KR102407773B1 - Facility reliablility calculation system and method reflecting real-time health of facilities - Google Patents

Abstract

A facility reliability calculation method reflecting facility health according to an embodiment includes an operation of acquiring raw data using at least one sensor, and a deep learning model based on the raw data. Acquiring score data through the operation, acquiring soundness through a machine learning model based on the score data, obtaining a reliability correction constant based on the soundness, and the reliability correction constant It may include an operation of correcting the equipment reliability based on the.

Description

A system and method for calculating equipment reliability reflecting real-time health of equipment

The present invention relates to a system and method for calculating equipment reliability reflecting equipment health. More particularly, it relates to an apparatus and method for predicting and calculating equipment reliability by reflecting AI-based equipment real-time health.

Reliability engineering includes temporal stability, such as failure, deterioration, etc., or integrity, such as repair and maintenance, in the function of parts, devices, or systems. However, the importance of reliability engineering is increasing.

In this case, reliability related to parts, etc. is defined as the probability that a specified function is performed for an intended period under specified conditions, and time elapses in each failure period such as initial, accidental, or wear, or system reliability, etc. It can be used as a quantitative analysis technique.

In such reliability engineering, the method of acquiring reliability is to obtain the failure state probability by randomizing past data, or by extracting n samples of the process state and using the average and standard deviation of the number of nonconforming products included in the sample. It is limited to deriving the probability that nonconforming products were found.

Korean Patent Registration No. 10-2309979 (2021.07.26.)

In reliability engineering, when a life model is obtained based on the failure history data of a facility and reliability is calculated based on the life model, the current state is not reflected, but only the operating time during which the facility has been operated may be reflected.

In addition, even when the probability of failure is obtained by randomizing past data or when the reliability is calculated using the data obtained by extracting a sample of the process state, there may be a problem that the current state of the equipment is not reflected in the reliability. .

In this case, even if the current state of the equipment is unstable, a problem of representing the same reliability value as that of the equipment in a steady state operated for the same time may occur.

The apparatus and method according to various embodiments may correct the reliability of the facility by reflecting the soundness of the current state of the facility.

A facility reliability calculation method reflecting facility health according to an embodiment includes an operation of acquiring raw data using at least one sensor, and a deep learning model based on the raw data. Acquiring score data through the operation, acquiring soundness through a machine learning model based on the score data, obtaining a reliability correction constant based on the soundness, and the reliability correction constant It may include an operation of correcting the equipment reliability based on the.

According to an embodiment, the facility reliability calculation system reflecting facility health is a life model analyzer that calculates a life model of the facility based on the failure history data of the facility, and determines the current state of the facility through a learning model An expected value calculation model learning unit, a health calculating unit that acquires the health of the equipment based on the current state of the equipment, and the life model calculated through the life model analysis unit and the healthiness obtained through the health calculator and a facility reliability calculator configured to calculate the facility reliability based on at least a part of the.

According to various embodiments of the present disclosure, by correcting the reliability by reflecting the soundness of the current state of the facility, the accuracy of the reliability may be increased.

In addition, according to various embodiments of the present disclosure, based on an artificial intelligence (AI)-based model, it is possible to provide an algorithm for securing high-accuracy reliability by acquiring the health of the current facility in real time and reflecting it in the reliability. can

The effects obtainable in the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those of ordinary skill in the art to which the present invention belongs from the following description. will be.

Other aspects, features and benefits as set forth above of certain preferred embodiments of the invention will become more apparent from the following description taken in conjunction with the accompanying drawings.
1 is a block diagram of a system for calculating equipment reliability according to an embodiment.
2 is a flowchart illustrating a method of calculating equipment reliability according to an exemplary embodiment.
3 illustrates a deep learning model for obtaining score data from raw data according to an embodiment.
4 illustrates a machine learning model for acquiring facility health based on score data, according to an embodiment.
5A illustrates a reliability correction constant according to a reflection ratio of soundness according to an exemplary embodiment.
5B illustrates facility reliability according to a reflection ratio of soundness according to an embodiment.
5C illustrates a failure probability function according to a reflection ratio of soundness according to an embodiment.
In connection with the description of the drawings, the same reference numerals may be used for identical or substantially identical components.

Advantages and features of the present invention, and a method for achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the technical field to which the present invention belongs It is provided to fully inform the possessor of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

In describing the embodiments of the present invention, if it is determined that a detailed description of a well-known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description thereof will be omitted. In addition, the terms described below are terms defined in consideration of functions in an embodiment of the present invention, which may vary according to intentions or customs of users and operators. Therefore, the definition should be made based on the content throughout this specification.

At this time, it will be understood that each block of the flowchart diagrams and combinations of the flowchart diagrams may be performed by computer program instructions. These computer program instructions may be embodied in a processor of a general purpose computer, special purpose computer, or other programmable data processing equipment, such that the instructions performed by the processor of the computer or other programmable data processing equipment are not described in the flowchart block(s). It creates a means to perform functions. These computer program instructions may also be stored in a computer-usable or computer-readable memory that may direct a computer or other programmable data processing equipment to implement a function in a particular manner, and thus the computer-usable or computer-readable memory. It is also possible that the instructions stored in the flow chart block(s) produce an article of manufacture containing instruction means for performing the function described in the flowchart block(s). The computer program instructions may also be mounted on a computer or other programmable data processing equipment, such that a series of operational steps are performed on the computer or other programmable data processing equipment to create a computer-executed process to create a computer or other programmable data processing equipment. It is also possible that instructions for performing the processing equipment provide steps for performing the functions described in the flowchart block(s).

Additionally, each block may represent a module, segment, or portion of code that includes one or more executable instructions for executing specified logical function(s). It should also be noted that in some alternative implementations it is also possible for the functions recited in blocks to occur out of order. For example, two blocks shown one after another may be performed substantially simultaneously, or the blocks may sometimes be performed in the reverse order according to a corresponding function.

In this case, the term '~ unit' used in this embodiment means software or hardware components such as field-programmable gate array (FPGA) or ASIC (Application Specific Integrated Circuit), and '~ unit' refers to what role carry out the However, '-part' is not limited to software or hardware. '~ unit' may be configured to reside in an addressable storage medium or may be configured to regenerate one or more processors. Thus, as an example, '~' refers to components such as software components, object-oriented software components, class components, and task components, and processes, functions, properties, and procedures. , subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. The functions provided in the components and '~ units' may be combined into a smaller number of components and '~ units' or further separated into additional components and '~ units'. In addition, components and '~ units' may be implemented to play one or more CPUs in a device or secure multimedia card.

In describing embodiments of the present invention in detail, an example of a specific system will be mainly targeted, but the main subject matter to be claimed in the present specification is to extend the scope disclosed herein to other communication systems and services having a similar technical background. It can be applied within a range that does not deviate significantly, and this will be possible at the discretion of a person with technical knowledge skilled in the art.

Hereinafter, the facility reliability calculation system 101 according to an embodiment of the present invention will be described with reference to the accompanying drawings.

1 is a block diagram of a system for calculating equipment reliability according to an embodiment.

Referring to FIG. 1 , a facility reliability calculation system 101 according to an exemplary embodiment includes a life model analysis unit 100 , an expected value calculation model learning unit 200 , a soundness calculation unit 300 , and a facility reliability calculation unit 400 . ) may be included.

According to an embodiment, the life model analysis unit 100 may calculate a life model of the equipment. The life model analyzer 100 according to an embodiment may calculate a life model based on data related to a failure history for each type of equipment and each manufacturer. For example, the life model analysis unit 100 may calculate a life model based on failure history data for each manufacturer and type of equipment related to the probability of occurrence of a failure of the equipment.

According to an embodiment, the life model analysis unit 100 may calculate a life model by utilizing data related to a failure history for each type of equipment and a manufacturer based on a Weibull distribution. However, the distribution of the lifetime model is not limited to the above-described example.

According to an embodiment, the expected value calculation model learning unit 200 may calculate desired data by inputting the acquired data by using a designated model. According to an embodiment, the expected value calculation model learning unit 200 may include deep learning or a machine learning model.

According to an embodiment, the expected value calculation model learning unit 200 may acquire score data from raw data through a deep learning model. For example, the expected value calculation model learning unit 200 may acquire score data by inputting raw data acquired through at least one sensor into the deep learning model. According to an embodiment, the score data may be obtained corresponding to each evaluation item, such as a failure probability and an average lifespan.

As evaluation items according to an embodiment, failure probability, reliability, and average lifespan may be referred to by the following equation.

는 고장확률이고, R(t)는 신뢰도이며, 수학식 3은 MTTF를 의미하는데, Mean Time To Failure로 평균 고장 수명, 혹은 평균수명을 의미한다.here

is the failure probability, R(t) is the reliability, and Equation 3 means the MTTF. Mean Time To Failure means the average failure life or average lifespan.

According to an embodiment, the expected value calculation model learning unit 200 may acquire soundness from the score data through the machine learning model. For example, the expected value calculation model learning unit 200 may acquire the health of the equipment by inputting the score data obtained through the deep learning model into the machine learning model. A detailed description thereof will be given later.

According to an embodiment, the health calculator 300 may determine a health index (HI) for determining the current state of the facility, and obtain the value.

According to an embodiment, the health calculation unit 300 receives the equipment measurement and inspection data and calculates the score for each evaluation item based on the real-time expected value calculation unit 310 for calculating the real-time expected value, the real-time measured value and the real-time expected value The evaluation item score calculation unit 320 and the evaluation item score calculation unit 320 may include a facility health calculation unit 330 for calculating the optimal health of the facility by adding up the scores.

According to an embodiment, the facility reliability calculation unit 400 may calculate the reliability according to the health status indicating the facility operation time and the current state, based on the life model calculated by the life model analysis unit 100 .

According to an embodiment, the facility reliability calculation unit 400 includes a reliability correction constant calculation unit ( 410), the facility reliability calculation unit 420 that calculates the reliability that reflects the health index in the reliability indicating the probability that the facility can perform its role without error for a specified time under the given conditions, and considers the health index after operation of the facility When the equipment failure probability calculation unit 430 for calculating the instantaneous failure rate at which a failure may occur at a specific point in time and the health index after operation of the facility are considered, the facility remaining life calculation unit 440 for calculating the remaining life compared to the expected lifespan may include

2 is a flowchart illustrating a method of calculating equipment reliability according to an exemplary embodiment.

Referring to FIG. 2 , the facility reliability calculation system 101 according to an embodiment may correct facility reliability by acquiring soundness from measurement data and using a reliability correction constant calculated based on the soundness. Devices utilized in the operation of FIG. 2 may be referred to as devices of FIG. 1 .

According to an embodiment, the facility reliability calculation system 101 may acquire raw data using at least one sensor in operation 201 . The facility reliability calculation system 101 according to an embodiment may acquire raw data through a measurement sensor connected to the facility.

According to an embodiment, the raw data may be referred to as data such as a failure frequency, a failure state occurrence probability, and a nonconforming product probability among random samples, but is not limited to the above-described examples.

According to an embodiment, the facility reliability calculation system 101 may acquire score data based on raw data through a deep learning model in operation 203 . According to an embodiment, in operation 203 , the facility reliability calculation system 101 may acquire score data by inputting the raw data acquired in operation 201 to the deep learning model. For example, in operation 203 , the facility reliability calculation system 101 checks data generated by inputting the raw data obtained in operation 201 into the deep learning model as expected value data, and the residual between the raw data and the expected value data ) based on the score data can be obtained.

According to an embodiment, the deep learning model used in operation 203 may be referred to as a learning model that imitates a human learning process and derives an optimal weight through repetitive training. A detailed description thereof will be given later.

According to an embodiment, the facility reliability calculation system 101 is a residual value of the raw data and the expected value for each evaluation item calculated in operation 203 , and a threshold and a limit value determined for each evaluation item. Based on the score data can be obtained. A method of obtaining score data according to an embodiment may be referred to by the following equation.

According to an embodiment, the facility reliability calculation system 101 may obtain a health index (HI) of a facility based on score data through a machine learning model in operation 205 . According to an embodiment, in operation 205 , the facility reliability calculation system 101 may acquire the health indicating the current state of the facility by inputting the score data obtained in operation 203 to the machine learning model.

According to an embodiment, the machine learning model used in operation 205 may be referred to as an algorithm using several decision trees. A detailed description thereof will be given later.

According to an embodiment, the facility reliability calculation system 101 may obtain a reliability correction constant based on the soundness in operation 207 . In operation 207 , the facility reliability calculation system 101 according to an embodiment may obtain a reliability correction constant by using the quality of the facility acquired in operation 205 .

Operation 207 of obtaining the reliability correction constant k based on the health HI according to an embodiment may be referred to by the following equation.

)에 신뢰도 보정 상수(k)를 곱함으로써 각 평가 항목 별 설비의 신뢰도를 보정하여 획득할 수 있다.According to an embodiment, the facility reliability calculation system 101 may correct the facility reliability based on the reliability correction constant in operation 209 . According to an embodiment, in operation 209 , the facility reliability calculation system 101 may acquire the corrected reliability of the facility by using the reliability correction constant obtained in operation 207 . For example, in operation 209, the facility reliability calculation system 101 determines the scale parameter (

) by the reliability correction constant (k), it can be obtained by correcting the reliability of the equipment for each evaluation item.

According to an embodiment, operation 209 of acquiring the reliability of the facility based on the reliability correction constant may be referred to by the following equations.

Here, R(t) denotes reliability, h(t) denotes an instantaneous failure rate, Equation 8 denotes RUL, and Remaing Useful Life.

According to an embodiment, the facility reliability calculation system 101 corrects the reliability of the facility for each evaluation item using the soundness indicating the current state of the facility or the reliability correction constant obtained through the soundness, thereby increasing the accuracy. reliability can be obtained.

3 illustrates a deep learning model for obtaining score data from raw data according to an embodiment.

Referring to FIG. 3 , the deep learning model 303 according to an embodiment may receive raw data 301 and output (or generate) expected value data 302 . According to one embodiment, score data generation through the deep learning model 303 of FIG. 3 may be referred to as operation 203 of FIG. 2 .

According to an embodiment, raw data 301 (eg, failure probability) measured through a measurement sensor connected to a facility may be input to the deep learning model 303 as input data.

The deep learning model 303 according to an embodiment may receive the raw data 301 and output (or generate) expected value data. According to an embodiment, the deep learning model 303 may acquire score data based on the raw data 301 and the expected value data. For example, the deep learning model 303 may acquire score data based on the residual between the raw data 301 and the expected value data.

The deep learning model 303 according to an embodiment may be referred to as a learning model that imitates a human learning process and derives an optimal weight through repetitive training. For example, the deep learning model 303 may be referred to as a gated recurrent unit-variational auto encoder (GRU-VAE). GRU-VAE adds the computational complexity of LSTM (long short-term memory), a recurrent neural network model, to a VAE model that receives raw data (e.g., raw data 301) as input and generates data (e.g., expected value data 302). As a model combined with reduced GRU, it can be suitable for forecasting measurement data of facilities with time-series characteristics. However, the deep learning model 303 according to an embodiment is not limited to the above-described example and may be referred to as various models.

4 illustrates a machine learning model for acquiring facility health based on score data, according to an embodiment.

Referring to FIG. 4 , the machine learning model 403 according to an embodiment may receive score data 401 and calculate (or output) the health 402 . The operation of calculating the health 402 by using the machine learning model 403 according to an embodiment may be referred to as operation 205 of FIG. 2 .

According to an embodiment, the machine learning model 403 receives the score data 401 obtained through the deep learning model 303 as input data, and outputs the health 402 indicating the current state of the equipment as output data. can be calculated.

The machine learning model 403 according to an embodiment may be referred to as eXtream Gradient Boosting (XGBoost). For example, the machine learning model 403 may be referred to as XGBoost as a type of an ensemble boosting algorithm that uses a combination of several decision trees. XGBoost is an algorithm that improves the error of multiple decision trees, and can be referred to as an algorithm with improved prediction and computational performance through distributed and parallel processing. However, the machine learning model 403 is not limited to the above-described example, and may be referred to as various learning algorithms.

According to an embodiment, the deep learning model 303 of FIG. 3 and/or the machine learning model 403 of FIG. 4 calculates the weight of a plurality of inputs in the function through deep learning through learning. can In addition, various models such as RNN (Recurrent Neural Network), DNN (Deep Neural Network), and DRNN (Dynamic Recurrent Neural Network) may be used as an AI network model used for such learning.

Here, RNN is a deep learning technique that considers current data and past data at the same time. Recurrent neural network (RNN) refers to a neural network in which connections between units constituting an artificial neural network constitute a directed cycle. Furthermore, various methods may be used for a structure capable of constructing a recurrent neural network (RNN), for example, a fully recurrent network, a hopfield network, an Elman network, an ESN (Echo). state network), long short term memory network (LSTM), bi-directional RNN, continuous-time RNN (CTRNN), hierarchical RNN, and secondary RNN are representative examples. In addition, as a method for learning a recurrent neural network (RNN), a method such as gradient descent, Hessian Free Optimization, or Global Optimization Method may be used.

5A illustrates a reliability correction constant according to a reflection ratio of soundness according to an exemplary embodiment. 5B illustrates facility reliability according to a reflection ratio of soundness according to an embodiment. 5C illustrates a failure probability function according to a reflection ratio of soundness according to an embodiment.

5A to 5C , facility reliability according to an embodiment may vary according to a reflection ratio of soundness indicating the current state of the facility.

Referring to FIG. 5A , the reliability correction constant k according to an embodiment may vary according to a reflection ratio of the health of the equipment. According to an embodiment, the reliability correction constant k may gradually increase as the facility soundness reflection ratio gradually decreases from 100%. For example, the reliability correction constant may increase from 1 to 2 as the reflection ratio of facility health is lowered from 100% to about 30%.

Referring to FIG. 5B , the reliability function R(t) according to use time according to an embodiment may vary according to a reflection ratio of the health of the equipment. According to an embodiment, in the reliability function according to the usage time, as the reflection ratio of soundness gradually decreases from 100%, the reliability for the same usage time may decrease. For example, as the reflection rate of soundness decreases from 100% to 70%, the reliability for 50,000 ms of use time may decrease from about 58% to about 38%.

)는 설비의 건전도의 반영 비율에 따라 달라질 수 있다. 일 실시 예에 따르면, 사용 시간에 따른 고장 확률 함수는 건전도의 반영 비율이 100%에서 점차 감소함에 따라, 점차 증가할 수 있다. 예를 들어, 건전도의 반영 비율이 100%에서 30%로 감소함에 따라, 사용 시간 약 75,000ms에 대한 고장 확률이 약 0.004%에서 약 0.023%로 증가할 수 있다.Referring to FIG. 5C , a failure probability function according to usage time (

) may vary depending on the reflection ratio of the soundness of the equipment. According to an embodiment, the failure probability function according to usage time may gradually increase as the reflection ratio of soundness gradually decreases from 100%. For example, as the reflection rate of soundness decreases from 100% to 30%, the failure probability for about 75,000 ms of use time may increase from about 0.004% to about 0.023%.

According to an embodiment, accurate reliability may be calculated by reflecting the soundness representing the current state of the facility for each item representing the reliability of the facility. For example, when applying a condition of a different degree of aging according to an installation environment of a facility and an operating load condition as the soundness, accurate reliability can be calculated.

A facility reliability calculation method reflecting facility health according to an embodiment includes an operation of acquiring raw data using at least one sensor, and a deep learning model based on the raw data. Acquiring score data through the operation, acquiring soundness through a machine learning model based on the score data, obtaining a reliability correction constant based on the soundness, and the reliability correction constant It may include an operation of correcting the equipment reliability based on the.

The facility reliability calculation method according to an embodiment may further include calculating at least one of a failure probability and an average lifespan of the facility, based on the corrected facility reliability.

According to an embodiment, the reliability correction constant may be referred to as a reciprocal of the soundness.

According to an embodiment, the deep learning model may be referred to as a gated recurrent unit-variational auto encoder (GRU-VAE).

According to an embodiment, the machine learning model may include an algorithm including a plurality of decision trees.

According to an embodiment, the facility reliability calculation system 101 reflecting facility health is the current state of the facility through a life model analysis unit that calculates a life model of the facility based on the failure history data of the facility, and a learning model An expected value calculation model learning unit to determine the health condition of the facility based on the current state of the facility, a health calculation unit for obtaining the health of the facility, and the life model calculated through the life model analysis unit and the health condition obtained through the and a facility reliability calculator configured to calculate the facility reliability based on at least a part of the soundness.

According to an embodiment, the life model of the facility calculated through the life model analyzer may be based on a Weibull distribution.

According to an embodiment, the learning model may include at least one of a deep learning model and a machine learning model.

The health calculation unit according to an embodiment, a real-time expected value calculation unit for calculating a real-time expected value through the measurement data and inspection data of the facility, a score calculation unit for obtaining score data based on the real-time expected value, and the score data Based on the , a facility health calculator for calculating the health of the facility may be included.

According to an embodiment, the facility reliability calculator may calculate at least one of a failure probability of the facility and a remaining life of the facility based on the facility reliability.

The embodiments of the present invention disclosed in the present specification and drawings are merely provided for specific examples in order to easily explain the technical contents of the present invention and help the understanding of the present invention, and are not intended to limit the scope of the present invention. That is, it will be apparent to those of ordinary skill in the art to which the present invention pertains that other modifications can be implemented based on the technical idea of the present invention. In addition, each of the above embodiments may be operated in combination with each other as needed.

In addition, the method for controlling the facility reliability calculation system 101 according to the present invention may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium.

As such, various embodiments of the present invention may be embodied as computer readable code on a computer readable recording medium in a particular aspect. A computer readable recording medium is any data storage device capable of storing data that can be read by a computer system. Examples of computer readable recording media include read only memory (ROM), random access memory (RAM), and compact disk-read only memory (CD-ROM). ), magnetic tapes, floppy disks, optical data storage devices, and carrier waves (such as data transmission over the Internet). The computer readable recording medium may also be distributed over network-connected computer systems, so that the computer readable code is stored and executed in a distributed manner. In addition, functional programs, codes, and code segments for achieving various embodiments of the present invention may be easily interpreted by programmers skilled in the field to which the present invention is applied.

In addition, it will be appreciated that the apparatus and method according to various embodiments of the present invention can be realized in the form of hardware, software, or a combination of hardware and software. Such software may include, for example, a volatile or non-volatile storage device such as a ROM, or a memory such as, for example, RAM, a memory chip, device or integrated circuit, whether erasable or rewritable, or For example, the storage medium may be stored in an optically or magnetically recordable storage medium such as a compact disk (CD), DVD, magnetic disk or magnetic tape, and at the same time, a machine (eg, computer) readable storage medium. The method according to various embodiments of the present invention may be implemented by a computer or portable terminal including a control module and a memory, and the memory is configured to store a program or programs including instructions implementing embodiments of the present invention It can be seen that this is an example of a machine-readable storage medium suitable for

Accordingly, the present invention includes a program including code for implementing the apparatus or method described in the claims of the present specification, and a machine (computer, etc.) readable storage medium storing such a program. Also, such a program may be transmitted electronically through any medium such as a communication signal transmitted through a wired or wireless connection, and the present invention suitably includes the equivalent thereof.

The embodiments of the present invention disclosed in the present specification and drawings are merely provided for specific examples to easily explain the technical content of the present invention and help the understanding of the present invention, and are not intended to limit the scope of the present invention. In addition, the above-described embodiments according to the present invention are merely exemplary, and it will be understood by those skilled in the art that various modifications and equivalent ranges of embodiments are possible therefrom. Accordingly, the true technical protection scope of the present invention should be defined by the following claims.

Claims (10)

In the facility reliability calculation method reflecting facility soundness,
acquiring raw data using at least one sensor;
acquiring score data corresponding to an evaluation item corresponding to a failure probability or an average lifespan through a deep learning model based on the raw data;
acquiring the health of the current state of the equipment through a machine learning model based on the score data;
obtaining a reliability correction constant based on the soundness; and
correcting the equipment reliability based on the reliability correction constant;
Comprising an operation of calculating at least one of a failure probability and an average lifespan of the equipment based on the corrected equipment reliability,
The score data is
Obtained based on the raw data, the residual value of the expected value for each evaluation item, and a threshold and a limit value determined for each evaluation item,
The health is
Calculates the real-time expected value by receiving the equipment measurement and inspection data, calculates the score for each evaluation item based on the real-time measured value and the real-time expected value, and calculates the final equipment soundness by summing the obtained scores,
Calculate the life model of the equipment based on the type of equipment and the failure history data, and calculate the equipment reliability based on at least a part of the life model and the health,
The lifetime model is characterized in that it can be based on the Weibull distribution (Weibull distribution), reliability calculation method. delete The method according to claim 1,
The reliability correction constant is a reciprocal of the soundness, the reliability calculation method. The method according to claim 1,
The deep learning model is a GRU-VAE (gated recurrent unit-variational auto encoder), a reliability calculation method. The method according to claim 1,
The machine learning model comprises an algorithm comprising a plurality of decision trees (decision tree), reliability calculation method. In the facility reliability calculation system reflecting facility health,
a life model analysis unit for calculating a life model of the equipment based on the type and failure history data of the equipment;
an expected value calculation model learning unit for determining the current state of the facility through a learning model;
a health calculator for acquiring the health of the equipment based on the current state of the equipment; and
a facility reliability calculation unit configured to calculate the facility reliability based on at least a part of the life model calculated through the life model analysis unit and the soundness obtained through the health condition calculation unit;
The life model of the facility calculated through the life model analysis unit is based on a Weibull distribution,
The health calculator includes:
A real-time expected value calculation unit that calculates a real-time expected value through the measurement data and inspection data of the facility;
A score calculation unit for obtaining score data corresponding to an evaluation item corresponding to a failure probability or an average lifespan based on the real-time expected value, and
Based on the score data, comprising a facility health calculation unit for calculating the health of the facility,
The facility reliability calculation unit calculates at least one of a failure probability of the facility and the remaining life of the facility based on the facility reliability,
The score data is
Obtained based on raw data, the residual value of the expected value for each evaluation item, and the threshold and limit value determined for each evaluation item,
The health is
Computing the real-time expected value by receiving the equipment measurement and inspection data, calculating the score for each evaluation item based on the real-time measured value and the real-time expected value, and calculating the final equipment soundness by summing the obtained scores , A facility reliability calculation system comprising a facility reliability calculation unit. delete 7. The method of claim 6,
Wherein the learning model includes at least one of a deep learning model and a machine learning model. delete delete

Images