JPH10171519A - Method and device for monitoring process data - Google Patents

Abstract

PROBLEM TO BE SOLVED: To quickly monitor process data by previously defining and monitoring inclinations related to plural process variables. SOLUTION: Monitoring scenario definition for defining an inclination monitoring pattern related to plural process variables to be monitored is used, and when a monitoring start condition is satisfied, monitoring to be executed while dynamically changing upper and lower limit values is started (step S2). If a process variable is deviated from the upper and lower limit values (step S6), whether a pattern having the same direction as the defined inclination monitoring pattern exists or not is judged (step S7). When the pattern exists, monitoring to be executed by shifting relative time up to the pattern and monitoring to be executed by returning the advanced time to the original state are successively executed (step S8), and when the pattern does not exist, an operator is informed of abnormality generation (step S12).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention utilizes an alarm scenario definition table that defines, as an alarm scenario, a time-varying behavior between a plurality of processes to be monitored in process operation monitoring in the industrial field. And a monitoring method.

[0002]

2. Description of the Related Art Conventionally, in this type of process control system, upper and lower limits are usually monitored as fixed values for process data during plant operation. For this reason, there may be a case where it is necessary to perform an operation of analyzing a temporal change of a plurality of process data using sufficient knowledge of a target process and then monitor operation of a temporal behavior. .

[0003]

That is, in the above-described method, it is extremely difficult to instantaneously determine normal or abnormal for a plurality of process data in operation monitoring of an industrial process. There is also a problem that there is an individual difference in the knowledge about the process, and it is difficult to uniformly monitor the process. Accordingly, an object of the present invention is to improve safety and reliability by enabling quick and appropriate monitoring.

[0004]

In order to solve such a problem, according to the first aspect of the present invention, when monitoring the time-varying trend of time-series process data, the following method is used.
A trend monitoring pattern relating to a plurality of process variables to be monitored is defined in advance as an alarm scenario, and when a predetermined monitoring start condition is satisfied, monitoring based on the alarm scenario is started, and the process variable is set to an upper and lower threshold. When it deviates, it is determined whether or not a trend monitoring pattern indicating the same change direction exists.If the same trend monitoring pattern exists, the trend monitoring pattern is moved to the existing position in the future time direction. Monitoring is performed, and monitoring is performed sequentially with the time advanced. If there is no trend monitoring pattern, it is determined that an error has occurred, and the operator is notified of the occurrence of the error. ,
When the monitoring end condition is satisfied without deviating from the upper and lower thresholds, it is determined that the monitoring has ended normally and the operator is notified of the normal end.

According to the second aspect of the present invention, an input means for inputting time-series process data, an alarm scenario definition table in which a trend monitoring pattern relating to a plurality of process variables to be monitored is defined as an alarm scenario, and an alarm scenario definition table. It is characterized by comprising processing means for executing a monitoring process for monitoring the time-series process data, and output means for outputting a processing result. According to the third aspect of the present invention, when the accumulated process data or online real-time data is input to the processing means and an allowable range is designated, the processing means outputs the history data,
Alternatively, a learning process of searching the online real-time data for a start point where the monitoring start conditions match, determining an upper limit and a lower limit in consideration of an allowable range for the data, and storing the upper limit and the lower limit in the alarm scenario definition table is executed. ing.

According to the fourth aspect of the present invention, the alarm scenario definition table can be interactively edited from an alarm scenario definition screen and a time-series scenario data editing screen (Gantt chart). Further, in the invention according to claim 5, regarding the abnormality detection information output to the operator via the output means, the date and time of occurrence, the cause of the abnormality, the action,
The monitoring threshold value and process data at that time are stored as notification history data. In the invention according to claim 6, a list of the trend monitoring pattern used for the abnormality detection based on the notification history data, a value at that time, and a movable time width of the trend monitoring pattern is displayed in a list to analyze an abnormal factor and correct the trend monitoring pattern. A scenario analysis screen that supports changes is output.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The first embodiment of the present invention is shown in FIGS.
Description will be given based on FIG. 1 is a flowchart for explaining an embodiment of the present invention, FIG. 2 is an explanatory diagram for explaining an example of a scenario definition table, and FIG. 3 is an explanatory diagram showing FIG. 2 in a graph. FIG. 3A shows the ignition flag, FIG. 3B shows the “generator output” as a monitoring target,
(C) “Main steam temperature” as a monitoring target, and (d) indicate evaluation (sampling) timing. In FIG. 1, an alarm scenario definition that defines a trend monitoring pattern relating to a plurality of process variables to be monitored is read in order to monitor a temporal change tendency of time-series process data (table acquisition in step S1 in FIG. 1). reference).
Thereafter, when a predetermined monitoring start condition is satisfied (see S2), the driver (operator) is notified that the trend monitoring has been started (see S3). The monitoring start condition is when the ignition flag changes from OFF to ON in the examples of FIGS.

At the same time, monitoring is started while dynamically changing upper and lower thresholds for a plurality of process variables to be monitored (see S4 and S5), and even if only one process variable is monitored, When the value deviates from the lower threshold value, it is determined whether the value deviates from the upper limit value or the lower limit value (see S6). Now, in the examples of FIGS. 2 and 3, for example, if the monitored generator output shows a measurement value of “8.5” at the sampling timing of t3,
This means that the upper limit deviates and the direction of change rises, and this direction of change is indicated by an arrow in FIG. 2 as a trend monitoring pattern. Although FIG. 2 shows an example of rising, equilibrium, and falling, other cases such as a sudden rise and a sudden fall can also be considered.

When the deviation exceeds the upper and lower thresholds, the pre-defined trend monitoring pattern is sequentially shifted (shifted) within the movable time width in the future time direction, and is shifted in the same direction as the deviation from the upper and lower thresholds. It is determined whether or not there is a trend monitoring pattern (see S7), and if so, it is determined that the timing of the change is earlier than the predefined trend monitoring pattern, and the position where the trend monitoring pattern matches is determined. As a starting point, the monitoring is continued while dynamically changing the upper and lower thresholds for the process variable (see S8), and the operator is notified of that (see S9). On the other hand, if it is determined in step S7 that the same trend monitoring pattern as the direction deviating from the upper and lower thresholds is not within the movable time width, it is determined that an abnormality has occurred and the operator is notified of the occurrence of the abnormality. (See S12).

In the examples of FIGS. 2 and 3, an abnormal rise (exceeding the upper and lower thresholds) occurred at t3, but the same pattern is present at t5, which is within the movable time width. In this case, the monitoring is continued with the pattern for monitoring the trend of the generator output being advanced (advancing in the time direction in the future). Further, the time of the generator output advanced to match the trend monitoring pattern coincides with the time at which monitoring was started (synchronous) when the trend monitoring pattern became equilibrium (→) (t7 in FIG. 2). Until equilibrium (→)
Monitoring is continued in the state of, and when synchronization is established, t8 and t9
... and advance the time. Note that the “trend monitoring relative time change” process in step S8 of FIG. 1 also includes such a process. Thereafter, it is determined whether or not all the process variables to be monitored satisfy the monitoring end condition without deviating from the upper and lower thresholds (see S10).
If satisfied, it is determined that the trend monitoring has been completed normally, and the normal termination is notified to the operator (see S11).

An embodiment of the second invention will be described with reference to FIG. Reference numeral 100 denotes a processing device such as a computer, which executes the monitoring process described with reference to FIG. Reference numeral 400 denotes an alarm scenario definition table as shown in FIG. 2, which defines a trend monitoring pattern for a plurality of process variables to be monitored. 200 is an input device and 300 is an output device. Therefore, when information on a plurality of processes to be monitored is input via the input device 200, the processing device 100 dynamically changes the upper and lower thresholds for the plurality of processes to be monitored based on the contents of the alarm scenario definition table 400. Monitoring while executing, and if it deviates from the upper and lower thresholds, it is determined whether there is the same trend monitoring pattern as the direction deviating from the upper and lower thresholds within the movable time width in the future time direction, In some cases, it is judged that the timing of the change is earlier than the predefined trend monitoring pattern, and the upper and lower thresholds for the process variable are dynamically changed starting from the position where the trend monitoring pattern matches. While monitoring while continuing, if it is determined that there is no trend monitoring pattern that is the same as the direction deviating from the upper and lower thresholds, it is considered that an abnormality has occurred. Te, and notifies the abnormality to the operator via the output device 300.

An embodiment of the third invention will be described with reference to FIG. This shows an example in which a learning function is provided, in which process data or online real-time data previously stored in a memory or a database 900 is read, and a period of history data, a monitoring start condition, When a process variable to be monitored, an allowable range, and the like are specified, the processing device 100
Searches the accumulated process data or online real-time data for a starting point that matches the monitoring start condition, and sets upper and lower thresholds in consideration of an allowable range with respect to the values of these data, and sets the upper and lower thresholds in the alarm scenario definition table 400. Output to

An embodiment of the fourth invention will be described with reference to FIGS. FIG. 6 shows the creation and editing of the alarm scenario definition table 400. In this case, an alarm scenario definition screen 600 as shown in FIG. 7 and a time-series scenario data editing screen (scenario Gantt chart) as shown in FIG. Or simply Gantt chart) 700 to facilitate interactive editing. In addition,
The result is output to the alarm scenario definition table 400. In the alarm scenario definition screen in Fig. 7, the scenario name is "boiler ignition monitoring", the notification content at the time of normal termination is "boiler ignition normal termination", and the notification content at the time of abnormal termination is "boiler ignition abnormal termination". The targets to be monitored are “ignition”, “generator output”, “main steam temperature”,
"Main steam pressure", "Reheat steam temperature", "Rotation speed",
"Ventilation signal", "combined signal", "full load", and the like. In the time-series scenario data editing screen (scenario Gantt chart) in FIG. 8, the monitoring trigger (monitoring start condition) is “ignition”, and the upper and lower limits of the monitoring target such as the generator output and the main steam temperature are displayed. Also, a trend monitoring pattern such as an ideal behavior curve (solid line) is shown.

An embodiment of the fifth invention will be described with reference to FIG. By giving a predetermined instruction from the input device 200, the processing device 100 can operate the output device 30 as shown in FIG.
0, and the date and time of occurrence, the cause of the error (such as upper limit error, lower limit error, etc.), the action (such as shifting the trend monitoring pattern and continuing the error monitoring, etc.),
The monitoring threshold value and the process data value at that time are output to the notification history data memory or the database 500.

An embodiment of the sixth invention will be described with reference to FIG. This is because the processing apparatus 100 reads the alarm scenario definition table and the notification history data accumulated as shown in FIG. 9, and uses the trend monitoring pattern used for abnormality detection, the value at that time, and the pattern shift time (movable A time range) is displayed as a list, and a scenario analysis screen 800 for supporting analysis of an abnormal factor or the like and correction and change of a trend monitoring pattern is output. FIG. 11 shows an example of the scenario analysis screen. In this example, the alarm scenario for which the analysis result is to be displayed is “boiler ignition” and the analysis period is “199
From October 17, 2006 to October 24, the same year. Further, the monitoring of the time-varying direction of the time-series process data based on the alarm scenario of “boiler ignition” is started 15 times, and two items are attached to the monitoring target “rotational speed” and one item “generator output”. Matter, `` Main steam output ''
The upper and lower threshold abnormalities are detected in 5 cases. In particular, the details of the “main steam output” exceed the set value (upper threshold) “14” and “15”, “16” And the like.

[0016]

According to the first and second aspects, the time behavior of a plurality of processes is defined and monitored in advance in the alarm scenario definition table, so that the time behavior of a plurality of process data can be quickly grasped. be able to. According to the third aspect, the alarm scenario definition table can be automatically generated by using the accumulated process data or online real-time data. According to the fourth aspect, the alarm scenario definition table can be used. The contents of the definition table can be edited by operating the operator. Further, according to the fifth invention, it is possible to accumulate the monitoring results that have been notified in the past, and according to the sixth invention, the analysis is performed from the accumulated monitoring results and the alarm scenario definition table, and the results are analyzed. , Etc. can be notified.

[Brief description of the drawings]

FIG. 1 is a flowchart for explaining an embodiment of the present invention.

FIG. 2 is an explanatory diagram showing an example of an alarm scenario definition table.

FIG. 3 is an explanatory diagram showing a graph of FIG. 2;

FIG. 4 is a block diagram showing a system configuration example for implementing FIG. 1;

FIG. 5 is a system configuration diagram showing another embodiment of the present invention.

FIG. 6 is a system configuration diagram showing still another embodiment of the present invention.

FIG. 7 is an explanatory diagram showing an example of an alarm scenario definition screen.

FIG. 8 is an explanatory diagram of a time-series scenario data editing screen (scenario Gantt chart).

FIG. 9 is a system configuration diagram showing another embodiment of the present invention.

FIG. 10 is a system configuration diagram showing still another embodiment of the present invention.

FIG. 11 is an explanatory diagram of an example of a scenario analysis screen.

[Explanation of symbols]

100 processing device, 200 input device, 300 output device, 400 alarm scenario table, 500 notification history data memory (database), 600 alarm scenario definition screen, 700 alarm scenario Gantt chart screen
800: scenario analysis screen; 900: process data memory (database).

 ──────────────────────────────────────────────────続 き Continuing on the front page (72) Inventor Mitsuhiro Watanabe 1st Fuji-Facom Control Co., Ltd., Hino-shi, Tokyo Inside (72) Inventor Masato Sasaki 1st Fuji-cho, Fuji-cho Co., Ltd., Hino-shi, Tokyo

Claims (6)

[Claims] 1. A method for monitoring a time-varying trend of time-series process data, wherein a trend monitoring pattern (hereinafter referred to as an alarm scenario) relating to a plurality of process variables to be monitored is defined in advance, and a predetermined monitoring is started. When the condition is satisfied, monitoring based on the alarm scenario is started, and when the process variable deviates from the upper and lower thresholds, whether or not a trend monitoring pattern indicating the same change direction exists within the movable time width. If the same trend monitoring pattern exists, the monitoring is sequentially performed by moving the trend monitoring pattern in the future time direction to the position where the pattern exists, and monitoring by returning the time advanced to the original position. If the pattern does not exist within the movable time width, it is assumed that an error has occurred and the operator is notified of the occurrence of the error, and all process changes to be monitored are processed. However, when the monitoring end condition is satisfied without deviating from the upper and lower thresholds, it is determined that the monitoring has been completed normally, and a normal end is notified to an operator. . 2. An input means for inputting time-series process data, an alarm scenario definition table in which a trend monitoring pattern relating to a plurality of process variables to be monitored is defined as an alarm scenario, and the time-series process based on the alarm scenario. An apparatus for monitoring process data, comprising: processing means for executing a monitoring process for monitoring data; and output means for outputting a processing result. 3. When the accumulated process data or online real-time data is input to the processing means and an allowable range is designated, the processing means sets a monitoring start condition from the history data or the online real-time data. 3. The learning process according to claim 2, wherein a matching start point is searched, an upper limit and a lower limit are determined in consideration of an allowable range for these data, and a learning process of storing the determined upper limit and lower limit in an alarm scenario definition table is executed. Monitoring device. 4. The process data according to claim 3, wherein the alarm scenario definition table can be edited interactively from an alarm scenario definition screen and a time-series scenario data editing screen (Gantt chart). Monitoring device. 5. The abnormality detection information output to the operator via the output means, the date and time of occurrence, the cause of the abnormality, the action, the monitoring threshold at that time, and the process data are accumulated as notification history data. The process data monitoring device according to claim 2, wherein: 6. A list of the trend monitoring pattern used for the abnormality detection based on the notification history data, the value at that time, and the movable time width of the trend monitoring pattern is displayed in a list to analyze the cause of the abnormality and modify the trend monitoring pattern. The process data monitoring apparatus according to claim 5, wherein a supporting scenario analysis screen is output.