JP2021179791A - Abnormality sign detection system, abnormality sign detection device and abnormality sign detection method - Google Patents

Abstract

To detect an abnormality sign of a roller with high accuracy.SOLUTION: An abnormality sign detection system includes: storage means for storing operation information for each roller as time-series data; calculation means for calculating a predetermined index value indicating the degree of variation in a value of the operation information in a predetermined period for each roller; and detection means for detecting the roller with an abnormality sign based on the predetermined index value calculated by the calculation means.SELECTED DRAWING: Figure 1

Description

The present invention relates to an abnormality sign detection system, an abnormality sign detection device, and an abnormality sign detection method.

The rolling process is known as one of the steelmaking processes. The rolling process is a process of processing a steel piece into a plate shape by passing the steel piece between rotating rollers. In the rolling process, for example, rolling is performed by arranging a plurality of rolling mills in which rollers are arranged one above the other and passing steel pieces in order between the rollers of these rolling mills.

Techniques for detecting roller abnormalities and signs thereof by collecting and analyzing rolling mill roller operation information (for example, acceleration, etc.) have been conventionally known (for example, Patent Documents 1 to 7). reference).

Japanese Unexamined Patent Publication No. 6-71320 Japanese Unexamined Patent Publication No. 2010-234422 Japanese Unexamined Patent Publication No. 4-123807 Japanese Unexamined Patent Publication No. 62-9709 Japanese Unexamined Patent Publication No. 7-12641 Japanese Unexamined Patent Publication No. 2013-81994 Japanese Unexamined Patent Publication No. 61-202701

However, in the past, the accuracy of detecting abnormal signs was not always high. It is considered that one of the reasons for this is that the timing at which the steel pieces pass between the rollers of each rolling mill is different, and it is difficult to create a normal model of the rollers regarding the operation information. It is also considered that one of the causes is that the pattern represented by the value of the roller operation information differs depending on the steel type.

The embodiment of the present invention has been made in view of the above points, and an object thereof is to detect an abnormality sign of a roller with high accuracy.

In order to achieve the above object, the abnormality sign detection system according to the present embodiment has a storage means for storing operation information for each roller as time-series data and a variation in the value of the operation information for each roller in a predetermined period. It has a calculation means for calculating a predetermined index value indicating a degree, and a detection means for detecting the roller having an abnormality sign based on the predetermined index value calculated by the calculation means.

It is possible to detect the abnormal sign of the roller with high accuracy.

It is a figure which shows an example of the whole structure of the abnormality sign detection system which concerns on this embodiment. It is a figure which shows an example of the hardware configuration of the abnormality sign detection apparatus which concerns on this embodiment. It is a figure which shows an example of the functional structure of the abnormality sign detection apparatus which concerns on this embodiment. It is a flowchart which shows an example of the abnormality sign detection processing which concerns on this embodiment. It is a figure (the 1) which shows an example of the calculation result of an index value. It is a figure (the 2) which shows an example of the calculation result of an index value.

Hereinafter, embodiments of the present invention will be described. In the present embodiment, an abnormality sign detection system 1 capable of detecting an abnormality sign of a roller of a rolling mill used in a rolling process of a steelmaking process with high accuracy will be described for a steelmaking plant. The roller abnormality sign is a sign before a failure occurs in the roller or a sign before some abnormality occurs in the roller.

However, the rolling process of the steelmaking process is an example and is not limited to this. The abnormality sign detection system 1 according to the present embodiment can be similarly applied not only to the iron making process but also to detect an abnormality sign of a rolling mill used in a rolling process of an arbitrary metal processing process. .. In addition to the metal processing process, it can also be applied to detect abnormal signs of rollers installed for any purpose (for example, rollers of a belt conveyor for transporting metal etc.). be.

In the abnormality sign detection system 1 according to the present embodiment, after collecting the operation information (for example, acceleration, etc.) of each roller by a sensor, an index value indicating the degree of variation in the value of the operation information is calculated for each roller, and the index value is calculated. By comparing the index values between the rollers, the abnormal sign of each roller is detected. As a result, the abnormality sign detection system 1 according to the present embodiment can detect the abnormality sign of the roller with high accuracy.

<Overall configuration of abnormality sign detection system 1>
First, the overall configuration of the abnormality sign detection system 1 according to the present embodiment will be described with reference to FIG. FIG. 1 is a diagram showing an example of the overall configuration of the abnormality sign detection system 1 according to the present embodiment.

As shown in FIG. 1, the abnormality sign detection system 1 according to the present embodiment includes an abnormality sign detection device 10, a plurality of sensors 20, and a plurality of rollers 30. Further, the abnormality sign detection device 10 and each sensor 20 are connected to each other so as to be communicable by an arbitrary communication method (for example, wired or wireless LAN (Local Area Network), short-range wireless communication such as Bluetooth (registered trademark), etc.). ing.

The sensor 20 generates operation information of the roller 30 by sensing the roller 30 contained in the rolling mill for rolling the steel piece, and transmits the generated operation information to the abnormality sign detecting device 10. The rolling equipment of a steelmaking plant is generally composed of a plurality of rolling mills. In general, the sensor 20 is also included in the rolling mill. However, the present invention is not limited to this, and for example, the sensor 20 may be installed outside the rolling mill and the roller 30 may be sensed from outside the rolling mill.

Here, in the present embodiment, as an example, it is assumed that one rolling mill includes two upper and lower rollers 30 and two sensors 20 that sense each of these rollers 30. Further, when the total number of rollers 30 is N (N is an even number) and each roller 30 is represented separately, it is represented as "roller 30 _n " (however, n = 1, ..., N) in one rolling mill. Includes rollers 30 _n and rollers 30 _{n + 1} (where n = 1,3, ..., N-1), respectively (that is, the number of rolling mills is N / 2). .. Further, the sensor 20 that senses the _{roller 30 n} with respect to n = 1, ..., N is referred to _{as "sensor 20 n} ".

The abnormality sign detection device 10 is a computer or a computer system that detects an abnormality sign of the roller 30 by using the operation information collected from the sensor 20. The abnormality sign detection device 10 calculates a predetermined index value (that is, an index value indicating the degree of variation in the value of the operation information) for each roller 30 using the operation information collected from each sensor 20, and then the index value. Is compared between the rollers 30 to detect abnormal signs of each roller 30.

The configuration of the abnormality sign detection system 1 shown in FIG. 1 is an example, and may be another configuration. For example, in the example shown in FIG. 1, one rolling mill contains two rollers 30, but the rolling mill is not limited to this, and one rolling mill contains an arbitrary number of rollers 30. You may. However, in general, in the rolling process, rolling is performed by passing steel pieces between the rollers 30 arranged above and below, so that the number of rollers 30 is an even number.

Further, for example, the abnormality sign detection system 1 according to the present embodiment may include a plurality of abnormality sign detection devices 10.

<Hardware configuration of abnormality sign detection device 10>
Next, the hardware configuration of the abnormality sign detection device 10 according to the present embodiment will be described with reference to FIG. FIG. 2 is a diagram showing an example of the hardware configuration of the abnormality sign detection device 10 according to the present embodiment.

As shown in FIG. 2, the abnormality sign detection device 10 according to the present embodiment includes an input device 11, a display device 12, an external I / F 13, a communication I / F 14, a processor 15, and a memory device 16. Have. Each of these hardware is connected to each other by a bus 17 so as to be able to communicate with each other.

The input device 11 is, for example, various buttons, a touch panel, a keyboard, a mouse, or the like, and is used to input various operations to the abnormality sign detection device 10. The display device 12 is, for example, a display or the like, and displays a processing result or the like of the abnormality sign detection device 10. The abnormality sign detection device 10 may not have at least one of the input device 11 and the display device 12.

The external I / F 13 is an interface with an external device. The external device includes a recording medium 13a and the like. The abnormality sign detection device 10 can read or write the recording medium 13a via the external I / F 13. The recording medium 13a includes, for example, an SD memory card, a USB memory, a CD (Compact Disk), a DVD (Digital Versatile Disk), and the like.

The communication I / F 14 is an interface for the abnormality sign detection device 10 to perform data communication with other devices, devices, and the like.

The processor 15 is, for example, a CPU (Central Processing Unit) or the like, and is an arithmetic unit that reads programs and data from the memory device 16 and executes various processes.

The memory device 16 is, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), a RAM (Random Access Memory), a ROM (Read Only Memory), or the like, and is a storage device for storing programs and data.

By having the hardware configuration shown in FIG. 2, the abnormality sign detection device 10 according to the present embodiment can realize various processes described later. The hardware configuration shown in FIG. 2 is an example, and the abnormality sign detection device 10 according to the present embodiment may have another hardware configuration. For example, the abnormality sign detection device 10 according to the present embodiment may have a plurality of processors 15 or a plurality of memory devices 16.

<Functional configuration of abnormality sign detection device 10>
Next, the functional configuration of the abnormality sign detection device 10 according to the present embodiment will be described with reference to FIG. FIG. 3 is a diagram showing an example of the functional configuration of the abnormality sign detection device 10 according to the present embodiment.

As shown in FIG. 3, the abnormality sign detection device 10 according to the present embodiment includes an index value calculation unit 101, an abnormality sign detection unit 102, and a notification unit 103. Each of these parts is realized, for example, by a process in which one or more programs installed in the abnormality sign detection device 10 are executed by the processor 15.

Further, the abnormality sign detection device 10 according to the present embodiment has a storage unit 104. The storage unit 104 can be realized by using, for example, a memory device 16. The storage unit 104 may be realized by using, for example, a storage device connected to the abnormality sign detection device 10 via a communication network or the like.

The storage unit 104 stores the operation information received from each sensor 20. Here, each operation information includes, for example, identification information (for example, a roller ID or the like) for identifying the roller 30 to be sensed, a date and time when the roller 30 is sensed, a speed, an acceleration, a rotation speed, and the like. In the present embodiment, as an example, the operation information includes a roller ID, a date and time, a speed, an acceleration, and a rotation speed.

In this way, the storage unit 104 stores the operation information for each roller 30 as time-series data. The above operation information is an example, and the operation information may include at least acceleration in addition to the identification information of the roller 30 and the date and time.

The index value calculation unit 101 calculates a predetermined index value for each roller 30 by using the operation information stored in the storage unit 104.

The abnormality sign detection unit 102 determines whether or not there is a roller 30 having an abnormality sign by using the index value calculated by the index value calculation unit 101. As a result, when it is determined that the roller 30 having the abnormality sign is present, the roller 30 is detected as the roller 30 having the abnormality sign.

When the abnormality sign detection unit 102 detects a roller 30 having an abnormality sign, the notification unit 103 notifies the roller ID and the like of the roller 30. The notification destination of the notification unit 103 can be arbitrarily set. For example, the notification unit 103 may display the notification content on the display of the abnormality sign detection device 10, or transmits the notification content to a terminal or the like connected to the abnormality sign detection device 10 via the communication network. May be good.

<Details of abnormality sign detection processing>
Next, the details of the abnormality sign detection process according to the present embodiment will be described with reference to FIG. FIG. 4 is a flowchart showing an example of the abnormality sign detection process according to the present embodiment.

Step S101: First, the index value calculation unit 101 calculates a predetermined index value for each roller 30 by using the operation information stored in the storage unit 104. Here, the index value calculation unit 101 calculates the index value by, for example, either or both of the following calculation method 1 and calculation method 2. When the index value is calculated by both the following calculation method 1 and the calculation method 2, two types of index values are calculated for each roller 30.

・ Calculation method of index value 1
The calculation method 1 calculates the average value of the operation information for each predetermined first period (for example, every day or every hour), and then calculates the average value of the average value in the past predetermined second period (for example, every hour). This is a method of calculating the variance between the past 6 months, the past 1 year, etc.) as an index value. Hereinafter, as an example, the predetermined first period will be described as “1 day”, and the predetermined predetermined second period in the past will be described as “the past 6 months”. Further, the index representing a predetermined first period (that is, "day" thereafter) is represented by d. As shown in the above example, it is assumed that the second period is a longer period than the first period (that is, for example, the first period is an hourly or daily period). On the other hand, the second period shall be a monthly or yearly period.)

At this time, in the index value calculation method 1, the index value is calculated by the following (1-1) to (1-2).

(1-1) First, the index value calculation unit 101 calculates the average value of each of the speed, acceleration, and rotation speed included in each operation information on the d-day for each roller 30. That is, the index value calculation unit 101 has, for each roller 30, the average value of the speed included in each operation information on the d day, the average value of the acceleration included in each operation information on the d day, and the average value of the acceleration on the d day. Calculate the average value of the number of revolutions included in each operation information.

After that, the average value of the speed included in each operation information on the d day is the "speed average value", the average value of the acceleration included in each operation information on the d day is the "acceleration average value", and each on the d day. The average value of the number of rotations included in the operation information is expressed as "average number of rotations".

(1-2) Next, the index value calculation unit 101 calculates the standard deviations of the speed average value, the acceleration average value, and the rotation speed average value for the past 6 months from the d day for each roller 30. That is, the index value calculation unit 101 has the standard deviation of the speed average value from the d day to the past 6 months, the standard deviation of the acceleration average value from the d day to the past 6 months, and the past 6 from the d day. The standard deviation of the average value of the number of rotations in a month is calculated as the index value on the d-day. For example, assuming that one month is 30 days, the standard deviation of 180 speed average values, the standard deviation of 180 acceleration average values, and the standard deviation of 180 rotation speed average values are respectively. It will be calculated as the index value on the dth day.

After that, the standard deviation of the speed average value from the d day to the past 6 months is the "speed standard deviation", the standard deviation of the acceleration average value from the d day to the past 6 months is the "acceleration standard deviation", and the d day. The standard deviation of the average rotation speed over the past 6 months is referred to as the "standard deviation of rotation speed".

As a result, the velocity standard deviation, the acceleration standard deviation, and the rotation speed standard deviation are calculated as index values on the d-day for each roller 30. As described above, in the index value calculation method 1, the index value is calculated for each predetermined first period by using the operation information during the past predetermined second period.

In the above (1-1), the average values of the speed, acceleration, and rotation speed included in each operation information on the d day are calculated, but the present invention is not limited to this, and for example, each operation information on the d day is calculated. You may calculate either the maximum value or the minimum value of each of the speed, acceleration and rotation speed included in. Further, in (1-2) above, the standard deviation of the velocity average value, the standard deviation of the acceleration average value, and the standard deviation of the rotation number average value during the predetermined second period in the past are calculated, respectively. For example, the variance may be calculated instead of the standard deviation.

・ Calculation method of index value 2
The calculation method 2 is a method of calculating the standard deviation of the operation information during a predetermined third period as an index value for each predetermined third period. Hereinafter, as an example, the predetermined third period will be described as “1 week”. Further, the index representing this predetermined third period is represented by m, and the period m is assumed to be a period of the past one week from the current date d (that is, a period from d-6 to d).

At this time, in the index value calculation method 2, the index value calculation unit 101 uses the standard of speed, acceleration, and rotation speed included in each operation information during the week for each roller 30 and for each week. The deviation is calculated as an index value in the period m.

As a result, for each roller 30, the standard deviation of velocity, the standard deviation of acceleration, and the standard deviation of the number of revolutions are calculated as index values in the period m, respectively. Here, when the index value is calculated by the index value calculation method 2, some sections may be superimposed on the period m and the period m + 1, or there may be no overlapping section. For example, when M is the length of a predetermined period, the period m is [d, d + M], and the period m + 1 is [d', d'+ M] (however, d <d'), d + M ≧ d'may be used. Then, d + M + 1 = d'may be used. d + M ≧ d'means that some sections overlap in the period m and the period m + 1, and d + M + 1 = d'means that there is no section in which the period m and the period m + 1 overlap. Further, for example, when d + M ≧ d', it is possible to arbitrarily set how much the period m and the period m + 1 are superimposed.

In the above, the standard deviations of the speed, acceleration, and rotation speed included in each operation information during the third period are calculated for each predetermined third period, but the standard deviation is not limited to this, for example. The variance may be calculated instead of the standard deviation. Further, for example, after calculating the average values of the speed, acceleration, and rotation speed included in each operation information for each date during the predetermined third period, the average of the speeds during the predetermined third period is calculated. Standard deviation (or variance) of the value, standard deviation (or variance) of the mean value of acceleration during a given third period, and standard deviation (or variance) of the mean number of revolutions during a given third period. ) May be calculated respectively. Further, at this time, the maximum value or the minimum value may be calculated instead of the average value.

Step S102: Next, the abnormality sign detection unit 102 determines whether or not there is a roller 30 having an abnormality sign by using the index value calculated in step S101 above. If there is a roller 30 with an abnormality sign, this roller 30 is detected as a roller 30 with an abnormality sign.

The abnormality sign detection unit 102 may use the index value calculated by either of the above calculation methods 1 or 2, or may use the index value calculated by both the above calculation methods 1 and 2. May be good. Hereinafter, the index value calculated by the above calculation method 1 will be referred to as a “first index value”, and the index value calculated by the above calculation method 2 will be referred to as a “second index value”.

Here, the abnormality sign detection unit 102 has detected an abnormality by, for example, calculating the Mahalanobis distance of each first index value on the d day and then performing abnormality detection using these Mahalanobis distances. The roller 30 corresponding to the first index value is detected as the roller 30 having an abnormality sign. At this time, the abnormality sign detection unit 102 performs abnormality detection for each of the first index value related to speed, the first index value related to acceleration, and the first index value related to the rotation speed. In the abnormality detection using the Mahalanobis distance, for example, the Mahalanobis distance of a certain first index value is L ₁ times the average value of the Mahalanobis distance of another first index value (L ₁ is predetermined). When the distance is greater than or equal to (value), the certain first index value is detected as abnormal.

Similarly, the abnormality sign detection unit 102 calculates, for example, the Mahalanobis distances of each second index value in the period m, and then performs abnormality detection using these Mahalanobis distances, whereby the abnormality is detected. The roller 30 corresponding to the index value of 2 is detected as the roller 30 having an abnormality sign. At this time, the abnormality sign detection unit 102 performs abnormality detection for each of the second index value related to speed, the second index value related to acceleration, and the second index value related to the rotation speed. In the abnormality detection using the Mahalanobis distance, for example, Mahalanobis distance of a certain second index value, L ₂ times the average value of the Mahalanobis distance of the other second index value (L ₂ are predetermined When the distance is greater than or equal to (value), the certain second index value is detected as abnormal.

As described above, abnormality detection is performed for each of the first index value related to speed, the first index value related to acceleration, and the first index value related to the rotation speed. Similarly, abnormality detection is performed for each of the second index value related to speed, the second index value related to acceleration, and the second index value related to the rotation speed. Therefore, for example, when an abnormality is detected in all the index values, the roller 30 corresponding to the index value may have an abnormality sign, or the roller 30 may have an abnormality sign by using the weighted sum of the index values. It may be determined whether or not there is.

For example, when the weighted sum of each index value is used, the abnormality detection result using the first index value related to the speed of a certain roller 30 is a ₁ , and the abnormality detection result using the first index value related to acceleration is a. _2, the abnormality detection result using the first index value related to the rotation speed as a _3, if the abnormality in the roller 30 is detected 1, otherwise it is assumed to take 0. Further, the weights are w ₁ , w ₂ and w ₃ . At this time, if w ₁ a ₁ + w ₂ a ₂ + w ₃ a ₃ exceeds a predetermined threshold value, it is determined that the roller 30 has an abnormality sign, and if not, it is determined that there is no abnormality sign. May be good. In addition, the weighted sum for each second index value for the roller 30 may be added to the weighted sum.

Further, for example, after narrowing down the roller 30 having an abnormal sign with a certain index value, it may be determined whether or not the roller 30 after the narrowing down has a further abnormal sign. For example, after narrowing down the rollers 30 having an abnormality sign by detecting an abnormality using the first index value related to the number of revolutions, the roller 30 after being narrowed down is subjected to abnormality detection using the first index value related to acceleration. May be good.

Step S103: Next, the abnormality sign detection unit 102 determines whether or not the roller 30 having the abnormality sign is detected in the above step S103. Then, if the roller 30 having an abnormality sign is detected, the process proceeds to step S104, and if not, the abnormality sign detection process is terminated.

Step S104: Then, the notification unit 103 notifies a predetermined notification destination such as the roller ID of the roller 30 detected that there is an abnormality sign. As a result, the user or the like of the abnormality sign detection device 10 can know the roller 30 having the abnormality sign, and can, for example, replace or repair the roller 30.

<Calculation result of index value>
Here, FIG. 5 shows an example of the first index value calculated by the calculation method 1 in the above step S102. FIG. 5 shows the first index values (velocity standard deviation, acceleration standard deviation, and rotation speed standard deviation) of each roller 30 on a certain day d. The horizontal axis in FIG. 5 represents the roller ID, and the vertical axis represents the value of the first index value. "1", "2", ..., "X ₁ ", ..., "X ₂ ", ..., "N" are roller IDs.

In the example shown in FIG. 5, _{the first index value regarding the acceleration of the roller 30 of the roller ID “X 2} ” is higher than the first index value regarding the acceleration of the other rollers 30. Therefore, in this case, in step S102 above, _{the roller 30 with the roller ID “X 2} ” is detected as the roller 30 having an abnormality sign. Further, since the first index value regarding the speed of the roller 30 having the roller ID “X ₁ ” is relatively high, the roller 30 has an abnormal sign depending on the set value of _{the parameter (for example, L 1 described above).} Detected.

Further, FIG. 6 shows an example of the second index value related to acceleration among the second index values calculated by the calculation method 2 in step S102. FIG. 6 shows a second index value regarding the acceleration of each roller 30. In FIG. 6, the horizontal axis represents time (that is, the period m), and the vertical axis represents the value of the second index value (that is, the standard deviation of acceleration) regarding acceleration.

In the example shown in FIG. 6, the roller 30 having the roller ID “X” has a large deviation of the second index value regarding acceleration from the other rollers 30 as time advances. Therefore, in this case, in step S102, the roller 30 having the roller ID “X” is detected as the roller 30 having an abnormality sign.

As described above, in the abnormality sign detection system 1 according to the present embodiment, after calculating a predetermined index value indicating the degree of variation in the value of the operation information for each roller 30, the index value is compared between the rollers 30. As a result, the abnormality sign detection system 1 according to the present embodiment can detect the abnormality sign of the roller 30 with high accuracy.

The present invention is not limited to the above-described embodiment disclosed specifically, and various modifications and modifications can be made without departing from the scope of claims.

1 Abnormal sign detection system 10 Abnormal sign detection device 11 Input device 12 Display device 13 External I / F
13a Recording medium 14 Communication I / F
15 Processor 16 Memory device 17 Bus 20 Sensor 30 Roller 101 Index value calculation unit 102 Abnormality sign detection unit 103 Notification unit 104 Storage unit

Claims (7)

A storage means for storing operation information for each roller as time-series data,
A calculation means for calculating a predetermined index value indicating the degree of variation in the value of the operation information in a predetermined period for each roller.
A detection means for detecting the roller having an abnormal sign based on a predetermined index value calculated by the calculation means, and a detection means.
Abnormal sign detection system with. The calculation means is
For each roller, the average value, maximum value, or minimum value of the operation information value for each predetermined first period is calculated, and the calculated average value, maximum value, or minimum value is standard in a predetermined second period. The abnormality sign detection system according to claim 1, wherein the deviation or variance is calculated as the predetermined index value. The calculation means is
The first or second claim, wherein the standard deviation or variance of the value of the operation information in the predetermined third period is calculated as the predetermined index value for each roller and for each predetermined third period. Abnormal sign detection system. The abnormality sign detection system according to any one of claims 1 to 3, wherein the value of the operation information is at least two or more of speed, acceleration, and rotation speed. The detection means
By detecting the abnormal value of the predetermined index value using the Mahalanobis distance of the predetermined index value, the roller corresponding to the predetermined index value in which the abnormal value is detected is detected as a roller having an abnormality sign. , The abnormality sign detection system according to any one of claims 1 to 4. A storage means for storing operation information for each roller as time-series data,
A calculation means for calculating a predetermined index value indicating the degree of variation in the value of the operation information in a predetermined period for each roller.
A detection means for detecting the roller having an abnormal sign based on a predetermined index value calculated by the calculation means, and a detection means.
Abnormal sign detection device with. A storage procedure for storing operation information for each roller as time-series data in the storage unit,
A calculation procedure for calculating a predetermined index value indicating the degree of variation in the value of the operation information in a predetermined period for each roller, and a calculation procedure.
A detection procedure for detecting the roller having an abnormal sign based on a predetermined index value calculated in the calculation procedure, and a detection procedure for detecting the roller having an abnormal sign.
A method of detecting abnormal signs that the computer performs.

Images