JP2022066762A - Abnormality diagnostic device and abnormality diagnostic method - Google Patents

Abstract

To provide an abnormality diagnostic device which can precisely detect an abnormality or an abnormal sign of a diagnosed object, minimizes a downtime of the diagnosed object, and decreases a maintenance cost of the diagnosed object.SOLUTION: An abnormality diagnostic device comprises: an MD value calculation unit 22 which calculates Mahalanobis distances between collected data from diagnosed objects and normal time data of the diagnosed objects and creates MD values; an abnormal wave form pattern storing unit 32 which classifies abnormalities about the diagnosed objects into a plurality of patterns and stores the patterns as abnormality wave form patterns; a wave form pattern comparison unit 23 which, when diagnosing diagnosed objects, compares wave forms of time series data of the generated MD value with the abnormal wave form patterns stored in the abnormal wave form storing unit 32; and a diagnosis unit 24 which, when the wave forms of the time series data of the MD value calculated from the collected data match the abnormality wave form patterns as a result of the comparison 23, generates a diagnosis result indicating that they match.SELECTED DRAWING: Figure 1

Description

The present invention relates to an abnormality diagnosis device and an abnormality diagnosis method using the MT method.

In recent years, an abnormality diagnosis technique using the MT method (Mahalanobis Taguchi method) has been known. The MT method diagnoses whether or not the condition is "as usual", and the Mahalanobis distance to be diagnosed is farther than the Mahalanobis space with reference to the center of the Mahalanobis space that represents the same as usual condition that is set in advance. It is a method to judge that it is abnormal when it becomes.

In the conventional MT method, when the calculated degree of abnormality (MD value) exceeds a preset threshold value, it is determined as "abnormal" or "abnormal sign".

Japanese Unexamined Patent Publication No. 2013-041490

However, actual plant equipment has many operation patterns depending on the influence of load conditions, weather conditions, and the like. Therefore, even if it is determined to be in a normal state by the MT method, there is a problem that the degree of abnormality (MD value) varies to some extent and it is difficult to set an appropriate threshold value.

Further, in the past, even if the waveform of the degree of abnormality (MD value) clearly behaves differently from the normal one, it is not determined to be abnormal unless the threshold value is exceeded, so that there is a possibility that the abnormality may be overlooked. If the threshold value is set low in order to avoid this, on the contrary, a normal state may be erroneously detected as an abnormality.

The present invention has been made in view of the above circumstances, and it is possible to accurately detect an abnormality or a sign of an abnormality of a diagnosis target, minimize downtime of the diagnosis target, and reduce maintenance costs. It is an object of the present invention to provide an abnormality diagnosis device and an abnormality diagnosis method.

One aspect of the abnormality diagnostic apparatus according to the present invention for achieving the above object is to calculate the Maharanobis distance between the data collected from the diagnosis target and the normal data of the diagnosis target to generate an MD value. The unit, the abnormal waveform pattern storage unit that classifies the abnormality related to the diagnosis target into a plurality of patterns and stores it as an abnormal waveform pattern, and the waveform of the time-series data of the generated MD value at the time of diagnosis of the diagnosis target. And the time-series data of the MD value calculated from the collected data as a result of the matching between the waveform pattern matching unit for collating the abnormal waveform pattern stored in the abnormal waveform pattern storage unit and the waveform pattern matching unit. This is an abnormality diagnostic device including a diagnostic unit that generates a diagnosis result to that effect when the waveform of the above and the abnormal waveform pattern match.

Further, in one aspect of the abnormality diagnosis method according to the present invention, a unit space is generated in advance from the data of the diagnosis target at the normal time, and a plurality of patterns of abnormal waveform patterns of the MD value at the time of the abnormality of the diagnosis target are stored in advance. At the time of diagnosis of the diagnosis target, the collected data from the diagnosis target, the time-series waveform of the MD value generated from the unit space, and the abnormal waveform pattern are collated, and the collected data is checked. This is an abnormality diagnosis method for diagnosing the presence or absence of an abnormality in the diagnosis target based on the collation result between the waveform of the time-series data of the MD value calculated from the above and the abnormality waveform pattern.

According to one aspect of the present invention, it is possible to accurately detect an abnormality or a sign of an abnormality of the diagnosis target, minimize downtime of the diagnosis target, and reduce maintenance costs.

It is a block diagram which shows the structure of 1st Embodiment of the abnormality diagnosis apparatus which concerns on this invention. It is a flowchart which shows the processing procedure of 1st Embodiment. It is a waveform diagram which shows the abnormal waveform pattern 1 of MD value. It is a waveform diagram which shows the abnormal waveform pattern 2 of MD value. It is a waveform diagram which shows the abnormal waveform pattern 3 of MD value. It is a waveform diagram which shows the abnormal waveform pattern 4 of MD value.

<< Configuration of Embodiment >>
FIG. 1 is a block diagram showing a configuration of an embodiment of an abnormality diagnostic device according to the present invention.

As shown in the figure, the abnormality diagnosis device 1 is composed of a computer, and includes an input device 10, an arithmetic processing unit (CPU) 20, and a storage device 30. Further, the alarm output device 40 for outputting an alarm to the outside is provided.

The input device 10 includes a data acquisition unit 11 for inputting collected data collected at the plant to be diagnosed. It also has input devices such as a keyboard, mouse, operation buttons, and touch panel.

The arithmetic processing unit 20 includes a unit space calculation unit 21, an MD value calculation unit 22, a waveform pattern matching unit 23, and a diagnostic unit 24. The storage device 30 includes a unit space data storage unit 31, an abnormal waveform pattern storage unit 32, and a history data storage unit 33.

The unit space calculation unit 21 generates a unit space that serves as a reference for data diagnosis in the normal state based on the acquired data acquired by the data acquisition unit 11 in the normal state of the plant. The generated unit space is stored in the unit space data storage unit 31.

The MD value calculation unit 22 calculates the Mahalanobis distance between the data acquired from the plant and the unit space as the MD value at the time of diagnosis of the plant.

The waveform pattern collating unit 23 collates the time-series data of the MD value with the abnormal waveform pattern and detects the presence or absence of an abnormality in the acquired data.

The diagnosis unit 24 diagnoses the abnormality of the plant based on the collation result of the waveform pattern collation unit 23.

The unit space data storage unit 31 stores the unit space, which is the normal data of the plant, calculated by the unit space calculation unit 21.

The abnormal waveform pattern storage unit 32 classifies abnormalities related to the plant to be diagnosed into a plurality of patterns and stores them as abnormal waveform patterns. Specifically, the abnormal waveform patterns 1 to 4 shown in FIGS. 3 to 6 are stored.

The history data storage unit 33 stores the diagnosis result as history data. This historical data is used for updating and correcting abnormal waveform patterns.

<< Processing procedure of the embodiment >>
Next, the processing procedure of the embodiment will be described with reference to the flowchart in FIG.

The data acquisition unit 11 inputs a data group (collected data) collected at the plant to be diagnosed. Prior to the diagnosis, the unit space calculation unit 21 calculates a unit space as a reference at the time of diagnosis and stores it in the unit space data storage unit 31 as unit space data.

At the time of abnormality diagnosis, data is acquired from the operating plant via the data acquisition unit 11 (step S1). The MD value calculation unit 22 calculates the MD value from the acquired data (step S2).

Specifically, first, the mean value and the standard deviation are obtained from the acquired data. Next, the data is standardized using the value of the acquired data and the mean value and standard deviation obtained for the acquired data, and the standardized value for each variable is obtained. Subsequently, the MD value calculation unit 22 obtains the correlation matrix for the acquired data and the inverse matrix of the correlation matrix by using the standardized value obtained for the acquired data. Finally, the Mahalanobis distance is calculated using the obtained inverse matrix. From the Mahalanobis distance (MD value) obtained in this way, it is possible to identify the difference between the current state and the usual state (normal state) of the data acquired from the diagnosis target.

Next, if there is a waveform pattern that exceeds a preset threshold value from the waveform data obtained from the time-series data of the calculated MD value, the waveform pattern matching unit 23 stores the waveform pattern in advance. It is collated with the abnormal waveform pattern (step S3). If there is no corresponding abnormal waveform pattern as a result of collation (step S4NO), the process returns to step S1 and the data acquisition process is continued.

When there is a corresponding waveform pattern (step S4YES), the diagnostic unit 24 determines which abnormal waveform pattern the waveform pattern corresponds to (step S5). If the waveform pattern is the abnormal waveform pattern 1, a warning to that effect is output (steps S5YES, S6). If the waveform pattern is not the abnormal waveform pattern 1, it is determined whether or not the waveform pattern corresponds to the abnormal waveform pattern 2 (step S7).

If the waveform pattern is the abnormal waveform pattern 2, a warning to that effect is output (steps S7YES, S8). If the waveform pattern is not the abnormal waveform pattern 2, it is determined whether or not the waveform pattern corresponds to the abnormal waveform pattern 3 (step S9).

If the waveform pattern is the abnormal waveform pattern 3, a warning to that effect is output (steps S9YES, S10). If the waveform pattern is not the abnormal waveform pattern 3, it is determined whether or not the waveform pattern corresponds to the abnormal waveform pattern 4 (step S11).

If the waveform pattern is the abnormal waveform pattern 4, a warning to that effect is output (steps S9YES, S10).

In this way, it is collated which of the abnormal waveform patterns the waveform pattern corresponds to, and if a matching abnormal waveform pattern is detected, a warning to that effect is output. If the waveform pattern is not any of the abnormal waveform patterns 1 to 4, there is a possibility of erroneous detection, and a report to that effect is made.

In the above process, if there is an abnormality or a sign of an abnormality, the alarm output device 40 is immediately notified to a person concerned such as a maintenance person by e-mail or the like. Further, the abnormality detection history is stored in the history data storage unit 33.

Next, the abnormal waveform pattern used in the above-mentioned processing procedure will be specifically described. In this embodiment, the following four patterns are assumed as the abnormal waveform patterns. The time axis of the following patterns 1 to 3 is, for example, several seconds to several hours or several days, whereas the pattern 4 is a unit of several months to several years.

<Abnormal waveform pattern 1 of MD value>
The abnormal waveform pattern 1 shown in FIG. 3 is not detected as an abnormality when the MD value exceeds the threshold value Th1 even though it is not originally abnormal such as erroneous detection of the sensor. That is, if the MD value exceeds the threshold value Th1 and then continues to be below the threshold value Th1, it is not determined to be abnormal.

However, when viewed over a slightly longer span and exhibiting behavior that exceeds the threshold Th1 periodically or a specified number of times, it is regarded as abnormal.

For example, abnormality detection of a conveyor belt is applicable. If damage or foreign matter adheres to a part of the belt, it may appear as an abnormality such as abnormal vibration. When a sensor such as a vibration meter is attached to somewhere on the conveyor, the MD value exceeds the threshold value Th1 only when the abnormality occurrence point passes in the vicinity of the sensor, which is periodically repeated. When the foreign matter adheres and exceeds the threshold value Th1, but the foreign matter immediately disappears and returns to the original state, the curve changes as shown by the alternate long and short dash line in FIG.

<Abnormal waveform pattern 2 of MD value>
The abnormal waveform pattern 2 shown in FIG. 4 is a case where the MD value exceeds the threshold value Th1 in a relatively short period of time and the state is maintained for a certain period of time.

For example, it is assumed that the vibration behavior of the equipment changes due to damage to parts such as bolts that fix the motor. Further, as another example, when a foreign substance is caught in the drive unit of the motor drive device, the load becomes large, and the motor current value fluctuates, the MD value may change in the curve as shown in FIG. be.

<Abnormal waveform pattern 3 of MD value>
The abnormal waveform pattern 3 shown in FIG. 5 is a case where the MD value is clearly larger than the threshold value. Even if the MD value fluctuates only for a moment, if the MD value shows an excessively large value, it is judged to be abnormal.

For example, as shown in FIG. 5, a value such that the MD value greatly exceeds the first threshold value Th1 and further exceeds the second threshold value Th2 (= Th1 × α, for example α> 10) even for a moment. When indicated, a warning is issued assuming that something is wrong with the equipment of the plant.

<Abnormal waveform pattern 4 of MD value>
The abnormal waveform pattern 4 shown in FIG. 6 is a waveform of a change in MD value due to aged deterioration. It is assumed that it will change exponentially over a long period of time. The time axis of FIG. 6 is set with reference to the general life of the device, judging from the description in the catalog or the like, and is, for example, a period of several months to several years.

As described above, in one embodiment of the present invention, the MT method is used to perform abnormality diagnosis of plant equipment. The time-series data of the degree of abnormality (MD value: Mahalanobis distance value) calculated by the MT method is classified into a plurality of patterns according to the shape of the data waveform, and the presence or absence of abnormality or a sign of abnormality is determined. As a result, the accuracy and accuracy of the abnormality diagnosis are improved as compared with the conventional abnormality determination based on the threshold value. Further, by classifying the abnormality in advance, the time required for investigating the cause can be shortened.

In the above embodiment, four abnormal waveform patterns 1 to 4 are assumed. However, these abnormal waveform patterns 1 to 4 are examples, and in addition to these, various abnormal waveform patterns can be assumed depending on the equipment / equipment to be diagnosed and the installation environment thereof.

Although the present invention has been described in detail using the embodiments described above, the present invention is not limited to the embodiments described in the present specification and the drawings. The scope of the present invention is determined by the description of the scope of claims and the scope equivalent to the description of the scope of claims.

1 ... Abnormality diagnosis device, 10 ... Input device, 11 ... Data acquisition unit, 20 ... Arithmetic processing device, 21 ... Unit space calculation unit, 22 ... MD value calculation unit is 23 ... Wave pattern matching unit, 24 ... Diagnosis unit, 30 ... storage device, 31 ... unit space data storage unit is 32 ... abnormal waveform pattern storage unit, 33 ... history data storage unit, 40 ... alarm output device (alarm output unit).

Claims (5)

An MD value calculation unit that calculates the Mahalanobis distance between the data collected from the diagnosis target and the normal data of the diagnosis target and generates an MD value,
An abnormal waveform pattern storage unit that classifies abnormalities related to the diagnosis target into a plurality of patterns and stores them as abnormal waveform patterns.
At the time of diagnosis of the diagnosis target, a waveform pattern matching unit that collates the waveform of the time-series data of the generated MD value with the abnormal waveform pattern stored in the abnormal waveform pattern storage unit.
If the waveform of the time-series data of the MD value calculated from the collected data and the abnormal waveform pattern match as a result of the collation by the waveform pattern matching unit, the diagnostic unit that generates a diagnostic result to that effect. ,
An abnormality diagnostic device equipped with. The abnormal waveform pattern is a first abnormal waveform pattern that periodically repeats a state exceeding a threshold value indicating an abnormality of the MD value, a second abnormal waveform pattern that maintains the state exceeding the threshold value for a certain period of time, and an MD value. Includes at least one of a third abnormal waveform pattern in which is clearly larger than the threshold value, and a fourth abnormal waveform pattern in which the MD value increases with the passage of time.
The abnormality diagnostic device according to claim 1. A history data storage unit for storing the diagnosis result as history data is further provided.
The abnormality diagnostic device according to claim 1 or 2. An alarm output unit that outputs an alarm based on the diagnosis result from the diagnosis unit is provided.
The abnormality diagnostic device according to any one of claims 1 to 3. A unit space is generated in advance from the normal data of the diagnosis target,
A plurality of patterns of abnormal waveforms of MD values at the time of abnormality of the diagnosis target are stored in advance.
At the time of diagnosis of the diagnosis target, the data collected from the diagnosis target, the time-series waveform of the MD value generated from the unit space, and the abnormal waveform pattern are collated.
Based on the collation result between the waveform of the time-series data of the MD value calculated from the collected data and the abnormal waveform pattern, the presence or absence of the abnormality of the diagnosis target is diagnosed.
Abnormal diagnosis method.

Images