JP5253039B2 - Early detection method for abnormal data in power plants - Google Patents

Description

  The present invention relates to an early detection method for abnormal data in a power plant that detects abnormal trends in the operation of equipment and facilities in the power plant and discovers data in the abnormal trend.

  Electricity generated at thermal power plants, hydroelectric power plants, nuclear power plants, etc. is boosted in voltage to be transmitted efficiently and sent to electrical stations such as substations via transmission lines, where it is stepped down and further converted to other substations via transmission lines. Power is transmitted to the station, and is sent to the customer through the transmission line and distribution line. Thus, it is desired that the power plant performs safe and efficient power generation, and that the substation transforms the voltage, that is, transports the power safely and economically.

  In order to fulfill the mission of electric power stations such as power plants and substations, it is of course necessary to have the necessary equipment, but the operation and maintenance of the power plants must be easy, reliable and safe. There is. For this purpose, power plants, substations, and the like are provided with monitoring control devices such as switchboards and control consoles so that they can be operated easily. For example, data relating to temperature, pressure, output, vibration, pH, conductivity, or flow rate is collected during operation of equipment and facilities in a power plant. Therefore, all the analog data of the unit computer is organized in units of one month and the trend is managed.

  For example, the data collected automatically at 11:40 in the morning for each generator such as Unit 1 to Unit 3 in the power plant is graphed around 4 teams on duty and combined with the data for the past 5 months for 6 months The abnormal data is checked by graphing the trend.

  However, the method of checking abnormal data by each group as described above, that is, the method of organizing in units of one month, lacks speediness. In addition, these data have a problem that after a large abnormal situation occurs, it can be used only as proof as evidence of the abnormal situation.

In the method to check the conventional abnormality data, de-select the data that is heading to early abnormal direction is intended to check every day by focusing on the data, there is a problem of lack of immediacy It was. In other words, the conventional method has a problem that it cannot be predicted that an abnormal situation will occur in the equipment or facility, and cannot be used to prevent the occurrence of the abnormality.

A technique for detecting an abnormality, estimating the abnormality, and guiding the operator is proposed. For example, as shown in Japanese Patent Application Laid-Open No. 3-293524 “Plant Abnormality Diagnosis Device” of Patent Document 1, an online inference processing unit for inferring an abnormal location and an abnormal cause of the plant from a process value input from the plant, In a plant abnormality diagnosis device having a display means for displaying a processing result by the online inference processing unit, an abnormality data saving mechanism for saving abnormal data of the plant, and using abnormal data saved by the abnormal data saving mechanism There has been proposed a plant abnormality diagnosis apparatus provided with an inference processor for evaluation that evaluates knowledge offline.
JP-A-3-293524

  In the “plant abnormality diagnosis device” of this Patent Document 1, it is disclosed how the data can be specifically processed and the abnormality can be detected and estimated if the processing is judged. There wasn't.

The present invention has been developed to solve such problems. An object of the present invention, in Succoth from operating data relating to a power plant and equipment of thermal power stations out to select the suspect data, abnormality occurs in the power plant and equipment, or whether that is a preceding state An object of the present invention is to provide an early detection method of abnormal data in a power plant that can be confirmed and accidents can be monitored.

The present invention is an early detection method for abnormal data in a power plant that detects abnormal trends related to the operation of equipment and facilities in a power plant and displays data in the abnormal trend, and operates the equipment and facilities in the power plant. Is automatically measured for operation information data related to temperature, pressure, output, vibration, pH, conductivity, and flow rate generated during the operation, and these are collected and stored as operation information data and calculated by the unit computer (1) Processed and stored in the database (12), the latest operation information data relating to the operation of equipment and facilities in the collected power plant, and the previous operation information data stored in the database (12) ( Compared with 17) , when the fluctuation range is greater than or equal to a predetermined value , the corresponding driving information data is selected as data requiring attention. And displayed on the monitor (3) and, then, the when the variation width is less than the predetermined value, compared with the data (18) of one week prior to the operation information, when the variation width is greater than the predetermined value, the corresponding data operation information is displayed on the monitor (3) and out chosen as suspicious data, further, said time variation width is less than a predetermined value, compared with the data (19) immediately after the regular inspection, the variation width when more than a predetermined value, corresponding with the data of the operation information is displayed on the monitor (3) and out to select the suspect data to get the data of each driving information for each stage, careful data sequentially automatically It is displayed on the monitor (3) de-select as, characterized in that.

For example, regarding the operation of equipment and facilities in a power plant, the latest collected operation information data, the previous data (17), the previous week data (18), and the data immediately after the regular inspection ( 19) are compared for each stage, when you leave chosen as suspicious data, in variation width of each data, 2.5%, sequentially when fluctuating 5.0% or 7.5% or more They are out to choose automatically.
Regarding the operation of equipment and facilities in the power plant, the latest operation information data collected, the previous data (17), the previous week data (18), and the data immediately after the regular inspection (19). comparing the door, when to output select as suspicious data, the fluctuation range, equipment in a power plant comprising a determination target is changed depending on the facility.

In the method of the above configuration, for the operation information stored in the database (12), the previous data (17), the previous week data (18) and the data immediately after the regular inspection (19) related to the same operation information as the latest data. ) and compares, in variation width of each data, 2.5%, to exit select data fluctuates 5.0% or 7.5% or more sequential automatic as suspicious data. Therefore, the operator does not need to confirm the screen while constantly checking the screen of the monitor (3) as in the prior art, and the work effort is reduced.

  In addition, the operator checks the data of interest and related data continuously displayed on the screen of the monitor (3) to determine whether an abnormality has occurred in the power plant or whether it is in the previous stage of the abnormality. You can concentrate your efforts. Therefore, the method of the present invention can be used not for the post-occurrence processing for an accident but for monitoring the accident in advance.

The variation width of the equipment in a power plant comprising a determination target, by changing depending on the facility, the operator may display the next variation width of after confirmation process, that writing Monri data number to be displayed at a time This makes it easier for the operator to confirm.

The abnormality data early detection method in the power plant according to the present invention includes the latest collected operation information data , the previous operation information data related to the operation of equipment and facilities in the power plant, and the one week previous operation information data. and it compares the fluctuation range of each data of each operation information of the data immediately after outage, and exits select data operation information vary more than a predetermined value as a suspicious data, a method of automatically displaying on a monitor or the like is there.

Hereinafter, preferred embodiments of the present invention will be described with reference to the drawings.
FIG. 1 is a schematic configuration diagram showing an early detection method for abnormal data in a power plant according to the present invention.
The method for early detection of abnormal data in a power plant according to the present invention is a method for detecting abnormal trends in the operation of equipment and facilities in a power plant and displaying the data of the abnormal trends. In this method, first, numerical data is automatically measured with respect to operation information related to temperature, pressure, output, vibration, pH, conductivity, and flow rate generated during operation of equipment and facilities in a power plant. For example, about temperature, it measures about the temperature of a boiler, the temperature of a turbine, the temperature of the intake seawater, the temperature of air | atmosphere, etc. Regarding the pressure, measure the atmospheric pressure in the boiler, the vacuum level of the condenser, etc. Thus, the operation information regarding the equipment and facilities that are operating in the power plant is measured. Of course, the type of measurement differs depending on the power plant.

Regarding these driving information, taken for data to be stored and managed, stored in a database 12 of the plant data management system 2, as described later with computing the unit computer 1. The collected driving information data is stored in the database 12 for each data such as the driving information data 17 of the previous time, the driving information data 18 of the previous week, and the data 19 immediately after the regular inspection. By collecting operation information data for each stage, it is easy to detect abnormal trends in the operation of equipment and facilities in a power plant. For example, as shown in the flowchart of FIG. 3 to be described later, the collected driving information data is compared with the data 19 immediately after the regular inspection, that is, the driving information data 17 one time before , that is, a long time has passed since the regular inspection. This is because it tends to show an abnormal tendency. Similarly, the data 17 of the previous driving information is more likely to show an abnormal tendency than the data 18 of the previous week. Data vary more than a predetermined value, and output select as suspicious data is displayed automatically so the monitor 3. As a result, it can be estimated that there may be an abnormality in the equipment and facility.

FIG. 2 is a block diagram of a plant data management system for implementing an early detection method for abnormal data in a power plant according to the present invention.
A plant data management system 2 for carrying out the early detection method of abnormal data in a power plant according to the present invention is mainly composed of an arithmetic processing unit 11, a database 12, and an input unit 13. The arithmetic processing unit 11 includes an input processing unit 14, a determination processing unit 15, and an output processing unit 16 . The database 12 is composed of data such as data 17 related to the previous driving information, data 18 related to the driving information one week ago, data 19 related to the driving information immediately after the regular inspection, and generator output data 20 related to the set output of the generator. Become.

The arithmetic processing device 11 executes various arithmetic processing based on the data calculated by the unit computer 1 and other conditions input from the input device 13, and the input processing units 14 and 1 for performing input processing. Data 17 related to previous driving information, data 18 related to driving information one week ago, data 19 related to driving information immediately after regular inspection, and determination processing unit 15 for determination based on generator output data 20, results of the determination, and output select as suspicious data, an output processing unit 16 that performs control processing for outputting display it on the monitor 3.

FIG. 3 is a flowchart of the early detection method for abnormal data in the power plant of the present invention. FIG. 4 is a reference diagram showing an example of display on the monitor.
An example of processing by the early detection method of abnormal data in the power plant of the present invention will be described. The collected latest operation information data related to the operation of the equipment and facilities in the power plant, and the previous operation information data 17 related to the operation of the equipment and facilities in the power plant stored in the database 12, one week before the data 18 and data 19 immediately after outage operation information, compares the data for each operation information, to output select data that the fluctuation range is varied above a predetermined value as a suspicious data.

For example, the data collected latest operation information, and data 19 of one previous data 17,1 weeks immediately prior to the data 18 and outage for the same operation information stored in the database 12 compares each data in the variation width of each 2.5%, to exit select data fluctuates 5.0% or 7.5% or more sequential automatic as suspicious data. These threshold values of 2.5%, 5.0%, or 7.5% are examples, and are determined according to the target device and facility. Moreover, it is not bound by these numerical values.

Since early detection method for abnormal data in a power plant of the present invention is configured as this, the operator, as in the prior art, gets rid need to check while checking the screen constantly monitor 3, reduced the labor required Is done. In addition, the operator checks the caution data and related data continuously displayed on the screen of the monitor 3 to determine whether an abnormality has occurred in the power plant or whether it is in the previous stage of the abnormality. Can concentrate. Moreover, it can be used for accident monitoring rather than post-event processing.

  In addition, by changing the fluctuation range for each data according to the equipment and facilities of the power plant to be judged, the operator can display the next fluctuation range after the confirmation process for this automatically displayed data. Since the number of data displayed on the screen can be imprinted, the operator can easily confirm it.

It should be noted that the present invention is, in Succoth from operating data related to the power generation plant and equipment of thermal power plants, etc. out to select the suspicious data, abnormality occurs in the power generation plant and equipment, or to confirm whether that is a pre-stage state Thus, as long as an accident can be monitored in advance, the present invention is not limited to the embodiment of the invention described above, and can be variously changed without departing from the gist of the invention.

  The method for early detection of abnormal data in a power plant according to the present invention also detects an abnormal trend of operation in a plant facility that needs to maintain and manage a manufacturing plant, a cleaning plant or equipment, and a facility in addition to the power plant. Can be used.

It is a schematic block diagram which shows the early detection method of the abnormal data in the power plant of this invention. It is a block diagram of a plant data management system for implementing an early detection method for abnormal data in a power plant according to the present invention. 4 is a flowchart of a method for managing output limits of a generator of the present invention. It is a reference figure which shows an example of a monitor display.

DESCRIPTION OF SYMBOLS 1 Unit computer 2 Plant data management system 3 Monitor 11 Arithmetic processing unit 12 Database 13 Input device 14 Input processing unit 15 Judgment processing unit 16 Output processing unit 17 Data before 1 time 18 Data before 1 week 19 Data 20 immediately after regular inspection Generator output data

Claims (3)

An early detection method of abnormal data in a power plant that detects abnormal trends related to operation of equipment and facilities in a power plant and displays data in the abnormal trend,
Equipment in the power plant, the temperature generated during the operation of the facility, the pressure, output, vibration, pH, conductivity, measured automatically for the data of the operation information about the flow, collecting the data to be stored and managed them as driving information The unit computer (1) performs calculation processing and stores it in the database (12).
Compared to the latest operation information data regarding the operation of the equipment and facilities in the collected power plant and the previous operation information data (17) stored in the database (12) , the fluctuation range is a predetermined value. at the time of the above, the data of the corresponding operation information is displayed on the monitor (3) de-select as suspicious data,
Then, the when the variation range is less than a predetermined value, compared with the data of the previous week operation information (18), when the variation range is equal to or greater than the predetermined value, the data of the corresponding operation information is as suspicious data and out you wish to display on the monitor (3),
Furthermore, the when the variation width is less than the predetermined value, compared with the data (19) immediately after outage, when the variation width is greater than the predetermined value, the data of the corresponding operation information is then output to select as suspicious data by displaying on the monitor (3), to retrieve the data of each driving information for each step, and displays on a monitor (3) and out chosen as suspicious data sequentially automatically abnormality in a power plant, characterized in that An early detection method for data. Regarding the operation of equipment and facilities in the power plant, the latest operation information data collected, the previous data (17), the previous week data (18), and the data immediately after the regular inspection (19). steal compared to the step-by-step, when you out to choose as suspicious data,
In the variation width of each data, 2.5%, abnormal data in the power generation plant of claim 1 sequentially to output select automatically, characterized in that when fluctuating 5.0% or 7.5% or more Early detection method. Regarding the operation of equipment and facilities in the power plant, the latest operation information data collected, the previous data (17), the previous week data (18), and the data immediately after the regular inspection (19). It compares the door, when you out to choose as suspicious data,
The variation width of the equipment in a power plant comprising a determination target is changed depending on the facility, early detection method for abnormal data in a power plant according to claim 2, characterized in that.