JP2954613B2 - Plant life diagnosis support equipment - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a remaining life diagnosis performed in a periodic inspection of a plant such as a thermal power boiler (otherwise, a remaining life evaluation is also used. The present invention relates to a plant life diagnosis support apparatus for performing these operations as necessary and sufficient.

[Conventional technology]

When performing a remaining life diagnosis in a periodic inspection of a plant such as a thermal power boiler, generally, before starting the work, an inspection location, an inspection item and a method (an inspection chart) are determined in advance, and the inspection is performed according to the inspection. If there is a problem at the time of the inspection, conduct a detailed inspection and repair / replacement. Inspection chart creation, detailed inspection, repair / replacement decision
Conventionally, experts have decided based on past experience, results of similar plants, and the like. Recently, with the development of computers, expert systems for supporting inspection and diagnosis have been constructed.

Specifically, for example, this is an inspection chart creation support expert system. In creating an inspection chart, the past history (particularly, the accident history) of the target plant and the accident data of similar plants are statistically processed, and the Inspection charts are created in consideration of probabilities. Therefore, in preparing the inspection chart, oversight is avoided.

[Problems to be solved by the invention]

In the above prior art, a sufficient inspection chart can be created, but in actual inspection work, insufficient results often occur in determining whether or not a detailed inspection is required or repair / replacement. . In such a case, it is conventionally necessary to determine whether a detailed inspection is required or repair /
Replacement was decided. Experts also judge the inspection of similar parts based on these results.

Furthermore, the results of the periodic inspection must be reflected at the time of the next periodic inspection. In that case, the results of the periodic inspection were referred to, or the knowledge of experts who attended the periodic inspection was referred to. The determination of the contents of such a series of periodic inspection work needs to be made not by one expert but by several experts in different fields. For example, strength and corrosion require different specialists. One of the problems is that while the number of target plants that require a service life diagnosis is increasing recently, specialists are running short.

An object of the present invention is to evaluate the degree of creep and fatigue damage of materials used in a plant in a periodic inspection (referred to as damage evaluation. Damage evaluation is generally referred to as damage diagnosis. In the specification, this will be referred to as damage evaluation), and the life of the material will be diagnosed (referred to as life diagnosis).
Although the service life diagnosis is generally referred to as service life evaluation, it is referred to as service life diagnosis in this specification. In other words, an object of the present invention is to provide a plant life diagnosis support device capable of performing appropriate diagnosis work without expert knowledge.

[Means for solving the problem]

The above object is to provide an inspection chart creating means for creating an inspection chart based on input information on a plant to be diagnosed, an inspection item selection database and an inspection chart creating database, and inspect the plant based on the inspection chart. Damage assessment means for assessing the damage of the plant based on the results and the damage assessment database; judgment means for judging the necessity of measures or additional inspection based on the result of the damage assessment and the reliability of the assessment results; If the judgment means determines that countermeasures are necessary, the optimal countermeasure method is output based on the countermeasure method selection database, and if the damage is small and no countermeasure is determined, the next inspection schedule is output based on the inspection schedule creation database. If the reliability of the evaluation result is low, It is achieved by providing a protection and inspection output means for instructing pressurization inspection.

[Action]

In the plant life diagnosis support apparatus according to the present invention, a damage evaluation inference / calculation section performs damage evaluation and life diagnosis based on the data / knowledge base by inputting an inspection result. As a result, if there is sufficient remaining life, a future optimal inspection schedule is created based on the result and the inspection schedule creation data / knowledge base. If there is almost no remaining life, an optimal countermeasure method is inferred based on the data / knowledge base and output.

Further, similar parts requiring inspection are inferred and output based on the data / knowledge base, and inspection is requested. Life can not be inferred or is unclear (reliability of life diagnosis is low)
In such a case, an inspection item for which a life diagnosis can be performed is inferred and output based on the data / knowledge base, and an additional inspection is requested.
Perform the life diagnosis by adding the inspection data, and again,
The same processing is performed. These results are fed back to the data / knowledge base and reflected in the creation of diagnostic charts and diagnostic work for the next periodic inspection. With the above-described device configuration, it is possible to support an appropriate diagnosis operation without any expert knowledge.

〔Example〕

FIG. 1 shows an overall configuration diagram of a plant life diagnosis support apparatus of the present invention. The plant life diagnosis support device 1 includes a processing device 2, a display device 5, an input device 6, and an output device 7,
The processing device 2 includes a data / knowledge base unit 3 and an inference / operation unit 4
Consists of The plant life diagnosis support apparatus 1 is connected to the host computer 8 and searches various plant operation histories, regular inspection records, accident histories and the like recorded in the host computer 8 and takes them into the data / knowledge base unit 3. It can be used as a database.

The inference / calculation unit 4 performs inference / operation based on the information input from the input device 6 and the data / knowledge base in accordance with the dialogue screen output to the display device 5 and outputs the result to the output device 7 and the display device 5. Output to The result is sent to the host computer according to a certain format and becomes a new database.

FIG. 2 shows a processing flow of the plant life diagnosis support apparatus 1. In the plant life diagnosis support apparatus 1, an inspection chart is created prior to the inspection. For this purpose, first, plant data to be diagnosed, such as a plant name and an inspection time, is input from an input device (keyboard) 6 according to an interactive screen output to a display device (CRT) 5 (target plant data input 9). Based on the information and the data and the knowledge base 25 for preparing the inspection chart in the data / knowledge base unit 3, the damaged part in the target plant is inferred and calculated (damaged part inference / calculation 10) to create the inspection chart. Output (check chart output 11).

Inspection chart creation data / knowledge base 25 contains various plant designs / structures, repairs / modifications, operation histories, accidents /
Data such as damage is stored. Searching for the target plant from the target plant data input 9, searching for similar (type, capacity and operation history) plants, statistical processing, calculating the probability of an accident, etc. Select a site with a high probability. The site with a high accident probability is a site with a statistically high accident probability from an accident example of a similar plant, or a region with a large damage estimated from an operation history or the like.

As the inspection chart output 11, as shown in the inspection chart (list of inspection parts) in Fig. 3, a list of parts that require a high probability of an accident, an inspection method, and a list of inspection parts with inspection methods and damage factors are listed. Is output to the display device 5 and the output device (printer) 7. Inspection method is Data / Knowledge Base 3
The minimum required inspection items are selected from the inspection method item selection database 26 in the above, and are input to the damaged part inference / calculation 10 in advance. For example, in the case of a heat transfer tube, external appearance inspection, outer diameter measurement, and the like are performed. The damage factor is output as expected for reference. In addition, if a site is selected from the input device 6 such as a mouse or a keyboard, the drawing can be selected from a database storing the drawings, and the target site can be indicated by the drawing, although not shown. Output device 7
Can be output. Further, the inspection result record sheet shown in FIG. 4 is output for each part, so that the inspection work can be performed smoothly.

Next, in the plant life diagnosis support apparatus 1, the inspection result is input from the input device 6 (inspection result input 12), and the damage state of each part is evaluated (hereinafter, referred to as damage evaluation). Inspection results may be entered in numerical values, such as hardness, or expressed in text, such as the scale status by visual inspection, etc. The input is distinguished, and in the case of text input, the result is selected from several options. Inference and calculation are performed based on such input information and the data / knowledge base 27 for damage evaluation in the data / knowledge base unit 3 to evaluate damage.

The damage evaluation data / knowledge base 27 is used to infer the damage state by fuzzy inference from data indicating the relationship between the evaluation parameters and the damage amount and several evaluation parameters including an ambiguity amount expressed as a sentence. Stores membership function data. FIG. 5 is an example of data indicating the relationship between the evaluation parameters and the damage amount, and is master data indicating the relationship between the cavity area ratio, which is a creep damage evaluation parameter, and the creep damage ratio. Based on the damage evaluation data / knowledge base 27 and the inspection result input 12, damage evaluation is performed by a calculation using master data, fuzzy inference, statistical calculation such as extreme value statistics, etc., and a life diagnosis is performed (damage amount inference / operation 13 ).

As a result of the damage evaluation performed by the damage amount inference / calculation unit 13, if the reliability of the evaluation is high and the damage is large, and conversely, if the remaining life is short, a measure is required. The optimal countermeasures such as replacement, repair, remodeling, etc. are inferred based on the countermeasure method selection data / knowledge base 28 in (3) (countermeasure method inference 16) and output to the display device 5 and output device 7 (countermeasure method). Output 17). An example of the output is shown in FIG.

The certainty of the evaluation is reliability. In general, when performing damage evaluation, a single evaluation data has a high risk, and thus various types of evaluation data are comprehensively diagnosed. If these evaluation results tend to match, it can be said that the reliability is high, and if they do not match, the reliability is low.

In addition, a part similar to a part having a large damage and, conversely, a part having a small remaining life is inferred based on the similar part selection data / knowledge base 29 in the data / knowledge base part 3 (similar part inference 18). An inspection request is output (similar part output 19).
Thereby, oversight can be eliminated.

As a result of the damage evaluation by the damage amount inference / calculation unit 13, if the confidence of the evaluation is high and no countermeasure is required, the data for creating the inspection schedule in the data / knowledge base unit 3
Based on the knowledge base 30, the next inspection schedule is inferred (inspection schedule inference 20), and the next inspection time, inspection items, inspection methods, etc. are output (inspection schedule output 21). Although not shown, if temperature measurement data is required at the next inspection, instructions for work that must be performed during the inspection period, such as a thermocouple installation request, etc., and instructions for collecting the necessary data at the next inspection Is output.

As a result of the damage evaluation by the damage amount inference / calculation unit 13, if the confidence of the evaluation is low due to lack of information or input of abnormal data, etc., an inspection item that enables a high confidence evaluation, Based on the inspection items in the data / knowledge base unit 3 and the method selection data / knowledge base 26, additional inspection parts / items are inferred (additional inspection parts / item inference 22) and output (additional inspection parts / items). Output 23). For example, although not shown, a request is made to collect a replica for detailed evaluation and obtain a cavity area ratio. In addition, re-inspection may be requested when it is estimated that the data is inappropriate as well as addition. In addition, if it is difficult to evaluate only by inspection of the target part, inspection of an adjacent part may be requested ( Additional inspection item input 24). In some cases, a sample may be collected and a destructive test such as a creep test may be required.
In this case, the result comes out several months later, and in this case, a replacement method is instructed and the process ends.

When the additional inspection result is input (additional inspection item input 24), the process returns to the damage amount inference / calculation unit 13, where the damage is evaluated again, and the same procedure is repeated. In this case, the damage amount inference / calculation is comprehensively evaluated including the inspection data collected earlier.

As described above, the plant life diagnosis support device according to the present invention can be used in an interactive manner, such as creation of an inspection chart, damage evaluation, countermeasure method, indication of similar parts, instruction of additional inspection items and methods, creation of a future inspection schedule, and the like. These data are input / output to support the service life diagnosis work of the printer. After the inspection, these data are stored in the host computer and used as a database.

〔The invention's effect〕

According to the present invention, in a printer life diagnosis operation,
This system can be used without unnecessary inspections and without overlooking.Also, judgments that previously depended on experts can be made by non-experts with this device, greatly improving work efficiency and accuracy, and technically Is also effective.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of a plant life diagnosis support device, FIG.
Fig. 3 is a flowchart of the process, Fig. 3 is a diagram showing a list of inspection sites output to the output device, Fig. 4 is a diagram showing an inspection result record table, and Fig. 5 is a diagram showing the cavity area ratio and the creep damage ratio. FIG. 6 is a diagram showing the relationship, and FIG. 6 is a diagram showing an example of the countermeasure inference result output to the display device. 1 ... Plant life diagnosis support device 2 ... Processing device 3
…… Data / knowledge base part, 4 …… Inference operation part, 5 ……
Display device 6 Input device 7 Output device

Claims (1)

(57) [Claims] 1. An inspection chart creating means for creating an inspection chart based on input information on a plant to be diagnosed, an inspection item selection database and an inspection chart creating database, and inspecting the plant based on the inspection chart. And the results
Damage evaluation means for evaluating the damage of the plant based on the damage evaluation database, judgment means for judging the necessity of a countermeasure or additional inspection based on the result of the damage evaluation and the reliability of the evaluation result, and the judgment means If it is determined that a countermeasure is necessary, the optimal countermeasure method is output based on the database for selecting the countermeasure method, and if it is determined that the damage is small and no countermeasure is necessary, the next inspection schedule is output based on the database for creating the inspection schedule and evaluated. A plant life diagnosis support device comprising: a countermeasure / inspection output means for instructing an additional inspection based on an inspection item selection database when the reliability of the result is low.