing system for processing an object described in the section of the problem to be solved by the invention (in this specification, the processing system is an object, that is, a workpiece to be processed). For example, a press working system for forming a material into a desired shape by using a mold including a fixed mold part and a movable mold part. , Will be described as an example.

  In this flowchart, vibration data of the press processing system is acquired as waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of the press processing system to the end of operation. Start (step S2). In this example, vibration data is acquired by a sensor provided in the mold.

  Further, in the present embodiment, a plurality (N) of vibration data is acquired in the plurality (N) of the predetermined period. This will be specifically described with reference to FIG. In FIG. 2, the operation period of the press working system is shown on the time axis, the left end of the time axis is when the press working system starts to operate, and the right end of the time axis is the time when the press working system operates. It is time to finish. Then, as shown in FIG. 2, a period (hereinafter, also referred to as an analysis time zone) for executing diagnosis of the press working system is determined from the operation period (for example, the movable mold is made of a material as the analysis time zone). The period from the contact to the end of the molding of the material is selected), and the above-mentioned plural (N) predetermined periods are selected from the determined analysis time zone. Here, in the present embodiment, the analysis time zone is divided into N, and each of the divided time zones is set as the predetermined period. That is, the plurality of predetermined periods are selected such that the end of one predetermined period among the plurality of predetermined periods coincides with the start of another predetermined period. A plurality (N) of vibration data is acquired in a plurality (N) of predetermined periods (that is, the first time zone, the second time zone, the third time zone,... N time zone). Is done.

  Next, as shown in FIG. 3, the vibration data acquired in the first time zone (FIG. 2) is divided into divided vibration data as an example of T pieces of divided waveform data (division in the time axis direction, step S4). ). Then, each of the T divided vibration data is Fourier transformed to obtain T frequency spectra (step S6). Further, the strength of each of the T frequency spectra is obtained for each of the divided frequency bands divided into P (division in the frequency axis direction, step S8).

  FIG. 4 and FIG. 5 are explanatory diagrams and tables showing how the vibration data acquired in the first time zone is divided in the time axis direction and the frequency axis direction. As is apparent from these figures, the strengths of T × P frequency spectra are obtained based on the vibration data acquired in the first time zone by the procedure from step S4 to step S8. Note that the strength of each of the T × P frequency spectrums is Xij (i = 1,... T, j = 1,... P), and the center frequency of each of the P divided frequency bands is , Fj (j = 1,... P).

Next, based on the frequency spectrum strength Xij obtained for each divided frequency band, a principal component score is obtained for each of the T frequency spectra (step S10).
In this case, first, the average (Ave) and variance (Var) of the intensity X11,..., X1p of the frequency spectrum are obtained, and the frequency normalized by the calculation formula: X′1j = (X1j−Ave) / Var The spectrum intensity X′11,... X′1p is obtained. Then, this procedure is performed for X21,..., X2p, X31,... X3p,..., Xt1,. Get ij.

If X′ij is represented by a vector and the vector is X ′, the principal component score Z is calculated by a calculation formula: Z = A (eigenvector) · X ′. Note that when Z (vector) = (Z1, Z2,... Zm) and X ′ (vector) = (X′1, X′2,... X′p), the calculation formula is written down. It becomes as follows.
First principal component score Z1 = a11 · X′1 + a12 · X′2... + A1p · X′p
Second principal component score Z2 = a21 * X'1 + a22 * X'2 ... + a2p * X'p
...
M-th principal component score Zm = am1 · X′1 + am2 · X′2... + Amp · X′p
In this way, the principal component score Z is calculated for every T frequency spectra.

  The principal component score Z is calculated by using the eigenvector A as described above, but in the present embodiment, the operation of the press working system is used as the eigenvector A used when obtaining the principal component score Z. Is normal, a reference eigenvector Ab previously obtained based on vibration data (also referred to as reference vibration data) acquired during the predetermined period, that is, the first time period is used.

  The above will be described more specifically. The eigenvector A is calculated from an eigen equation CA = λA including a correlation matrix C obtained from the intensity Xij of the frequency spectrum of the acquired vibration data. Therefore, it is possible to obtain an eigenvector (referred to as a diagnostic eigenvector Ad in order to distinguish the eigenvector from the reference eigenvector Ab described above) based on the vibration data acquired in the first time zone in step S2. . However, in the present embodiment, the reference that has been obtained in advance based on the reference vibration data acquired in the first time zone when the operation of the press working system is normal without obtaining the diagnostic eigenvector Ad. The eigenvector Ab is used when obtaining the principal component score Z (that is, the principal component score Z is calculated not by the calculation formula: Z = Ad · X ′ but by the calculation formula: Z = Ab · X ′). .

  The reference eigenvector Ab is obtained in advance based on the vibration data acquired when the operation of the breath processing system is normal (in other words, after the vibration data is acquired, the operation of the breath processing system is normal). The reference eigenvector Ab is determined based on the acquired vibration data), but the vibration data is always acquired in the first time zone described above. For example, when the first time zone is a time zone that starts after t1 seconds from the start of operation of the press working system and ends after t2 seconds, based on the vibration data acquired in the time zone, Eigenvector Ab is obtained. Further, when the reference eigenvector Ab is obtained, the reference principal component score Z is obtained together with the frequency spectrum strength Xij obtained based on the reference vibration data and the reference eigenvector Ab.

  Next, the principal component score obtained in step S10 (hereinafter, the principal component score is also referred to as a diagnostic principal component score Zd) is represented by the reference principal component score (hereinafter, the reference principal component score is represented by the symbol Zb). ) And a comparison (step S12). An example of the comparison method will be described below.

  As described above, the diagnostic principal component score Zd is calculated for each of the T frequency spectra in step S10. That is, T diagnostic principal component scores Zd are obtained. Also, T reference principal component scores Zb are obtained for every T frequency spectra.

  In this situation, the DI value as an index of the degree of difference between the two is obtained from the diagnostic principal component score Zd and the reference principal component score Zb. The DI value is the absolute value of the difference between the average value of the T diagnostic principal component scores Zd and the average value of the T reference principal component scores Zd, and the variance of the T diagnostic principal component scores Zd and the T values. Is calculated by dividing (dividing) by the square root of the sum of the variance of the reference principal component score Zd. Then, by comparing the calculated DI value with a preset threshold value, it is possible to determine whether or not an abnormality has occurred in the press working system in the first time zone. That is, if the DI value> threshold, an abnormality occurs in the press processing system in the first time zone (for example, cracks or cracks occur in the material), and if the DI value <threshold, press in the first time zone. It is determined that no abnormality has occurred in the machining system.

  As described above, there are a plurality of principal component scores, such as a first principal component score and a second principal component score, and the procedure is performed for one of them (for example, the first principal component score). Or the above procedure may be performed for a plurality (from the first principal component score to the m-th principal component score). In the above description, the DI value is used as an index of the degree of difference between the diagnostic principal component score Zd and the reference principal component score Zb. However, the present invention is not limited to this. May be used.

  Next, for each of the vibration data acquired in the second time zone, the vibration data acquired in the third time zone,... The procedure up to S12 is performed (step S14). Thus, the diagnostic principal component score Zd is obtained based on each of the plurality of vibration data, and the obtained diagnostic principal component score Zd is compared with the reference principal component score Zb.

  FIG. 6 shows the diagnosis results when the diagnosis method according to the present embodiment is actually applied to the diagnosis of the press working system. In this diagnosis, the diagnosis is performed 16 times (more specifically, the press working system is operated 16 times, and the DI value is obtained for each of the 16 operations). The data (line graph) is displayed in the figure. In the figure, the horizontal axis is from the first time zone to the n-th time zone (n = 69 in the figure), and the above-mentioned DI value is on the vertical axis. In this diagnosis, a plurality of principal component scores (from the first principal component score to the m-th principal component score) are calculated, and DI values are obtained for each, so that a plurality of DI values per one time zone are calculated. However, the maximum DI value among these DI values is extracted for each time period, and the extracted DI value is displayed in the figure. In this diagnosis, the threshold value of the DI value is set to 3.

  As is clear from FIG. 6, in this diagnosis, a result that the DI value exceeded the threshold value was obtained only in the eighth diagnosis out of the 16 diagnoses. Since the DI value exceeds the threshold in the 64th time zone, it can be seen that an abnormality has occurred in the press working system in the time zone of the eighth diagnosis.

  Further, when the molded product obtained by press working was visually observed after completion of the diagnosis, only the molded product produced by the eighth press working was found to have cracks (the other molded products were found to have cracks, etc. Not) This shows that the diagnostic method according to the present embodiment is effective.

  As described in the section of the problem to be solved by the invention, equipment having characteristics that can cause a significant change in vibration data from the initial operation to the late operation even if the operation of the equipment is normal (for example, When the conventional method (the method disclosed in Japanese Patent No. 3382240) is used for the diagnosis of the press working system described above, even if the operation of the equipment is normal, the operation is started from the initial operation to the late operation. Since the vibration data can change significantly between times, the main component score obtained based on the vibration data acquired at the initial stage of operation, despite the normal operation of the equipment at the middle stage and late stage of operation. In comparison, there is a possibility that the principal component scores obtained based on the vibration data acquired at the middle stage of the operation or at the latter stage of the operation are significantly different. Therefore, when a conventional method is used for diagnosing equipment having the characteristics, even if the equipment is diagnosed, the accuracy thereof is insufficient.

  On the other hand, when the method of the present case is adopted for diagnosis of equipment having the characteristics, the main component score Zd at diagnosis is based on vibration data acquired in a predetermined period of the operation period of the equipment. Is obtained using the reference eigenvector Ab obtained based on the reference vibration data acquired in the same period as the predetermined period, and the diagnostic principal component score Zd is obtained based on the reference vibration data. Compared with the reference principal component score Zb, when the operation of the equipment is abnormal during the predetermined period, the diagnosis principal component score Zd is conspicuous compared with the reference principal component score Zb. In contrast, if the operation of the equipment is normal during the predetermined period, the diagnostic principal component score Zd is not so different from the reference principal component score Zb. Therefore, when adopting the method of the present case, the diagnosis of the target equipment is performed more accurately.

In the above, for each of vibration data acquired in the second time zone, vibration data acquired in the third time zone,... Vibration data acquired in the nth time zone in step S14. The procedure from step S4 to step S12 is executed, but step 14 may be omitted. That is, in the above, in the step of acquiring the vibration data of the press working system in the predetermined period, in the step of acquiring a plurality of vibration data in the plurality of predetermined periods and obtaining the principal component score, a plurality of The principal component score is determined based on each of the vibration data of, but is not limited to this, and one of the first time period to the n-th time period is set as the predetermined time and is only once. The procedure from step S4 to step S12 may be performed.
However, the above-described embodiment is more preferable in that it is possible to reliably grasp whether or not an abnormality has occurred in the press working system.

<<< Modification of First Embodiment According to Target Equipment Diagnosis Method >>>
Here, the modification of 1st embodiment which concerns on the diagnostic method of object installation is demonstrated using FIG. 7 thru | or FIG. 8B. FIG. 7 is a flowchart illustrating a modification of the first embodiment according to the target facility diagnosis method. 8A and 8B will be described later.

  As shown in FIG. 7, when this modification is compared with the first embodiment described above, the procedure up to step S8 described above is the same for both. That is, in this modified example, first, the same procedure as the above-described step S2 to step S8 is performed (step S102 to step S108), and T × P pieces are obtained based on the vibration data acquired in the first time zone. The frequency spectrum strength Xij (i = 1,... T, j = 1,... P) is obtained.

  Next, in this modification, instead of calculating the principal component score, the frequency spectrum strength corresponding to one divided frequency band among the frequency spectrum strengths Xij obtained for each divided frequency band and the other divisions. A correlation coefficient with the intensity of the frequency spectrum corresponding to the frequency band is obtained (step S110).

  The method for obtaining the correlation coefficient will be specifically described. First, frequency spectrum strengths X11, X21,... Xt1 corresponding to the first divided frequency band (center frequency f1) and frequency spectrum strength X12 corresponding to the second divided frequency band (center frequency f2), From X22,..., Xt2, a correlation coefficient R12 (both covariances divided by the respective standard deviations) is obtained. Similarly, the intensity of the frequency spectrum corresponding to the first division frequency band (center frequency f1), the third division frequency band (center frequency f3), the fourth division frequency band (center frequency f4),. From the intensity of the frequency spectrum corresponding to the frequency band (center frequency fp), correlation coefficients R13, R14,... R1p are obtained, and further, the second divided frequency band (center frequency f2), the third divided frequency band ( Center frequency f3),... The same calculation is performed with reference to the (p-1) -th divided frequency band (center frequency fp-1), and correlation coefficients R23,... R2p, R34. .., R (p-1) p is obtained. Finally, P · (P−1) / 2 correlation coefficients are calculated.

  Next, the obtained correlation function (hereinafter, the correlation function is also referred to as a diagnostic correlation function Rd) is set to the predetermined period, that is, the first time zone when the operation of the press working system is normal. Based on the acquired vibration data (also referred to as reference vibration data), a comparison is made with a reference correlation coefficient Rb obtained in advance using a procedure similar to the above procedure (step S112). An example of the comparison method will be described below.

As described above, P · (P−1) / 2 diagnosis correlation coefficients Rd are calculated in step S110. Also, P · (P−1) / 2 pieces of reference correlation coefficients Rd are obtained.
In such a situation, a difference (that is, Rd−Rb) as an index of the degree of difference between the two is obtained from the correlation function Rd during diagnosis and the reference correlation coefficient Rb. The difference is calculated for each of P · (P−1) / 2 correlation coefficients (that is, Rd12−Rb12, Rd13−Rb13,...). Then, by comparing the calculated difference with a preset threshold value, it is possible to determine whether or not an abnormality has occurred in the press working system in the first time zone. That is, if the difference> threshold, an abnormality occurs in the press processing system in the first time zone (for example, a crack or a crack occurs in the material), and if the difference <threshold, the press processing system in the first time zone. It is determined that no abnormality has occurred. In the above, the difference is used as an index of the degree of difference between the correlation function Rd at diagnosis and the reference correlation coefficient Rb. However, the present invention is not limited to this. A correlation coefficient may be used.

  Next, for each of vibration data acquired in the second time zone, vibration data acquired in the third time zone,... The procedure up to S112 is performed (step S114). Thus, the diagnostic correlation function Rd is obtained based on each of the plurality of vibration data, and the obtained diagnostic correlation function Rd is compared with the reference correlation coefficient Rb.

  FIG. 8A and FIG. 8B show the diagnostic results when the diagnostic method according to this modification is actually applied to the diagnosis of the press working system. In this diagnosis, the diagnosis is performed twice (more specifically, the press working system is operated twice and the correlation coefficient is obtained for each of the two operations). The diagnosis result is shown in FIG. 8A, and the second diagnosis result is shown in FIG. 8B. In these figures, 43 differences (Rd67) based on the sixth division frequency band among the differences of P · (P−1) / 2 pieces (P = 49 in this example) described above. -Rb67, Rd68-Rb68,... Rd6p-Rb6p) are shown in one of the N time zones. In the present diagnosis, the threshold value of the absolute value of the difference is set to 1.

  As is clear from FIGS. 8A and 8B, in this diagnosis, a result in which the absolute value of the difference exceeds the threshold value was obtained only in the first diagnosis among the diagnoses performed twice. Therefore, it can be seen that an abnormality occurred in the press processing system in the first diagnosis, and no abnormality occurred in the press processing system in the second diagnosis.

  Further, when the molded product obtained by press working was visually observed after the diagnosis was completed, cracks were found only in the molded product produced by the first press working (in the molded product produced by the second press working) No cracks were observed). This shows that the diagnostic method according to the present modification is effective.

  As described above, also in the modification, the diagnosis parameter (that is, the diagnosis correlation function Rd) is the reference parameter (that is, the reference correlation) obtained based on the reference vibration data acquired in the same period as the predetermined period. Since it is compared with the function Rb), the same effect as the effect according to the first embodiment is exhibited.

In the above description, in step S114, for each of vibration data acquired in the second time zone, vibration data acquired in the third time zone,... The procedure from step S104 to step S112 is executed, but step S114 may be omitted. That is, in the above, in the step of obtaining the vibration data of the press working system in the predetermined period, a plurality of vibration data is obtained in the plurality of the predetermined periods, and a plurality of vibration data is obtained in the step of obtaining the correlation coefficient. However, the present invention is not limited to this, and one of the first time period to the n-th time period is set as the predetermined time only once. The procedure from step S104 to step S112 may be performed.
However, the above-described embodiment is more preferable in that it is possible to reliably grasp whether or not an abnormality has occurred in the press working system.

=== Second Embodiment According to the Diagnosis Method of Target Equipment ===
Next, a second embodiment according to the diagnosis method for the target facility will be described with reference to FIGS. 9 to 12B. FIG. 9 is a flowchart illustrating a second embodiment according to the target facility diagnosis method. 10 to 12B will be described later.

  In this section as well, the processing system for processing an object described in the section of the problem to be solved by the invention, for example, a mold composed of a fixed mold section and a movable mold section is used as the target equipment. A press working system for forming the material so that the material has a desired shape will be described as an example.

  This flowchart acquires the first vibration data of the press processing system as the waveform data of the target equipment in the first predetermined period of the operation period from the start of the press processing system to the end of the operation. And the second vibration data of the press working system is a second predetermined period that is a part of the operation period, and the start of the second predetermined period coincides with the end of the first predetermined period. The process starts with acquisition in two predetermined periods (step S202).

  In the present embodiment, vibration data of the press working system is acquired in the analysis time zone shown in FIG. 2, and first, the first time zone in the analysis time zone is set as the first predetermined period (first time period). The vibration data corresponding to the band is the first vibration data described above), and the second time period is the second predetermined period (vibration data corresponding to the second time period is the second vibration data described above).

  And the procedure similar to step S4-step S8 is implemented with respect to 1st vibration data and 2nd vibration data. That is, the first vibration data and the second vibration data are divided into T first divided vibration data and T second divided vibration data (division in the time axis direction, step S204), and T first division data is obtained. Each of the waveform data and the T second divided waveform data is Fourier transformed to obtain T first frequency spectrums and T second frequency spectra (step S206). Further, the strength of each of the T first frequency spectra is obtained for each of the divided frequency bands divided into P pieces, and the strength of each of the T second frequency spectra is divided into P pieces. Obtained for each frequency band (division in the frequency axis direction, step S208). Here, each of the obtained strengths of the first frequency spectrum is represented by Xij (i = 1,... T, j = 1,... P) and each of the second frequency spectrum strengths. , Yij (i = 1,... T, j = 1,... P).

  Next, based on the strength Xij of the first frequency spectrum obtained for each divided frequency band, the principal component score is calculated for each of the T frequency spectrums using the same procedure as that shown in the first embodiment. (Step S210). That is, first, the standardized first frequency spectrum intensity X′ij is obtained from the first frequency spectrum intensity Xij. Then, the principal component score Z is calculated by a calculation formula: Z = A (eigenvector) · X ′. The eigenvector A used here is obtained based on the strength Xij of the first frequency spectrum, and needs to be obtained before the principal component score Z is obtained.

  Next, based on the intensity Yij of the second frequency spectrum obtained for each divided frequency band, the principal component score is calculated for each of the T frequency spectrums using the same procedure as that shown in the first embodiment. (Step S212). That is, first, from the intensity Yij of the second frequency spectrum, the standardized intensity Y′ij of the second frequency spectrum is obtained, and the principal component score Z is calculated by the formula: Z = A (eigenvector) · Y ′. ,calculate. Here, the eigenvector used when obtaining the principal component score Z is not obtained based on the intensity Yij of the second frequency spectrum, but the eigenvector A described above, that is, the intensity of the first frequency spectrum. The eigenvector A obtained based on Xij is used.

  Next, the principal component score obtained in step S212 (the principal component score obtained based on the strength Yij of the second frequency spectrum) is used as the principal component score obtained in step S210 (the intensity of the first frequency spectrum). The principal component score obtained based on Xij) is compared with the same procedure as the procedure shown in the first embodiment (step S214). That is, the DI value is calculated from both principal component scores, and the calculated DI value is compared with a preset threshold value. If the DI value> threshold, an abnormality occurs in the press processing system in the second time zone (for example, cracks or cracks occur in the material), and if the DI value <threshold, press in the second time zone. It is determined that no abnormality has occurred in the machining system.

  The procedure may be performed on one of a plurality of principal component scores such as a first principal component score and a second principal component score (for example, a first principal component score), It is the same as that of the first embodiment that the above procedure may be performed for 1 main component score to m-th main component score). As in the first embodiment, a correlation coefficient may be used instead of the DI value as an index of the degree of difference between the two principal component scores.

  Next, the second time zone in the analysis time zone described above is set as the first predetermined period (vibration data corresponding to the second time zone becomes the first vibration data described above), and the third time zone is set as the first time period. As the second predetermined period (vibration data corresponding to the third time period becomes the second vibration data described above), the procedure from step S204 to step S214 described above is performed. Then, the same procedure is repeated until a + 1 = n, with the a-th time zone (a = 3, 4,...) As the first predetermined period and the a + 1-th time zone as the second predetermined period (step 1). S216). Thus, it is determined whether or not an abnormality has occurred in the press working system in each time zone from the third time zone to the nth time zone.

  FIG. 10A and FIG. 10B show the diagnosis results when the diagnosis method according to the present embodiment is actually applied to the diagnosis of the press working system. In this diagnosis, the diagnosis is performed twice (more specifically, the pressing system is operated twice to obtain the DI value for each of the two operations), and the first diagnosis The result is shown in FIG. 10A, and the second diagnosis result is shown in FIG. 10B. In the figure, the horizontal axis is from the second time zone to the n-th time zone (n = 69 in the figure), and the above-mentioned DI value is on the vertical axis. In this diagnosis, a plurality of principal component scores (from the first principal component score to the m-th principal component score) are calculated, and DI values are obtained for each, so that a plurality of DI values per one time zone are calculated. However, the maximum DI value among these DI values is extracted for each time period, and the extracted DI value is displayed in the figure. In this diagnosis, the threshold value of the DI value is 2.

  As is clear from FIGS. 10A and 10B, in this diagnosis, among the diagnoses performed twice, only the first diagnosis yielded a result that the DI value exceeded the threshold value. Since the DI value exceeds the threshold value in the seventh, tenth, and eleventh time zones, it can be seen that an abnormality has occurred in the press processing system in the time zone of the first diagnosis.

  Further, when the molded product obtained by press working was visually observed after the diagnosis was completed, cracks were found only in the molded product produced by the first press working (in the molded product produced by the second press working) No cracks were observed). This shows that the diagnostic method according to the present embodiment is effective.

  As described above, in the first embodiment, as the principal component score to be compared with the eigenvector for obtaining the principal component score corresponding to the vibration data acquired in a predetermined period, and the obtained principal component score, In the second embodiment, the data corresponding to the reference vibration data acquired in advance when the operation of the facility is normal in the same period as the predetermined period (eigenvectors and principal component scores) is used. An eigenvector for obtaining a principal component score corresponding to vibration data (that is, the second vibration data) acquired in a predetermined period (that is, the second predetermined period), and a comparison target of the obtained principal component score As the principal component score, the end of which coincides with the start of the second predetermined period, in other words, one time before the second predetermined period and adjacent to the second predetermined period Predetermined period (i.e., the first predetermined period) that was using the vibration data acquired in (i.e., the first vibration data) to those obtained on the basis of (eigenvectors and principal component score). Even if it does in this way, the subject mentioned above can be solved like 1st embodiment.

  That is, for the diagnosis of equipment (for example, the above-described press working system) having characteristics that can cause a significant change in vibration data between the initial operation time and the late operation time even if the operation of the equipment is normal. When adopting the conventional method (the method disclosed in Japanese Patent No. 3382240), even if the operation of the facility is normal, the vibration data can change significantly from the initial operation to the late operation. It is acquired at the middle stage of operation or at the latter stage of operation compared to the principal component score obtained based on the vibration data obtained at the initial stage of operation, even though the operation of the equipment is normal at the middle stage or the latter stage of operation. The principal component scores obtained based on the vibration data may be significantly different. Therefore, when a conventional method is used for diagnosing equipment having the characteristics, even if the equipment is diagnosed, the accuracy thereof is insufficient.

  On the other hand, when the method of the present case is adopted for diagnosis of equipment having the characteristics, the first predetermined period is one before the second predetermined period and is adjacent to the second predetermined period. Therefore, as long as the operation of the facility is normal, there is no significant change in the vibration data between the first predetermined period and the second predetermined period. Therefore, the operation of the facility does not occur during the second predetermined period. In the case of normality, the principal component score determined based on the second vibration data is not so different from the principal component score determined based on the first vibration data. On the other hand, when the operation of the equipment is abnormal in the second predetermined period, the principal component score obtained based on the second vibration data is significantly higher than the principal component score obtained based on the first vibration data. Since it is different, when the method of this case is adopted, the diagnosis of the target equipment will be performed more accurately.

<<< Modification of Second Embodiment According to Diagnostic Method for Target Equipment >>>
Here, the modification of 2nd embodiment which concerns on the diagnostic method of object installation is demonstrated using FIG. 11 thru | or FIG. 12B. FIG. 11 is a flowchart showing a modified example of the second embodiment according to the target facility diagnosis method. 12A and 12B will be described later.

  As shown in FIG. 11, when this modification is compared with the second embodiment described above, the procedure up to step S208 described above is the same for both. That is, in the present modification, first, the same procedure as the above-described step S202 to step S208 is performed (step S302 to step S308), and based on the first vibration data acquired in the first time zone, T × Based on the second vibration data acquired in the second time zone, and the strengths Xij (i = 1,..., T = 1,..., P) of the P first frequency spectra are obtained. , T × P second frequency spectrum intensities Yij (i = 1,... T, j = 1,... P) are obtained.

  Next, in this modification, instead of calculating the principal component score, the frequency spectrum strength corresponding to one divided frequency band of the frequency spectrum strengths obtained for each divided frequency band and the other divided frequencies A correlation coefficient with the intensity of the frequency spectrum corresponding to the band is obtained.

  The method for obtaining the correlation coefficient will be specifically described. First, the first frequency spectrum intensities X11, X21,... Xt1 corresponding to the first divided frequency band (center frequency f1) and the first frequency spectrum corresponding to the second divided frequency band (center frequency f2). A correlation coefficient R12 is obtained from the strengths X12, X22,... Xt2. Similarly, the strength of the first frequency spectrum corresponding to the first divided frequency band (center frequency f1), the third divided frequency band (center frequency f3), the fourth divided frequency band (center frequency f4),. Correlation coefficients R13, R14,... R1p are obtained from the intensity of the first frequency spectrum corresponding to the p-divided frequency band (center frequency fp), and the second divided frequency band (center frequency f2), third The same calculation is performed using the divided frequency band (center frequency f3),..., The (p-1) th divided frequency band (center frequency fp-1) as a reference, and correlation coefficients R23,. , R3p,..., R (p−1) p. Finally, P · (P−1) / 2 correlation coefficients (the correlation coefficient is referred to as a first correlation coefficient) are calculated (step S310).

  Next, the same procedure as in step S310 is performed on the intensity Yij of the second frequency spectrum, and P · (P−1) / 2 correlation coefficients (the correlation coefficients are expressed as the second phase). (Referred to as the relation number) (step S312).

  The correlation coefficient obtained in step S312 (correlation coefficient obtained based on the second frequency spectrum intensity Yij) is used as the correlation coefficient obtained in step S310 (first frequency spectrum intensity Xij). The correlation coefficient obtained based on the above is compared with a procedure similar to the procedure shown in the modification of the first embodiment (step S314). That is, the difference between the two is calculated, and the calculated difference is compared with a preset threshold value. If the difference is greater than the threshold, an abnormality occurs in the press system in the second time zone (for example, cracks or cracks occur in the material), and if the difference is less than the threshold, the press system in the second time zone. It is determined that no abnormality has occurred. In addition, it is the same as that of the modified example of 1st embodiment that you may use DI value and a correlation coefficient instead of the said difference.

  Next, the second time zone in the analysis time zone described above is set as the first predetermined period (vibration data corresponding to the second time zone becomes the first vibration data described above), and the third time zone is set as the first time period. As the second predetermined period (vibration data corresponding to the third time zone becomes the second vibration data described above), the procedure from step S304 to step S314 described above is performed. Then, the same procedure is repeated until a + 1 = n, with the a-th time zone (a = 3, 4,...) As the first predetermined period and the a + 1-th time zone as the second predetermined period (step 1). S316). Thus, it is determined whether or not an abnormality has occurred in the press working system in each time zone from the third time zone to the nth time zone.

  12A and 12B show the diagnostic results when the diagnostic method according to this modification is actually applied to the diagnosis of the press working system. In this diagnosis, the diagnosis is performed twice (more specifically, the press working system is operated twice and the correlation coefficient is obtained for each of the two operations). The diagnosis result is shown in FIG. 12A, and the second diagnosis result is shown in FIG. 12B. Also, in these figures, among the above-mentioned differences of P · (P−1) / 2 (in this example, P = 49), 43 differences based on the sixth division frequency band (the first) N correlation coefficient R67-first correlation coefficient R67, second correlation coefficient R68-first correlation coefficient R68,..., Second correlation coefficient R6p-first correlation coefficient R6p) It is shown in one of the time zones. In the present diagnosis, the threshold value of the absolute value of the difference is set to 1.

  As is clear from FIGS. 12A and 12B, in this diagnosis, a result in which the absolute value of the difference exceeds the threshold value was obtained only in the first diagnosis among the diagnoses performed twice. Therefore, it can be seen that an abnormality occurred in the press processing system in the first diagnosis, and no abnormality occurred in the press processing system in the second diagnosis.

  Further, when the molded product obtained by press working was visually observed after the diagnosis was completed, cracks were found only in the molded product produced by the first press working (in the molded product produced by the second press working) No cracks were observed). This shows that the diagnostic method according to the present modification is effective.

  Thus, also in the modified example, the second parameter (that is, the second correlation function) obtained based on the second vibration data acquired in the second predetermined period is acquired in the first predetermined period. Since it is compared with the first parameter (that is, the first correlation function) obtained based on one vibration data, the same effect as the effect according to the second embodiment is exhibited.

=== Device for diagnosing the target equipment ===
Next, an example of a device for diagnosing a target facility for realizing the above-described target facility diagnosis method (hereinafter, this device is also referred to as a target facility diagnosis device) will be described with reference to FIG. FIG. 13 is a conceptual diagram illustrating an example of the target facility diagnosis apparatus.

As described above, in the present embodiment, the target facility is the press processing system 2, and thus the target facility diagnosis device is the press processing system diagnosis device 102.
The press working system diagnostic apparatus 102 includes an AE sensor (Acoustic Emission Sensor) 104 as an example of a sensor, a computer 114, and a display device 116.

  The AE sensor 104 has a function of acquiring waveform data (that is, vibration data) of the press working system 2. The sensor is not limited to the AE sensor 104 as long as it has a function of acquiring vibration data, and any sensor (for example, an acceleration sensor) may be used.

  The computer 114 has a computer program 114a for realizing the above-described diagnosis method, and the above-described diagnosis method is executed by processing the computer program 114a by a CPU provided in the computer 114. The computer program 114a is composed of codes for executing the above-described diagnostic method. Note that the computer 114 may include a plurality of devices instead of a single device. In such a case, the computer program 114a may be provided separately for each of a plurality of devices.

  The display device 116 has a function of giving various information to the operator of the press working system diagnostic device 102. Further, it is more preferable that the graphs shown in FIGS. 6, 8A and 8B, 10A and 10B, 12A and 12B are displayed on the display device 116.

=== Other Embodiments ===
As mentioned above, although the diagnosis method of the object equipment concerning the present invention etc. based on the above-mentioned embodiment was explained, the above-mentioned embodiment is for making an understanding of the present invention easy, and limits the present invention. It is not a thing. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes the equivalents thereof.

  In the above, equipment having characteristics that can cause significant changes in vibration data between the initial operation time and the late operation time even if the equipment operation is normal (for example, a processing system for processing an object). ), The diagnosis method for the target equipment according to the first embodiment, its modification, the second embodiment, and its modification is employed, but is not limited thereto. The diagnosis method of the target equipment according to the first embodiment, its modification, the second embodiment, and its modification is vibration data between the initial operation time and the late operation time even if the operation of the equipment is normal. This is particularly effective when diagnosing equipment that has characteristics that can cause significant changes in the process (for example, a processing system for processing an object), but of course, as long as the operation of the equipment is normal. Can also be applied when diagnosing equipment having characteristics that do not cause much change in vibration data from the initial stage of operation to the late stage of operation (for example, a pump system in which the pump impeller keeps rotating at a constant speed). .

  Further, in the above, as a processing system for processing an object, a press processing system has been described as an example, but the present invention is not limited to this, for example, a welding system for welding an object, etc. You may apply the diagnostic method of the target installation which concerns on 1st embodiment, its modification, 2nd embodiment, and its modification.

  In the above description, vibration data is described as an example of waveform data. However, the present invention is not limited to this. For example, the waveform data may be voltage waveform data or current waveform data.

  In the first embodiment and its modification, as shown in the lower diagram of FIG. 2, the end of one predetermined period of the plurality of predetermined periods coincides with the start of another predetermined period. However, the present invention is not limited to this. For example, as shown in FIG. 14, one predetermined period among the plurality of predetermined periods, A plurality of the predetermined periods may be selected such that a part of the predetermined period overlaps a part of another predetermined period.

  Considering the merits of both, the former has the advantage in that the time required for diagnosis is shortened, and the latter is diagnosed appropriately when an abnormality occurs in the target equipment at the boundary of each time zone It is advantageous in that it is possible to prevent inconveniences that are not made in the past.

  Similarly, in the second embodiment and its modification, as shown in the lower diagram of FIG. 2, the first vibration data is a part of the operation period from the start of operation of the equipment to the end of operation. In the first predetermined period, and the second vibration data is a part of the second predetermined period of the operation period, and the start of the second predetermined period is the end of the first predetermined period. Although it was decided to acquire in the second predetermined period of coincidence, it is not limited to this, for example, as shown in FIG. 14, the first vibration data from the start of operation of the equipment to the end of operation A second predetermined period that is acquired after the first predetermined period and that is acquired in a part of the first predetermined period of the operation period and the second vibration data is acquired , A portion of the second predetermined period is the first predetermined period Second predetermined time period that overlaps with the parts, may be acquired in.

  As for both merits in such a case, the time required for diagnosis is shortened for the former (the former is appropriately diagnosed when an abnormality occurs in the target equipment at the boundary of each time zone. In the latter case, the vibration data changes between the first predetermined period and the second predetermined period when the operation of the equipment is normal compared to the former. Therefore, there is a merit that diagnosis of the target equipment can be performed more accurately.

  Further, in the above description, after the vibration data for all the time periods has been acquired (step S2, step S102, step S202, step S302), the analysis for diagnosis is performed. However, the present invention is not limited to this. However, the vibration data acquisition and the analysis for diagnosis may be performed in parallel.

It is a flowchart which shows 1st embodiment which concerns on the diagnostic method of object equipment. It is explanatory drawing for demonstrating the predetermined period which concerns on acquisition of vibration data. It is explanatory drawing for demonstrating the division | segmentation of vibration data into division | segmentation vibration data. It is explanatory drawing which showed the mode of the division | segmentation in the time-axis direction and a frequency-axis direction of the vibration data acquired in the 1st time slot | zone. It is the table | surface which showed the mode of the division | segmentation in the time-axis direction and a frequency-axis direction of the vibration data acquired in the 1st time slot | zone. It is the figure which showed the diagnostic result when the diagnostic method which concerns on 1st embodiment is applied to the diagnosis of a press work system. It is a flowchart which shows the modification of 1st embodiment which concerns on the diagnostic method of object installation. 8A and 8B are diagrams illustrating diagnosis results when the diagnosis method according to the modification of the first embodiment is applied to the diagnosis of the press working system. It is a flowchart which shows 2nd embodiment which concerns on the diagnostic method of object equipment. FIG. 10A and FIG. 10B are diagrams showing diagnosis results when the diagnosis method according to the second embodiment is applied to the diagnosis of the press working system. It is a flowchart which shows the modification of 2nd embodiment which concerns on the diagnostic method of object equipment. 12A and 12B are diagrams showing diagnosis results when the diagnosis method according to the modification of the second embodiment is applied to the diagnosis of the press working system. It is a conceptual diagram which shows an example of object facility diagnostic apparatus. It is explanatory drawing for demonstrating the other example of the predetermined period which concerns on acquisition of vibration data.

Explanation of symbols

2 Pressing System 102 Pressing System Diagnosis Device 104 AE Sensor 114 Computer 114a Computer Program 116 Display Device

Claims (23)

Obtaining waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation;
Dividing the waveform data acquired in the predetermined period into T pieces of divided waveform data;
Fourier transforming each of the T divided waveform data to obtain T frequency spectra;
Obtaining the strength of each of the T frequency spectra for each of the divided frequency bands divided into P pieces;
Based on the strength of the frequency spectrum obtained for each of the divided frequency bands, a principal component score is obtained for each of the T frequency spectra, and as an eigenvector used when obtaining the principal component score, Using a reference eigenvector obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal;
A method for diagnosing a target facility, comprising: In the diagnostic method of the object equipment according to claim 1,
The obtained principal component score is
Comparing with a reference principal component score obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal,
A method for diagnosing a target facility, comprising: In the diagnostic method of the object equipment of Claim 1 or Claim 2,
In the step of acquiring the waveform data of the target equipment in the predetermined period, a plurality of the waveform data is acquired in the plurality of the predetermined periods,
In the step of obtaining the principal component score, a principal component score is obtained based on each of the plurality of waveform data. Obtaining waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation;
Dividing the waveform data acquired in the predetermined period into T pieces of divided waveform data;
Fourier transforming each of the T divided waveform data to obtain T frequency spectra;
Obtaining the strength of each of the T frequency spectra for each of the divided frequency bands divided into P pieces;
A step of obtaining a correlation coefficient between a frequency spectrum intensity corresponding to one of the divided frequency bands and a frequency spectrum intensity corresponding to another divided frequency band among the intensity of the frequency spectrum obtained for each of the divided frequency bands; When,
Comparing the obtained correlation coefficient with a reference correlation coefficient obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal;
A method for diagnosing a target facility, comprising: In the diagnosis method of the object equipment according to claim 4,
In the step of acquiring the waveform data of the target equipment in the predetermined period, a plurality of the waveform data is acquired in the plurality of the predetermined periods,
In the step of obtaining the correlation coefficient, the correlation coefficient is obtained based on each of the plurality of waveform data. In the diagnostic method of the object equipment of Claim 3 or Claim 5,
A plurality of the predetermined periods are selected such that the end of one of the plurality of predetermined periods coincides with the start of the other predetermined period. Diagnosis method. In the diagnostic method of the object equipment of Claim 3 or Claim 5,
A plurality of the predetermined periods are selected such that a part of the predetermined period among a plurality of the predetermined periods overlaps a part of the other predetermined period. Diagnosis method. As the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and
Acquiring the second waveform data in a second predetermined period of a part of the operation period, wherein the second predetermined period coincides with the end of the first predetermined period. When,
Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data;
Fourier transforming each of the T first divided waveform data and the T second divided waveform data to obtain T first frequency spectra and T second frequency spectra;
Division of each of the T first frequency spectrums obtained for each P divided frequency band, and division of each of the T second frequency spectra into P divisions Obtaining each frequency band; and
Obtaining an eigenvector to be used when obtaining a principal component score based on the strength of the first frequency spectrum obtained for each of the divided frequency bands;
A step of obtaining a principal component score for each of the T second frequency spectra based on the strength of the second frequency spectrum obtained for each of the divided frequency bands, wherein the principal component score is obtained. Using an eigenvector obtained based on the strength of the first frequency spectrum as the eigenvector to be used;
A method for diagnosing a target facility, comprising: In the diagnostic method of the object installation of Claim 8,
In the step of obtaining the eigenvector, a principal component score is obtained for each of the T first frequency spectra together with the eigenvector based on the strength of the first frequency spectrum,
Comparing the principal component score determined in the step of determining the principal component score based on the strength of the second frequency spectrum with the principal component score determined based on the strength of the first frequency spectrum;
A method for diagnosing a target facility, comprising: As the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and
Acquiring the second waveform data in a second predetermined period of a part of the operation period, wherein the second predetermined period coincides with the end of the first predetermined period. When,
Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data;
Fourier transforming each of the T first divided waveform data and the T second divided waveform data to obtain T first frequency spectra and T second frequency spectra;
Division of each of the T first frequency spectrums obtained for each P divided frequency band, and division of each of the T second frequency spectra into P divisions Obtaining each frequency band; and
Of the intensity of the first frequency spectrum obtained for each of the divided frequency bands, the intensity of the first frequency spectrum corresponding to one divided frequency band and the intensity of the first frequency spectrum corresponding to another divided frequency band. Obtaining a first correlation coefficient, and
Of the intensity of the second frequency spectrum obtained for each of the divided frequency bands, the intensity of the second frequency spectrum corresponding to one divided frequency band and the intensity of the second frequency spectrum corresponding to another divided frequency band. Obtaining a second correlation coefficient;
Comparing the second correlation coefficient with the first correlation coefficient;
A method for diagnosing a target facility, comprising: In the diagnostic method of the object installation in any one of Claims 8 thru | or 10,
As the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and
Acquiring the second waveform data in a second predetermined period of a part of the operation period, wherein the second predetermined period coincides with the end of the first predetermined period. Instead of
As the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and
The second waveform data is a second predetermined period that comes after the first predetermined period of a part of the operation period, and a part of the second predetermined period is a part of the first predetermined period. And a step of acquiring in a second predetermined period that overlaps with the target facility. In the diagnostic method of the object installation in any one of Claims 1 thru | or 11,
The target facility is a processing system for processing an object. In the diagnostic method of the object installation in any one of Claims 1 thru | or 12,
The method for diagnosing a target facility, wherein the waveform data is vibration data. Obtaining waveform data of the target equipment in a predetermined period of at least a part of the operation period from the start of operation of the equipment to the end of operation;
Dividing the waveform data acquired in the predetermined period into T pieces of divided waveform data;
Fourier transforming each of the T divided waveform data to obtain T frequency spectra;
Obtaining the strength of each of the T frequency spectra for each of the divided frequency bands divided into P pieces;
Based on the strength of the frequency spectrum obtained for each of the divided frequency bands, a principal component score is obtained for each of the T frequency spectra, and as an eigenvector used when obtaining the principal component score, Using a reference eigenvector obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal;
The obtained principal component score is
Comparing with a reference principal component score obtained in advance based on the waveform data acquired during the predetermined period when the operation of the facility is normal,
Have
In the step of acquiring the waveform data of the target equipment in the predetermined period, a plurality of the waveform data is acquired in the plurality of the predetermined periods,
In the step of obtaining the principal component score, a principal component score is obtained based on each of the plurality of waveform data,
A plurality of the predetermined periods are selected such that the end of the predetermined period of the plurality of the predetermined periods coincides with the start of the other predetermined period;
The target equipment is a processing system for processing an object,
The method for diagnosing a target facility, wherein the waveform data is vibration data. As the waveform data of the target equipment, the first waveform data is acquired in a part of the first predetermined period of the operation period from the start of operation of the equipment to the end of operation, and
Acquiring the second waveform data in a second predetermined period of a part of the operation period, wherein the second predetermined period coincides with the end of the first predetermined period. When,
Dividing the first waveform data and the second waveform data into T first divided waveform data and T second divided waveform data;
Fourier transforming each of the T first divided waveform data and the T second divided waveform data to obtain T first frequency spectra and T second frequency spectra;
Division of each of the T first frequency spectrums obtained for each P divided frequency band, and division of each of the T second frequency spectra into P divisions Obtaining each frequency band; and
Obtaining an eigenvector to be used when obtaining a principal component score based on the strength of the first frequency spectrum obtained for each of the divided frequency bands;
A step of obtaining a principal component score for each of the T second frequency spectra based on the strength of the second frequency spectrum obtained for each of the divided frequency bands, wherein the principal component score is obtained. Using an eigenvector obtained based on the strength of the first frequency spectrum as the eigenvector to be used;
Have
In the step of obtaining the eigenvector, a principal component score is obtained for each of the T first frequency spectra together with the eigenvector based on the strength of the first frequency spectrum,
Comparing the principal component score determined in the step of determining the principal component score based on the strength of the second frequency spectrum with the principal component score determined based on the strength of the first frequency spectrum;
Have
The target equipment is a processing system for processing an object,
The method for diagnosing a target facility, wherein the waveform data is vibration data.   The computer program for implement | achieving the diagnostic method of the object installation of Claim 1.   The computer program for implement | achieving the diagnostic method of the object installation of Claim 4.   The computer program for implement | achieving the diagnostic method of the object installation of Claim 8.   The computer program for implement | achieving the diagnostic method of the object installation of Claim 10.   An apparatus for diagnosing a target facility, comprising: a sensor for acquiring waveform data of the target facility; and a computer comprising the computer program according to claim 16.   An apparatus for diagnosing a target facility, comprising: a sensor for acquiring waveform data of the target facility; and a computer comprising the computer program according to claim 17.   An apparatus for diagnosing a target facility, comprising: a sensor for acquiring waveform data of the target facility; and a computer including the computer program according to claim 18. An apparatus for diagnosing a target facility, comprising: a sensor for acquiring waveform data of the target facility; and a computer comprising the computer program according to claim 19.