Status monitoring device, status monitoring method, and status monitoring program

Abstract

To identify an abnormal period in a method other than discrimination of whether a change rate of a state quantity exceeds a prescribed threshold value.SOLUTION: A state monitoring device 10 comprises: a time-series data acquisition unit 11B acquiring moving average time-series data showing a time change of a state quantity of a field device; and an abnormality identification unit 11C configured so as to identify a period during which a vibration amplitude of a time change of a state quantity increases with time as an abnormal period during which a time change of a state quantity is abnormal on the basis of the moving average time-series data. The abnormality identification unit 11C identifies a section in which the number of standard deviations increasing against the immediately preceding standard deviation became a prescribed standard or higher as an abnormal period among a plurality of sections sequentially shifted along a time-series in a time change of standard deviation of a moving average value.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present invention relates to a status monitoring device, a status monitoring method, and a status monitoring program for monitoring the status quantities of field devices in an oil plant or the like. [Background technology]

[0002] Patent Document 1 discloses a technique for determining that a time-series data of a state quantity of a field device includes an abnormal period when the rate of change of the state quantity exceeds a predetermined threshold based on the time-series data of the state quantity. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2018-180703 A Summary of the Invention [Problem to be solved by the invention]

[0004] The technique described in Patent Document 1 has the disadvantage that it is not possible to identify an abnormal period that occurs when the rate of change does not exceed a predetermined threshold value.

[0005] An object of the present invention is to identify an abnormal period by a method other than judging whether or not the rate of change of a state quantity exceeds a predetermined threshold value. [Means for solving the problem]

[0006] In order to solve the above problem, a condition monitoring device according to a first aspect of the present invention comprises an acquisition unit configured to acquire time series data indicating a time change in a state quantity of a field device, and an identification unit configured to identify, based on the time series data, a period in which the vibration amplitude of the time change in the state quantity increases over time, as an abnormal period in which the time change in the state quantity is abnormal.

[0007] The identification unit may be configured to derive each standard deviation of the set of state quantities for each of a plurality of first intervals sequentially shifted along a time series in the time series data, and to identify, as the abnormal period, a second interval in which, in the time change of the standard deviation obtained by the derivation, a number of the standard deviations that is increased relative to the previous standard deviation is equal to or greater than a first criterion, among a plurality of second intervals sequentially shifted along the time series.

[0008] The identification unit may be configured to identify, as the abnormal period, a third section in which a number of the state quantities increasing relative to the previous state quantity is equal to or greater than a second criterion, among a plurality of third sections sequentially shifted along the time series in the time series data.

[0009] The identification unit may be configured to identify, as the abnormal period, a fourth section in which a number of the state quantities decreasing relative to the previous state quantity is equal to or greater than a third criterion, among a plurality of third sections sequentially shifted along the time series in the time series data.

[0010] The state quantity may be a moving average value based on an original state quantity.

[0011] The apparatus may further include a display control unit configured to display a graph of the time series data on a display device, and the display control unit may be configured to display a portion of the graph representing the abnormal period in a display manner different from a portion of the graph representing a normal period other than the abnormal period.

[0012] A condition monitoring method according to a second aspect of the present invention includes a step of identifying, based on time series data indicating a time change in a state quantity of a field device, a period in which a vibration amplitude of the time change in the state quantity increases over time, as an abnormal period in which the time change in the state quantity is abnormal.

[0013] A status monitoring program according to a third aspect of the present invention causes a computer to execute a step of identifying, based on time series data indicating a time change in a state quantity of a field device, a period in which a vibration amplitude of the time change in the state quantity is increasing over time, as an abnormal period in which the time change in the state quantity is abnormal. Effect of the Invention

[0014] According to the present invention, it is possible to specify an abnormal period by a method other than judging whether or not the rate of change of a state quantity exceeds a predetermined threshold value. [Brief description of the drawings]

[0015] [Figure 1] FIG. 1 is a configuration diagram of a condition monitoring device according to an embodiment of the present invention and an oil plant that is the object of monitoring by the device. [Diagram 2] FIG. 2 is a diagram showing the configuration of the state monitoring device. [Diagram 3] FIG. 3 is a diagram showing an example of the contents of the original time series data and the moving average time series data. [Figure 4] FIG. 4 is a flowchart of the abnormality identifying process executed by the abnormality identifying unit in FIG. [Diagram 5] FIG. 5 is a diagram showing the contents of the original time series data, the moving average time series data, the normal / abnormal discrimination result, and the like, which are stored in the storage device. [Figure 6] FIG. 6 is a graph showing each of the time series data in FIG. [Figure 7] FIG. 7 is a diagram showing the contents of the original time series data, the moving average time series data, the normal / abnormal discrimination result, and the like, which are stored in the storage device. [Figure 8] FIG. 8 is a graph of each time series data of FIG. [Figure 9] FIG. 9 is a diagram showing the contents of the original time series data, the moving average time series data, the normal / abnormal discrimination result, and the like, which are stored in the storage device. [Figure 10] FIG. 10 is a graph showing each of the time series data in FIG. [Figure 11] FIG. 11 is a graph of the standard deviation of FIG. [Figure 12] FIG. 12 is a diagram illustrating an example of the configuration of the abnormal period database. [Figure 13] FIG. 13 is a diagram showing an example of an image displayed on the display device. [Figure 14] FIG. 14 is a diagram showing an example of an image displayed on the display device. [Figure 15] FIG. 15 is a diagram showing an example of transition of an image displayed on a display device. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0016] Hereinafter, a condition monitoring device according to an embodiment of the present invention will be described with reference to the drawings.

[0017] 1, a status monitoring device 10 according to this embodiment is connected to an oil plant P. The status monitoring device 10 monitors the statuses of a first valve V1 to a fifteenth valve V15 (hereinafter simply referred to as valves V1 to V15), which are field devices possessed by the oil plant P.

[0018] The oil plant P is a facility that produces oil by refining heavy oil. The oil plant P includes valves V1 to V15 and a controller P1 that controls the valves V1 to V15. The valves V1 to V15 are provided to control the flow rate of oil, such as heavy oil and petroleum, flowing through pipes (not shown) that the oil plant P includes. The controller P1 acquires various measured values ​​that indicate the state of each of the valves V1 to V15 measured by the various sensors included in the oil plant P. The controller P1 feedback-controls the valve opening degree of each of the valves V1 to V15 using at least a part of the acquired various measured values ​​as feedback values. The valves V1 to V15 may be flow control valves, and in this case, the controller P1 supplies a control amount that specifies the flow rate to the valves V1 to V15. The controller P1 supplies the acquired measured values ​​and the output control amount to the state monitoring device 10 in real time. In other words, the controller P1 outputs the measured values ​​and the control amount to the state monitoring device 10 at the time when the measured values ​​and the control amount are acquired or output.

[0019] The condition monitoring device 10 is composed of a computer or the like, and includes a processor 11 such as a CPU (Central Processing Unit), a RAM (Random Access Memory) 12 functioning as a main memory for the processor 11, and a non-volatile storage device 13. The condition monitoring device 10 further includes a display device 14 that displays various images, an I / O (Input / Output) 15 connected to the oil plant P, and an operation device 16 that accepts input operations from a user of the condition monitoring device 10 who is an operator of the oil plant P. The I / O (Input / Output) 15 may be connected to the oil plant P via a network such as a LAN (Local Area Network) or the Internet. The operation device 16 includes an operation device such as a keyboard, a mouse, or a touch panel.

[0020] The storage device 13 includes a hard disk, a solid state drive (SSD), a flash memory, and the like, and stores a status monitoring program. The storage device 13 can also be said to be a non-transitory computer-readable storage medium that stores the status monitoring program. The processor 11 executes the status monitoring program stored in the storage device 13. As a result, the processor 11 operates as a status quantity acquisition unit 11A, a time-series data acquisition unit 11B, an abnormality identification unit 11C, and a display control unit 11D shown in FIG. 2. Note that the hardware configuration of each unit 11A to 11D is arbitrary, and at least a part of each unit 11A to 11D may be configured by other circuits such as an application specific integrated circuit (ASIC) or a field programmable gate array (FPGA).

[0021] The state quantity acquisition unit 11A shown in FIG. 2 acquires the above-mentioned measured values ​​and controlled variables supplied from the controller P1 of the oil plant P via the I / O 15. The state quantity acquisition unit 11A acquires state quantities A to H indicating the state of each of the valves V1 to V15 based on the acquired information. Any state quantities A to H are adopted as the state quantities A to H for detecting abnormalities in the valves V1 to V15. The state quantities A to H are acquired by the state quantity acquisition unit 11A every day. The state quantities A to H include, for example, an integrated value of the sliding distance of the valve in one day, the number of times the valve is fully closed in one day, the maximum frictional force of the valve in one day, the number of times the valve is reversed in one day, and the maximum magnitude of stick-slip of the valve in one day. The state quantities A to H have little or no fluctuation under normal conditions.

[0022] For example, the state quantity acquisition unit 11A acquires, via the I / O 15, the rotation angle of the disc of the valve V1 that is measured and output by the controller P1 using a rotary encoder or the like, and derives the sliding distance of the disc of the first valve V1 based on the acquired rotation angle. Note that derivation includes both calculation using an equation and acquisition by referring to a table prepared in advance (the same applies to derivation below). The state quantity acquisition unit 11A calculates the daily total of the derived sliding distances, and thereby acquires an integrated sliding distance value. Similarly, the state quantity acquisition unit 11A acquires integrated sliding distance values ​​for each of the valves V2 to V15.

[0023] For example, the state quantity acquiring unit 11A acquires a control amount indicating that the first valve V1 has been fully closed by the controller P1 via the I / O 15, counts the number of times this control amount is acquired per day, and acquires the counted number of times as the number of times the first valve V1 has been fully closed. Similarly, the state quantity acquiring unit 11A acquires the number of times each of the valves V2 to V15 has been fully closed.

[0024] For example, the state quantity acquisition unit 11A acquires the kinetic friction force between the valve disc and the valve seat of the first valve V1, which is measured and output by the controller P1 using a sensor that measures friction, via the I / O 15 each time the valve disc moves. The state quantity acquisition unit 11A acquires the maximum of the acquired kinetic friction forces for one day as the maximum friction force. Similarly, the state quantity acquisition unit 11A acquires the maximum friction force for each of the valves V2 to V15.

[0025] Similarly to the above, the state quantity acquiring unit 11A acquires the number of reversal operations of each of the valves V1 to V15, the maximum magnitude of stick-slip in one day, and the like, based on the measured values ​​and controlled variables from the controller P1.

[0026] The state quantity acquisition unit 11A acquires the state quantities A to H every day, and records the acquired state quantities A to H individually in the storage device 13 in chronological order. For example, the state quantity acquisition unit 11A records the state quantities A to H in the storage device 13 in association with the date data at the time of recording. By such processing, the storage device 13 records time series data indicating the time change of the state quantities for each of the valves V1 to V15 and each of the state quantities A to H. As described later, a moving average value is derived based on the state quantities A to H. This moving average value also represents the state of the valves V1 to V15. Therefore, this moving average value is also a type of state quantity. The state quantities A to H are values ​​on which the moving average value is based, and can be called original state quantities. This time series data of the original state quantities is also called original time series data. FIG. 3 shows the contents of the original time series data indicating the time change of the state quantity C as the maximum friction force, stored in the storage device 13. 3, here, original time series data of state quantity C from June 3 to June 24, 2021 is recorded in the storage device 13. Here, for other state quantities such as state quantity A, original time series data of each state quantity from June 3 to June 24, 2021 is also recorded in the storage device 13.

[0027] The time series data acquiring unit 11B shown in FIG. 2 acquires original time series data stored in the storage device 13 from the storage device 13 at any timing (for example, at the timing when an instruction to display a monitoring window 90 described later is given). The time series data acquiring unit 11B calculates a moving average value of state quantities for each of a plurality of fixed intervals sequentially shifted along the time series based on the acquired original time series data. Here, the plurality of fixed intervals are periods each including three days, i.e., three state quantities, and are intervals sequentially shifted by one day. For example, based on the original time series data of the state quantity C shown in FIG. 3, the time series data acquiring unit 11B derives a simple moving average value of the state quantity C for three consecutive days as the moving average value of the last day of the three days, and records it in the storage device 13 corresponding to the last day as shown in FIG. 3. For example, the time series data acquisition unit 11B records the average value (20.4) of the three state quantities C (20.8, 19.9, 20.5) from June 3 to June 5, 2021 in the storage device 13 as the moving average value on June 5, 2021. The average value (20.0) of the three state quantities C (19.9, 20.5, 19.6) from June 4 to June 6, 2021 in the storage device 13 as the moving average value on June 6, 2021. In this way, the time series data acquisition unit 11B acquires time series data indicating the time change of the moving average value of the state quantity C as moving average time series data and records it in the storage device 13. Similarly, the time series data acquisition unit 11B acquires time series data indicating the time change of each moving average value of other state quantities such as the state quantity A for each of the state quantities A to H of each of the valves V1 to V15 as moving average time series data and records it in the storage device 13. As described above, this moving average value is also a state quantity that represents the state of the valve, like the state quantities A to H.

[0028] 2 identifies, among the multiple moving average time series data acquired by the time series data acquisition unit 11B, moving average time series data having an abnormal period in which the time change of the state quantity (here, the moving average value) satisfies a predetermined criterion as abnormal time series data, and identifies moving average time series data having no abnormal period as normal time series data. An abnormal period is also a period of trend change in the state quantity (for example, a certain trend changes to an upward or downward trend).

[0029] 4 at an arbitrary timing (for example, at the timing when an instruction to display a monitoring window 90 described later is given), the abnormality identification unit 11C performs the above identification. This abnormality identification process is performed by repeatedly executing the moving average time series data of each of the state quantities A to H of the valves V1 to V15 individually and sequentially for the number of pieces of moving average time series data.

[0030] In the abnormality identification process of FIG. 4, the abnormality identification unit 11C first detects an interval in which the number of moving average values ​​that is increased relative to the previous moving average value (state quantity) is four or more from a plurality of intervals sequentially shifted along the time series in the moving average time series data to be processed (step S11).

[0031] When the moving average time series data to be processed is as shown in FIG. 5, in step S11, the abnormality identifying unit 11C identifies whether each moving average value is rising or falling relative to the moving average value immediately above (the previous one), and stores the identified results in the storage device 13 in association with each moving average value (see "Change Type" in FIG. 5). After that, for each moving average value, the abnormality identifying unit 11C counts the rise / fall between that moving average value and the four moving average values ​​immediately preceding that moving average value. For example, in the section of indexes 8 to 12 in FIG. 5, two rises and three falls are counted. The count results are recorded in the storage device 13 in association with the "moving average value" (see the "Rise / Fall Count Result" column in FIG. 5).

[0032] Thereafter, the abnormality identifying unit 11C identifies the above-mentioned section as "abnormal" and identifies sections other than this section as "normal" (step S12). The abnormality identifying unit 11C associates the moving average values ​​(indexes 17 to 20) that have risen by four or more in the counting result with information of "abnormal" and stores them in the storage device 13 (see the "Normal / Abnormal" column). The abnormality identifying unit 11C associates the other moving average values ​​with information of "normal" and stores them in the storage device 13 (see the "Normal / Abnormal" column).

[0033] FIG. 6 is a diagram showing each graph of the original time series data and the moving average time series data in FIG. 5. Each graph is displayed on the display device 14 in the manner shown in FIG. 6 by the display control unit 11D based on the original time series data, the moving average time series data, and the above-mentioned "normal" / "abnormal" data. The black circle in the moving average time series data indicates the moving average value associated with "abnormal", and indicates that a certain period before this moving average value (a period of 5 days including this moving average value) is an abnormal period. Black circles may be added to all moving average values ​​in this certain period (similar to other graphs). As shown in FIG. 6, in a section with four or more increases (a section with a high frequency of increases), the moving average value has an upward trend. That is, steps S11 and S12 can detect a section in which the moving average value has an upward trend as an abnormal period.

[0034] Thereafter, the abnormality identifying unit 11C detects an interval in which the number of moving average values ​​that are lower than the previous moving average value (state quantity) is four or more from a plurality of intervals sequentially shifted along the time series in the moving average time series data to be processed (step S13). Thereafter, the abnormality identifying unit 11C identifies the detected interval as "abnormal" and identifies intervals other than this interval as "normal" (step S14).

[0035] A specific example of steps S13 and S14 is similar to the description of steps S11 and S12 (just change "up" to "down"). When the moving average time series data to be processed is as shown in FIG. 7, the information "abnormal" is associated with the moving averages of indexes 16 to 20. FIG. 8 is a diagram showing each graph of the original time series data and the moving average time series data of FIG. 7. Each graph is displayed on the display device 14 in the manner shown in FIG. 8 based on the original time series data, the moving average time series data, and the above-mentioned "normal" / "abnormal" data by the display control unit 11D. The black circle of the moving average time series data indicates the moving average value associated with "abnormal", and indicates that a certain period before this moving average value (a period of 5 days including this moving average value) is an abnormal period. As shown in FIG. 8, in a section with four or more declines (a section with a high frequency of declines), the moving average value has a downward trend. That is, steps S13 and S14 can detect a section in which the moving average value has a downward trend as an abnormal period.

[0036] Thereafter, the abnormality identifying unit 11C derives a standard deviation based on the moving average value of the moving average time series data (step S15). The abnormality identifying unit 11C derives each standard deviation of a set of moving average values ​​for each of a plurality of intervals sequentially shifted along the time series in the moving average time series data, similar to the derivation of the moving average value. Here, the plurality of intervals are periods each including five days, i.e., five moving average values, and are intervals sequentially shifted by one day. When the moving average time series data shown in FIG. 9 is to be processed, the abnormality identifying unit 11C derives the standard deviation of the moving average values ​​for five consecutive days based on the moving average time series data as the standard deviation of the last day of the five days, and records it in the storage device 13 in association with the moving average value of the last day. For example, the abnormality identifying unit 11C derives the standard deviation of the five moving average values ​​from June 5 to June 9, 2021 as the standard deviation of June 9, 2021.

[0037] Thereafter, the abnormality identifying unit 11C detects an interval in which the number of standard deviations that increases relative to the previous standard deviation is four or more among a plurality of intervals sequentially shifted along the time series in the time change (daily transition) of the derived standard deviation (step S16).Then, the abnormality identifying unit 11C identifies the detected interval as "abnormal" and identifies intervals other than this interval as "normal" (step S17).

[0038] A specific example of steps S16 and S17 is similar to the description of steps S11 and S12 (just change "moving average" to "standard deviation"). Each of the above-mentioned multiple intervals is an interval of five standard deviations, and is an interval shifted by one day. When the standard deviation to be processed is as shown in FIG. 9, the moving averages of indexes 10 and 16 to 20 are "abnormal". FIG. 10 is a diagram showing each graph of the original time series data and the moving average time series data in FIG. 9. As described later, each graph of the original time series data and the moving average time series data is displayed on the display device 14 in the manner shown in FIG. 10 by the display control unit 11D based on the original time series data, the moving average time series data, and the above-mentioned "normal" / "abnormal" data. The black circle of the moving average time series data indicates a moving average value associated with a count result of four or more standard deviations of increase, that is, "abnormal", and indicates that a certain period before this moving average value (a period of five days including this moving average value) is an abnormal period. As shown in Fig. 10, in a section where the standard deviation increases four or more times, that is, in a section where the standard deviation increases frequently as shown in Fig. 11, the degree of dispersion of the moving average value tends to increase, and when the moving average value oscillates over time, its amplitude (for example, the difference between an extreme value in the time change of the state quantity and the average of the moving average value under normal conditions) gradually increases. Steps S16 and S17 make it possible to detect a section where the amplitude gradually increases over time as an abnormal period. The graph in Fig. 11 may be displayed on the display device 14 together with the graph in Fig. 10.

[0039] Thereafter, the abnormality identifying unit 11C updates the abnormality period database recorded in the storage device 13 (step S18). As shown in FIG. 12, the abnormality period database stores abnormality period information indicating "present" or "absent" of an abnormality period in association with each of the state quantities A to H of the valves V1 to V15. If the above-mentioned interval is not detected in all of steps S11, S13, and S16, that is, if there is no "abnormality," the moving average time series data to be processed this time is normal time series data that does not include an abnormality period (time series data in which the fluctuation of the moving average is within an allowable range, for example, time series data with no fluctuation). In this case, the abnormality identifying unit 11C updates the abnormality period information corresponding to the state quantity of the moving average time series data to be processed in the abnormality period database to "absent." On the other hand, if the above-mentioned interval is detected in at least one of steps S11, S13, and S16, that is, if there is "abnormality," the moving average time series data to be processed this time is abnormal time series data that includes an abnormality period. In this case, the abnormality identifying unit 11C updates the abnormality period information corresponding to the state quantity of the moving average time series data to be processed in the abnormality period database to "present".

[0040] Through the above series of processes, the abnormality identifying unit 11C identifies the moving average time series data to be processed as either normal time series data or abnormal time series data. Note that any of steps S11 to S12, steps S13 to S14, and steps S15 to S17 may be omitted.

[0041] The display control unit 11D of FIG. 2 displays the monitoring window 90 shown in FIG. 13 on the display device 15 when the user issues an instruction to the operation device 16 to display a monitoring window 90 for monitoring the states of the valves V1 to V15. The monitoring window 90 includes a dashboard image 91 and an input area 95 for inputting processing conditions for time-series data. Note that the eighth valve is missing in FIG. 13. The display control of the images described below is performed by the display control unit 11D. It is assumed that data necessary for the display control is stored in the storage device 13.

[0042] Dashboard image 91 displays a pie chart 91A that displays the percentage of valves that had an abnormal period (valves corresponding to the time-series data in which "abnormal" was recorded above) and the like.

[0043] The dashboard image 91 further includes a list display image 91B displaying a list of a plurality of state quantity images (cells such as "present" and "absent") 91BA representing each of the state quantities A to H of the valves V1 to V15. The list display image 91B is displayed so as to be scrollable by a scroll bar. The plurality of state quantity images 91BA are displayed in a matrix with the vertical axis representing the type of valve and the horizontal axis representing the type of state quantity. On the vertical axis, device images (images of device names such as "first valve") 91BB representing each of the valves V1 to V15 by their names are displayed in a vertical row. On the horizontal axis, type images (images such as "state quantity A") 91BC representing the type of state quantity by its name are displayed in a horizontal row. The state quantity images 91BA represent the state quantities by their relative positions with respect to the vertical and horizontal axes. That is, the state quantity represented by the state quantity image 91BA is identified as the state quantity of which valve and which of the state quantities A to H it is, based on the relative positions. The vertical and horizontal axes may be reversed.

[0044] Here, the state quantity image 93BA is expressed by "Yes" or "No" which indicates whether or not the time series data of the state quantity represented by the state quantity 93BA includes an abnormal period. The display control unit 11D identifies "Yes" or "No" by referring to the abnormal period database in the storage device 13. The state quantity image 93BA may be expressed by a character string indicating at least one of the valve name and the state quantity name. In this case, at least one of the vertical axis and the horizontal axis may be omitted. The state quantity image 93BA may be plain. In this case, the vertical axis and the horizontal axis are required to represent the state quantity by the above-mentioned relative position. Note that the state quantity image 91BA with "Yes" (for example, the state quantity image 91BA representing the state quantity C of the sixth valve V6), that is, the state quantity image 91BA representing the state quantity with an abnormal period, is displayed in a first display mode. The state quantity image 91BA with "No" (for example, the state quantity image 91BA representing the state quantity A of the sixth valve V6), that is, the state quantity image 91BA representing the state quantity without an abnormal period, is displayed in a second display mode different from the first display mode. Different display modes include different colors of characters (essential elements for the image) such as "Yes" and "No" or different backgrounds (colors, patterns, etc.) other than the characters in the cell. For example, the first display mode is a mode in which the background color in the cell is white, and the second display mode is a mode in which the background color in the cell is pink. In this way, by changing the display mode depending on the presence or absence of an abnormal period, the presence or absence of an abnormal period is notified to the user in an easy-to-understand manner.

[0045] The user can use the operation device 16 to select one of the multiple state quantity images 91BA and the multiple device images 91BB. Here, it is assumed that the state quantity image 91BA or device image 91BB related to the sixth valve V6 is selected. In this case, as shown in FIG. 14, a tab 92A for the sixth valve V6 is newly generated, and the sixth valve image 92 including the tab 92A is displayed instead of the dashboard image 91. Note that the display is switched to the dashboard image 91 by selecting a tab 91C of the dashboard image 91.

[0046] The sixth valve image 92 includes a state quantity type image 92B representing each of the state quantities A to H, and a graph image 92C displaying a list of graphs of the time series data of each of the state quantities A to H of the sixth valve image 92 (original time series data and moving average time series data) arranged vertically. The graph image 92C is displayed in a scrolling manner. This makes it easy for the user to compare the graphs displayed in a list. When any of the state quantity type images 92B is selected by the user, the graph image 92C is scrolled so that a graph of the time series data of the state quantity represented by the selected state quantity type image 92B is displayed.

[0047] As shown in FIG. 15, when the device image 91BB representing the sixth valve V6 is selected and the sixth valve image 92 is displayed, the top portion of the graph image 92C is displayed first. On the other hand, when the state quantity image 91BA corresponding to the sixth valve V6 is selected and the sixth valve image 92 is displayed, the graph of the time series data of the state quantity represented by the selected state quantity image 91BA is displayed first for the graph image 92C. As shown in FIG. 15, when the state quantity image 91BA representing the state quantity C is selected, the graph of the time series data of the state quantity C is displayed first. This display reduces the user's effort to manually search for the graph of the state quantity represented by the selected state quantity image 91BA. Here, the state quantity image 91BA representing the state quantity C of the sixth valve V6 is set to "present", and the time series data of the state quantity C is displayed in the manner shown in FIG. 6, 8, and 10. That is, a graph in which the abnormal period and the other normal period are displayed in different manners by black circles is displayed. This allows the user to easily identify the abnormal period. The user checks this graph and considers subsequent measures. Any specific means may be used to differentiate the display mode of the abnormal period from the display mode of the other normal period. For example, the color and / or shape of the graph itself may be differentiated. More specifically, in addition to adding the black circle, an image with a different shape from the black circle may be added, or at least one of the color, thickness, and pattern (solid line, dotted line, dashed line, etc.) of the graph may be changed.

[0048] A maintenance designation image 93A is arranged next to the graph of the time series data of the state quantity C. When selected by the user, the image transitions to a screen for changing control parameters (such as set flow rates) indicating the control modes of the valves V1 to V15, which are set in the controller P1 of the oil plant P. When the maintenance designation image 93A is designated by the user, the display control unit 11D may display a parameter setting image, and when the control parameters have been changed, set the new control parameters in the controller P1 via the I / O 15.

[0049] Next to the graph of the time series data of the state quantity C, an image 93B indicating that it is necessary to take action on the sixth valve V6 in the near future, an image 93C indicating that the progress of the state quantity C is to be observed, and an image 93D indicating that it is not necessary to observe the progress of the state quantity C and that no action is required are also displayed. When the image 93B is selected by the user, the display mode of at least the former of the state quantity image 91BA representing the state quantity C of the sixth valve V6 and the device image 91BB representing the sixth valve V6 in the list display image 91B that is displayed again by selecting the tab 91C of the dashboard image 91 is changed to a third display mode different from the first and second display modes. For example, each display mode is set to the color of a cell, the second display mode is set to white, the first display mode is set to orange or pink, and the third display mode is set to red. In this way, the urgency can be notified to the user by setting the third display mode to the darkest color. When the image 93C is selected, the display mode of the state quantity image 91BA representing the state quantity C of the sixth valve V6 and the device image 91BB representing the sixth valve V6 is not changed. When the image 93D is selected, the display mode of the state quantity image 91BA representing the state quantity C of the sixth valve V6 and the device image 91BB representing the sixth valve V6 is changed to the first display mode. In this case, the state quantity image may be changed to "none". This prevents the continuation of notification of an abnormal period with no problem. Note that when there is an abnormal period for another state quantity, it is preferable not to change the display mode of the device image 91BB to the first display mode.

[0050] In the input area 95 of FIG. 13, the user inputs processing conditions to be processed by the time series data acquisition unit 11B, the abnormality identification unit 11C, and the like. The input is performed by the operation device 16. The "analysis period" is determined by inputting the "period" and the "end date and time". The time series data acquisition unit 11B, the abnormality identification unit 11C, and the like perform processing using the time series data of this analysis period. The "number of moving average samples" is the number of original state quantities that are the basis for deriving the moving average value. The "number of standard deviation samples" is the number of moving average values ​​that are the basis for deriving the standard deviation. By appropriately managing these input values, it is possible to switch between monitoring the state from a long-term perspective or a short-term perspective, and the accuracy of identifying the abnormal period, etc. can be improved.

[0051] As described above, in this embodiment, a period (FIG. 11) in which the vibration amplitude of the time change of the state quantity (e.g., the difference from the average value of the state quantity in a normal state) increases over time is identified as an abnormal period in which the time change of the state quantity is abnormal, based on time-series data indicating the time change of the state quantity. Therefore, an abnormal period is identified by a method other than determining whether the rate of change of the state quantity exceeds a predetermined threshold value. For example, when the state quantity is the maximum frictional force of a valve, the valve may be damaged, causing the maximum frictional force to become unstable and its vibration amplitude to increase over time. In this case, the rate of change of the maximum frictional force may not increase so much. In this way, according to this embodiment, it is also possible to identify an abnormal period that cannot be identified by comparing the rate of change with a threshold value.

[0052] In this embodiment, the abnormality identifying unit 11C derives each standard deviation of a set of state quantities in each of a plurality of first sections sequentially shifted along the time series in the time series data, and identifies, as the abnormal period, a second section in which the number of standard deviations increasing relative to the previous standard deviation is equal to or greater than a first criterion among a plurality of second sections sequentially shifted along the time series in the time change of the standard deviation obtained by the derivation (steps S16 and S17 in FIG. 4). In this way, the abnormal period is identified by a method other than determining whether or not the rate of change of the state quantity exceeds a predetermined threshold. With this configuration, a period in which the slope of the change of the state quantity increases over time may also be identified as an abnormal period. The period in which the vibration amplitude of the time change of the state quantity increases over time may be identified by other methods. For example, the abnormal period may be identified by detecting that the maximum value of the time change of the state quantity increases over time or that the minimum value decreases over time.

[0053] In this embodiment, the abnormality identifying unit 11C identifies, as the abnormality period, a third section in which the number of the state quantities increasing relative to the previous state quantity is equal to or greater than a second reference among a plurality of third sections sequentially shifted along the time series in the time series data (steps S11 and S12 in FIG. 4). This makes it possible to identify an abnormality period in which the state quantities are on an upward trend.

[0054] In this embodiment, the abnormality identifying unit 11C identifies a fourth section in which the number of the state quantities that are decreasing relative to the previous state quantity is equal to or greater than a third criterion as the abnormal period among a plurality of third sections sequentially shifted along the time series in the time series data (steps S13 and S14 in FIG. 4). This makes it possible to identify an abnormal period in which the state quantities are tending to decrease.

[0055] Although the state quantity may be the original state quantity, it is preferable to use a moving average value based on the original state quantity, which can prevent an abnormal period from being erroneously detected due to a noisy state quantity.

[0056] Furthermore, the display control unit 11D displays a graph of the time-series data, and displays the portion of the graph that represents the abnormal period in a different display mode (with a black circle) from the portion that represents a normal period other than the abnormal period on the display device. This allows the abnormal period in the graph to be easily notified to the user. Note that the "different display mode" includes adding a figure such as the black circle (for example, a point that indicates a physical quantity and has a different shape from that in a normal state), as well as using different colors, different thicknesses, different line types such as dotted lines and solid lines.

[0057] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-mentioned embodiments. For example, the present invention includes various modifications to the above-mentioned embodiments that can be understood by a person skilled in the art within the scope of the technical concept of the present invention.

[0058] For example, the present invention is applicable to general techniques for monitoring field devices other than valves. Field devices include plant devices constituting various plants. The field devices may be various sensors such as flow meters. The state quantity of the field device to be monitored may be, for example, a value obtained by measuring the state of the field device using a sensor attached to the field device, a control amount or information input to the field device, or a measurement value measured by the field device itself. Furthermore, the state quantity may be a continuously changing physical quantity such as a flow rate, pressure, temperature, or vibration, or a moving average value thereof. By identifying a period in which the vibration amplitude of the time change of such a physical quantity increases over time (see steps S16 and S17 in FIG. 4), an abnormal period due to a hunting phenomenon, resonance, or the like is identified. The state quantity may be a value derived from the original state quantity as a value other than the moving average value. The time series data acquisition unit 11B may acquire time series data stored in a storage unit provided outside the state monitoring device. The graph of the time series data may be only a graph of the moving average time series data. Based on the original time series data, the above-mentioned various processes such as identification of the abnormal period may be performed. In this case, the graph of the time series data may be only the graph of the original time series data. [Explanation of symbols]

[0059] 10...condition monitoring device, 11...processor, 11B...time series data acquisition unit, 11C...abnormality identification unit, 11D...display control unit, 13...storage device, 14...display device, 16...operation device, 91B...list display image, 91BA...status quantity image, 91BB...equipment image, 91BC...type image, 92C...graph image, 93B to 93D...images.

Claims

1. An acquisition unit configured to acquire time series data indicating a time change in a state quantity of a field device; an identification unit configured to derive a standard deviation of the set of state quantities for each of a plurality of first sections sequentially shifted along the time series of the time series data based on the time series data, and to identify a second section, among a plurality of second sections sequentially shifted along the time series in the time change of the standard deviation obtained by the derivation, in which the number of the standard deviations increasing relative to the previous standard deviation is equal to or greater than a first criterion, as an abnormal period in which an oscillation amplitude of the time change of the state quantity increases over time and the time change of the state quantity is abnormal, the state quantity is a moving average value based on an original state quantity, The number of the first criterion is a part of the number of the standard deviations in each of the second intervals, and when the number of the standard deviations is equal to or greater than the number of the first criterion, the number is a number that can be said to indicate that the dispersion of the moving average values ​​is tending to increase. Condition monitoring device.

2. An analysis period, which is the length of a period of time series data acquired by the acquisition unit, the number of state quantities on which the moving average value is based, and the number of state quantities on which the standard deviation is based can be designated by a user. The condition monitoring device according to claim 1 .

3. The identification unit is further configured to identify, as the abnormal period, a third section in which a number of the state quantities increasing relative to the previous state quantity is equal to or greater than a second reference, among a plurality of third sections sequentially shifted along the time series in the time-series data. The condition monitoring device according to claim 1 or 2.

4. The identification unit is further configured to identify, as the abnormal period, a fourth section in which a number of the state quantities decreasing relative to the immediately preceding state quantity is equal to or greater than a third criterion among a plurality of third sections sequentially shifted along the time series in the time-series data. The condition monitoring device according to any one of claims 1 to 3.

5. a display control unit configured to display a graph of the time series data on a display device; The display control unit is configured to display, on a display device, a portion of the graph representing the abnormal period in a display mode different from a portion of the graph representing a normal period other than the abnormal period. The condition monitoring device according to any one of claims 1 to 4.

6. A computer is caused to function as the condition monitoring device according to claim 1. Condition monitoring program.

Images