JPH0315768A - Method and device for diagnosing remaining life, information displaying method for remaining life and device therefor, and expert system - Google Patents

Abstract

PURPOSE:To attain a secure diagnosis for life remainder by obtaining the remaining life of parts assembled body from test results for both parts deterioration characteristic and function of the assembled body including these parts and from a correlation for both, and adopting the lowest value. CONSTITUTION:An expert system 1 is constituted with a knowledge acquisition supporting device 2, inference device 3, user interface 4, outer system interface 5 and knowledge base 6, and the interface 4 is connected to a data system 7 for plant data management and a terminal system 8. The data input, revision, and debugging of the base 6 in which data for expert and experiential prevention maintenance work are stored capable of inference processing, are controlled by the device 2, and a practice of the inference using these data is controlled by the device 3. Then, in accordance with the remaining lives L1, L2, L3 obtained in each analyzing part of parts deterioration (a) for system 1, device function (b) and correlation (c), the minimum value is made to the most suitable remaining life L by a life remainder evaluation part (d). With this arrangement, the life remainder of an object to be diagnosed is predicted securely and rapidly, thereby the perfect preventive maintenance can be expectedly performed.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a method and apparatus for diagnosing the remaining life of a parts assembly that is composed of a plurality of parts and in which the life of each part is related to the life of the whole. In particular, we will develop a suitable remaining life diagnosis method and device for accurately diagnosing remaining life, a display method and device for displaying the diagnosis results, and infer what measures need to be taken based on the obtained remaining life. Regarding expert systems. [Conventional technology] For example, parts that make up equipment such as power generation plants are subjected to external forces at high temperatures, so if they are used for a long time, they will suffer life damage and material deterioration, and at some point they will need to be replaced. It becomes necessary to do so. Therefore, in order to predict the replacement time, it is necessary to diagnose the remaining life. Conventionally, for example, JP-A-62
- As described in Publication No. 276470, the remaining life of the equipment is diagnosed using the set lifespan value set by the manufacturer during equipment manufacturing or the estimated lifespan value obtained from accelerated life test data. In addition, the deterioration characteristics of the parts that make up the equipment are determined from the deterioration characteristics test data, and the remaining life of the equipment is predicted from the limit values of the parts and these deterioration characteristics. Furthermore, we conduct functional tests on the equipment and use this functional test data to predict the remaining life of the equipment. [Problems that Gekmei aims to solve] Each of the above-mentioned conventional technologies has a problem in that it is not possible to accurately predict the remaining life of the device. For example, in the conventional method of determining the deterioration characteristics of a component from deterioration characteristic test data and predicting the remaining life from this deterioration characteristic, in order to obtain an accurate deterioration characteristic equation, There is a problem in that it requires a large number of files (in which it is necessary to destroy the This is due to the fact that the equation that approximates the deterioration characteristic equation is not appropriate. In addition, the conventional method of predicting the remaining life of equipment from functional test data performed during periodic inspections has the problem of relying heavily on the experience of experts, as many equipment do not show any functional deterioration during testing. If it is not possible to accurately predict the remaining life of equipment, equipment that does not need to be replaced will end up being replaced. Newer equipment does not mean fewer failures; the initial failure rate is inevitably higher than the failure rate of equipment in operation.
Therefore, unnecessarily replacing the equipment with a new one not only increases costs but also poses safety issues. Furthermore, conventional remaining life diagnosis techniques have a problem in that they do not take into consideration displaying the diagnosis results in a manner that allows the operator to easily understand trends in the overall remaining life.

A first object of the present invention is to provide a remaining life diagnosis method, apparatus, and expert system that can accurately diagnose the remaining life of a parts assembly.

A second object of the present invention is to provide a method and apparatus for displaying remaining life data that clearly displays trends in the remaining life of a large number of equipment components. [Means for Solving the Problem] The first objective is to determine the remaining life of a parts assembly from the correlation between the deterioration characteristic test data of the part and the equipment test data of the parts assembly including the part. Achieved. The first purpose above is to determine the reliability of the component by performing a Weibull distribution reliability analysis on the deterioration characteristic test data of the component, to determine the remaining life of the component assembly from this reliability, and to calculate the remaining life of the component assembly from the reliability test data of the component. This can also be achieved by determining the remaining life of the parts assembly and adopting the shorter remaining life.

The first purpose above is to convert the component deterioration characteristic equation to σ(t)==+
y0exp{--f(T)x tα) −(1)
Here, σ0; initial value of deterioration T; process amount t that promotes deterioration; time f (T) = x T2+ y T + zα+X+3
'+Z; It can also be achieved by approximating with coefficients.

The second purpose is to display the component parts as figures and to display the remaining life of each component corresponding to each figure.
achieved.

The second purpose mentioned above is to display the remaining life displayed for structural parts in different colors depending on the length of the remaining life. achieved.

[Effect]

Since the remaining life is determined from the correlation between the deterioration characteristic test data of the parts and the functional test data of the parts assembly, it is possible to accurately diagnose the remaining life.6 In addition, the remaining life determined from the reliability of the equipment and the parts Reliability can be increased by using the shorter remaining life determined from the deterioration characteristic test data. Furthermore, σ(t)=cr
Since the approximation formula 0exp(-f(T)xtα} approximates the deterioration characteristics of the component regardless of its type, the remaining life determined from this approximation formula is highly reliable.

When displaying the calculated remaining life, it is displayed graphically or color-coded for each component, allowing the operator to judge the trends in the remaining life of each component at a glance, and taking appropriate measures. It becomes possible to know. [Embodiment] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.6 Fig. 1 is a configuration diagram of an expert system for diagnosing the remaining life of a power plant, which is an example of a component assembly. This expert system 1 includes a knowledge acquisition support device 2, an inference device 3, a user interface 4, an external system interface 5, and a knowledge base 6. The user interface 4 is connected to a data system 7 for managing plant data, and a terminal system 8 consisting of input/output devices such as a keyboard and a display device.

Knowledge base 6 is knowledge data (specifications of component equipment and parts,
performance, limit values, accident and defect information, maintenance information, etc.) are stored in a format that can be inferred. Knowledge acquisition support device 2
inputs, changes, and debugs knowledge data. The inference device 3 executes various controls for inference using knowledge data stored in the knowledge base 6. The user interface 4 facilitates the input and maintenance of knowledge data obtained from experts and the user's responses, and the external system interface 6 is provided for importing data from the outside.

The inference device 3 executes software for diagnosing the remaining life of the power plant, and its features are as follows.

(1) Knowledge representation is a hybrid type of knowledge that can handle both rule-type knowledge expressed in the form of if-then type production rules and fact-type knowledge, that is, frame-type knowledge, in which the truth or falsity of the expression is determined. expression is possible.

(2) A flexible inference method capable of forward inference and backward inference can be executed, multiple conflict resolution strategies are provided, and the condition part of the rule, meta-rule, and debugger can be freely selected.

(3) Knowledge data stored in the knowledge base is converted into a format that can be processed at high speed before inference processing, and checking of rules unnecessary for inference is omitted to speed up the execution speed of inference processing. . Furthermore, it is now possible to speed up the process by narrowing down the number of rule groups used using meta-rules. Figure 2 shows the functional structure of the expert system with the above features. When diagnosing the remaining life of a device (component assembly) using this expert system, the remaining life Ll obtained by the parts deterioration analysis section, the remaining life L2 obtained by the equipment performance analysis section, and the remaining life L2 obtained by the correlation analysis section are used. Based on the lifespan L8,
The remaining life evaluation section calculates the optimal remaining life L. That is, the component deterioration analysis section determines the deterioration characteristic value of the component parts of the device and determines the remaining life L1 from this deterioration characteristic value. The device performance analysis section calculates the time required for the device to reach its limit value from the functional test data of the device composed of each component, and sets this time as the remaining life L2.The correlation analysis section calculates the deterioration characteristics of the component parts. The remaining life L3 is calculated from the correlation between the value and the equipment's functional test data. Then, the remaining life evaluation section calculates the remaining life L 1 + L
Let the minimum value of 2 and L3 be the optimal remaining life L. The processing of each part shown in FIG. 2 is performed by the inference device 3.

FIG. 3 is a flowchart showing the processing procedure of the parts deterioration analysis section. The parts deterioration analysis section sets the remaining life of the equipment obtained from the deterioration characteristic values of the parts as LL', the remaining life of the equipment obtained from the equipment reliability as L1'', and sets the shorter of the two as the remaining life L1. .Of course, it goes without saying that Ll' = Lt or .Lt' = Lz.

This component deterioration analysis section predicts the remaining life of the device based on reliability analysis results of accelerated life test data of the components constituting the device. In this case, the remaining life of the equipment can be predicted by evaluating changes in deterioration characteristic values such as bending strength and hardness of the equipment components under certain operating conditions.

First, in the first step, failure information or accelerated life test data of the components of the power plant stored in the database 7 is read. In the next step, reliability analysis such as Weibull distribution analysis is performed using the read data. Through this reliability analysis, the Weibull form parameter ml (in case of initial failure, m
+ < 1, m (= 1 if it is a random failure, m I > 1 if it is a wear-out failure), and a scale parameter η1 indicating the characteristic life.

In the next step, the reliability of the component is calculated using the following equation (2) from both of the above parameters and the operation history time t of the target device.
seek.

Ri=exp(-(t/ηt)")
- (2) This reliability R will be used later.

Next, the operating environment condition history data such as temperature is read in, the deterioration trend of component life is analyzed, and the deterioration characteristic value σ of the component is determined using equations (1) and (3). a = cy0exp{- - f (T)x tα
) −(1)f (T) = x T”+ y
T + z − (3) where σ. is the initial value of the deterioration characteristic value, T is the temperature, x, y, and z are experimental constants, and f (T) is an approximate expression of life data. Generally α=1.

Here, we will take up, for example, the control drive device (hereinafter referred to as CRD), which is one of the components that makes up a power generation plant.
One of the many components that make up this CRD is a carbon seal. The bending strength of this carbon seal tends to decrease as the operating temperature increases. Equation (1) is an approximate expression indicating the tendency of deterioration of bending strength. Figure 7 shows a characteristic diagram of the carbon seal. - Next, set the deterioration characteristic limit value σC of the part, and calculate the time to reach the limit value of the part, that is, the remaining life Lll of the part, from the deterioration tendency and deterioration characteristic limit value of the part, as shown in Fig. 8, for example. Calculate using equation (4).

Lxt=log(σo/ac)/f(T)t
-(4) Repeat the above processing steps until the analysis of all parts included in the device is completed. After completing the analysis for all parts, perform the following processing using the reliability R and remaining life Lzt found for each part. First, the shortest remaining life is selected from among the remaining lives L11 of each component and is designated as Li'. In CHD, the part with the shortest life is the carbon seal, so there is a high probability that L, I
The remaining life of the carbon seal is selected as Next, the reliability Re of the device and the reliability R1 of each component are determined by equation (5).

R e ” IT R I = R t ・ Rz・ ... ・ R.Then, set the device reliability limit value and calculate backwards from equation (5) to find the time to reach the limit value, that is, the remaining life L1' of the device. .

Finally, the remaining life LL' and the remaining life Ll' are compared, and the shorter remaining life is set as L1.

FIG. 4 is a flowchart showing the processing procedure of the device performance analysis section. The device performance analysis section analyzes the functional test data to determine the remaining life Lx of the device.

Figure 9 is a characteristic diagram of functional test data for determining the remaining life L2 of the CHD. First, functional test data is read. For example, in the case of a CRD, past data on the amount of driving water leakage (the larger the leakage amount, the greater the amount of driving water required to push up the control shaft. Therefore, the data on the driving water flow rate is sufficient) is carried out during periodic inspections. Load data. Then, perform a regression analysis such as the least squares method to find an approximate formula (the formula indicated by the broken line in Figure 9), set the limit value of the equipment, and calculate the time required to reach the limit value based on the approximate formula, that is, the remaining life. Lz
seek. Note that if there are multiple types of functional test data for the device, the remaining life may be determined using each functional test data, and the shortest one may be selected. In addition, the optimal remaining life L is calculated by taking into account the weighting determined by importance analysis to the remaining life obtained from each functional test data.
You can also find 2.

Figure 5 is a flowchart showing the processing procedure of the correlation analysis section. Moreover, FIG. 10 and FIG. 11 are diagrams explaining this correlation analysis. First, functional test data and component deterioration data are read. For example, read the CRD drive water flow rate data and carbon seal bending strength data. Using these data, regression analysis is performed using the least squares method, a linear regression model, etc. to obtain an approximate equation (the equation shown by the solid line in FIG. 10). Using this approximation formula, determine the deterioration characteristic value σ (of the component for the current functional test data Ft).Then, the deterioration characteristic value and progress predicted from the past process quantities that promote deterioration, such as the operating history of temperature in CHD. Relationship with time (C
This relationship in RD is shown in Figure 11. ), the current virtual elapsed time t' for the deterioration characteristic value σ of the component is determined. Then, the time required to reach the limit value is calculated from the predicted relationship between the deterioration characteristic value and the elapsed time and the limit value of the component, and the difference between this time to reach the limit value and the virtual elapsed time t' is defined as the remaining life L35. is a flowchart showing the processing procedure of the remaining life evaluation section 6 Here, the remaining life LL, L! obtained as described above is shown. , [, 3, the most reliable remaining life L is determined, and the equipment to be inspected is selected based on the result. The equipment to be inspected is selected as follows.

Assume that the total number of devices to be inspected is N. The number of devices that showed signs of failure between the previous periodic inspection and the current periodic inspection is N1, and the number of devices whose functional test data is expected to exceed the control value by the next periodic inspection is Nz. Suppose that Therefore, (N1+Nz) pieces of equipment will be intensively inspected during this periodic inspection, regardless of their remaining service life.
For the other (N-(Nt+Nz)) devices, the following processing is performed individually. First, remaining lifespans L1, Lx, and L3 are determined using the remaining lifespan diagnosis device described above. Then, the remaining life L of the device is set to be the shortest among these remaining lives L, L2, and L3. Then, it is determined whether or not the remaining life is within a specified period, for example, one year, and equipment with one year or less is subject to this periodic inspection, and other equipment is scheduled to be inspected in consideration of its remaining life. Decide on the periodic inspection period for inspection. Of course, if the number of devices that need to be inspected this time is small, the devices with the shortest remaining life will be inspected first until the number of devices that can be inspected in this periodic inspection is reached. When executing the above processing, especially the remaining life evaluation function,
Put your reasoning device to work. For example, a knowledge base with rules in the if=then format stores the following production rules.

if (Scrum time is 3.5 seconds or more) then (Replace the relevant CRD.) if (The remaining life of the motorized valve is 1 year or less) then (Replace the relevant motorized valve.) if (There are no defects and the limit value (The need for the current inspection of the CRD is small.) Then (The need for this inspection of the CRD is small.) When applying the above-mentioned expert system to a power plant equipped with, for example, four types of equipment to be inspected, first, For example, a menu screen shown in FIG. 12 is displayed on the display screen of the terminal system shown in FIG. If the subject of periodic inspection is CRD, select No. 1 on this menu screen. 6. Next, enter the year of diagnosis. Failures, reliability, and remaining life are then calculated for each diagnosis year and displayed at the bottom of the menu screen in Figure 12. Furthermore, as preventive maintenance guidance, the periodic inspection interval based on the required reliability during periodic inspection is presented. If the remaining life of a CRD is simply expressed as a numerical value, the number of CRDs used in nuclear power plants is large. It is difficult for operators to understand if each item is simply displayed as a numerical value. Therefore, if the remaining life of each CRD is displayed together with the arrangement position of the CRD as shown in FIG. 14, it becomes easier to compare the remaining life of each CRD and to understand the overall trend. Furthermore, by displaying the CRD in different colors according to the length of remaining life, trends in remaining life can be seen at a glance. FIG. 13 is a diagram explaining the concept when this expert system is applied to an electric valve.

In the case of electric valves, the mechanical strength of the gland packing and stem nut corresponds to the deterioration characteristic values of the component parts, and the process quantities that cause these deteriorations include the environmental temperature and fluid pressure6.In addition, the equipment performance data There is the amount of fluid leakage and the amount of wear on the threads of the stem nut. Based on this data, the remaining lifespan of many electric valves is predicted, and safety can be further improved by selecting the electric valves to be disassembled and inspected during this or subsequent periodic inspections. According to the embodiments described above, by using process quantities such as operating temperature history that deteriorate the characteristics of the equipment to be inspected, it is possible to non-destructively predict the deterioration tendency of parts, and based on that data, each CRD and The remaining life of electric valves, etc. can be predicted. From this result, the failure rate, reliability, periodic inspection cycle, etc. of each CHD, electric valve, etc. can be predicted quickly and with high accuracy, so the time required to formulate a preventive maintenance plan can be shortened. Furthermore, the effects of improving the reliability and economic efficiency of power plants are significant.

In the above-mentioned embodiment, a power generation plant was explained as an example, but the present invention can be applied to all systems in which a target to be diagnosed is composed of a plurality of parts, and the lifespan of each component is related to the lifespan of the whole. Needless to say, it is a thing.

Other embodiments of the present invention will be described below with reference to the drawings.

FIG. 15 is a configuration diagram of a typical example of an expert system according to an embodiment, and is a configuration diagram of an expert system 11 for diagnosing the remaining life of a component assembly of a power generation plant (for example, a nuclear power plant) as an example of a component assembly. be.

Expert system 1 is knowledge acquisition support system 11F2
, an inference device 1f3, a user interface 4, an external system interface 5, and a knowledge base 16.
The user interface l4 is a database system for managing plant data. A terminal system 18 having a system 17 and input/output devices such as a keyboard and a hard copy device.
It is connected to the. Further, a display device, for example, a CRTIO, is connected to the terminal system 18.

The following three types of data 21, 22, and 23 are input into the terminal system 18 using a keyboard (not shown) or the like. The functional test data 21 is functional test data at the time of periodic inspection of plant component equipment (parts assembly), and is input at each periodic inspection. The data 22 is part deterioration characteristic data obtained by an accelerated life test regarding the components of the component equipment, and is inputted in advance and is part deterioration characteristic data inputted in a timely manner with respect to the component.

The knowledge data 23 is knowledge data related to preventive maintenance work obtained from experts and past experience (specifications, performance, limit values of component equipment and parts, accident and malfunction information, maintenance information, etc.), and forecasts are also input. .

The external system interface 5 is provided online with data of the plant in operation (for example, data indicating the environment of the component equipment (for example, temperature (T)), etc.) from an external sensor (not shown) as historical data 24.

The data 21 and 22 are stored as databases in the database system 17 via the terminal system 18 and the user interface 4 in files 70 and 72, respectively. The historical data 24 is stored in the file 76 of the database system 7 via the external system interface 5 and the user interface 4.

The knowledge data 23 is stored via the terminal system 18, the user interface 4, and the knowledge acquisition support device 2 in the knowledge data file 64 of the technology mt<--base 16 in a format that allows inference processing. The inference device 13 includes a component deterioration analysis section 36, a device performance analysis section 32, a correlation analysis section 34, and a remaining life evaluation section 38.
Like the inference device 3, the inference device 13 performs (1) to (3)
It has the following characteristics and also determines the optimal remaining life L. FIG. 16 shows a flowchart of the procedure for diagnosing the remaining life of a device (parts assembly) in this embodiment.

First, for example, a menu screen shown in FIG. 12 is displayed on the display screen of the CRTIO (step 200).

Next, select the device to be diagnosed from the displayed menu, for example, the CRD.
is specified (step 202).

Then, for CRD, first, partial deterioration analysis processing (2
04), equipment performance analysis processing (step 206), correlation analysis processing (step 208), and remaining life evaluation (step 210) are sequentially performed.

In the example below, the remaining life L obtained by the remaining life evaluation process is output and displayed, but it is also possible to output and display the remaining life obtained by one of the parts deterioration analysis, equipment performance analysis, and correlation analysis. It's okay.

FIG. 17 shows a CH which is an example of the equipment to be diagnosed in this embodiment.
A longitudinal section of D is shown. In the figure, 42 is P-bon seal, 44
is a spud, 46 is the bottom of the reactor pressure vessel, 48 is a cylinder, 50 is a housing, 52 is a drive piston, 54 is a collet spring, 56 is a collet piston, 58 is a stop piston, 60 is a collet tube, 62 is a piston tube, 64 is an index tube, 66 is a drive water extraction pipe, 67 is a drive water insertion pipe, and 68 is a Paul check valve. The arrow in the figure indicates the direction in which the driving water flows when the control shaft is pulled out. First, the processing procedure for parts deterioration analysis will be explained using the flowchart shown in FIG. The parts deterioration analysis process in this example is as follows. Let Ll' be the remaining life of the CHD obtained from the component deterioration characteristic data of each component that makes up the CRD, for example, accelerated life test data, and let Ll' be the remaining life of the CHD obtained from the component deterioration characteristic data of each component constituting the CRD, and the failure data or component deterioration characteristic data of each component, for example, accelerated life test data. The remaining life of the CRD is determined from the reliability of each component determined based on Ll.
5, the shorter of the two is set as the remaining life L1.

Of course, as mentioned above, even if LL' = Ll, Ls'
It goes without saying that L 1 may also be used. In this case, the remaining life of the device (CRD) can be predicted by evaluating changes over time in the bending strength, hardness, etc. of each component of the device under certain operating conditions. can.

In other words, the bending strength of the carbon seal 42, which is one of the components constituting the CRD, decreases as the operating environment conditions (e.g., temperature, pressure, number of times of use, etc.) increase. Therefore, by examining the history change characteristics of the bending strength of the carbon seal 42 as a function of operating temperature, its deterioration characteristics can be easily evaluated and predicted. , failure information of the CRD stored in the file 72 of the data bus system (for example, abnormally high temperature of the CRD, deformation of the coupling between the CRD and the control shaft, etc.) or acceleration of each component of the CRD (carbon seal, etc.) Load life test data.

The failure information is used in the remaining life evaluation provided from the terminal system 18 to the database system 17 at any time.

In step 502, reliability analysis such as Weibull distribution analysis is performed using the read data, for example accelerated life test data.

Although there are other reliability analysis methods based on normal distribution, lognormal distribution, exponential distribution, etc., Weibull distribution analysis will be explained here.

First, we will analyze the carbon seal data as accelerated life test data. Figure 7 shows an example of accelerated life test data for carbon seals. The Weibull distribution function is given by the following equation. (t≧O,'l+”>O,mi>O) −(
6) Also, the unreliability Fi(t) is given by the following equation, and the reliability R1(t) is given by equation (2). Here, mi is the Weibull shape parameter representing the failure mode of the component (m l < 1 for initial failure, m+=1 for random failure, m+>1 for wear-out failure), and η is It is a scale parameter indicating the characteristic life.

Based on the accelerated life test data of the carbon seal 42 shown in FIG. 7, the shape parameter m1 and the scale parameter η at the expected temperature after the present time are determined from the distribution function equation (1).

In the next step 504, the reliability R1 of the component at the expected temperature is determined using equation (3) from both of the above parameters and the operation history time t of the target component (carbon seal).

FIG. 22 is a characteristic diagram of the unreliability F(t) at each temperature (50, 100, 200.285 and 300° C.) of the carbon seal 42 obtained from the deterioration characteristic diagram of FIG. 7.

In FIG. 22, the shape parameter mi at each temperature is determined from the slope of the characteristic straight line at each temperature, and the point in time when the straight line reaches an unreliability of 63.2% corresponds to the characteristic life η1. In addition, in the figure, E on the horizontal axis means an index, and IE-1=1
0″″l=Q,↓ ,IE+0=10'=1,IE+1
=10'-40.

The reliability R1 obtained in this way is temporarily stored in a memory (not shown) in the inference device 13 and used in step 520 later.

Next, in step 506, accelerated life test data for the carbon seal and historical data of operating environmental conditions (eg, operating temperature) up to this point are read from the file 76.

In step 508, the deterioration tendency of the carbon seal is analyzed based on these data to determine the deterioration characteristic value of the carbon seal.

As is clear from Fig. 7, the rate of deterioration of bending strength σ tends to increase as the operating temperature increases, and bending strength It turns out that it can be expressed as an exponential function of temperature.

a = a 0exp{- - f (T) x tα
) =-(1)f(T)=aT'+bT"-
"-+xT"+yT+z:xT2+yT+z...(8
) Here, σ. is the initial value (experimental value) of the deterioration characteristic value, T is the process amount that promotes deterioration, here is the operating temperature, αl a
, b...X1 yIZ is an experimental constant, and f(T) is an approximate expression for life data. Generally α=1.

Therefore, the constant X*YrZ is determined by, for example, the least squares method using temperature history data and accelerated life test data.

Therefore, if the predicted pattern T of the operating temperature is known from equations (1) and (8), the predicted deterioration characteristic value σ(1) can be determined as a function of time t.

Note that equations (1) and (8) can be applied not only to carbon seals but also to other parts. For example, the thread wear amount σ(1
) is determined as a function of the number of uses T and time t. However, in that case, the experimental constants x, y, and 2 have different values.

The curve shown by the solid line in FIG. 23 is the deterioration characteristic value data of the carbon seal calculated by equations (1) and (8) in consideration of the historical temperatures T1 and Tz up to the current time tl.
The initial value σ0 of the bending strength is stored in the file 72, and the limit value σC is also stored in advance in the file 64 as knowledge data.

The process amount T, that is, the temperature at the present moment 11 is T3 (
'C), and assuming that the current temperature will continue in the future, the expected pattern of deterioration characteristic values will be found as shown by the dotted line.

Generally, the expected pattern of the history of the process amount, here the ambient temperature, is selected from the following three types.

(i) Continuation of constant temperature: Continuation at the same temperature as the current temperature.

(n) Continuation of constant weighted average temperature; continues from now on at the weighted average temperature up to the current point.

(m) Temperature change pattern: Changes periodically in the same pattern as the temperature change up to the present time.

Therefore, if the remaining life of the carbon seal is the operating time required from the present moment until the expected value of the characteristic value reaches the limit value σC, the remaining life Lll is determined by equation (9) (step 512
, 514). Ltt=log(c+o/σc)/f(T)-tt
...(9) Here, T is one selected from the above three types of prediction patterns, and T(2) . (6
) and (7) are determined by this selection prediction pattern. The above processing steps 502 to 514 are repeated until the analysis is completed for all n parts constituting the CRD (step 516). Reliability Rt determined for each part
The following processing is performed using the remaining life Ls+ and the remaining life Ls+.

First, extract the shortest remaining life from the remaining life Lll (Lll to Ltn) of each component and set it as LX' (
step 518). In a CHD, the part with the shortest life is the carbon seal, so the remaining life of the carbon seal is selected with high probability as L1/. Next, the reliability Ra of the device (CRD) is calculated using equation (5) using the reliability Ri of each component obtained in step 504 (step 5
20). Next, the CHD reliability limit value ReC is read from the knowledge data file 64 (step 522). (5
) by substituting R e =R ec into the equation, t is calculated using a successive approximation method such as the Newton-Rabson method.

Here, FIG. 24 is a characteristic diagram of the reliability Re of the CRD, and the value of the reliability Re up to the current time tl is determined from equations (2) and (5) according to the predicted operating temperature T.

Assuming that the predicted operating temperature T continues from the current value T3, the predicted pattern of the future reliability Re is (2)
, From equation (5), it can be predicted as shown by the dotted line that Re = ReC
The time point tc at which tc becomes can be calculated by the above-mentioned successive approximation method.

Therefore, the remaining life LL' of the CRD is determined as L1' = tc-t+ (step 526).

Finally, the remaining life span L1/ is compared with the remaining life span Ls', and the shorter remaining life span is set as L1 (step 528).

FIG. 19 is a flowchart showing the processing procedure of the device performance analysis section 32. The equipment auxiliary analysis unit 32 analyzes equipment (CR
Analyze the functional test data in D) to determine the remaining life L2 of the CRD. First, in step 600, functional test data is read from the file 70. As the functional test data, for example, in the case of a CRD, past data on the amount of leakage of drive water that is carried out during regular inspections is read.

The driving water is used to push up and lower the control shaft as shown in FIG. Other carbon seals 42
Leak water flows between them as shown by arrow 40. As the amount of leak water increases, the flow rate of drive water required to push up the control rod increases.

Therefore, the driving water flow rate can be used as a quantity that indicates the deterioration of the CHD function. Therefore, regression analysis such as the least squares method is performed based on the temporal change trend of the data of drive water flow rate < Ql minute during past periodic inspections, which is indicated by the X mark in Fig. The broken line equation) is calculated (step 602).

F=Pt”+qt+r−
(10) Here, −P+ q+ r is a constant determined from test data. Next, the limit value Fc of the driving water flow rate F is read from the file 64 (step 604). Based on approximate formula (10), find the time tc when the water amount F reaches the limit value Fc, and (tc
-t1), the remaining life L2 is determined (step 606,
608).

If the CRD has multiple types of functional test data, calculate the remaining life using each functional test data.
The shortest one among them may be selected.

Alternatively, the optimal remaining life Lz may be determined by the following equation (1l), taking into consideration the weighting determined by the importance analysis on the remaining life determined from each functional test data.

L2=(ΣβJLzJ)/Σα1β, ...
(1l) where: j: Functional test item number β: Weighting coefficient FIG. 20 is a flowchart showing the processing procedure of the correlation analysis section 34. Also, the characteristics shown in Figures 10 and 11 are used in this processing procedure as well. First, functional test data of the device and deterioration data of the component parts of the device are read (step 700). That is, for example, the CRD drive water flow rate data (FIG. 9) and the carbon seal bending strength data (FIG. 7) are read from files 70 and 72, respectively. Figure 10 shows the correlation of these data. and,
Approximate equation (12) (solid line equation in FIG. 10) is obtained by regression analysis using the least squares method, linear regression model, etc. (step 702).

a = SF + So
”・(12) Here, S and So are constants determined by the above data.

Next, the deterioration characteristic value σ of the component for the functional test data Ft at the current time t1 is determined from this approximate expression, and σt=
SFt+So is obtained (step 704).

Next, a predicted pattern of the deterioration characteristics of the carbon seal similar to that shown in Fig. 23 is derived from the operating history data of the operating temperature, which is a process variable in file 74, and the accelerated life test data of the bending strength of the carbon seal (Fig. 7). Obtained as shown in the solid curve in Figure 11. That is, the experimental constant x in equations (1) and (8),
Determine y and z.

Next, from the above component deterioration characteristic value σ, the current virtual elapsed time t' for the deterioration characteristic value σ, based on equation (1), is calculated as t' = log(co/cr t)/f(
T) (step 706).

Furthermore, from the predicted pattern of deterioration characteristics and the component limit value σ0, find the limit value arrival time tc as tc=log(σo/ac)/f(T), and use this difference (tc-t') as the remaining life L8. (step 708).

Note that if there are multiple types of at least one of the component deterioration data and the functional test data, the remaining life may be determined for all combinations of the functional test data and the component deterioration data, and the shortest one may be set as the remaining life L8. Further, in the above example, the virtual elapsed time t' was determined from the current drive water flow rate Ft, but t' may be determined from the current bending strength σ, and L3 may be determined.

Based on the processing results of the analysis units 32 to 36 as described above, the remaining life evaluation unit 38 evaluates and displays the remaining life.

FIG. 21 is a flowchart showing the processing procedure of the remaining life evaluation section 38. Here, the most reliable remaining life L is determined from the remaining lives LX, L2, and L3 obtained as described above, and the equipment to be inspected is selected based on the result.
This is used to display inspection results, etc. First, in step 800, the remaining lives Ll and L2 obtained are
1. Preferably the shortest Ls is connected to the device (CHD
)'s remaining life is L.

If there are multiple CRDs that are diagnostic target devices, all CRs
Perform the above analysis on D to find the remaining life L. Next, it is determined whether the remaining life L of each CRD obtained is less than a predetermined time (eg, periodic inspection period), for example, one year (step 802). Each CRD that is less than one year old is subject to inspection at the current periodic inspection, and for each CRD that is not older, a check is made to see if any signs of malfunction have appeared between the previous inspection and the current inspection (step 804). ).

Here, the current periodic inspection time means the next periodic inspection if the current inspection is a periodic inspection, or the nearest periodic inspection if it is a normal inspection.

Here, malfunctions include, for example, sudden changes in CRD operating temperature, CRD
This is a modification of the coupling between the HD and the control rod, etc., and can be detected by checking the history data in the file 76.

CRDs that are found to be defective will be subject to inspection during this periodic inspection.

For each CRD in which no defects have occurred, it is determined whether the functional test data is expected to exceed the limit value at the time of the next periodic inspection (step 806). That is, it is checked whether the remaining life L of the CRD obtained through the functional performance analysis is shorter than the period up to the current periodic inspection, and if it is shorter, the CRD is subject to inspection. For other CRDs, it is determined that no inspection is necessary during the current periodic inspection (step 808), and the next periodic inspection timing is determined by taking into consideration their remaining life (step 810). For example, if the remaining life is 2 years, it is decided to conduct an inspection after 1 year, and if the remaining life is 3 years, it is decided to conduct an inspection at the regular inspection after 2 years.

On the other hand, for CHDs that are determined to require inspection, the CRD
It is checked whether the number of inspections is equal to or greater than a predetermined number of inspections that can be performed at the time of inspection, and if the number is above a predetermined number of inspections, for example, CHDs with short remaining life are selected preferentially until the predetermined number is reached.

On the other hand, if the number of CHDs judged to need to be inspected this time is small, the CRDs with the shortest remaining life will be inspected in order of priority until the selected number of CHDs that can be inspected this time is reached. The above diagnosis results are sent to the terminal system 18, and
Information on CRDs that are determined to require inspection are stored in the file 70 of the database system 7 as inspection history data.

When executing the above processing, especially the remaining life evaluation function (steps 802 to 806, 810, etc.), the inference function is activated. For example, the following production rules are stored in the knowledge base 16 that has rules in the if"then format: if (the remaining life of the CHD is one year or less) then (replace the CHD 9) if ( There is no problem with the CHD, it is smaller than the limit value, and the remaining life is over 1 year.) then (The need for this inspection of the CRD is small.
) if (the remaining life of the CRD is 3 years) then (the CRD will be inspected in 2 years) if (
Then (Replace the CRD.) Next, in step 816, operate the keyboard of the terminal system to display the output selection menu screen on the CRT 20 and select the output of the diagnosis result. The diagnosis result menu is as follows:
For example, "Remaining life map""Selected CRD map JT%l
Reasons for selection'', etc. If you select "Remaining life map", the remaining life of all CRDs is displayed, preferably along with the CHD placement position, as shown in Figure 14, allowing you to compare and compare the remaining life of each CRD. The overall trend is easier to understand.Also, by displaying the CRD in multiple colors according to the length of the remaining life, you can understand the trend of the remaining life at a glance.In addition, in the diagram, the horizontal axis The station on the vertical axis indicates the coordinate position of CRD. If you select "Selected CRD map", preferably the 25th
As shown in the figure, all CRDs are displayed, and among them, the CRDs selected in step 814 as inspection targets are displayed in different colors. That is, for example, the CHDs may be sequentially assigned numbers as shown in the figure, and the selected CRDs may be displayed in different colors as shown below depending on the reason for their selection.

Red: CRD determined to have a defect Purple: CRD determined to have a remaining life of one year or less Yellow: CRD whose functional test data is expected to exceed the limit value by the next regular inspection The above remaining life map and selection The CRD map may specify the CRD using the keyboard and display only the remaining life of the specified CHD, whether it has been selected, and the reason thereof.

If you select ``Reason for selection'' and specify the shame of the selected CRD, the reason for selection will be displayed as shown in Figure 26. In addition to this, past operating temperatures of each CHD can be read out from the file 74 and displayed, functional test data can be read out from the file 70 and displayed, and component deterioration data can be read out and displayed from the file 72. Also good. It should be noted that since the information of the CRD targeted for inspection is stored in the file 70, it can be read out and displayed at any time. This embodiment can also achieve the same effect as the embodiment shown in FIG. 1. In the above embodiment, the shortest one among LL, Lz, and La is selected as the remaining life, but the remaining life L may also be Ll/ obtained in step 518, or Li' obtained in step 526. Alternatively, it may be Ls obtained in step 528, L2 obtained in step 608, or L3 obtained in step 708. Also, the shorter of remaining life L1' and L8 is L.
It's good as well.

Selection of these remaining life analysis methods is performed at menu selection step 202 in FIG. 16.

Similarly, in step 202, one item of the functional test may be selected from, for example, scram time, driving water flow rate, etc.

In addition, in step 202, the selection of the type of parts (e.g., carbon seal, collet spring, etc.) used for part deterioration analysis, the selection of deterioration parameters (e.g., bending strength, hardness, etc.), and the specification of their limit values are also carried out in step 202. conduct.

Furthermore, the amount of process that promotes deterioration (used in component deterioration analysis) is
For example, the operating temperature, etc.) are specified, and an expected pattern of the history of the specified process amount is selected in step 202.

For example, there are three types of prediction patterns: i) Constant process amount: The current process amount continues.

ii) Constant weighted average process amount: The weighted average process amount up to the current point continues.

City) Process amount change pattern: T! ! ．． Thereafter, it changes periodically in the same pattern as the process amount change up to that point.

〔Effect of the invention〕

According to the present invention, the remaining lifespan of the object to be diagnosed can be accurately and quickly predicted, making it possible to ensure preventive maintenance.

[Brief explanation of the drawing]

Fig. 1 is a block diagram of a remaining life diagnosis device according to an embodiment of the present invention, Fig. 2 is a functional block diagram of the inference device of Fig. 1, and Fig. 3 is a processing of the parts deterioration analysis section of Fig. 2. FIG. 4 is a flowchart showing an example of the processing procedure of the equipment performance analysis section in FIG. 2; FIG. 5 is a flowchart showing an example of the processing procedure of the correlation analysis section of FIG. 2; The figure is a flowchart showing an example of the processing procedure of the remaining life evaluation section in Fig. 2, Fig. 7 is a graph showing the deterioration characteristics of the carbon seal, Fig. 8 is an explanatory diagram for calculating the remaining life Ll, and Fig. 9 is the remaining life. An explanatory diagram for calculating L2, Figure 10 is a diagram of the relationship between force and bending strength of the Bonseal and drive water flow rate, and Figure 11 is the remaining life La.
Fig. 12 is an explanatory diagram of the menu screen of a remaining life diagnosis device that diagnoses the remaining life of a power generation plant, Fig. 13 is a conceptual illustration of remaining life diagnosis of an electric valve in a power generation plant, and Fig. 14 is a CRD FIG. 15 is an overall configuration diagram of a remaining life diagnosis device according to another embodiment of the present invention, FIG. 16 is a flowchart showing the remaining life diagnosis processing procedure in the embodiment of FIG. 15, Figure 17 shows the control shaft drive device (C
HD), FIG. 18 is a flowchart showing an example of the processing procedure of the partial deterioration analysis section of FIG. 15, and FIG. 19 is a flowchart showing an example of the processing procedure of the equipment performance analysis section of FIG. Fig. 20 is a flowchart showing an example of the processing procedure of the correlation analysis section of Fig. 15, Fig. 21 is a flowchart showing an example of the processing procedure of the remaining life evaluation section of Fig. 15, and Fig. 22 is a flowchart showing an example of the processing procedure of the correlation analysis section of Fig. 15. A characteristic diagram showing the unreliability of the carbon seal obtained from the characteristics, Figure 23 is a predicted diagram of the deterioration characteristics of the carbon seal, Figure 24 is a diagram showing the reliability characteristics of CRD, and Figure 25 is a diagram showing the reliability of the carbon seal as an inspection target. Selected C
R. A diagram showing an example of the display of D, Figure 26 is the selected CRD
FIG. 4 is a diagram showing an example of displaying reasons for selection. 1, l1...Remaining life diagnosis device, 2...Knowledge acquisition support device, 3,↓3...Inference device, 6.16...Knowledge base.

Claims (1)

[Scope of Claims] 1. A method for diagnosing the remaining life of an assembly of parts comprising a plurality of parts and having at least one function, based on deterioration test data of one characteristic regarding at least one part of the assembly. obtaining a first remaining life of the assembly; obtaining a second remaining life of the assembly based on test data for at least one function of the assembly; relating to at least one component of the assembly; obtaining a third remaining life of the aggregate based on deterioration test data of one property and test data of at least one function of the aggregate, and during the first, second and third remaining lives; A method for diagnosing remaining lifespan, characterized in that the shortest value is obtained as the remaining lifespan of the aggregate. 2. A method for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, comprising deterioration test data of one characteristic regarding at least one component of the assembly and at least one of the assembly. Determining a first elapsed time from the start of deterioration of the one characteristic or at least one function to the present time based on test data regarding the one characteristic or test data regarding the at least one function; predicting future deterioration characteristics for the value of said one characteristic or at least one function based on said predicted deterioration characteristic until said value of said one characteristic or at least one function reaches a limit value; A method for diagnosing remaining life, comprising: determining a second elapsed time from the start of deterioration; and determining the difference between the second elapsed time and the first elapsed time as the remaining life of the partial assembly. 3. In the step of determining the first elapsed time, the deterioration data of the one characteristic is shown with respect to the test data of the function by regression analysis from the deterioration test data of the one characteristic and the test data of the at least one function. a step of obtaining a first approximation formula; a step of determining a deterioration data value of the one characteristic at the current time corresponding to a test data value of the function at the current time based on the first approximation formula; obtaining a second approximation formula for the deterioration data by regression analysis based on the characteristic deterioration test data; 3. The remaining life diagnosing method according to claim 2, further comprising the step of determining the first elapsed time from an approximation equation. 4. The one characteristic regarding the at least one component is a function of the amount of process that accelerates the deterioration of the component and the elapsed time from the start of deterioration, and the step of predicting the future deterioration characteristic includes the step of predicting the future deterioration characteristic regarding the at least one component. a step of predicting a future process amount based on historical data of a process amount that promotes deterioration; and a step of obtaining a predicted pattern of the future deterioration characteristics by applying the predicted process amount to the second approximation formula. and the step of obtaining the second elapsed time is a step of obtaining the elapsed time from the start of the deterioration required for the predicted value of the one characteristic to reach the limit value from the predicted pattern as the second elapsed time. Claim 3 comprising
How to diagnose remaining lifespan. 5. Execute the above steps for all combinations of one characteristic of each of the plurality of parts and each test data regarding the plurality of functions of the above assembly, and select the one with the shortest remaining life as the above set. The method for diagnosing remaining lifespan according to claim 2, wherein the remaining lifespan of the body is measured. 6. A method for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, comprising performing a Weibull distribution reliability analysis on deterioration test data of one characteristic regarding at least one part of the assembly. determining a predicted pattern of reliability of the component as a function of time, and determining a first arrival time from the current point until the reliability reaches a predetermined first limit value from the obtained reliability predicted pattern; Determining a predicted pattern of deterioration characteristics of the component from the deterioration test data; Determining a second arrival time from the present moment required for the one characteristic to reach a predetermined second limit value from the predicted deterioration characteristic pattern. and obtaining the shorter of the first and second arrival times as the remaining life of the aggregate. 7. Determine the first and second arrival times for each of the plurality of parts of the assembly, and determine the shortest of the plurality of obtained first arrival times and the shortest of the plurality of obtained second arrival times. 7. The remaining life diagnosis method according to claim 6, wherein the shorter of the two is determined as the remaining life of the aggregate. 8. The reliability of the above-mentioned part is a first function of the first process quantity indicating the operating environment regarding the part and the elapsed time from the start of deterioration, and the step of obtaining the predicted pattern of the above-mentioned reliability includes the following: a step of predicting a future first process amount based on historical data; and a step of applying the predicted first process amount to the first function to obtain a reliability prediction pattern, One characteristic of the component is a second function of a second process amount that accelerates the deterioration of the component and a second elapsed time from the start of deterioration, and the step of obtaining the predicted pattern of the deterioration characteristic includes determining the deterioration of the at least one component. a step of predicting a future process amount based on historical data of a second process amount to be promoted; and a step of obtaining a predicted pattern of the future deterioration characteristics by applying the predicted process amount to the second function. The method for diagnosing remaining lifespan according to claim 6. 9. A method for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, wherein deterioration test data of one characteristic regarding at least one part of the assembly is analyzed by regression analysis of the part. Obtaining an approximation formula showing the deterioration data of the above one characteristic which is a function of the process amount that promotes the deterioration and the elapsed time from the start of the deterioration; applying the predicted process amount to the approximation formula to obtain a predicted pattern of the deterioration characteristics; A method for diagnosing remaining lifespan, characterized in that the elapsed time from the start of deterioration is determined, and the difference between the obtained elapsed time and the current time is obtained as the remaining lifespan of the aggregate. 10. The above approximate formula is: σ(t)=σ_0exp(-f(T)t). Here, σ_0; one characteristic value T at the start of the deterioration; process amount t that promotes deterioration; time f(T)=xT^2+yT+z x, y, z; coefficient 11, 10. The method for diagnosing remaining lifespan according to claim 9, wherein the remaining lifespan is determined for each of them, and the shortest one of them is obtained as the remaining lifespan of the aggregate. 12. The remaining life diagnosis method according to claim 9, wherein the assembly is a control rod drive mechanism of a power generation plant, the component is a carbon seal, and the process amount is an operating temperature of the drive mechanism. 13. In an apparatus for diagnosing the remaining life of a component assembly consisting of a plurality of components and having at least one function, means for receiving deterioration test data of one characteristic regarding at least one component of the assembly, and the deterioration test described above. means for obtaining a first remaining life of the aggregate based on data; means for receiving test data for at least one function of the aggregate; and means for obtaining a second remaining life of the aggregate based on the test data for the function. means for obtaining a lifespan; and means for obtaining a third remaining life of the assembly based on deterioration test data for one property regarding at least one component of the assembly and test data for at least one function of the assembly; , and means for obtaining the shortest value among the first, second, and third remaining lives as the remaining life of the aggregate. 14. In a device for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, deterioration test data of one characteristic regarding at least one part of the assembly and at least one of the assembly means for receiving test data regarding the function; and means for determining a first elapsed time from the start of deterioration of the one characteristic or at least one function to the present time based on the deterioration test data of the one characteristic and the test data of the function. and means for predicting future deterioration characteristics of the value of the one characteristic or at least one function based on deterioration test data of the one characteristic or test data of the at least one function; and the predicted deterioration characteristics. means for determining a second elapsed time from the start of deterioration required for the value of the one characteristic or at least one function to reach a limit value from the above, and the difference between the second elapsed time and the first elapsed time. and means for determining the remaining life of the partial assembly. 15. The means for determining the first elapsed time indicates the deterioration data of the one characteristic with respect to the test data of the function by regression analysis from the deterioration test data of the one characteristic and the test data of the at least one function. means for obtaining a first approximation formula; means for determining a deterioration data value of the one characteristic at the current time corresponding to the test data value of the function at the current time based on the first approximation formula; means for obtaining a second approximation equation for the deterioration data by regression analysis based on the deterioration test data of the one characteristic; and means for determining the first elapsed time from an approximate expression. 16. One characteristic of the at least one component is a function of the amount of process that accelerates the deterioration of the component and the elapsed time from the start of deterioration, and the means for predicting the future deterioration characteristic of the at least one component includes: means for predicting future process quantities based on historical data of process quantities that promote deterioration; and means for obtaining a predicted pattern of the future deterioration characteristics by applying the predicted process quantities to the second approximation. and the means for determining the second elapsed time is means for obtaining the elapsed time from the start of the deterioration required for the predicted value of the one characteristic to reach the limit value from the predicted pattern as the second elapsed time. The remaining life diagnostic device according to claim 15. 17. Find the remaining life for all combinations of one characteristic of each of the plurality of parts and each test data regarding the plurality of functions of the above assembly, and calculate the shortest remaining life of the above assembly. Claim 14 configured to have a remaining life.
Remaining life diagnosis device. 18. In an apparatus for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, means for receiving deterioration test data of one characteristic regarding at least one part of the assembly, and the deterioration test described above. means for performing a Weibull distribution reliability analysis on data to obtain a predicted pattern of reliability of the part as a function of time; means for determining a first arrival time from the deterioration test data; means for determining a predicted pattern of deterioration characteristics of the component from the deterioration test data; A remaining lifespan diagnosis device comprising: means for determining a second arrival time from the present time required to reach the above, and means for obtaining the shorter of the first and second arrival times as the remaining lifespan of the aggregate. 19. Determine the first and second arrival times for each of the plurality of parts of the assembly, and determine the shortest of the plurality of obtained first arrival times and the shortest of the plurality of obtained second arrival times. 19. The remaining life diagnosing device according to claim 18, further comprising means for outputting the shorter of the two as the remaining life of the aggregate. 20. The reliability of the above-mentioned parts is the first indicative of the operating environment of the parts.
is a first function of the process amount and the elapsed time from the start of deterioration, and the means for obtaining the predicted pattern of reliability includes means for predicting the future first process amount based on the historical data of the process amount, and the prediction applying a first process quantity determined to the first function to obtain a predicted pattern of reliability, wherein one characteristic regarding the at least one component is a second process quantity that promotes deterioration of the component. The means for obtaining the predicted pattern of the deterioration characteristic, which is a second function of elapsed time from the start of deterioration, predicts a future process amount based on historical data of a second process amount that promotes deterioration of the at least one component. 19. The remaining life diagnosis apparatus according to claim 18, further comprising: means for applying the predicted process amount to the second function to obtain a predicted pattern of the future deterioration characteristics. 21. In a device for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, means for receiving deterioration test data of one characteristic regarding at least one component of the assembly, and the deterioration test described above. means for regression-analyzing the data to obtain an approximate formula representing deterioration data of the one characteristic that is a function of the process amount that accelerates the deterioration of the part and the elapsed time from the start of deterioration; and the process amount that promotes the deterioration. means for predicting a future process amount based on historical data; means for applying the predicted process amount to the approximation formula to obtain a predicted pattern of the deterioration characteristics; means for determining the elapsed time from the start of the deterioration required to reach a predetermined limit value; and means for obtaining the difference between the obtained elapsed time and the current time as the remaining life of the aggregate. 22. The means for obtaining the predicted pattern is as follows: σ(t)=σ_0exp{−f(T)t} as the above approximate expression. Here, a predicted pattern of the deterioration characteristics is obtained using σ_0; one characteristic value T at the start of the deterioration; a process amount that promotes the deterioration; time f(T)=xT^2+yT+z. Remaining life diagnostic device. 23. The remaining life diagnosis according to claim 21, wherein the means for obtaining the remaining life is a means for obtaining the remaining life for each of the plurality of parts of the assembly, and obtaining the shortest among them as the remaining life of the assembly. Device. 24. In a method of determining the remaining life of the components of a parts assembly and displaying the remaining life on the screen, each component of the parts assembly is displayed as a figure corresponding to the actual placement position, A method for displaying remaining life information, characterized by displaying remaining life in correspondence with graphic representations of individual component parts. 25. In the method of determining the remaining life of the components of a parts assembly and displaying the remaining life on the screen, the remaining life of each component is displayed corresponding to the component, and the remaining life of each component is displayed according to the length of the remaining life. A method of displaying remaining life information characterized by color-coded display. 26. In a device for determining and displaying the remaining life of the components of a parts assembly on a screen, a means for displaying each component of the parts assembly as a figure corresponding to the actual placement position, and a means for displaying the remaining life of each component of the parts assembly on a screen. and means for displaying information corresponding to graphical representations of individual component parts. 27. In a device for determining and displaying the remaining life of the components of a parts assembly on a screen, a means for displaying the determined remaining life of each component corresponding to the component and displaying the remaining life in different colors depending on the length of the remaining life. Remaining life information display device having means. 28. In an expert system for diagnosing the remaining life of an assembly of parts consisting of a plurality of parts and having at least one function, deterioration test data of one characteristic regarding at least one part of the assembly and at least one of the assembly means for receiving test data for one function; and a database for storing deterioration test data for one characteristic regarding the at least one component and test data for at least one function from the receiving means; The remaining life diagnosis device according to claim 13, which reads data and calculates the remaining life of the aggregate; a knowledge base that stores information regarding at least inspection of the aggregate according to the value of the remaining life as knowledge data; and the remaining life diagnosis device. An expert system comprising means for displaying information regarding the remaining life of the aggregate obtained in the above. 29. The expert system according to claim 29, wherein the knowledge data includes information indicating that inspection of the assembly is necessary if the remaining life is within a predetermined period. 30. The expert system of claim 29, wherein the knowledge data includes information indicating that the aggregate needs to be replaced if the test data value for the function is outside a predetermined range.