TWI607328B - Operational auxiliary device and wind power generation system - Google Patents

Operational auxiliary device and wind power generation system Download PDF

Info

Publication number
TWI607328B
TWI607328B TW106104632A TW106104632A TWI607328B TW I607328 B TWI607328 B TW I607328B TW 106104632 A TW106104632 A TW 106104632A TW 106104632 A TW106104632 A TW 106104632A TW I607328 B TWI607328 B TW I607328B
Authority
TW
Taiwan
Prior art keywords
failure
t1
data
risk
maintenance
Prior art date
Application number
TW106104632A
Other languages
Chinese (zh)
Other versions
TW201734870A (en
Inventor
Norio Takeda
Hiroshi Shintani
Kazuo Muto
Tomoaki Yamashita
Original Assignee
Hitachi Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to JP2016061725A priority Critical patent/JP2019113883A/en
Application filed by Hitachi Ltd filed Critical Hitachi Ltd
Publication of TW201734870A publication Critical patent/TW201734870A/en
Application granted granted Critical
Publication of TWI607328B publication Critical patent/TWI607328B/en

Links

Classifications

    • FMECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03DWIND MOTORS
    • F03D17/00Monitoring or testing of wind motors, e.g. diagnostics
    • FMECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03DWIND MOTORS
    • F03D80/00Details, components or accessories not provided for in groups F03D1/00 - F03D17/00
    • F03D80/50Maintenance or repair
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06QDATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00Administration; Management
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02EREDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00Energy generation through renewable energy sources
    • Y02E10/70Wind energy
    • Y02E10/72Wind turbines with rotation axis in wind direction
    • Y02E10/722Components or gearbox
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02EREDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00Energy generation through renewable energy sources
    • Y02E10/70Wind energy
    • Y02E10/72Wind turbines with rotation axis in wind direction
    • Y02E10/723Control of turbines

Description

Operation aids and wind power systems

The invention relates in particular to an operation aid for an arbitrary product and a wind power generation system.

A large number of sensors are assembled in products such as factories, and the information measured by the sensors is used for reliability analysis based on the control of stable operation and the components of the products (diagnosis of failure signs, remaining life) Diagnosis, etc.) The maintenance plan is prepared. For example, in a power plant, the main three types of measurements are implemented by a variety of sensors. The three types of measurement are referred to as SCADA (Supervisory Control And Data Aquisition), Condition Monitoring (CMS, Condition Monitoring System), and Structure Monitoring (SHM, Structural Health Monitoring). In the case of wind power plants, in the control measurement (SCADA), the purpose is to grasp the environmental conditions or operating conditions of the windmill to properly control the windmill, and to measure the wind speed, wind direction, power generation, number of revolutions of the generator, temperature, etc. Various physical quantities. In the state monitoring (CMS), the purpose is to detect the symptoms of the windmill failure, and to minimize the damage caused by the failure, and to perform the measurement. In addition to structure monitoring (SHM) In order to evaluate the soundness of the blade of the windmill, etc., the deformation of the structure and the like are measured. In a general windmill, all or part of such control measurement, state monitoring, and structure monitoring is performed to control the windmill and evaluate the reliability to achieve stable operation of the windmill.

Patent Document 1 discloses an operation control program and a windmill, which is a type of operation control device for a windmill, which includes: a fatigue deterioration schedule, and an accumulated operation time of the wind turbine is associated with an optimum fatigue deterioration degree of the wind turbine; and fatigue degradation The calculation means calculates the current fatigue deterioration degree of the windmill; and the operation control means, the fatigue deterioration degree of the wind speed calculated by the fatigue deterioration calculation means, and the current optimum fatigue obtained from the fatigue deterioration schedule The operation of the windmill is controlled by the relationship of the degree of deterioration (request 1).

Patent Document 2 discloses "an operation control system for a wind power plant, which is an operation control system for a wind power plant having a plurality of windmills, and is characterized in that: a remaining life prediction unit is provided, and the remaining life of the components is predicted for each windmill; The electric revenue prediction unit predicts the sales revenue under the plurality of output restriction conditions for each windmill; and the maintenance cost prediction unit predicts the maintenance cost under each output restriction condition for each windmill based on the remaining life of the component; and the output restriction condition selection For each windmill, based on the sales revenue and maintenance cost predicted for each of the aforementioned output constraints, the choice of each windmill will maximize the output gains from the wind power plant; and the operation command, based on the selection The output restriction condition is sent to each windmill" (request item 5).

Patent Document 3 discloses an operation diagnostic device for a plant machine. "With: the maximum probability of damage to each damage phenomenon in the current operating environment of the machine and each component that constitutes the plant, multiplied by the weight coefficient of each damage form determined in advance for each component, Means for estimating the value of the plant risk; and means for calculating the operation limit value of the plant by calculating the operating conditions, which are used to multiply the probability of destruction of each failure phenomenon under the assumed operating conditions of the aforementioned components The plant risk value obtained by determining the maximum value of the damage form of the damage form does not exceed the set value; and the remaining life evaluation information from the current time point of the member is calculated, and the future life consumption is calculated according to the operation plan. Means; and based on the remaining life information at the current time point and the future remaining life prediction calculated based on the operational plan, the evolution of the damage probability of each application plan is evaluated, and each damage form of each of the aforementioned components is Multiply the weighting factor by the evolution data of the above-mentioned damage probability, and calculate the means by which the plant uses the risk estimation value. Yes).

[Previous Technical Literature] [Patent Document]

[Patent Document 1] Japanese Patent Laid-Open Publication No. 2006-241981

[Patent Document 2] Japanese Patent Laid-Open Publication No. 2013-170507

[Patent Document 3] Japanese Patent Laid-Open Publication No. 2002-73155

In the above Patent Documents 1 and 2, the fatigue damage rate and the remaining life are used as the evaluation criteria for reliability, but in these documents, it is not disclosed. In the case where the fatigue damage rate or the remaining life is obtained from a plurality of parts, the operation control or maintenance plan of the plant is implemented based on which part reliability (fatigue damage rate or remaining life). On the other hand, in Patent Document 3, for each of a plurality of members, a numerical value obtained by multiplying the probability of destruction by a weighting coefficient is calculated, and the maximum value is set as a factory risk estimation value to operate the plant, thereby revealing the Focus on the reliability of which parts to use the plant and plan maintenance. However, in the maintenance plan of the plant, not only the parts whose risk value becomes the maximum value, but also the risk value of other parts to be maintained must be considered to formulate the maintenance plan. However, none of the above-mentioned patent documents discloses or teaches a reliability evaluation value (risk value, etc.) of a plurality of parts to formulate an operation and maintenance plan to enable the plant to operate efficiently and stably. Further, Patent Document 3 assumes a thermal power plant such as a steam turbine, and it is conceivable that objects used in windmills, construction machines, and the like that are not exposed to harsh outdoor environments are not considered.

In addition, the reliability evaluation obtained from the fatigue damage rate, the remaining life, the probability of failure, the risk value, and the like has a problem that, in particular, when the target product is a newly developed product, reliability is lacking due to failure data, etc. The accuracy of the evaluation may not be good. Any of the above patent documents also discloses the evaluation of reliability from the material of the object product and uses it for maintenance or application, but in contrast, it does not disclose or teach examples such as the use of products of the same type as the object product, similar articles. The reliability evaluation value is used to update the reliability evaluation value of the object product, and the product is planned and maintained according to the high-accuracy evaluation value.

The present invention has the above problems in view of the above problems, and aims to achieve a consideration. The high-precision reliability evaluation value of the multiple parts of the product is used to implement the operation and maintenance plan of the product.

According to a first aspect of the present invention, a service assisting device includes: a failure risk assessment unit; and a maintenance/operation scenario creation unit; and the failure risk assessment unit uses environmental information input from a plurality of sensors of the target product and The operation data, the design data and the material data set in advance, the failure probability F(t1) at the time t1 of the part p regarded as the object, and the probability of destruction at the time t1 of the part p at the time t1 (t1) Calculate the failure risk RS(t1, p) of the part p contained in the product by multiplying the influence degree C(p) of each part p which is determined in advance in the case where the part p is damaged. The maintenance/operation scenario development unit stores the failure risk data (t1, p) of the component p at the time t1 sent from the failure risk assessment unit, and the failure risk data that has been sent from the failure risk assessment unit. The plurality of failure risks from the past to the time t1, and the physical quantity x and the time t that affect the failure risk selected from the environmental data and the operation data input from the target product are set as variables, and the failure is generated. The trend curve of risk, based on the trend curve of the fault risk, is obtained from the current time point to the previous The predicted value of the failure risk set in advance is determined according to the predicted value of the failure risk, manually set according to the input from the input unit by the maintenance/operator, or by the predetermined processing. Automatically setting the maintenance/application script of the product, and generating a fault risk prediction model with the physical quantity y that is affected by the failure risk by the time t, the physical quantity x, and the maintenance data and/or the operation data as a variable. Predict the future failure risk of the part p, and sort out the future failure risk of each part according to the higher predicted value of the failure risk and display it on the display part and/or in the memory part, and set it in advance. The plurality of thresholds are divided into groups, and the maintenance period including each part is automatically set for each group manually by the input from the input unit by the maintenance/operator or manually by the predetermined processing. The maintenance content is used internally, and the maintenance application content is memorized in the memory unit and/or displayed on the display unit.

According to a second solution of the present invention, there is provided a wind power generation system comprising: the operation aid device as described above; and the wind power generator, which is a target product having a plurality of sensors. According to a third aspect of the present invention, there is provided a operation assisting device comprising: a failure risk assessment/update unit; and a maintenance/operation script development unit; a fault data base for accumulating a plurality of constituent products and their identical machines and/or the like Fault data of parts; The fault risk assessment/update unit described above calculates the fault data based on the probability density function of the life of the target component p and the fault data calculated from the fault data contained in the fault database. The probability density function of the updated life is obtained by using the probability density function of the updated life, using environmental data and operation data input from a plurality of sensors of the target product, and design data and material materials set in advance. The calculation is regarded as the updated failure probability F(t1)' at time t1 of the part p of the object, and at time t1, the updated failure probability F(t1)' at time t1 of the part p and the part p In the case of damage, the product of the influence degree C(p) of each part p which is determined in advance is calculated to calculate the updated failure risk RS(t1, p)' of the part p contained in the product, the aforementioned maintenance/ The script creation unit is based on the updated failure risk RS(t1, p)' of the part p at the time t1 sent from the failure risk assessment/update unit, and has been sent from the failure risk assessment/update unit. Memory in faulty wind The plurality of updated failure risks from the past to the time t1 of the database, and the physical quantity x and the time t that are selected to affect the failure risk by the environmental data and the operation data input in advance by the target product are set as variables. The trend curve of the fault risk is generated, and the predicted value of the fault risk from the current time point to the previous predetermined time is obtained according to the trend curve of the fault risk, and according to the predicted value of the fault risk, according to the maintenance/operator made The maintenance/application script of the product that is manually set by input from the input unit or automatically set by the predetermined processing, and generated at time t, physical quantity x, and maintenance data and/or operation The data selects the physical quantity y that affects the risk of failure as a variable risk prediction model to predict the future failure risk of the part p, and memorizes the predicted value and the trend curve together in the memory and/or on the display.

According to a fourth aspect of the present invention, there is provided a wind power generation system comprising: the operation assisting device as described above; and the first wind power generator, which is a target product having a plurality of sensors; and the second wind power generator having A plurality of sensors are the same type of machine as the first wind turbine or the like.

According to the present invention, it is possible to realize a high-accuracy reliability evaluation value of a plurality of components constituting a product, and to implement an operation and maintenance plan of the product.

1‧‧‧ object products

2‧‧‧Fault Risk Assessment Department

3‧‧‧Maintenance/Using Script Development Department

4‧‧‧Fault Risk Forecasting Department

5‧‧‧Design/Materials Database

6‧‧‧Destruction probability calculation

7‧‧‧ Risk calculation

8‧‧‧Impact database

10‧‧‧ Failure risk trend analysis

11‧‧‧Fault risk prediction

12‧‧‧Maintenance/application scripting

13‧‧‧The same type of machine, similar machine

14‧‧‧Fault Risk Assessment/Update Department

15‧‧‧ Fault Database

16‧‧‧The probability density function of the lifetime under the equivalent stress amplitude

17‧‧‧External database

18‧‧‧Destruction of the probability assessment department

19‧‧‧Damage Assessment Department

20‧‧‧ probability density function of fatigue life

21‧‧‧stress frequency distribution

22‧‧‧Fatigue life curve with probability of failure P%

30‧‧‧Fault Risk Database

31‧‧‧Destroy probability database

32‧‧‧Damage database

100‧‧‧Operational support system

Fig. 1 is a schematic block diagram showing the main components of the operation support system according to the first embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

[Fig. 2] A schematic block diagram showing the main functions of the failure risk assessment unit in the operation support system according to the first embodiment of the present invention.

[Fig. 3] The operation assisting system according to the first embodiment of the present invention In the case of the stress history of the parts contained in the object product, the P-S-N line diagram necessary for the probability of destruction is calculated by the remaining life evaluation.

[Fig. 4] A schematic diagram showing a method of calculating the degree of damage by the stress frequency distribution generated in the component included in the target product and the P-S-N line diagram in the operation assisting system according to the first embodiment of the present invention.

[Fig. 5] A schematic diagram of a method for calculating the probability of destruction using the probability density function of the damage life in the operation support system according to the first embodiment of the present invention.

[Fig. 6] A schematic diagram of a method for determining the probability density function of the life by selecting and destroying the physical quantity associated with the measurement data of the product according to the measurement data of the first embodiment of the present invention.

[Fig. 7] A schematic block diagram showing the main functions of the maintenance/operation scenario creation unit in the operation support system according to the first embodiment of the present invention.

[Fig. 8] Fig. 8 is a schematic diagram showing an example of a trend curve of a failure risk generated by a failure risk trend analysis of a failure risk prediction unit in the operation support system according to the first embodiment of the present invention.

[Fig. 9] Fig. 9 is a diagram showing an example of risk prediction processing in the case where the maintenance/operation script executed by the maintenance/application scenario creation unit is changed in the operation support system according to the first embodiment of the present invention.

[Fig. 10] In the operation support system according to the first embodiment of the present invention, the main components of the product are explained as parts, and the example in which the predicted value of the failure risk is displayed in a high order is shown and the group is divided.

[Fig. 11] In the operation support system according to the first embodiment of the present invention, the predicted value of the failure risk is calculated for each of the main components of the subsidiary product, and the parts are displayed in accordance with the order in which the predicted values are high. Figure.

[Fig. 12] A schematic diagram of a procedure from the destruction probability calculation to the maintenance/application script development in the operation support system according to the first embodiment of the present invention.

Fig. 13 is a schematic block diagram showing the main components of the operation support system according to the second embodiment of the present invention, and the products, the same type machines, the similar machines and the data banks which provide information to the operation support system, and their relationships.

[Fig. 14] In the operation assisting system according to the second embodiment of the present invention, when the probability of failure is calculated by the residual life evaluation of the stress history of the components included in the product, the probability density of the life is plotted with the equivalent stress amplitude. A schematic diagram of the function and updating it based on Bayesian statistics.

[Fig. 15] In the operation assisting system according to the second embodiment of the present invention, when the probability of failure is calculated by the residual life evaluation from the stress history of the components included in the product, the calculation from the equivalent stress amplitude to the update is broken. A schematic diagram of the procedure for the probability density function of the broken life.

[FIG. 16] In the operation assisting system according to the second embodiment of the present invention, when the probability of failure is calculated based on the probability density function of the breaking life of the component included in the product, the life of the pre-set is updated according to the Bayesian statistics. A schematic diagram of an example of a probability density function.

[Fig. 17] In the operation assisting system according to the second embodiment of the present invention, the failure rate is expressed by a probability density function including a variable amount of time. In the case of a life-destroying probability, the example of the probability density function set in advance is updated according to the Bayesian statistics.

Fig. 18 is a schematic block diagram showing the main components of the operation support system according to the third embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

Fig. 19 is a schematic block diagram showing main components of a maintenance/operation scenario creation unit of the operation support system according to the third embodiment of the present invention, and flow of information between elements.

Fig. 20 is a schematic block diagram showing the main components of the operation support system according to the fourth embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

Fig. 21 is a schematic block diagram showing main components of the maintenance/operation scenario creation unit of the operation support system according to the fourth embodiment of the present invention, and the flow of information between the elements.

[Fig. 22] In the operation support system according to the fourth embodiment of the present invention, the main components of the maintenance/operation scenario creation unit in the case of the tendency curve of the destruction probability, and the flow of the data between the elements are schematically illustrated. Block diagram.

Fig. 23 is a schematic block diagram showing the main components of the operation support system according to the fifth embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

[Fig. 24] In the operation support system according to the fifth embodiment of the present invention, the main components of the maintenance/operation scenario creation unit in the case of the trend curve of the damage degree, and the flow of information between the elements. Slightly schematic block diagram.

A. Summary

This embodiment includes a plurality of means for solving the above problems, but as an example, it can be an operation assisting system for an arbitrary product.

It is characterized in that, in addition to the risk of failure of a plurality of parts constituting the aforementioned product, that is, information including at least one of environmental materials, operation materials, design materials, and material materials of the aforementioned products from the past to the present, The information on the fault data of the same type machine and the similar machine is used to perform the evaluation and update of the fault risk, and the fault risk estimation value of the plurality of the aforementioned parts that may change when the maintenance/application plan of the product is changed now is determined. The calculation device includes means for assigning the use of the product or a plurality of maintenance periods of the plurality of parts to each of the estimated risk risk estimates of the plurality of parts.

According to the present embodiment, it is possible to provide an operation support system for a product capable of setting up an operation and maintenance plan for a highly reliable product and performing stable operation.

B. Operation aids and wind power systems

Hereinafter, embodiments of the present invention will be described using the drawings.

[First Embodiment]

Hereinafter, the first embodiment of the present invention will be described with reference to Figs. Operational assistance system. In the present embodiment, the product 1 is a wind power plant, but the operation of the present invention and/or the present embodiment is not limited to a wind power plant.

According to the first embodiment, the plurality of components constituting the product are grouped, whereby a highly accurate reliability evaluation value of a plurality of components constituting the product can be achieved, and the operation and maintenance plan of the product can be implemented.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic block diagram showing the main components of the operation support system of the product of the present embodiment, and the products and databases for providing the data used in the operation support system, and their relationships. The operation support system 100 shown in FIG. 1 includes a failure risk assessment unit 2 and a maintenance/operation scenario creation unit 3. The maintenance/application scenario creation unit 3 includes a failure risk prediction unit 4 that targets a plurality of parts included in the product and predicts the risk of failure of the part that changes when the maintenance/operation plan is changed. Although omitted in the figure, the operation support system 100 includes an input unit, a display unit, and an output unit for other devices.

A plurality of sensors for measuring the use environment or operating state are assembled in the product 1. The environmental data and the operational data measured by the sensors are sent to the failure risk assessment unit 2 of the operation support system 100, and are used for evaluation. The environmental information referred to here is information related to the environment exposed by the product. In the case of a wind power plant, wind conditions such as wind speed and wind direction of the windmill are included in the environmental data. In the case of a wind power plant installed on the ocean, in addition to the wind condition data, sea state data such as wavelength or wave height is also the category of environmental data. In addition, the so-called operating data is the speed, acceleration, rotation speed, rotation angle, etc. Even the information. In the case of a wind power plant, the amount of power generated by the windmill, the rotational speed of the generator, the azimuth angle, and the nacelle angle are the categories of operational data. In the plant, environmental data and operational data are mostly measured as control measurement (SCADA). However, for the plant containing state monitoring (CMS) or structure monitoring (SHM), if the data measured by them is also related to the use environment or operational status of the product, the environmental data included in the embodiment is included. Or operational data.

In addition to environmental data or operational data, design data or material data are also used in the failure risk assessment department 2. Here, the design information, such as the drawing containing the product, and the shape of the product. Further, the material information includes characteristics of materials constituting the product, or characteristics of a structure such as a bolt fastening or a welding pipe.

FIG. 2 is a schematic block diagram showing the processing executed in the failure risk assessment unit 2.

The failure risk assessment unit 2 calculates the failure risk using at least one of environmental data, operational data, design data, and material data. At time t1, the failure risk RS(t1, p) of a certain part p included in the product can be caused by the destruction probability F(t1) at the time t1 of the target part p, and when the target part p is damaged. The degree of influence C(p) is calculated according to the following formula. In addition, the information of the degree of influence C(p) of each part p is accumulated in advance in the influence degree database 8.

RS(t1,p)=C(p)×F(t1)‧‧‧(1)

The fault risk RS(t1,p) of equation (1) is set as an indicator of reliability, so that it can be considered when a certain part fails. Respond to the size, to plan the maintenance / use of parts.

The probability of destruction F(t1) can be calculated by residual life assessment or symptom diagnosis.

First, the remaining life assessment is explained.

The fatigue phenomenon that occurs when the load is applied to the product is considered as the object. If the damage probability F(t1) is to be obtained by the remaining life assessment, first, the environmental data, the operational data, and the design data are used to calculate the time from the past to the present. The stress history of the part at point t1. Next, in this stress history, a frequency analysis method such as a rainflow method is used to create a stress frequency distribution obtained by sorting out a certain magnitude of stress. Then, using the stress frequency distribution and the material data associated with the target part, the probability of failure F(t1) is obtained.

Figure 3 is a P-S-N diagram necessary to calculate the probability of failure by the residual life assessment from the stress history of the parts contained in the target article.

The material data used herein is ideally the fatigue life curve of Fig. 3, which is called a P-S-N line diagram, and is accumulated in the design/material database 5 in this embodiment. The PSN line diagram is a probability density function 20 of the fatigue life obtained by performing a fatigue test on each of the stress amplitudes of the vertical axis, and the number of times of the failure probability P% is obtained, and these are connected as shown in the figure (Fig. 3). ).

Fig. 4 is a schematic view showing the method of calculating the degree of damage by the stress frequency distribution and the P-S-N line diagram of the parts included in the object product.

As shown in Figure 3, when it is possible to assume the fatigue life of each stress amplitude When the rate density function 20 is all the same, the probability of destruction F(t1) can be obtained by the following procedure. First, as shown in FIG. 4, the stress frequency distribution 21 obtained by applying the frequency analysis method to the stress history occurring from the past to the current time point t1, and the fatigue life curve 22 having the predetermined breaking probability P% can be set. And the damage degree D(t1) with respect to the fatigue damage P% is calculated according to the following formula.

D(t1)=(n 1 /N 1 )+(n 2 /N 2 )+...+(n m /N m ) ‧‧‧(2)

Here, n 1 , n 2 , and n m are the repetition times (m is an integer) of the stress amplitudes S 1 , S 2 , and S m obtained by frequency analysis of the stress history, respectively. Further, N 1 , N 2 , and N m are the number of times of breaking and the occurrence of fatigue fracture at the failure probability P% when the load stress amplitudes S 1 , S 2 , and S m are repeated. When it is assumed that the probability density function 20 of the fatigue life is not the same as the stress amplitude, the number of times of the occurrence of the damage degree D(t1) N(t1) can be obtained by the following equation.

N(t1)=D(t1)×Np‧‧‧(3)

Np is the number of times of the failure probability P% set in advance, and the stress amplitude is set to an appropriate Si (for example, an average value, an intermediate value, a value close to the value, etc.) by an input unit (not shown). As shown in FIG. 4, the probability of failure F(t1) can be obtained from the probability density function 20 of N(t1) and fatigue life by the equation of FIG. In other words, if the elapsed time from the start of the product operation to the present is set to t1, the probability of destruction F(t1) is a value obtained by integrating the density function f(N) from 0 to N(t1).

Next, another example of the remaining life evaluation will be described.

Figure 5 is a schematic diagram of another method for calculating the probability of failure using the probability density function of the damage life.

In addition, the probability of failure F(t1) up to the damage can be obtained by directly defining the probability density function f(t1) of the life as shown in FIG. In other words, if the elapsed time from the start of the product operation to the present is set to t1, the probability of destruction F(t1) is a value obtained by integrating the density function f(t) from 0 to t1. Here, if the time to reach the failure rate of 50% is defined as the life T, the difference between the current time point t1 and the life time T can be regarded as the remaining life. Therefore, the method of directly defining the probability density function of the life can also be used. Interpreted as residual life assessment. Further, the information of the probability density function f(t1) of the life can be stored in an appropriate memory such as the internal memory of the failure risk assessment unit 2 in advance for each target product. In addition, the data of the probability density function f(t1) of the life can be determined in advance using environmental data, operation data, and the like.

Next, the symptom diagnosis will be explained.

Fig. 6 is a schematic view showing a method for selecting and destroying the physical quantity associated with the measured data of the product and appropriately converting the probability density function of the life.

In the symptom diagnosis of one type, the operation data of the normal and abnormal time of the target part is sorted in advance, and the operation data at the current time point is monitored to determine the failure of the target part. If the operational data group at the time of the abnormality is understood as the probability density function of the failure as shown in FIG. 6, by describing the position of the operation data at the current time point in FIG. 6, the destruction machine at the current time point t1 can be obtained. Rate F(t1). There is a physical quantity related to the destruction of the part selected and focused by the operation data at the time of the abnormality, and the physical quantity is selected as the probability density function of the probability variable, or as shown in Fig. 6, the operation data is included. The physical quantity is converted into a form related to the destruction of the component under consideration, and the converted physical quantity x1', x2' is used as the probability variable to create a probability density function. Here, the physical quantity to be converted includes, for example, a spectral value of a specific frequency among acceleration spectra obtained by performing fast Fourier transform on acceleration data included in the operation data. Further, the probability density function can be stored, for example, in an appropriate memory such as the internal memory of the failure risk assessment unit 2 or the design/material database 5 for each target product. In addition, the probability density function can be determined in advance using environmental data, operation data, and the like.

Next, C(p) will be described.

The influence degree database 8 stores information on the degree of influence C(p) of each part p included in the equation (1). As the degree of influence C(p), if the cost of all or a predetermined range required in the case where a certain part p is actually damaged is used, the risk of failure RS(t1, p) can be considered as It is an expected value of the loss cost of all or a predetermined range which is generated by the failure at the current time point t1. The cost of a certain part is damaged, including the cost of the new part body, the replacement cost of the part, the handling cost of the part, and the loss of the generator due to the operation stoppage of the product. On the other hand, instead of a specific cost, an integer may be assigned due to the influence of the influence of the damage of the part, and the integer is adopted as the influence degree C(p). Such a situation In the next case, it is possible to compare the failure risk RS(t1, p) between the parts and determine which part of the reliability should be noted.

The fault risk RS (t1, p) calculated in this way is input to the maintenance/application script development unit 3 together with the maintenance/application data, the environmental data, the operation data, the design data, and the material data.

Fig. 7 is a schematic block diagram showing the processing executed in the maintenance/operation scenario creation unit. The maintenance/application scenario creation unit 3 includes a failure risk prediction unit 4, and the failure risk prediction unit 4 is the failure risk RS(t1, p) at the time t1 sent from the failure risk assessment unit 2, and the failure fault. The plurality of failure risks sent by the risk assessment unit 2 and memorized in advance in the failure risk database 30 from the past to the time t1 (n=0, -1, -2, -3, ‧‧) RS (tn, p), and environmental data, operational data, to analyze the trend of failure risk.

FIG. 8 is a diagram showing an example of a trend curve of a failure risk generated by a failure risk trend analysis of a failure risk prediction unit.

For example, in the case of a wind power plant in Japan, there is a tendency for winds to be strong in winter. Therefore, depending on the components, as shown in Fig. 8, a trend curve in which the horizontal axis is set as time and the vertical axis is set as a risk of failure can be drawn. However, even if there is a seasonal change, the distribution shape of the P-S-N line or the probability density function used in the calculation of the probability of destruction is different, and there are many cases where a regular trend curve as shown in Fig. 8 cannot be obtained. In addition, even if it is a component that constitutes the same wind power plant, the fluctuation risk RS (t1, p) due to the short-term wind disturbance is still the same as the seasonal change or larger than the seasonal change. Each part must be selected from environmental data or operational data. The physical quantity with the trend change, in addition to the time, the selected physical quantity is set as a variable to describe the trend curve of the failure risk. In other words, when the time limit is set to t and the selected physical quantity is set to x, the failure risk prediction unit 4 selects the following items (4), (5), etc. to determine the risk of failure in response to the product or the component constituting the product. Trend curve RS.

RS(t,p)=g1(t,p)‧‧‧(4)

RS(t,x,p)=g2(t,x,p)‧‧‧(5)

Equation (4)(5) is a trend curve of the risk of failure with time t or physical quantity x as a variable, but it can also adopt an autoregressive moving average containing physical quantities or error terms from the past and the present time to a certain point in time. A model, or a neural network that is learned by inputting time, environmental data, and operational data as a trend curve.

In addition, as a simple formula, for example, RS(t, x, p) = α(p) ‧ t + β (p) ‧ x + γ (p) (α, β, γ is a constant, but related to the part ), etc., but not limited to this.

The failure risk prediction unit 4 implements the failure risk prediction 11 based on the trend curve of the failure risk, the maintenance/operation data, and the maintenance/application script specified in the maintenance/application script. The so-called maintenance/use information is information about the past, such as information on periodic inspections of products, information on changes in control used to change operations, and inspections performed on failures. The so-called maintenance/application scripts, for example, refer to plans such as which parts are replaced in what month and how they are used. The maintenance/application script development 12 can be automatically created as a maintenance/application script by a predetermined method, or a maintenance/application script can be manually input from the input unit.

FIG. 9 is a diagram showing an example of risk prediction processing in the case where the maintenance/operation script executed by the maintenance/application scenario development unit is changed.

In the case where the current periodic inspection and application method (product control method, etc.) will be continued in the future, the future failure risk (predicted risk) a from the current time point to the prior time is set in advance. It will become the trend curve along the past as shown by the predicted risk a in Figure 9.

Next, as shown in FIG. 7, the maintenance/operation scenario creation unit 3 performs maintenance/operation script development 12 using the predicted risk. That is to say, based on the value of the predicted risk, the maintenance or the change of the method of the product is automatically reviewed by the maintenance of the input from the input unit by the operator or the like manually or by a predetermined process. For example, when the predicted risk is higher than the pre-defined threshold, the maintenance/application scripting 12 will propose a change to a more stable method of use, when the predicted risk is lower than the pre-defined threshold. More intense use of the product or frequent maintenance inspections will be proposed. The proposed maintenance/application script is again sent to the failure risk prediction 11 as the future risk b or c is predicted as in Figure 9. In order to predict the future risk b or c in the case of changing maintenance or operation, in addition to the physical quantity x (selected by environmental data and operational data) that affects the risk of failure, the risk of failure is also selected from the maintenance/operation data. The physical quantity y of the influence is caused, and the failure risk prediction model g3 or g4 having the time t, the physical quantity x, and y as the variable expression can be utilized.

RS(t+△T,y,p)=g3(t+△T,y,p)‧‧‧(6)

RS(t+△T,x,y,p)=g4(t+△T,x,y,p)‧‧‧(7)

Equations (6) and (7) are trend curves with time t + ΔT or future physical quantities x and y as variables, but physical quantities or errors including from 遡 and a certain time point to time t + ΔT can also be used. The autoregressive moving average model of the item or the neural network learned by inputting time, environmental data, operational data, and maintenance/utilization data as a predictive model.

Next, FIG. 10 is a schematic diagram showing an example in which the main components of the product are interpreted as parts, and the predicted values of the failure risk are ranked in a high order, and the group is divided. Consider the grouping to perform the maintenance/application script formulation 12 of FIG. For example, plan which group to maintain and other plans after a few years.

The failure risk prediction of Fig. 9 can be implemented by considering all the components constituting the product as objects. However, only parts with high risk of failure (or predicted risk or future risk) are considered as objects to predict the risk of failure, thereby reducing the number of projects necessary for prediction and efficiently planning maintenance or operation. The maintenance/application scenario development unit 3 predicts the risk of failure, for example, as a component having the largest predicted value, together with the maximum value thereof, as predicted from the failure risk (or predicted risk or future risk). Those with higher values are sorted in order and displayed on the display. For specific processing of grouping, for example, the group and maintenance period can be assigned by the threshold of the upper and lower limits of each group, so that the risk of failure of all parts does not exceed a certain threshold. Such assignments can be achieved, for example, by a genetic algorithm. The maintenance/application script creation 12 of the maintenance/application scenario creation unit 3 is performed by the maintenance operator, etc. with reference to FIG. From the input of the input unit, the maintenance operation content is automatically scheduled for each group, such as which part is scheduled, manually or by a predetermined process. Alternatively, a maintenance/application script such as maintenance operation contents may be manually input from the input unit. In the maintenance/operation scenario creation 12 of the maintenance/application scenario creation unit 3, the planned maintenance operation content is stored in an appropriate storage unit and/or displayed on the display unit. If the predicted value of the risk of failure (predicted risk or future risk, etc.) is sorted according to the size, for example, as shown in FIG. 10, the group of parts can be grouped into group A by setting a plurality of thresholds in advance. Group B, group C, etc. plan maintenance. With such a maintenance plan due to grouping, it is possible to match the planned maintenance schedule with the pre-planned maintenance schedule, and to replace the parts with higher predicted risk (predicted risk or future risk), for example, in the scheduled Replacement of parts belonging to group A is performed from the current maintenance period. The grouping of such maintenance periods can be implemented by a combination optimization method such as a gene algorithm or a branch and bound method.

In addition, FIG. 11 is a diagram showing the predicted values of the risk of failure of each component of the main components of the belonging product, and sorting the parts in the order in which the predicted values are higher.

Figure 10 above is a figure in which the components included in the product are interpreted as parts, and the relationship between the parts and the predicted value of the failure risk is sorted. For example, the file of which component is included in the predetermined memory is stored in the appropriate memory unit, and the maintenance/application scripting unit 3 can display the component, the component, and the risk of failure on the display unit with reference to it. Treat components as objects to sort out the predicted value of the risk of failure (predicted risk or future risk), This makes it possible to clarify the components that affect the product. On the other hand, as shown in Fig. 11, the predicted value of the failure risk can also be sorted in units of parts included in the assembly. In this way, the predicted value of the failure risk is sorted in units of parts, whereby the parts to be prepared for maintenance can be made clear. In order to obtain such an effect, the operation assisting system of the present embodiment has means for sorting and displaying the predicted value of the risk of failure as shown in Figs.

The failure risk assessment in the failure risk assessment unit 2 of the operation support system 100, the failure risk prediction in the maintenance/operation scenario creation unit 3, and the maintenance/operation script creation are performed at certain time intervals. This time interval can be the same as or different from the time interval between the measurement environment data and the operation data. In the control measurement (SCADA) used in a wind power plant, the statistical values (maximum value, minimum value, average value, etc.) of the environmental data or the operational data are calculated, for example, at intervals of 10 minutes, and the statistical values are accumulated in the PC or the like. The server that constitutes it. For example, if the fault risk assessment and the fault risk prediction are performed in accordance with the 10-point interval, and the fault risk prediction ΔT is set to, for example, several months, the wind power plant can be sufficiently predicted by the operation assisting system of the present embodiment. The failure of the parts contained in it also stabilizes the windmill.

Fig. 12 is a flow chart showing the operation assisting system according to the first embodiment. First, the failure risk assessment unit 2 calculates the failure probability F(t1) and the failure risk RS(t1, p) at the time t1 of the component p which is the target (S11, S12). Next, the failure risk prediction unit 4 calculates a trend curve RS(t, x, p) of the failure risk and a future failure risk RS (t + ΔT, x, y, p) (S13, S14). Then, based on the calculated future risk of failure, by automatically comparing with a predetermined threshold, etc. Alternatively, it is determined whether or not the maintenance/operation is changed by input from the input unit by manual operation such as maintenance/operator. Then, if it is determined that there is a need for the change, the maintenance/application script is manually set/inputted automatically by the predetermined processing or by the input from the input unit by the maintenance/operator or the like. The maintenance/application scenario creation unit 3 creates a maintenance/operation script in accordance with the setting/input (the parameters of the physical quantity/condition of the change) (S15, S16, and S14). Such procedures are repeated at intervals of, for example, 10 minutes as described above to assist in the stable operation of the article.

[Second Embodiment]

Next, an operation assisting system according to a second embodiment of the present embodiment will be described with reference to Figs.

According to the second embodiment, it is possible to achieve a reliability evaluation value of a product similar to the target product and a similar product, and to evaluate the high-accuracy reliability evaluation value of the plurality of components constituting the product, and to implement the operation and maintenance plan of the product.

In the present embodiment, as shown in FIG. 13, in addition to the information of at least one of the environmental data, the operation data, the design data, and the material data of the products from the past to the present, the product 1 and its like machine and the similar machine 13 are used. The fault information is updated, and the fault risk assessment/update unit 14 performs the evaluation and update of the fault risk. The failure data of the plurality of components constituting the product 1 and its homogenizer and similar machine 13 is stored in the fault database 15. The fault data here includes, for example, the operation time from the start of the operation to the failure, or the environmental data and operation from the start of the operation to the time of the failure. Transfer data, maintain/use data. In addition, the so-called homogenizer is a product which is different from the place where the product of the product 1 is of the same type, and the so-called similar machine comprises a product of a different type from the product 1.

Fig. 14 is a diagram showing an example of a probability density function for depicting the life with an equivalent stress amplitude when the probability of failure is calculated by the residual life evaluation of the component included in the article, and updating it according to the Bayesian statistics.

In the failure risk assessment/update unit 14, the fatigue phenomenon that occurs when the product is repeatedly subjected to the load is taken as a target, and when the failure probability F(t1) is obtained by the remaining life evaluation, the rain flow counting method or the like is used. The frequency analysis method is applied to the stress history occurring in the part from the past to the current time point t1, and the stress frequency distribution is obtained (see Fig. 4). In fact, the stress history contains various kinds of stresses, so the stress frequency distribution 21 is used to express it, but when it is assumed that only a certain magnitude of stress occurs, the equivalent stress amplitude Seq can be used, for example. The equation in 14 is calculated from the stress frequency distribution. As long as the equivalent stress amplitude Seq is obtained, the probability density function 16 of the life of the component in the equivalent stress amplitude Seq can be depicted. The update of the risk of failure can be achieved by updating the probability density function that destroys the lifetime. That is to say, based on the probability calculated by the fault data included in the fault database 15 and the density function of the pre-renewed life before the update, the Bayes' theorem of the following formula can be used to obtain the fault data. The density function of the updated lifetime.

Updated density function = probability × density function beforehand ‧ ‧ (8)

Figure 15 reveals an updated flow chart of the density function for such breaking life. First, the failure risk assessment/update unit 14 calculates the stress frequency distribution obtained by performing frequency analysis on the stress history from the past to the current time point t1 of the target part p, and calculates the equivalent stress amplitude Seq using the equation in FIG. (p) (S21). Then, the failure risk assessment/update unit 14 refers to the material data (PSN line diagram) of the target part p previously stored in the design/material database 5, etc., and finds the breaking life under the equivalent stress amplitude Seq(p). Probability density function f(N) (S22). Here, it is assumed that there are k fault data in the fault database 15 and the target part p is the same part, and the fault data (j=1 to k) of the part p j mounted on the same machine or the like. The failure risk assessment/update unit 14 obtains the stress history and the stress frequency distribution from the time of the failure, the environmental data, the operation data, the maintenance/operation data, and the design/material data from the start of the operation of the components to the failure time. The breaking life N f j (S23) at the equivalent stress amplitude Seq(p j ) is calculated in accordance with the equation on the right side in Fig. 15 . The failure risk assessment/update unit 14 can calculate the probability L from the k breaking lifes N f j thus obtained in accordance with the equation on the left side in FIG. 15 (S24). Next, the failure risk assessment/update unit 14 can obtain the probability density function f(N)' of the updated breaking life in accordance with the equation (8) from the prior probability density function f(N) and the degree L (S25). . In the case of dealing with a destructive life, in general, the probability density function of the prior probability density and the probability distribution shape of the probation will obey the Weber distribution or the lognormal distribution. Referring to the updated probability density function f(N)' of the breaking life, the updated PSN line graph can be drawn.

Next, as shown in FIG. 4, it is possible to apply the frequency analysis method to the stress frequency distribution obtained from the stress history occurring from the past to the current time point t1, and the fatigue life curve obtained by setting the damage probability P% in advance. The damage degree D(t1) with respect to the fatigue damage P% is calculated according to the following formula.

D(t1)=(n 1 /N 1 )+(n 2 /N 2 )+...+(n m /N m )‧‧‧(2)

Here, n 1 , n 2 , and n m are the repetition times (m is an integer) of the stress amplitudes S 1 , S 2 , and S m obtained by frequency analysis of the stress history, respectively. Further, N 1 , N 2 , and N m are the number of times of breaking and the occurrence of fatigue fracture at the failure probability P% when the load stress amplitudes S 1 , S 2 , and S m are repeated. Further, the number of times of repetition N(t1) which causes the degree of damage D(t1) to occur can be obtained by the following equation.

N(t1)=D(t1)×Np‧‧‧(3)

Np is the number of times of the failure probability P% set in advance, and the stress amplitude is set to a predetermined stress amplitude Si or Seq(p), as shown in FIG. 4, which can be obtained by N(t1) and fatigue life. The probability density function, by the equation of Fig. 4, finds the probability of failure F(t1)'. In other words, if the elapsed time from the start of the product operation to the present is set to t1, the probability of destruction F(t1)' is a value obtained by integrating the density function f(N)' from 0 to N(t1). .

The updated PSN line graph is used to update the probability of destruction, and multiply the updated failure probability F(t1)' by the influence degree C(p), thereby being able to follow the updated (1) algorithm to calculate the risk of failure after updating. RS(t1,p)'. image Such an update method based on the stress frequency distribution, the P-S-N line graph, and the fault risk of the fault data is an example. For example, the probability of the damage after the update may be calculated without transmitting the equivalent stress amplitude. In addition, the update formula of the density function based on the damage life of the Bayes' theorem can be changed in response to the use of the product 1 as the object, the use environment of the similar machine 13, the operation state, the similarity of the structure, and the like.

In addition, FIG. 16 is a diagram showing an example of a probability density function for updating the life set in advance according to Bayesian statistics in the case where the probability of failure is calculated according to the probability density function of the breaking life of the components included in the product.

In the failure risk assessment/update unit 14, when the probability of failure is directly calculated by using the density function of the life of FIG. 5 without referring to the stress frequency distribution or the PSN line diagram, the shell represented by the equation (8) can also be utilized. Theorem. That is to say, the density function of the lifetime and the fault data can be substituted into the equation (8) to update the density function of the lifetime, and the probability of failure F(t1)' can be obtained by the updated probability density function (Fig. 16). And multiply it by the impact to update the risk of failure. However, in the case where the product 1 is the same type machine or the like machine 13 and the use environment or the operation state is greatly different, it is conceivable that even if the operation time is the same, the parts of the product 1 are the same type machine, the parts of the similar machine 13 Destructive life will still vary significantly. In such a case, in addition to the time, several physical quantities affecting the damage life are selected as the variables from the environmental data and the operational data, and the probability density function of the multivariate life is created.

In addition, the data of the probability density function f(t1) of the lifetime can be stored, for example, in the internal memory of the failure risk assessment unit 2 for each object product. Appropriate memory such as body or design/material database. In addition, the data of the probability density function f(t1) of the life can be determined in advance using environmental data, operation data, and the like.

Fig. 17 is a diagram showing an example of updating the probability density function set in advance according to Bayesian statistics in the case where the failure life is expressed by the probability density function including the variable of time.

Fig. 17 is an example of the probability density function of the life of the two variables. In this example, the value x' obtained by converting the physical quantity x selected from the environmental data or the operation data is set as the probability variable together with the time. This transformation includes a statistic for a physical quantity that changes with time (a mean value, a maximum value, etc. at a certain time interval), or a transformation of an equivalent physical quantity according to frequency analysis, and the like. Even if the density function of the lifetime is multivariate, as shown in Fig. 17, as in the case of the 1 variable, the density function of the lifetime can be updated in accordance with Bayes' theorem. In addition, it is also possible to create a function r(t, x') which includes a physical quantity which greatly affects the damage life or converts it into a quantity x' and a time t, and interprets the function as a probability variable to create a failure life. Probability density function. Further, in the failure risk assessment/update unit 14, when the probability of destruction is calculated by the symptom diagnosis, the probability density function of the failure can be updated by using the Bayes' theorem represented by the equation (8) in the same manner. The probability of damage is determined by the updated probability density, and the probability of failure is multiplied to increase the risk of failure.

Further, the probability density function of the life of the two variables can be stored, for example, in an appropriate memory such as the internal memory of the failure risk assessment unit 2 or the design/material database 5 for each target product. In addition, the life of 2 variables The probability density function can be determined in advance using environmental data, operational data, and the like.

Then, the failure risk assessment/update unit 14 and the maintenance/operation scenario creation unit 3 perform the failure risk trend analysis using the updated failure risk RS(t1, p) as in the first embodiment shown in FIG. S13), future risk prediction (S14), maintenance/operation change (S15), and maintenance/application script development (S16).

By using the failure data of the product 1 and the similar machine and the similar machine 13 in the present embodiment to update the failure risk, the calculation of the destruction probability can be made highly accurate, so that the components included in the product 1 can be implemented with higher precision. Assessment and prediction of failure risk. The time interval for updating the risk of failure may be the same as the interval for assessing the risk of failure, or it may be different. By increasing the frequency of the update of the risk of failure, the risk of failure of the parts contained in the product can be assessed with greater precision.

[Third embodiment]

Next, an operation assisting system according to a third embodiment of the present embodiment will be described with reference to Figs.

Fig. 18 is a schematic block diagram showing the main components of the operation support system according to the third embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

In the maintenance/operation scenario creation unit 3 of the present embodiment, in addition to the environmental data and the operation data measured by the sensor incorporated in the product 1, or the design data, the material data, and the maintenance/application data for the product 1, Information from the external database 17 that is not related to the article 1 is also utilized. The external data included in the external database 17 includes, for example, weather forecasting data of a seahorse calculated by a large computer, prediction data of resources, and burial prediction data of resources. Such external data does not affect the operation state of the product 1, and therefore the external data is not related to the product 1.

Fig. 19 is a schematic block diagram showing the main components of the maintenance/operation scenario creation unit of the operation support system according to the third embodiment of the present invention, and the flow of information between the elements.

As shown in Fig. 19, the failure risk prediction 11 of the maintenance/operation scenario creation unit 3 calculates the risk of future failure by using the trend curve of the failure risk, the maintenance/operation script, and the external data. That is to say, the physical quantity x contained in the environmental data or the operational data, the physical quantity y included in the maintenance/application data, and the physical quantity z included in the external data can be calculated according to the following formula.

RS(t+△T, y, z, p)=g5(t+△T, y, z, p)‧‧‧(9)

RS(t+△T,x,y,z,p)=g6(t+△T,x,y,z,p)‧‧‧(10)

By utilizing the use of external data, it is possible to more widely consider the environment in which the product is exposed for one week, and thus it is possible to predict the risk of future failure more accurately.

[Fourth embodiment]

Next, an operation assisting system according to a fourth embodiment of the present embodiment will be described with reference to Figs.

Figure 20 is a diagram showing the main function of the operation assisting system according to the fourth embodiment of the present invention. A schematic block diagram of the components and the products and databases that provide the elements used in the operational support system and their relationships.

In addition, FIG. 21 is a schematic block diagram showing the main components of the maintenance/operation scenario creation unit of the operation support system according to the fourth embodiment of the present invention, and the flow of information between the elements.

In the present embodiment, the environmental data and the operation data measured by the product 1 are input to the destruction probability evaluation unit 18, and the destruction probability F of the components included in the product 1 is calculated therein. The calculated failure probability F becomes an input to the maintenance/operation scenario creation unit 3, where the probability of failure is multiplied by the degree of influence to create a trend curve of the risk of failure.

Fig. 22 is a schematic block diagram showing the flow of the main components of the maintenance/operation scenario creation unit in the case of the tendency curve of the destruction probability and the flow of information between the elements.

Therefore, in the present embodiment, as shown in Fig. 21, the influence degree database 8 is placed in the maintenance/operation scenario creation unit 3. On the other hand, as shown in FIG. 22, it is also conceivable to form a trend curve which does not cause a risk of failure, but to form a trend curve which destroys the probability. In such a case, the part p can be regarded as an object, and the time t and the physical quantity x which affects the probability of destruction or the probability of destruction are formed as a trend curve of the destruction probability F as follows.

F(t,p)=h1(t,p)‧‧‧(11)

F(t,x,p)=h2(t,x,p)‧‧‧(12)

Equation (11) is a case where only the time t affects the probability of destruction, and the equation (12) is a case where the time t and the measured other physical quantity x have an influence on the probability of destruction or the probability of destruction. In the failure risk prediction 11, the future failure probability F in the case of changing maintenance/utilization is calculated according to the following formula.

F(t+△T,y,p)=h3(t+△T,y,p)‧‧‧(13)

F(t+△T,x,y,p)=h4(t+△T,x,y,p)‧‧‧(14)

Here, y is the physical quantity that affects the failure risk RS in the maintenance/operation data. By multiplying the future probability of destruction or the probability of failure by the influence degree C(p), the future failure risk RS can be determined as follows.

RS(t+△T,y,p)=C(p)×h3(t+△T,y,p)‧‧‧(15)

RS(t+△T,x,y,p)=C(p)×h4(t+△T,x,y,p)‧‧‧(16)

[Fifth Embodiment]

Fig. 23 is a schematic block diagram showing the main components of the operation support system according to the fifth embodiment of the present invention, and the products and databases for providing the data used in the operation support system, and their relationships.

Hereinafter, an operation assisting system according to a fifth embodiment of the present embodiment will be described with reference to Fig. 23 . In the present embodiment, the environmental data and the operation data measured by the product 1 are input to the damage degree evaluation unit 19, and the damage degree of the components included in the product 1 is calculated there. As shown in FIG. 4, when the damage probability is obtained from the stress history and the remaining life evaluation, the P-S-N diagram of FIG. 3 can be obtained by selecting a fatigue life curve of a certain probability of destruction. The calculated degree of damage is input to the maintenance/application scripting unit 3, and the material probability is calculated by referring to the material data, and multiplied by the influence degree to create a trend curve of the risk of failure. Therefore, in the present embodiment, as shown in Fig. 21, the influence degree database 8 is placed in the maintenance/operation scenario creation unit 3. On the other hand, as shown in Fig. 22, it is also possible to assume a trend curve that does not pose a risk of failure, but to create a trend of destruction probability. The shape of the line.

In addition, FIG. 24 is a schematic block diagram showing the main components of the maintenance/operation scenario creation unit in the case of the trend curve of the damage degree, and the flow of the data between the elements. As shown in Fig. 24, it is also conceivable to form a trend curve of the degree of damage. In other words, the future damage degree d3, d4 can be multiplied by the influence degree C(p), and the trend curve of the failure risk is determined as follows.

F(t+△T,y,p)=K(p)×d3(t+△T,y,p)‧‧‧(17)

F(t+△T,x,y,p)=K(p)×d4(t+△T,x,y,p)‧‧‧(18)

RS(t+△T,y,p)=C(p)×K×d3(t+△T,y,p)‧‧‧(19)

RS(t+△T,x,y,p)=C(p)×K×d4(t+△T,x,y,p)‧‧‧(20)

Here K(p) is the conversion constant necessary to calculate the probability of damage from the damage degree. K(p) can also be accumulated in the damage database, impact database or other database. In the present embodiment, since the failure probability or the damage degree is temporarily calculated independently, there is an advantage that the failure probability or the damage degree is visualized (displayed by the display unit).

The risk of failure according to the formula (1) is a risk of failure that is included in the damage accumulated in the product from the start of the operation of the product to the time t1. On the other hand, it is also possible to calculate the risk of failure due to the damage accumulated in the product from the current time point t1 to a certain time point t1 + ΔT in the future, and use it as a risk of failure. Predictive value. In such a case, first calculate the period from the current time point t1 to a certain time point t1 + ΔT in the future according to the following equation Medium, the probability of damage to a part contained in the product.

P(t1, t1+△T)=(F(t1+△T)-F(t1))/(1-F(t1))

‧‧‧(twenty one)

Here, F(t1) and F(t1+ΔT) are the probability of destruction at time t1 and t1+ΔT, respectively. Then, the risk of failure from the current time point t1 to a certain time point t1 + ΔT in the future can be determined by the influence degree C(p) of the target component and the probability of destruction P(t1, t1 + ΔT). Calculation.

RS(t1+△T)=C(p)×P(t1,t1+△T)‧‧‧(22)

The failure rate and failure risk of equations (21) and (22) will change according to environmental data, operation data, and maintenance/application data. Therefore, the probability of failure and the risk of failure are not only a function of time, but equation (21), 22) A function described as time for simplicity.

C. Notes

Further, the present invention is not limited to the above embodiment, and various modifications are also included. For example, the above-described embodiments are described in detail to explain the present invention in an easy-to-understand manner, and are not necessarily limited to having all of the configurations described. Further, a part of the configuration of a certain embodiment may be replaced with a configuration of another embodiment, and a configuration of another embodiment may be added to the configuration of a certain embodiment. Further, other components may be added, deleted, or replaced for a part of the configuration of each embodiment.

Further, each of the above-described configurations, functions, processing units, processing means, and the like may be designed by an integrated circuit or the like, for example, by a part or all of them. Implemented in hardware. Further, each of the above-described configurations, functions, and the like can be realized by software by interpreting and executing a program that realizes each function by a processor. Information such as programs, tables, and files for each function can be placed in a memory, or a recording device such as a hard disk or an SSD (Solid State Drive), or a recording medium such as an IC card, an SD card, or a DVD.

In addition, if the control line or information line reveals that it is necessary, it may not reveal all the control lines or information lines on the product. In fact, it is also conceivable that almost all of the components are connected to each other.

1‧‧‧ object products

2‧‧‧Fault Risk Assessment Department

3‧‧‧Maintenance/Using Script Development Department

4‧‧‧Fault Risk Forecasting Department

5‧‧‧Design/Materials Database

100‧‧‧Operational support system

Claims (15)

  1. An operation assisting device comprising: a failure risk assessment unit; and a maintenance/application script development unit; the failure risk assessment unit uses environmental data and operation data input from a plurality of sensors of the target product, and a predetermined design The data and material data are used to calculate the probability of failure F(t1) at time t1 of the part p regarded as the object. At time t1, the probability of failure F(t1) at time t1 of part p is damaged when part p is In the case of the predetermined product C (p) of each part p, the risk of failure of the part p contained in the product is calculated (t1, p), and the maintenance/application scripting unit is based on The failure risk RS (t1, p) of the part p at the time t1 sent from the failure risk assessment unit, and the memory that has been sent from the failure risk assessment unit and memorized in the failure risk database from the past to the time t1 A plurality of failure risks, and the physical quantity x and the time t which are selected from the environmental data and the operation data input in advance by the target product are selected as variables, and a trend curve of the failure risk is generated, The trend curve of the risk, and the predicted value of the risk of failure from the current time point to the predetermined time in advance is obtained, and the predicted value of the risk is determined according to the input from the input unit according to the maintenance/operator. Set it up, or set it in advance The process of automatically setting and maintaining the maintenance/application script of the product, and generating a fault risk prediction with the physical quantity y that affects the failure risk selected by the maintenance data and/or the operation data as the variable at time t, the physical quantity x, and the maintenance data and/or the operation data Model to predict the future failure risk of part p, and to sort out the future failure risk of each part according to the higher predicted value of the failure risk and display it on the display part and/or in the memory part, and A plurality of thresholds are set in advance to divide the group, and each part is automatically set for each group by manual input by the maintenance/operator from the input unit or by a predetermined process. The maintenance content is maintained during the maintenance period, and the maintenance application content is memorized in the memory unit and/or displayed on the display unit.
  2. The operation assisting device according to claim 1, wherein the fault risk assessment unit calculates the stress history of the component from the past to the current time point t1 using the environmental data, the operation data, and the design data. In the course of the process, the rain flow counting method or other frequency analysis method is used to create the stress frequency distribution, and the stress frequency distribution and the fatigue life curve determined in advance are calculated, and the damage degree D(t1) is calculated, and the damage degree D(t1) is determined. The number of times of repetition N (t1) is integrated from 0 to N (t1) as a function of the probability density of the fatigue life defined in advance as the material data, and the probability of failure F(t1) is obtained.
  3. Operation aids as described in claim 1 or 2 In the above-described maintenance/application scenario development unit, the memory includes a memory, and one or a plurality of components included in each component for classifying the product are stored in advance, and the components of the component are referred to by referring to the memory. The predicted value of the future failure risk is sorted and displayed on the display unit and/or stored in the memory unit.
  4. The operation support device according to the first aspect of the invention, comprising: an external database, in which external data irrelevant to the product is stored in advance, and the maintenance/application script development unit uses a trend curve of failure risk and maintenance/ Use scripts and external data to calculate future failure risks.
  5. The operation assisting device according to claim 1 or 2, wherein the maintenance/operation scenario creating unit creates a trend curve for destroying the probability F and/or a trend curve for the damage degree.
  6. A wind power generation system comprising: the operation auxiliary device according to claim 1 or 2; and the wind power generator, which is an object product having a plurality of sensors.
  7. An operation auxiliary device having: a failure risk assessment/update unit; and a maintenance/application scripting unit; a fault database for accumulating fault data of a plurality of parts constituting the target product and the same type machine; the failure risk assessment/update unit, a probability density function based on the life of the target part p, and the aforementioned The probability calculated from the fault data contained in the fault database, using the Bayes' theorem, finds the probability density function that takes into account the updated life of the fault data, and uses the probability density function of the updated lifetime to utilize the target product. The environmental information and operation data input by the plurality of sensors, and the design data and material materials set in advance, to calculate the updated probability of destruction F(t1)' at time t1 of the part p regarded as the object, At time t1, the product of the failure probability F(t1)' at the time t1 of the part p and the influence degree C(p) of each part p which is previously determined in the case where the part p is damaged is multiplied. The updated failure risk RS(t1,p)' of the part p included in the product is calculated, and the maintenance/application scripting unit is based on the time t1 sent from the failure risk assessment/update unit. The updated failure risk RS(t1, p)' of the part p, and the plurality of updated failure risks from the past to the time t1 that have been received from the failure risk assessment/update unit and memorized in the failure risk database The physical quantity x and the time t which influence the risk of failure selected from the environmental data and the operation data input by the object product in advance are set as variables, and a trend curve of the risk of failure is generated. According to the trend curve of the fault risk, the predicted value of the fault risk from the current time point to the time set in advance is obtained, and the input from the input unit according to the maintenance/operator is followed according to the predicted value of the fault risk. And the maintenance/application script of the product that is manually set or automatically set by the predetermined processing to generate the pair selected by the time t, the physical quantity x, and the maintenance data and/or the operation data The physical quantity y affected by the failure risk is used as a variable risk prediction model to predict the future failure risk of the part p, and the prediction and trend curves are memorized together in the memory and/or displayed on the display.
  8. The operation assisting device according to claim 7, wherein the failure risk assessment/update unit obtains a stress frequency distribution obtained by performing frequency analysis on a stress history from a past part to a current time point t1 of the part p, Calculate the equivalent stress amplitude Seq(p) using the following equation, n i : frequency m of stress amplitude S i : slope of fatigue life curve Referring to the material data of part p, the probability density function f(N) of the breaking life under the equivalent stress amplitude Seq(p) is obtained, if the above failure In the database, there are k and the object parts p are the same parts, and the fault data (j=1~k) of the parts p j loaded on the same machine, the environment from the start of the operation of the parts to the time of failure Data, operation data, maintenance/application data, and design/material data, find the stress history and stress frequency distribution up to the fault, and calculate the breaking life under the equivalent stress amplitude Seq(p) according to the following formula , n i : frequency m of the stress amplitude S i : slope of the fatigue life curve from the obtained k breaking lives , calculate the probability L according to the following formula, From the probability density function f(N) and the probability L beforehand, the probability density function f(N)' of the updated breaking life is obtained according to the following equation, and the updated probability density function f(N)'=probability× The ex ante probability density function f(N) calculates the updated failure probability F(t1)' by the updated probability density function f(N)' of the breaking life, and the updated failure probability F(t1)' The influence degree C(p) of each part p is multiplied in advance, thereby calculating the updated failure risk RS(t1, p)'.
  9. The operation assisting device according to claim 7 or 8, wherein the fault risk assessment/update unit uses the environmental data, the operation data, and the design data to calculate the stress occurring in the part from the past to the current time point t1. History, the stress history, using the rain flow counting method or other frequency analysis method, the stress frequency distribution, the stress frequency distribution, and the fatigue life curve of the pre-determined failure probability P%, calculated relative to the fatigue damage P% The damage degree D(t1) is obtained by determining the number of times of overshoot N(t1) which causes the damage degree D(t1) to occur, and the probability density function f(N)' of the obtained fatigue life is obtained from 0 to N(t1) The integral is obtained, and the probability of destruction F(t1)' after the update is obtained.
  10. The operation assisting device according to claim 7 or 8, wherein the failure risk assessment/update unit utilizes the Bayesian theorem by using the probability density function f(t1) of the lifetime that is memorized in advance and the aforementioned generality. , update the density function of the life, and integrate the updated probability density function f(t)' from 0 to t1 to obtain the updated probability of failure F(t1)', and multiply it by the influence degree C(p) Update the risk of failure.
  11. The operation auxiliary device as described in claim 7 or 8, wherein The foregoing failure risk assessment/update unit selects and destroys a plurality of physical quantities related to each other based on the measurement data of the product in advance, and uses the probability density function of the damage according to the operation data group at the time of the abnormality and the aforementioned generality, The theorem, update the probability density function of the damage, and describe the position of the running data at the current time point for the updated probability density function, thereby obtaining the updated probability of failure F(t1)' at the current time point t1.
  12. For example, the operation auxiliary device described in item 7 of the patent application; and having an external database, which previously stores external data irrelevant to the product, the maintenance/application scripting department, using the trend curve of the risk of failure, and maintenance/operation Scripts, and external materials, calculate the risk of future failures.
  13. The operation auxiliary device as described in claim 4 or 12, wherein the external data includes pre-calculated meteorological and/or walrus future prediction data, resource supply prediction data, and resource burial prediction data. One or plural.
  14. The operation assisting device described in claim 7 of the patent application; and the maintenance/application scripting unit, the trend curve for destroying the probability and/or the trend curve for the damage degree.
  15. A wind power generation system comprising: the operation auxiliary device as described in claim 7; and The first wind power generator is a target product having a plurality of sensors; the second wind power generator has a plurality of sensors, and the first wind power generator is a same type machine.
TW106104632A 2016-03-25 2017-02-13 Operational auxiliary device and wind power generation system TWI607328B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP2016061725A JP2019113883A (en) 2016-03-25 2016-03-25 Utilization assisting apparatus and wind power generation system

Publications (2)

Publication Number Publication Date
TW201734870A TW201734870A (en) 2017-10-01
TWI607328B true TWI607328B (en) 2017-12-01

Family

ID=59899924

Family Applications (1)

Application Number Title Priority Date Filing Date
TW106104632A TWI607328B (en) 2016-03-25 2017-02-13 Operational auxiliary device and wind power generation system

Country Status (3)

Country Link
JP (1) JP2019113883A (en)
TW (1) TWI607328B (en)
WO (1) WO2017163561A1 (en)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP6526362B1 (en) * 2018-03-13 2019-06-05 三菱電機株式会社 Failure degree calculation device and failure degree calculation system
JP2019160128A (en) * 2018-03-16 2019-09-19 株式会社日立製作所 Fault probability evaluation system and method

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103914791A (en) * 2014-03-17 2014-07-09 中国南方电网有限责任公司超高压输电公司检修试验中心 Electrical equipment state maintenance system
TW201532400A (en) * 2013-11-27 2015-08-16 Apple Inc Provisioning of credentials on an electronic device using passwords communicated over verified channels
CN102637289B (en) * 2012-04-16 2016-02-24 天津大学 Containing the security value appraisal procedure of the Power System Planning scheme of large-scale wind power

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4528533B2 (en) * 2004-02-03 2010-08-18 株式会社東芝 Nuclear power plant equipment inspection method and nuclear power plant equipment inspection system
JP4977064B2 (en) * 2008-03-12 2012-07-18 株式会社東芝 Maintenance plan support system
US8326577B2 (en) * 2011-09-20 2012-12-04 General Electric Company System and method for predicting wind turbine component failures

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102637289B (en) * 2012-04-16 2016-02-24 天津大学 Containing the security value appraisal procedure of the Power System Planning scheme of large-scale wind power
TW201532400A (en) * 2013-11-27 2015-08-16 Apple Inc Provisioning of credentials on an electronic device using passwords communicated over verified channels
CN103914791A (en) * 2014-03-17 2014-07-09 中国南方电网有限责任公司超高压输电公司检修试验中心 Electrical equipment state maintenance system

Also Published As

Publication number Publication date
JP2019113883A (en) 2019-07-11
TW201734870A (en) 2017-10-01
WO2017163561A1 (en) 2017-09-28

Similar Documents

Publication Publication Date Title
Carroll et al. Failure rate, repair time and unscheduled O&M cost analysis of offshore wind turbines
US8401833B2 (en) Method for predicting power usage effectiveness and data center infrastructure efficiency within a real-time monitoring system
Wang et al. Fault prognostics using dynamic wavelet neural networks
US9471452B2 (en) Adaptive handling of operating data
US6871160B2 (en) Intelligent condition-based engine/equipment management system
Faulstich et al. Wind turbine downtime and its importance for offshore deployment
Meeker et al. Reliability meets big data: opportunities and challenges
Vachtsevanos et al. Fault prognosis using dynamic wavelet neural networks
Shafiee Maintenance logistics organization for offshore wind energy: Current progress and future perspectives
Smith Reliability, maintainability and risk: practical methods for engineers
Byon et al. Season-dependent condition-based maintenance for a wind turbine using a partially observed Markov decision process
US20180210479A1 (en) Microgrid model based automated real time simulation for market based electric power system optimization
Garcia et al. SIMAP: Intelligent System for Predictive Maintenance: Application to the health condition monitoring of a windturbine gearbox
US20040049715A1 (en) Computer networked intelligent condition-based engine/equipment management system
US5210704A (en) System for prognosis and diagnostics of failure and wearout monitoring and for prediction of life expectancy of helicopter gearboxes and other rotating equipment
Hameed et al. Condition monitoring and fault detection of wind turbines and related algorithms: A review
Lapira et al. Wind turbine performance assessment using multi-regime modeling approach
US8370046B2 (en) System and method for monitoring a gas turbine
US20110020122A1 (en) Integrated condition based maintenance system for wind turbines
Nielsen et al. On risk-based operation and maintenance of offshore wind turbine components
US7895016B2 (en) System and method for wind turbine health management
Shafiee A fuzzy analytic network process model to mitigate the risks associated with offshore wind farms
CN104254810A (en) Method and system for condition monitoring of a group of plants
RU2313815C2 (en) Device and method for controlling technical plant, which contains a set of systems, in particular, electric power plant
JP6397330B2 (en) Measure the remaining service life of rotating machinery such as drivetrains, gearboxes, and generators