JPH02136713A - Diagnostic system for plant facility - Google Patents

Abstract

PURPOSE:To contrive the improvement of the efficiency of a plant operation and the improvement of safety by automatizing the decision of a deterioration factor diagnosis, based on the reliability and the transition time of a variation from data which is processed and brought to calculation analysis in the course of plant operation. CONSTITUTION:In this system 20, data in the course of plant operation is processed by a data processing part 15, and the calculation analysis of its data is executed by a performance calculating part 16. Subsequently, a main point diagnostic part 21 evaluates the reliability of the data and decides a factor, and thereafter, evaluates the reliability of the diagnosis. These measured data and analysis result information are reserved in a data bank 17, and displayed by a display part 18. In this diagnostic part 21, the state of a sudden variation is decided by a fact that calculated data exceeds the mean value of data in a prescribed period in the past stored in the bank 17 and a sudden variation criterion based on the degree of variance. Also, a slow variation is decided by a fact that a predicting value by the calculated data exceeds a criterion value based on initial data of a deterioration diagnostic start time point. Also, by installing this diagnostic part 21, the deterioration can be decided automatically.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Field of Application) The present invention relates to a plant equipment diagnosis system for diagnosing signs of deterioration of equipment in thermal power generation plan 1, for example, and supporting preventive maintenance.

<Conventional technology> In thermal power plants, etc., operational data is periodically collected to assess the performance and evaluate the efficiency of each component of the plant, as important indicators of economic efficiency. Analysis and evaluation are performed. Through this, we can grasp the deterioration status of the entire plant over time and the performance of major equipment, estimate abnormal conditions of equipment from the operation plan including other plants and the deterioration status, and carry out detailed inspections and repairs for preventive maintenance. The plan is to carry out the following.

However, although periodic performance evaluation work uses scheduled stoppage machines at each stage such as data collection, analysis evaluation, and performance calculation, it is difficult to set up performance tests for data collection and details of calculation data. , YJI cutting, etc. are done manually and require a great deal of effort.

Therefore, various diagnostic systems have been proposed to automate and support the monitoring and abnormality diagnosis functions of thermal power plants. For example, a management computer system periodically saves measurement data from various sensors that are recorded for controlling and monitoring steam temperature, pressure, temperature of each extraction section, water supply flow rate, generator output, etc. Based on this data, performance evaluations are performed as required.

FIG. 7 shows a conventional example of an anchoring diagnosis system for a thermal power plant. In FIG. 7, the thermal power plant 1 includes a boiler 2, a turbine 3, a generator 4, a condenser 5, a condensate pump 6, a low-pressure feed water heater 7, a deaerator 8, a feed water pump 9,
It is composed of a high-pressure feed water heater 10 and the like. A control computer 12, an operation panel 13, and a management computer 14 are connected to this thermal power plant 1. The management computer 14 includes a data processing section 15, a performance measurement section 16, a data bank 17,
It is composed of a result display section 18, an external information input section 19, and the like.

The data processing unit 15 extracts the necessary measurement data from the database of the data bank 17 that is stored at the time of request and periodically, performs averaging processing to prevent sensor measurement errors, and calculates data dispersion using standard deviation, etc. It has functions such as processing, detecting and eliminating unreasonable values that occur due to sensor abnormalities such as breakage.

In addition, the performance measurement section 16 calculates entropy, etc. based on steam tables from steam temperature, pressure, etc., analyzes based on performance meter functions given in advance, and calculates correction coefficients, etc. based on measurement conditions such as atmospheric pressure and cooling water temperature. Perform calculations.

Further, the result display section 18 displays and outputs the performance calculation results and each scale data in the form of a list or the like.

Based on this display result, if the deterioration of each device is progressing, for example, measures that can be taken during operation include backwashing to clean the condenser tubes, and from a preventive maintenance standpoint. From then on, detailed inspection plans and repair plans for deteriorated equipment are made.

(Problem to be solved by the invention) In order to understand long-term deterioration trends through plant performance evaluation and perform preventive maintenance of each equipment that makes up Plan 1-, it is necessary to determine deterioration as well as the state of change. It is essential to do so. In other words, even if the deterioration is of the same degree,
The cause of the abnormality inside the device differs depending on whether the change progresses gradually (hereinafter referred to as gradual change) or rapidly after a certain period of time (hereinafter referred to as sudden change). . For example, in addition to the internal efficiency of a steam turbine, the first
At the stage of gradual change in which the internal efficiency of the stage gradually changes, it is thought that the first stage nozzle's construction or scale adhesion is the cause. In this case, it is necessary to take countermeasures based on a relatively long-term plan based on the history of past periodic inspections, and to take measures such as water quality management, which is the cause of the occurrence. On the other hand, in the case of sudden changes, damage to the first stage nozzle, damage to the rotor blades, clogging of foreign objects,
Or, there may be an abnormality in the valves, etc., and it is urgent! A chastity test is required.

Judgments based on such measurement data and performance evaluation data have conventionally been made by skilled drivers or repair personnel, and are influenced by individual judgment criteria. In addition, regarding the algorithm for determining the state of change in abnormal signs, for example, in an abnormal diagnosis system that diagnoses abnormal signs based on shaft vibration changes, after detecting abnormal signs in the data, it is repeatedly checked whether the data changes. Because it can be done,
For example, there is a method disclosed in Japanese Patent Application No. 62-260193.

However, even if periodic data is processed, such as deterioration diagnosis, similar algorithms cannot be used when the measurement interval is over a long period of time, such as 1 day or 1M. We had no choice but to rely on human judgment.

The present invention has been made in view of these circumstances, and by automating the determination of change status in deterioration diagnosis, which has been mainly performed by humans, it is possible to evaluate the performance of each plant component on a computer-based performance evaluation system. The purpose of the present invention is to provide a plant equipment diagnosis system that makes it possible to determine the causes of deterioration, ensure support for preventive maintenance of each piece of equipment, and improve the efficiency and safety of plant operation.

[Structure of the invention]

(Means for Solving the Problems) The present invention provides a data processing means for processing data during operation in a plant equipment diagnosis system that performs performance evaluation of a plant based on data during operation in Plan 1. , a total zero analysis means for performing calculation analysis of the processed data, a factor diagnosis means for diagnosing deterioration factors based on reliability of the data and data change transition time, and a data bank for storing measurement data and analysis result information. , and a data display means for displaying these data.

(Operation) In the present invention, the operating state values of the plant are constantly taken in as data from sensors, and when operating conditions are satisfied, diagnosis and evaluation are performed at regular intervals.

When performing performance evaluation, etc., first, necessary operating state data is sent from the data bank to the data processing section, and necessary preprocessing such as average value calculation for performance calculation is performed. The results are stored in a data bank and sent to a factor diagnosis tool, which determines whether the change is sudden or gradual, based on the reliability of the data and the time of data change. Diagnoses such as: Further, the results are displayed on the data display means.

(Example) Hereinafter, an example of the present invention will be described with reference to FIGS. 1 to 5. Note that this embodiment is intended for a thermal power plant, and FIG. 7 is also referred to for the parts common to the conventional one, and redundant explanation will be omitted.

FIG. 1 shows the procedure of the plant membrane width diagnosis system. That is, the performance evaluation system 20 that implements this method includes a data processing section 15, a performance (efficiency) measuring section 16,
, a key diagnosis section 21 that evaluates the reliability of the data and determines the factors, and then evaluates the reliability of the diagnosis; a result display section 18; and a data recalculation section 22 that recalculates the data based on the displayed results or performs the completion procedure after confirmation. , a data bank 17 that stores measurement data and analysis results and provides comparative data if necessary.
, a sensor management unit 23 that manages revision information of the system external environment, such as sensor replacement information and equipment replacement information.

Cause diagnosis! l1M21 is composed of data reliability evaluation means 24, sudden change/gradual change determination f stage 25, factor determination means 26, and diagnosis reliability evaluation means 27.

The sudden change determination means determines if the current calculated data exceeds the sudden change determination criterion 10 determined from the average value and degree of dispersion of past fixed period data stored in the data bank. Similarly, gradual change is determined when the predicted value (-th order approximation value) based on the above data exceeds the determination reference value determined by the initial data at the time of starting the deterioration diagnosis,
By adding this determination means, automatic determination of deterioration becomes possible.

In other words, as shown in Figure 2, sudden changes are determined by using data from a certain period in the past (one month in Figure 2) and extrapolating the current data using the method of least squares. ′ and the sudden change judgment standard margin DA determined based on the degree of data variation, etc.
The determination is made based on whether or not the data A that has been output this time exceeds the range defined by O.

This standard tolerance is calculated using standard deviation, etc., and has the following conditions.

+A-AO8' I≧DAO On the other hand, as shown in Figure 3, the judgment of gradual change is based on the predicted value AO of the current data obtained by extrapolating based on the data of a certain period in the past.
8' and the current data 8 is within the range defined by the sudden change judgment standard tolerance DAO, and the diagnosis target value AO8 calculated including the current data is the initial value defined when starting the deterioration diagnosis. AO and a certain period in the initial stage (1 in Figure 3)
Criteria tolerance DA for determining the presence or absence of change based on data from
When it falls outside the range defined in step 1, it is determined that the change is gradual.

That is, IA-AO8'l<DAOANDIAO-A
O8l≧DA1.

Next, the factor determination means 21 determines a method for estimating the factors that led to the deterioration based on the determination of the deterioration status of each part.

FIG. 4 shows an outline of an embodiment of the determination means. That is, after evaluating the overall efficiency change of a series such as high pressure turbine section or feed water heater performance for each block, if an improvement is recognized, the modification/repair judgment routine supported by the sensor management section 23 is executed. If there is no deterioration or no change, evaluate the change in efficiency of each part.

Here, too, if there is an improvement, a modification/repair determination routine is executed, if deterioration is observed, a routine to cause determination is executed, and if there is no change, a factor determination routine B is executed to check for sensor abnormality, etc. The efficiency of each part is changed sequentially, and when the overall judgment is completed, the result is output.

FIG. 5 shows an example of the criteria used for factor determination A.

Determining the state of change is the main item in determining the cause.

The diagnostic reliability evaluation means 27 performs a comprehensive evaluation by sifting the reliability evaluation of the measurement data and the reliability of factor determination described above.

The above deterioration diagnosis and factor determination results are available in Data Bank 1.
7 and display the results via a man-machine interface.

Based on this display result, it is possible to complete the procedure after confirmation or to change the judgment criteria or unreasonable data and recalculate the diagnosis.

In addition, the sensor management unit 23 inputs/outputs and continuously manages sensor transshipment information, instrumental error correction data, and historical data of modification/repair of the equipment to be diagnosed, which are important factors in diagnosis. .

As described above, in the plant design diagnosis system of this embodiment, by adding the sudden change/gradual change determining means, it is possible to determine the state of change in the deterioration tendency without human intervention. , it becomes possible to estimate the factors that led to the deterioration.

In addition, in the gradual change determination criterion shown in FIG. 3, it was determined that all data for a certain period of time can be used as data for calculating the sudden change determination criterion margin DAO n. However, it can be seen that the standard data of this system, which requires relatively long-term historical data, shows abnormal values such as sudden changes during a predetermined period. FIG. 6 illustrates such a state, and the data in state B is a value that is diagnosed as a sudden change. Therefore, when it is determined that such an abnormal value is present, a predetermined flag can be set and saved when storing it in a data bank, and usage can be restricted when creating reference data.

In addition, although the previous embodiment targeted performance evaluation, the present invention is not limited to this, but can be applied to various automated systems that manage long-term change trends such as remaining life management and determine factors based on the state of change. You can do that.

〔Effect of the invention〕

As described above, according to the present invention, it becomes possible to automatically determine the state of change in the deterioration tendency, such as determining sudden change/gradual change, and it is possible to reliably obtain indicators for determining factors essential for diagnosis.・The effect is that driving support and preventive maintenance can be performed effectively.

[Brief explanation of the drawing]

Figure 1 is a diagram of a plant performance evaluation system showing an embodiment of the present invention, Figures 2 and 3 are graphs for explaining methods for determining sudden changes and gradual changes, and Figure 4 is a means for determining factors. FIG. 5 is a diagram for explaining factor determination criteria, FIG. 6 is a graph for explaining another embodiment of the present invention, and FIG. 7 is a functional block diagram for explaining a conventional example. be. DESCRIPTION OF SYMBOLS 1... Thermal power plant, 2... High/medium pressure turbine, 3... Low pressure turbine, 4... Generator, 5... Condenser, 6... Condensate pump, 7...・・Low pressure water heater,
8... Deaerator, 9... Water supply pump, 10... ^ Pressure test water heater, 11... Boiler, 12... Controller i machine, 13... Operation panel, 14...・・Management meter, 15・
...Data processing section, 16...Performance (efficiency) total n parts, 1
7...Data bank, 18...Display section, 19...
External information input section, 20... Performance evaluation system, 21.
...Cause diagnosis new department, 22... Recalculation section, 23... Sensor management section, 24... Data reliability evaluation means, 25...
- Sudden change/gradual change determination means, 26... Factor determination means, 27
...Diagnostic reliability evaluation means. Figure 63-11゜餉5'-6-1l. Figure Figure

Claims (1)

[Claims] In a plant equipment diagnosis system that evaluates the performance of a plant based on data during plant operation,
a data processing means for processing data during operation; a calculation analysis means for performing calculation analysis of the processed data; a factor diagnosis means for diagnosing deterioration factors based on reliability of the data and transition time of data changes; A plant equipment diagnosis system comprising a data bank for storing data and analysis result information, and a data display means for displaying these data.