TWI683058B - Failure probability assessment system - Google Patents

Abstract

[Problem] The problem of the present invention is to perform a high-precision failure probability assessment or residual life assessment on a plurality of constituent elements of a mechanical system. [Solution Means] The failure probability evaluation system of the present invention is a system for evaluating the failure probability of a plurality of mechanical elements included in a mechanical system, and is characterized by having a physical quantity that changes according to the fatigue damage of the aforementioned mechanical elements or changes over time. , A means of evaluating the state quantity indicating the soundness of the mechanical element; a means of evaluating the cumulative fatigue damage degree of the mechanical element based on the physical quantity 2 that changes according to the load or load on the mechanical element or the operating data of the mechanical element ; A storage unit that stores the state quantity and the fatigue damage degree; and based on the state quantity and the fatigue damage degree of the mechanical element that has failed among the plurality of mechanical elements, calculates the machine that has not failed among the plurality of mechanical elements The failure probability evaluation section of the element's failure probability.

Description

Failure probability assessment system

The present invention relates to a system for evaluating the probability of failure of a plurality of homogeneous mechanical groups included in a mechanical system.

In the past, several methods of residual life assessment for fatigue failure of mechanical systems have been proposed. As a representative example, there are those who use a linear cumulative damage law (Non-Patent Document 1). Specifically, for example, there is a method described in Patent Document 1. On the other hand, as other methods for evaluating the soundness of mechanical components, a technique called abnormality diagnosis or warning sensing is known (Patent Document 2). [Prior Technical Literature] [Patent Literature]

[Patent Document 1] Japanese Patent Laid-Open No. 2015-229939 　　 [Patent Document 2] WO2016/117021 [Non-Patent Document]

[Non-Patent Document 1] M.A. Miner: Cumulative Damage in Fatigue, J. Appl. Mech., 12(3), ppA159-A164

(Problems to be solved by the invention)

In a mechanical system, mechanical elements or mechanical structures (hereinafter collectively referred to as mechanical components) exposed to dynamic load are designed for fatigue life in such a way as to satisfy the required life. However, in actual use, the dynamic load acting on these varies depending on the use environment or conditions. At the same time, the fatigue life of each body of these mechanical components is also potentially uneven. Therefore, in the design stage of fatigue life, these uneven ranges are assumed, and then the design on the safety side is carried out in a way that will not cause fatigue damage/fatigue failure. However, in recent years, from the perspective of cases where the mechanical system is exposed to an unexpected environment, or from the viewpoint of saving resources/energy, etc., the operation/maintenance of the mechanical system that seeks to safely exhaust the life of the mechanical components is sought. In view of the situation shown above, it is extremely important to more accurately grasp the residual life of each component of the mechanical system that is actually in the activated state. In addition, as mentioned above, the lifetime of mechanical components is unevenly distributed according to a certain probability. Therefore, the residual life is defined by the probability theory, and the evaluation of the residual life is equivalent to the evaluation of the probability of failure at a point in time when any time (residual life) passes from a certain point. If the soundness of the object can be evaluated as the probability of failure, by multiplying the amount of loss that occurs when the object fails and the probability of failure, the expected value of the amount of loss can be theoretically calculated. From the result, from an economic point of view See, you can easily make decision-making on operation/maintenance.

Under the background shown above, several methods of residual life assessment for fatigue failure of mechanical systems have been proposed. A typical example is the use of a linear cumulative damage law (Non-Patent Document 1). For example, if it is a mechanical structure, a strain gauge or other sensor that measures strain or stress is installed on the target part to obtain its time history data. For the obtained time history data, waveform counting methods such as Rain Flow Method are applied to obtain the frequency distribution of strain or stress waveforms. With regard to this occurrence frequency distribution, the fatigue damage degree is obtained by referring to the fatigue diagram of the material constituting the target part, and the linear cumulative damage law. Here, the degree of fatigue damage refers to a physical quantity representing the consumption rate of fatigue life relative to the average fatigue life of the object (Patent Document 1). In addition, the degree of fatigue damage can be obtained, and the probability of failure in any degree of fatigue damage can be obtained by referring to the uneven life defined by the fatigue diagram. In addition, if some methods are used, the degree of fatigue damage after any time can be predicted, and the probability of failure at this time can also be obtained, and the residual life can be evaluated by probability theory. However, the fatigue life of general mechanical components varies depending on the object, and there are cases where the uneven width is about 1/10 to 10 times. Therefore, for mechanical components with relatively large unevenness in relative fatigue life as shown above, it is also difficult to provide residual life or failure probability assessment with the accuracy required for operation/maintenance only by this method.

On the other hand, as other methods for evaluating the soundness of mechanical components, there is a method called abnormality diagnosis or warning detection (Patent Document 2). Quantitatively define or learn the operating state of the mechanical components in a sound state in advance. Here, the activation state refers to physical quantities such as vibration acceleration, frequency, temperature, etc. of the object that are expected to change depending on soundness, or state quantities expressed by a combination of these. During the operation, there are application examples in which the state quantity is evaluated from time to time, and an alarm is issued according to the degree of deviation from the sound state, or the mechanical system is automatically stopped. In this method, since the soundness is directly monitored, it is possible to expect the detection of soundness changes with relatively high sensitivity. However, in order to quantitatively calculate the residual life or failure probability using this method, the relationship between the change in state quantity and the residual life or failure probability when a certain state quantity changes must be obtained in advance, but it must be a mechanical component level or a mechanical system level It is hard to say that it is realistic to consider the time or cost of pre-testing other than material grade.

In addition, for example, in a wind power generation system, in a mechanical system that changes from moment to moment, the method of calculating the residual life in units of time initially is not practical. For example, in a wind power generation system, the average unit time load acting on each mechanical component varies depending on wind conditions or control conditions. Therefore, it is appropriate to change the residual life or failure probability according to these assumed conditions, but in the evaluation of the residual life in units of time, the requirements shown above cannot be met.

As mentioned above, in order to evaluate the probability of failure or residual life of mechanical systems due to fatigue, a method based on a linear cumulative damage law is known, but most of its accuracy is not practically sufficient. In addition, in the soundness evaluation method based on abnormality diagnosis, soundness changes can be sensed with relatively high accuracy, but in order to quantitatively evaluate the probability of failure, it is necessary to conduct a pre-test that requires time/cost. In addition, the method of assessing the residual life by time, especially in mechanical systems where the operating state is not constant, often has a bad situation in practice. Therefore, for mechanical systems used in environments with strong uncertain factors such as wind power generation systems, the emergence of systems that provide practical and highly accurate failure probability assessment or residual life assessment. (Means to solve the problem)

In order to solve the above-mentioned problems, for example, the structure described in the patent application scope is adopted. This case contains a plurality of means to solve the above-mentioned problems. If one example is given, it is a failure probability assessment system. It is a system for assessing the failure probability of a plurality of mechanical elements included in a mechanical system. It is characterized by: The physical quantity of the element’s fatigue damage or change over time 1. The method of evaluating the state quantity indicating the soundness of the aforementioned mechanical element; the physical quantity 2 or the operation of the aforementioned mechanical element that changes according to the load or load on the aforementioned mechanical element Data, a means for evaluating the cumulative fatigue damage degree of the mechanical element; a storage unit that stores the state quantity and the fatigue damage degree; and the state quantity and the fatigue damage degree of the mechanical element that has failed among the plurality of mechanical elements , A failure probability evaluation unit that calculates the failure probability of mechanical elements that have not failed among the aforementioned plural mechanical elements. (Effect of invention)

The present invention has the function of evaluating the fatigue damage degree of the object based on the measurement or simulation by the sensor as well as the well-known technology, and also has the soundness evaluation function measured by the sensor. In addition, the residual life of the mechanical system is evaluated on the axis of fatigue damage rather than the time axis, thereby providing an effective residual life assessment for a mechanical system whose operating state is not constant. In addition, by assuming statistical modeling of the probability distribution of the residual life unevenness in the condition where a certain soundness state is observed, it is possible to provide a future failure probability with a more cumulative degree of damage.

The following describes the embodiments using illustrations. [Example 1]

FIG. 1 is a diagram illustrating the operation of the failure probability evaluation system 100 in the present invention as a mechanical component (mechanical element) using the rotating bearing 1 as an example. The present invention is directed to plural, preferably mechanical component groups of the same type. Each mechanical component that is the object of observation is equipped with: a sensor that measures the physical quantity that changes to reflect the soundness; and for the purpose of evaluating the degree of fatigue damage, the physical quantity that changes according to the load or load on the mechanical component is measured Sensor. In this embodiment, the former corresponds to the acceleration sensor 2, and the latter corresponds to the load cell 3 and the rotometer 4, respectively. That is, the acceleration sensor 2 measures the rotational vibration acceleration generated by the damaged bearing, and obtains the amplitude and the number of repetitions of the load acting on the bearing by the load meter 3 and the rotation meter 4. Among them, the physical quantity system for evaluating the soundness must be directly measured, so at least one sensor for it is necessary in this embodiment, but the load on the object may not necessarily be directly measured. For example, if it is a rotating bearing used in a wind turbine or a car, if it is the former, it can also be estimated from the operating data of mechanical components such as the history of wind conditions or power generation, and if it is the latter, the history of the speed or the number of engine rotations. The history of the load carried by the bearing. In addition, in this embodiment, it is premised on the soundness evaluation by the acceleration sensor 2, but it is not intended to limit the type of the sensor to the acceleration sensor. For example, an AE sensor may also be used Or a temperature sensor, or a combination of multiple types of sensors. Moreover, the sensor which measures the physical quantity which changes according to load or load may also be a strain sensor. Various sensors can be reset as part of the failure probability evaluation system 100, and the failure probability evaluation system 100 can have a signal receiving unit (not shown) to receive the detection value of the sensor installed in the mechanical component.

The vibration acceleration data obtained by the acceleration sensor 2 is transmitted to the state evaluation unit 6 via the A/D conversion unit 5. Here, the vibration acceleration data system is converted into a state quantity for evaluating the soundness of the bearing 1. The state quantity expressing soundness considers several methods. For example, in the case of a bearing, with a repeated load, a slight crack or flaking may occur in the inner wheel or the outer wheel. Every time the internal rotating body passes through these damaged positions, vibration occurs. Therefore, it is more effective to use an acceleration effective value of a frequency band corresponding to the frequency corresponding to the number of rotations multiplied by the number of rotating bodies multiplied by the number of rotating bodies. That is, in the present embodiment, the acceleration effective value of the specific frequency band is used as the state quantity indicating the soundness (hereinafter referred to as the state quantity) in the following evaluation. As mentioned above, if a complex type of sensor is used, for example, by applying cluster analysis, etc., the obtained complex physical quantity data is converted into one state quantity for use. The obtained state quantity is transferred to the state quantity change temporary storage unit 8 and temporarily saved as time series data. Here, the storage period of the time series data can be set arbitrarily, but it is set to the same period as the lead time required for the maintenance or exchange operation of the object, from the viewpoint of suppressing the required memory area without compromising the function , The most effective.

On the other hand, the load amplitude and the number of repetitions obtained using the load meter 3 and the rotation meter 4 are transmitted to the fatigue damage degree evaluation unit 7. The fatigue damage degree evaluation unit 7 calculates the fatigue damage degree based on the linear cumulative damage law (Non-Patent Document 1). In this embodiment, the load obtained by the load meter 3 is used as the load amplitude, and the number of rotations obtained by the rotometer 4 is used as the number of iterations to calculate the degree of fatigue damage. However, if a mechanical structure is used as an object, strain can also be used. Instead of the load cell 3, a strain sensor such as a gauge is formed to directly measure the change in stress time series of the evaluation site. The calculated fatigue damage degree is transmitted to the damage degree change temporary storage unit 9 in the same manner as the aforementioned state quantity, and is temporarily stored here. The storage period here is also the same as the storage period of the aforementioned state quantity change temporary storage unit 8, and it is the same as the lead time set for the maintenance or exchange operation, which is the most effective.

The state evaluation unit 6 and the fatigue damage evaluation unit 7 shown above are shown as separate components in FIG. 1, but the present invention is not particularly limited to these configuration forms. For example, even if each piece of software is composed of software in a single computer system, there is no problem in function realization. In addition, the state quantity change temporary storage unit and the damage degree change temporary storage unit are also shown as independent constituent elements, but they may be configured on the same memory device. In addition, the above constituent elements are prepared for all the mechanical components formed as the target (that is, all of the first to nth bearings) as shown in FIG. 1, but the condition evaluation unit 6 or the fatigue damage degree evaluation unit 7, The state quantity change temporary storage unit 8 or the damage degree change temporary storage unit 9 may be configured to be formed on a common hardware.

FIG. 2 is a model illustrating a system for evaluating the probability of failure of another bearing (in this case, individual number 1) in a non-failure state after the point of failure of the bearing of the individual number n in the operating system of the configuration of FIG. 1 Diagram of the action. First, regarding the individual that has a failure, the data transfer from each sensor is stopped, but when the failure occurs, the state quantity change and damage that were originally stored in the state quantity change temporary storage unit 9 and the damage degree change temporary storage unit 8 The time series data of the degree change is transmitted to the fault history evaluation unit 12. As will be described later in detail, when an arbitrary state quantity is observed in the failure history evaluation unit 12, a statistical model 20 (relationship) that defines the probability of failure (probability of failure) from that time to the time when any damage degree is more cumulative . This statistical model is transmitted to the failure probability evaluation unit 11 to estimate the failure probability after the accumulation of an arbitrary damage degree at the current point of the state of the unfaulted individual with the state quantity. At this time, regarding the same individual, it is also possible to adopt the damage degree change prediction unit 10 to predict the subsequent damage degree change because the change in the damage degree so far is stored in the temporary damage degree change storage unit 9. (That is, to define the relationship between the change in the damage degree and the passage of time) In the failure probability evaluation unit 11, a method of evaluating the relationship between the change in the failure probability and the passage of time.

The failure probability estimated by the failure probability evaluation unit 11 is formed as a function defined in relation to the cumulative damage degree or time afterwards, and is not set to one value. Therefore, in the display unit 13, as shown in FIG. 3, the cumulative state of the damage degree after the current point or the change in the probability of failure with the passage of time is displayed in the form of a graph, so that the user can easily perform Decision-making for the use/maintenance of For example, in a mechanical system that can switch the operation mode, if the load on the target varies according to the operation mode, as shown in FIG. 3, if the failure probability during operation after each operation mode is performed is displayed, it can be While determining the maintenance period, it also efficiently supports the decision of the operating policy.

If the result of the failure probability evaluation, a high failure probability is calculated in the relatively near future, not only to display the result, but also to have the function of transmitting a stop command or a retract operation command to the target control device. If the object that is not the closest to the fault and the state quantity is to be detected, if the user waits for the judgment of the user, there is a possibility that the fault will occur at an early stage. At this time, the function of automatically stopping or shrinking the system according to the evaluation result can easily prevent the occurrence of a failure.

The schematic diagram of FIG. 4 is used to describe the construction method of the statistical model that defines the probability of failure after the accumulation of any damage degree. First, in step S1, regarding the bearing of the faulted individual number n, the state quantity time series data 19 (S=f(t)) and the damage degree time series data 18 (D =f(t)).

Next, in step S2, the two time series data are formed into a relationship X (△D=f(S _t )) between the damage degree increment (△D) and the state quantity up to the fault to establish a relationship . That is, it is equivalent to expressing the observed state quantity as a variable and the increase in the degree of damage up to failure as a function. Specifically, in the state quantity time series data 19 (S=f(t)) and the damage degree time series data 18 (D=f(t)), the state quantity recorded at the same time t1 and the t1 The difference in damage degree up to the occurrence of a failure is used as a data set, and data indicating the relationship between the state quantity and the increment of the damage degree in the process from the number n of the faulted individual to the failure is generated. Next, statistical modeling is performed on the data assuming uneven distribution according to a certain probability, and the statistical model 20 is obtained. In terms of statistical modeling, at its simplest, the least squares method that assumes a normal distribution can also be used, but the damage degree increment system that should be applied to the probability distribution is defined as a non-negative value, so the normal distribution is assumed to be rigorous and Words are not appropriate. In the review by the inventors, who are targeting bearings, it is known that the generalized linear model (GLM) that uses the gamma distribution defined as a non-negative distribution as the probability density function (PDF) exhibits a relatively good data distribution. In this case, it is appropriate to use the inverse function in the link function of GLM.

The relationship between the damage degree increment ΔD up to the failure and the observed state quantity S in the individual number n where the failure occurred is defined as the statistical model 20. Therefore, according to this, the probability of failure is followed by step S3 The relationship between F and the damage degree increment ΔD up to failure is defined. In any individual whose failure has not yet occurred, a certain state quantity S1 is observed, after which it is assumed that any damage degree ΔD _a increases. Here, the PDF of the increase in the degree of damage until failure in the state quantity S1 is represented as P=f(S1). At this time, the damage degree increment ΔD _{a is} followed, and the failure probability F at the time when the damage degree increases is expressed as Equation 1.

This is equivalent to calculating the cumulative probability using the cumulative distribution function (CDF) for PDF obtained by statistical modeling. That is, by the individual is not yet failed to be observed in any state quantity S, it should be determined with reference to the PDF, and then, after the increase of the degree of damage imaginary increment △ D _a, corresponding to substituting the CDF, whereby It is possible to calculate the failure probability F in individuals whose failure has not yet occurred. Therefore, when a certain state quantity S is observed, the change in failure probability F accompanying the increase in the damage degree increment 21 is entirely the CDF itself corresponding to the PDF obtained by the statistical model.

Through the above sequence, using the statistical model 20 of the relationship between the increase in damage degree ΔD up to the failure and the observed state quantity S in the individual number n where the failure occurred, based on a similar environment, especially the same type of machinery, Moreover, the state quantity S observed by other individuals whose failure has not yet occurred can be calculated for each other individual.

In addition, based on the relationship of the average time damage degree ΔD _a when operating under different load conditions, the failure probability F is expressed according to time, thereby predicting the variation of the failure probability F when operating under different load conditions in the future. The display of the failure probability prediction 16 in FIG. 3 is based on the current state quantity S of the individual number 1 selected in the menu 151, which is caused by the operating time when operating in the high output mode, the normal mode, and the retracted mode, respectively. Change in the probability of failure F. The overall status summary 17 in FIG. 3 shows the failure probability F after 10 days in the operation mode selected for each of the individual numbers 1 to 10. In this way, the user can easily make subsequent use/maintenance decision-making.

[Example 2]

In Embodiment 1, as shown in FIG. 3, a method of displaying the change in the probability of failure from now to the future is adopted. This method can be said to be an effective method from the viewpoint of the support made by the operation/maintenance decision. However, the state quantity calculated by the state evaluation unit 6 in FIG. 1 is calculated based on, for example, the number of vibrations measured by the bearing, so it is different from a large tendency change, and there are short-term changes observed by the operating conditions. Possibility. When the time variation of the state quantity that is the premise of calculating the failure probability F of each body is relatively large, there is a possibility that the entire chart of the failure probability future prediction display 16 in FIG. 3 is updated too frequently. At this time, as shown in the failure probability prediction history display unit 22 in FIG. 5, it may be formed such that the future time point of the failure probability is estimated in advance with the increment value of time or damage degree, and the transition of the failure probability evaluation result up to this point is displayed. form. In FIG. 5, for example, the prediction period displayed on the failure probability prediction history display unit 22 can be selected in the menu 152. By using the display format shown above, you can easily confirm the faults up to now The history of the probability evaluation results, so the user can easily estimate the future tendency.

[Example 3]

In Example 1 and Example 2, using the failure of mechanical components of the same type as an opportunity, by statistically processing the data up to the failure, the failure probability of the unfaulted object is evaluated. In this method, since the analysis is based on the state quantity, high-precision evaluation can be performed, but on the other hand, the failure probability cannot be defined until the actual failure data is obtained. Therefore, until any failure occurs in the target mechanical component group, as described in the prior art, the cumulative fatigue damage degree obtained by the fatigue damage degree evaluation unit 7 in FIG. 1 may also be used to define the failure probability . During this period, it is impossible to expect high prediction accuracy as predicted by other individual data of the same group of actual failures, but even if the failure data cannot be obtained, the system configuration can be evaluated without significantly changing the probability of failure .

In addition, the failure history evaluation unit 12 may be provided separately from the failure evaluation system. At this time, based on the individual data of the fault at the time of failure, a statistical model for evaluation is created in the fault history evaluation unit 12 and stored in the fault evaluation system. Based on the statistical model, the failure probability evaluation unit 11 performs a mechanical system including a plurality of mechanical components. Failure assessment.

1‧‧‧Bearing 2‧‧‧Acceleration sensor 3‧‧‧Load gauge 4‧‧‧Rotometer 5‧‧‧‧A/D conversion section 6‧‧‧ State evaluation section 7‧‧‧Fatigue damage evaluation section 8‧‧‧Temporary state change temporary storage unit 9‧‧‧Damage degree change temporary storage unit 10‧‧‧Damage degree change prediction unit 11‧‧‧Fault probability evaluation unit 12‧‧‧Fault history evaluation unit 13‧‧‧Display Part 14‧‧‧Control part 15‧‧‧‧Display content in the display part 16‧‧‧Fault probability future forecast display 17‧‧‧Fault probability overall status summary 18‧‧‧Damage degree time series data 19‧‧‧ State quantity Time series data 20 ‧ ‧ ‧ statistical model 21 ‧ ‧ ‧ failure probability prediction 22 ‧ ‧ ‧ failure probability prediction history display section 100 ‧ ‧ ‧ failure probability evaluation system 151 ‧ ‧ ‧ menu

FIG. 1 is a schematic diagram illustrating the operation when the present invention is applied to a bearing group. FIG. 2 is a schematic diagram illustrating the operation of the present invention applied to a bearing group and after a certain bearing fails. FIG. 3 is a display example by the display unit in one of the embodiments of the present invention. FIG. 4 is a schematic diagram illustrating a method of calculating the failure probability in the present invention. FIG. 5 is a display example by the display unit in one of the embodiments of the present invention.

1‧‧‧bearing

2‧‧‧Acceleration sensor

3‧‧‧Loadmeter

4‧‧‧Rotometer

5‧‧‧A/D Conversion Department

6‧‧‧ State Assessment Department

7‧‧‧Fatigue Damage Evaluation Department

8‧‧‧Temporary state change temporary storage department

9‧‧‧Temporary Department of Damage Change

10‧‧‧Damage prediction department

11‧‧‧ Failure Probability Evaluation Department

12‧‧‧Fault History Evaluation Department

13‧‧‧Display

14‧‧‧Control Department

100‧‧‧ Failure Probability Evaluation System

Claims (11)

A failure probability evaluation system, which is a system for evaluating the failure probability of a plurality of mechanical elements included in a mechanical system, and is characterized by having: a first physical quantity that changes according to the fatigue damage of the aforementioned mechanical element or changes over time, and the evaluation expression Means of the state of soundness of the mechanical element; means of evaluating the cumulative fatigue damage of the mechanical element based on the second physical quantity that changes according to the load or load on the mechanical element or the operating data of the mechanical element; The storage unit of the state quantity and the fatigue damage degree; based on the state quantity and the fatigue damage degree of the machine element that has failed among the plurality of machine elements, calculates the failure of the machine element that has not failed among the plurality of machine elements A probability failure evaluation section; and a failure history evaluation section that statistically establishes the relationship between the state quantity in the mechanical element that has failed among the plurality of mechanical elements and the increase in the fatigue damage degree until the failure occurs . As in the failure probability evaluation system of claim 1 of the patent application, the storage unit is a temporary storage unit characterized by temporarily storing the state quantity and cumulative fatigue damage degree for an arbitrary period. For example, in the system for evaluating the probability of failure according to item 1 of the patent application scope, the aforementioned second physical quantity is load or strain. For example, the failure probability assessment system of the first item of the patent scope, in which the aforementioned mechanical elements are bearings. For example, the system for evaluating the probability of failure according to item 1 of the patent application scope, wherein the means for evaluating the cumulative fatigue damage degree is to calculate the fatigue damage degree based on the linear cumulative damage law. A failure probability evaluation system as claimed in item 1 of the patent application, wherein the state quantity is a collection of complex physical quantities, and in the means for evaluating the state quantity, the state quantity is calculated based on the complex physical quantity. A failure probability evaluation system as claimed in item 1 of the patent scope, wherein the failure history evaluation unit uses the time series data of the state quantity and the fatigue damage degree of the mechanical element where the failure occurred to generate the fatigue until the failure occurs The increment of the damage degree is used as a function, and the statistical model using the state quantity as a variable is used to calculate the cumulative distribution function corresponding to the uneven probability density function representing the increment of the fatigue damage degree in the statistical model. For example, the system for evaluating the probability of failure in the 7th scope of the patent application, wherein the aforementioned statistical model is a generalized linear model. For example, the failure probability evaluation system of patent application item 7, wherein the probability density function uses gamma distribution. A failure probability evaluation system as claimed in any one of the first to ninth items of the patent application scope, which has a display unit that includes: among the plurality of mechanical elements calculated by the failure probability evaluation unit The relationship between the probability of any one or more failures and the future increase in fatigue damage or time is displayed as a function of the graph. A failure probability evaluation system as claimed in any one of the first to ninth items of the patent application scope, which has a display unit that is one or more of the plurality of mechanical elements calculated by the failure probability evaluation unit Among the probability of failure, the probability of failure evaluated for at least one state in the future determined by the increment or time of the assumed fatigue damage in the future will be the predicted value up to the present point and the aforementioned fatigue damage degree or The relationship of time is shown as a graph.

Images