TWI646476B - Fault risk index estimation device and fault risk index estimation method - Google Patents

Fault risk index estimation device and fault risk index estimation method Download PDF

Info

Publication number
TWI646476B
TWI646476B TW106114251A TW106114251A TWI646476B TW I646476 B TWI646476 B TW I646476B TW 106114251 A TW106114251 A TW 106114251A TW 106114251 A TW106114251 A TW 106114251A TW I646476 B TWI646476 B TW I646476B
Authority
TW
Taiwan
Prior art keywords
failure
information
unit
risk index
previous
Prior art date
Application number
TW106114251A
Other languages
Chinese (zh)
Other versions
TW201822080A (en
Inventor
藤野友也
Original Assignee
三菱電機股份有限公司
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
Application filed by 三菱電機股份有限公司 filed Critical 三菱電機股份有限公司
Publication of TW201822080A publication Critical patent/TW201822080A/en
Application granted granted Critical
Publication of TWI646476B publication Critical patent/TWI646476B/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0635—Risk analysis of enterprise or organisation activities
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0637—Strategic management or analysis, e.g. setting a goal or target of an organisation; Planning actions based on goals; Analysis or evaluation of effectiveness of goals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
    • G06Q10/063—Operations research, analysis or management
    • G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
    • G06Q10/06395—Quality analysis or management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/04—Manufacturing
    • Y—GENERAL 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
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • General Business, Economics & Management (AREA)
  • Theoretical Computer Science (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • Tourism & Hospitality (AREA)
  • Physics & Mathematics (AREA)
  • Development Economics (AREA)
  • Educational Administration (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Manufacturing & Machinery (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

模型式構築部(4a)，基於從FMEA結果DB(2)讀出的FMEA結果和統計評估用資訊A，構築表示依據統計分布的故障風險指標之推移的模型式。參數推定部(4b)，基於從FMEA結果DB(2)和既定作業間隔DB(3)讀出的資訊，統計推定由模型式算出的保養作業間隔和既定的保養作業間隔之差分為最小的前記模型式的參數值。 The model construction part (4a) constructs a model expression representing the transition of the failure risk index according to the statistical distribution based on the FMEA result read from the FMEA result DB (2) and the information A for statistical evaluation. The parameter estimation unit (4b), based on the information read from the FMEA result DB (2) and the predetermined operation interval DB (3), statistically estimates that the difference between the maintenance operation interval calculated by the model formula and the predetermined maintenance operation interval is the smallest Model-style parameter values.

Description

故障風險指標推定裝置以及故障風險指標推定方法 Fault risk index estimation device and fault risk index estimation method

本發明係關於推定設備發生故障之風險的指標的故障風險指標推定裝置及故障風險指標推定方法。 The present invention relates to a device and a method for estimating a risk index of a failure risk index for estimating an index of the risk of a device failure.

故障模式影響解析(以下記載為FMEA)為解析手法，得出就各個故障相關項目將故障對於對象設備之影響標以層級之結果。例如，故障相關項目有設備發生故障的頻度、故障對於設備之影響的大小等。另外，當尚未累積足夠的設備運轉資料，使得難以從運轉資料評估設備之保養或修理的必要性時，在評估的時候可以參照FMEA的解析結果。 Failure mode effect analysis (hereinafter referred to as FMEA) is an analysis method, and the result of labeling the effect of the failure on the target device for each failure-related item is graded. For example, failure-related items include the frequency of equipment failures and the magnitude of the impact of the failures on the equipment. In addition, when sufficient equipment operation data has not been accumulated, making it difficult to evaluate the necessity of equipment maintenance or repair from the operation data, the analysis results of FMEA can be referred to during the evaluation.

例如，專利文獻1中記載運轉率預測裝置，其利用FMEA，對於預測對象機械系統的運轉率進行預測。上記運轉率預測裝置，具有規定與故障率相關的複數評估項目的層級和故障係數的對應關係的對應，參照此對應，確定預測對象機械系統的構成要素的評估項目之層級所對應的故障係數。上記運轉率預測裝置，從已確定的故障係數推定構成要素的故障率，就複數構成要素中每一者基於推定的故障率對於預測對象機械系統的運轉率進行預測。再者，上記對應係經過調整以使其符合類似機械系統的運轉率的實績。 For example, Patent Document 1 describes an operation rate prediction device that uses FMEA to predict the operation rate of a mechanical system to be predicted. The operation rate prediction device described above has a correspondence that defines the correspondence between the levels of the complex evaluation items related to the failure rate and the failure coefficients, and refers to this correspondence to determine the failure coefficients corresponding to the levels of the evaluation items of the constituent elements of the predicted mechanical system. The above-described operation rate prediction device estimates the failure rate of the constituent element from the determined failure factor, and predicts the operation rate of the mechanical system to be predicted based on the estimated failure rate for each of the plural constituent elements. In addition, the above correspondences are adjusted to match the actual performance of similar mechanical systems.

另一方面，在想要正確評估對象設備發生故障的 風險的階段中，常有該設備的保養管理已經在實施中，已對應於該設備之狀況設定了保養作業間隔的情況。因此，過去已提出一種技術，基於對象設備的各個構成零件的故障要因，就各個構成零件算出期望的保養作業間隔。 On the other hand, when you want to correctly assess the failure of the target device In the risk stage, the maintenance management of the equipment is often implemented, and the maintenance operation interval has been set according to the status of the equipment. Therefore, a technique has been proposed in the past to calculate a desired maintenance work interval for each component based on the cause of failure of each component of the target device.

例如，專利文獻2記載的可動管理裝置中，考慮到處於停止狀態的設備的構成零件的故障要因和故障的影響度，選定該構成零件的零件等級係數，考慮該構成零件的預測壽命以選定該構成零件的劣化等級係數。上記可動管理裝置，基於已選定的零件等級係數和劣化等級係數，算出構成零件期望的檢查週期，基於已算出的檢查週期，決定該構成零件的有效保全週期(亦即保養作業間隔)。 For example, in the movable management device described in Patent Document 2, considering the cause of failure of the component parts of the equipment in the stopped state and the degree of influence of the fault, the component rank coefficient of the component part is selected, and the predicted life of the component part is considered to select the Deterioration level coefficient of component parts. The movable management device described above calculates the expected inspection cycle of a component based on the selected component level coefficient and degradation level coefficient, and determines the effective maintenance cycle (that is, maintenance work interval) of the component based on the calculated inspection cycle.

先行技術文獻 Advanced technical literature

專利文獻： Patent Literature:

專利文獻1：日本特開2016-126728號公報 Patent Literature 1: Japanese Patent Laid-Open No. 2016-126728

專利文獻2：日本特開2004-252549號公報 Patent Document 2: Japanese Patent Laid-Open No. 2004-252549

專利文獻1記載的運轉率預測裝置中，為了取得上記對應，必須要有能夠精確再現與預測對象機械系統類似的機械系統的運轉率之實績的實績資料。 In the operation rate prediction device described in Patent Document 1, in order to obtain the above correspondence, it is necessary to have actual performance data that can accurately reproduce the actual performance of the operation rate of a mechanical system similar to the prediction target mechanical system.

另外，上記運轉率預測裝置中，只考慮到評估項目的彼此相異的層級間之故障率的大小關係。例如，基於發生故障的頻度層級低的故障的故障率低於發生故障的頻度層級高的故障的故障率的限制條件，預測運轉率。因此，無法適當評估實際 的機械系統中的故障率的變動。 In addition, in the above operation rate prediction apparatus, only the magnitude relationship of the failure rate between different levels of evaluation items is considered. For example, the operation rate is predicted based on the limit condition that the failure rate of a failure with a low frequency of failure levels is lower than the failure rate of a failure with a high frequency of failure occurrences. Therefore, it is impossible to properly evaluate the actual The change in the failure rate in the mechanical system.

專利文獻2記載的可動管理裝置中，係以構成零件的期望檢查週期為現行檢查週期的定數倍作為前提，決定有效保全週期，但若故障的風險相對於經過時間未呈現線性推移的話，則此前提不成立。 The movable management device described in Patent Document 2 assumes that the expected inspection cycle of component parts is a fixed multiple of the current inspection cycle, and the effective maintenance cycle is determined. This premise is not established.

由於實際設備中發生故障的風險一般係呈現非線性推移，所以無法將上記可動管理裝置直接適用於實際設備。 Since the risk of failure in actual equipment generally exhibits a non-linear transition, the movable management device described above cannot be directly applied to actual equipment.

本發明係為了解決上記課題，其目的在於獲致故障風險指標推定裝置及故障風險指標推定方法，即使在沒有或缺少設備的保養實績資料的情況下，也能夠適當推定設備發生故障的風險之指標。 The present invention is to solve the above problem, and its purpose is to obtain a failure risk index estimation device and a failure risk index estimation method, even if there is no or lack of equipment maintenance performance data, it is possible to appropriately estimate the risk index of equipment failure.

本發明的故障風險指標推定裝置具備模型式構築部及參數推定部。模型式構築部，基於表示構成設備的零件的各檢查項目的FMEA結果之資訊及表示故障風險指標的推定使用的統計分布的資訊，構築表示依據統計分布的故障風險指標之推移的模型式。參數推定部，基於表示構成設備的零件的各檢查項目的FMEA結果之資訊及表示零件的各檢查項目之既定的保養作業間隔的資訊，統計推定由模型式算出的保養作業間隔和既定的保養作業間隔之差分為最小的模型式的參數值。 The failure risk index estimation device of the present invention includes a model construction unit and a parameter estimation unit. The model construction unit constructs a model expression representing the transition of the failure risk index based on the statistical distribution based on the information indicating the FMEA result of each inspection item constituting the parts of the equipment and the information indicating the statistical distribution of the estimated use of the failure risk index. The parameter estimation unit statistically estimates the maintenance operation interval calculated by the model formula and the predetermined maintenance operation based on the information indicating the FMEA result of each inspection item of the component constituting the equipment and the information indicating the predetermined maintenance operation interval of each inspection item of the part The difference between the intervals is the smallest model-type parameter value.

依據本發明，構築表示依據統計分布的故障風險指標的推移之模型式，統計推定模型式的參數值作為故障風險指標的推定值。藉此，即使在沒有或缺少設備的保養實績資料 的情況下，也能夠適當推定設備發生故障的風險之指標。 According to the present invention, a model formula representing the transition of the failure risk index according to the statistical distribution is constructed, and the parameter value of the statistically estimated model formula is used as the estimated value of the failure risk index. With this, even when there is no or lack of equipment maintenance performance data Under the circumstances, it is also possible to properly estimate the risk index of equipment failure.

1、1A‧‧‧故障風險指標推定裝置 1. 1A‧‧‧Estimation device for failure risk index

2‧‧‧FMEA結果DB 2‧‧‧FMEA result DB

3‧‧‧既定作業間隔DB 3‧‧‧Predetermined operation interval DB

4、11‧‧‧指標推定部 4. 11‧‧‧ Indicator Estimation Department

4a‧‧‧模型式構築部 4a‧‧‧Model building department

4b‧‧‧參數推定部 4b‧‧‧Parameter estimation section

5‧‧‧第1記憶部 5‧‧‧ First Memory Department

6‧‧‧設備資訊DB 6‧‧‧Equipment Information DB

7‧‧‧保養實績DB 7‧‧‧ Maintenance performance DB

8‧‧‧故障實績DB 8‧‧‧ Failure performance DB

9‧‧‧限縮部 9‧‧‧Restriction Department

10‧‧‧限縮資料記憶部 10‧‧‧Restricted Data Memory Department

12‧‧‧第2記憶部 12‧‧‧ 2nd Memory Department

13‧‧‧合併部 13‧‧‧ Merger Department

100‧‧‧資料庫 100‧‧‧Database

101‧‧‧DB輸出入介面 101‧‧‧DB I / O interface

102‧‧‧資訊輸入介面 102‧‧‧Information input interface

103‧‧‧資訊輸出介面 103‧‧‧Information output interface

104‧‧‧處理電路 104‧‧‧ processing circuit

105‧‧‧記憶體 105‧‧‧Memory

106‧‧‧處理器 106‧‧‧ processor

第1圖為表示本發明之實施形態1的故障風險指標推定裝置的功能構成的方塊圖。 FIG. 1 is a block diagram showing the functional configuration of the failure risk index estimation device according to Embodiment 1 of the present invention.

第2圖之第2A圖為表示記憶於FMEA結果資料庫(以下記載為DB)的資訊項目之圖。第2B圖為表示記憶於既定作業間隔DB之資訊項目的圖。第2C圖為表示統計評估用資訊的圖。 Figure 2A of Figure 2 is a diagram showing information items stored in the FMEA result database (hereinafter referred to as DB). Fig. 2B is a diagram showing information items stored in a predetermined operation interval DB. Figure 2C is a diagram showing information for statistical evaluation.

第3圖為表示實施形態1中記憶於第1記憶部的資訊項目的圖。 FIG. 3 is a diagram showing information items stored in the first storage unit in the first embodiment.

第4圖之第4A圖表示實現實施形態1的故障風險指標推定裝置之功能的硬體構成的方塊圖。第4B圖表示執行實現實施形態1的故障風險指標推定裝置的功能的軟體之硬體構成的方塊圖。 FIG. 4A and FIG. 4A are block diagrams showing the hardware configuration that realizes the function of the failure risk index estimation device according to the first embodiment. FIG. 4B is a block diagram showing the hardware configuration of the software that executes the function of the device for estimating the risk index of failure according to the first embodiment.

第5圖為表示實施形態1的故障風險指標推定裝置的動作的流程圖。 Fig. 5 is a flowchart showing the operation of the failure risk index estimation device in the first embodiment.

第6圖為表示故障風險指標和中間評估指標之關係的圖。 Figure 6 is a diagram showing the relationship between the failure risk index and the intermediate evaluation index.

第7圖為表示FMEA結果的評估項目和中間評估指標的關聯性的圖。 Figure 7 is a diagram showing the correlation between the evaluation items of the FMEA results and the intermediate evaluation indicators.

第8圖為表示模型式構築部及參數推定部的動作的流程圖。 FIG. 8 is a flowchart showing the operation of the model-based construction unit and the parameter estimation unit.

第9圖為表示將FMEA結果DB和既定作業間隔DB的表格資料合併的結果之圖。 Fig. 9 is a diagram showing the result of merging the table data of the FMEA result DB and the predetermined operation interval DB.

第10圖為表示將中間評估指標的參數分配給FMEA結果的評估項目之結果的圖。 Figure 10 is a diagram showing the results of assigning the parameters of the intermediate evaluation index to the evaluation items of the FMEA results.

第11圖為表示從模型式估計的保養作業間隔的圖。 FIG. 11 is a diagram showing the maintenance work interval estimated from the model formula.

第12圖為表示本發明之實施形態2的故障風險指標推定裝置的功能構成的方塊圖。 FIG. 12 is a block diagram showing the functional configuration of the failure risk index estimation device according to Embodiment 2 of the present invention.

第13圖之第13A圖為表示記憶於設備資訊DB的資訊項目之圖。第13B圖為表示記憶於保養實績DB的資訊項目的圖。第13C圖為表示記憶於故障實績DB的資訊項目的圖。 Fig. 13A of Fig. 13 is a diagram showing information items stored in the device information DB. Fig. 13B is a diagram showing information items stored in the maintenance performance DB. FIG. 13C is a diagram showing the information items stored in the failure record DB.

第14圖之第14A圖為表示記憶於限縮資料記憶部的資訊項目的圖。第14B圖為表示記憶於第2記憶部的資訊項目的圖。第14C圖為表示故障風險指標的推定值的圖。 FIG. 14A of FIG. 14 is a diagram showing information items stored in the limited data storage unit. FIG. 14B is a diagram showing information items stored in the second storage unit. FIG. 14C is a diagram showing the estimated value of the failure risk index.

第15圖為表示實施形態2的故障風險指標推定裝置的動作之流程圖。 Fig. 15 is a flowchart showing the operation of the failure risk index estimation device in the second embodiment.

第16圖為表示實施形態2的限縮部的動作之流程圖。 Fig. 16 is a flowchart showing the operation of the constriction unit in the second embodiment.

第17圖為表示實施形態2中指標推定部的動作的流程圖。 Fig. 17 is a flowchart showing the operation of the index estimation unit in the second embodiment.

第18圖為表示實施形態2的合併部的動作的流程圖。 Fig. 18 is a flowchart showing the operation of the merging unit in the second embodiment.

以下，為了更詳細說明本發明，依據附圖說明用以實施本發明的形態。 In the following, in order to explain the present invention in more detail, a mode for carrying out the present invention will be described based on the drawings.

實施形態1. Embodiment 1.

第1圖為表示本發明之實施形態1的故障風險指標推定裝置的功能構成的方塊圖。故障風險指標推定裝置1為推定設備的故障風險指標的裝置，具備：FMEA結果DB2、既定作業間隔DB3、指標推定部4及第1記憶部5。指標推定部4，基於分別由FMEA結果DB2及既定作業間隔DB3讀取的資訊和統計評估用資訊A，推定設備的故障風險指標。所謂的故障風險指標為，將構成設備的零件發生故障的風險大小予以定量化的資訊。 FIG. 1 is a block diagram showing the functional configuration of the failure risk index estimation device according to Embodiment 1 of the present invention. The failure risk index estimation device 1 is a device for estimating a failure risk index of equipment, and includes an FMEA result DB2, a predetermined operation interval DB3, an index estimation unit 4 and a first memory unit 5. The index estimation unit 4 estimates the equipment failure risk index based on the information read from the FMEA results DB2 and the predetermined operation interval DB3 and the information A for statistical evaluation. The so-called failure risk index is information that quantifies the magnitude of the risk of failure of parts that constitute equipment.

FMEA結果DB2為記憶構成設備的零件的各檢查項目之FMEA結果的DB。FMEA結果DB2中，記憶了例如第2A圖所示的項目的資訊。“零件ID”的項目中設定了零件的識別資訊。“檢查項目ID”的項目中設定了檢查項目的識別資訊。檢查項目中有例如外觀檢查、通電檢查、絕緣檢查、摩擦檢查等。通電檢查和絕緣檢查都是檢查電氣的導通狀態的，因此可稱之為類似的檢查項目。 The FMEA result DB2 is a DB that stores the FMEA result of each inspection item of parts constituting the equipment. In the FMEA result DB2, for example, information about the items shown in FIG. 2A is stored. The identification information of the part is set in the item of "Part ID". The identification information of the inspection item is set in the item of "Exam Item ID". The inspection items include, for example, visual inspection, power-on inspection, insulation inspection, and friction inspection. Both the power-on inspection and the insulation inspection check the electrical conduction state, so it can be called a similar inspection item.

“故障頻度層級”、“影響大小層級”和“檢出可能性層級”為FMEA的評估項目。“故障頻度層級”的項目中設定了關於故障頻度的FMEA的評估層級。“影響大小層級”的項目中設定了關於故障對零件造成之影響的大小的FMEA的評估層級。“檢出可能性層級”的項目中設定了關於故障的檢出容易度的FMEA的評估層級。 "Fault frequency level", "influence level" and "detection possibility level" are FMEA evaluation items. The "Frequency Frequency Level" item sets the FMEA evaluation level for the frequency of failure. In the item of “level of influence size”, the evaluation level of FMEA about the size of the influence of the fault on the part is set. The "detection possibility level" item sets the FMEA evaluation level regarding the ease of detection of the fault.

既定作業間隔DB3為，記憶表示零件的各檢查項目之既定的保養作業間隔之資訊的DB。既定作業間隔DB3中記憶了例如第2B圖所示的項目的資訊。“零件ID”及“檢查項目ID”與第2A圖所示者相同。“既定作業間隔月數”的項目中，設定了零件的各檢查項目之既定的保養作業間隔的月數。 The predetermined operation interval DB3 is a DB that stores information indicating the predetermined maintenance operation interval of each inspection item of a part. The predetermined operation interval DB3 stores, for example, information on items shown in FIG. 2B. "Part ID" and "inspection item ID" are the same as those shown in Fig. 2A. In the item "months of scheduled work interval", the number of months of the scheduled maintenance work interval for each inspection item of the parts is set.

統計評估用資訊A為，表示用於故障風險指標的統計推定的統計分布之資訊。統計評估用資訊A中記憶了例如第2C圖所示的項目之資訊。“統計分布”的項目中，設定了用於故障風險指標的統計推定之統計分布的種類。統計分的種類中有韋伯分布、γ分布、對數常態分布等的理論推導出的分布。 The information A for statistical evaluation is information indicating the statistical distribution used for statistical estimation of failure risk indicators. Information A for statistical evaluation stores information such as items shown in FIG. 2C. In the item of "statistical distribution", the type of statistical distribution used for statistical estimation of the failure risk index is set. The types of statistical scores include theoretically derived distributions such as Weber distribution, gamma distribution, and lognormal distribution.

“容許誤差”的項目中，設定了指標推定部4所推定 的故障風險指標的推定值相對於標準偏差的容許誤差的比率。 In the item of "tolerable error", the estimation by the index estimation unit 4 is set The ratio of the estimated value of the failure risk index to the allowable error of the standard deviation.

“信賴率”的項目中，設定了指標推定部4所推定的故障風險指標的信賴率。評估為故障風險指標的推定值落在容許誤差以內的結果失準的確率P1、以及評估為故障風險指標的推定值超過容許誤差的結果施準的確率P2兩方都在設定於信賴率”的值α以上(P1≧α、P2≧α)。 In the item of “trust rate”, the trust rate of the failure risk index estimated by the index estimating unit 4 is set. The accuracy rate P1 of the result miscalculation that the estimated value of the failure risk index falls within the allowable error and the accuracy rate P2 of the result that the estimated value of the failure risk index exceeds the allowable error are both set at the trust rate Above the value α (P1 ≧ α , P2 ≧ α ).

再者，“容許誤差”和“信賴率”為用於實施形態2中後述的合併處理的資訊，所以在實施形態1中的統計評估用資訊A中亦可沒有上述資訊。 In addition, "tolerable error" and "reliability" are information used in the merge processing described later in Embodiment 2, so the above information may not be included in the information A for statistical evaluation in Embodiment 1.

指標推定部4具備模型式構築部4a及參數推定部4b。 The index estimation unit 4 includes a model construction unit 4a and a parameter estimation unit 4b.

模型式構築部4a，基於從FMEA結果DB2讀取的FMEA結果和統計評估用資訊A，構築表示故障風險指標之推移(時間變化)的模型式。 The model construction unit 4a constructs a model expression indicating the transition (time change) of the failure risk index based on the FMEA result read from the FMEA result DB2 and the information A for statistical evaluation.

參數推定部4b，基於從FMEA結果DB2和既定作業間隔DB3讀取的資訊，統計推定上記模型式的參數值，使得從模型式算出的保養作業間隔和既定保養作業間隔的差分為最小。 The parameter estimation unit 4b statistically estimates the parameter value of the model formula based on the information read from the FMEA result DB2 and the predetermined operation interval DB3 so that the difference between the maintenance operation interval calculated from the model expression and the predetermined maintenance operation interval is minimum.

上記模型式為，將依據統計評估用資訊A所示之統計分布的故障風險指標的推移模型化的結果，此模型式的參數值為第1推定值。 The above model formula is the result of modeling the progress of the failure risk index of the statistical distribution shown in the statistical evaluation information A. The parameter value of this model formula is the first estimated value.

第1記憶部5，記憶作為指標推定部4所推定的參數值之第1推定值。第1記憶部5中記憶了例如第3圖所示的各項目的資訊。“零件ID”及“檢查項目ID”與第2A圖所示者相同，“統計分布”與第2C圖所示者相同。 The first memory unit 5 memorizes the first estimated value as the parameter value estimated by the index estimating unit 4. The first memory section 5 stores information such as items shown in FIG. 3. "Part ID" and "inspection item ID" are the same as those shown in Fig. 2A, and "statistic distribution" is the same as those shown in Fig. 2C.

“統計分布參數”的項目中，設定了規定用於第1推定值 的統計推定的統計分布的參數。統計分布是依據韋伯分布的情況下，設定了規定韋伯分布的累積密度函數的形狀參數γ及尺度參數。 In the item of “statistical distribution parameter”, a parameter specifying the statistical distribution used for the statistical estimation of the first estimated value is set. The statistical distribution is based on the Weber distribution, and the shape parameter γ and the scale parameter that specify the cumulative density function of the Weber distribution are set .

“時間規模係數”、“風險重要性係數”和“安全邊際”為上記的模型式之參數，設定了指標推定部4所推定的參數值。 The "time scale coefficient", "risk importance coefficient" and "safety margin" are the parameters of the model formula described above, and the parameter values estimated by the index estimation unit 4 are set.

“時間規模係數”為，在上記模型式中，關於故障風險增加速度的參數。“風險重要性係數”為，在上記模型式中，關於故障風險的重要性程度的參數。“安全邊際”為，表示依上記模型式中推定的作業間隔從保養作業日時回溯之時間間隔的參數。依照此參數將零件的保養作業提前實施。 The "time scale factor" is, in the above model formula, the parameter about the speed at which the risk of failure increases. The "risk importance coefficient" is a parameter regarding the importance degree of the failure risk in the above model formula. "Safety margin" is a parameter indicating the time interval from the maintenance operation date and time back to the operation interval estimated in the above model formula. According to this parameter, the maintenance work of the parts will be implemented in advance.

FMEA結果DB2及既定作業間隔DB3為第4A圖及第4B圖所示之資料庫100。分別記憶在FMEA結果DB2及既定作業間隔DB3的資訊，透過DB輸出入介面101輸入到指標推定部4。 The FMEA result DB2 and the predetermined operation interval DB3 are the database 100 shown in FIGS. 4A and 4B. The information stored in the FMEA result DB2 and the predetermined operation interval DB3 are respectively stored and input to the index estimation unit 4 through the DB input / output interface 101.

透過資訊輸入介面102，將統計評估用資訊A輸入故障風險指標推定裝置1。透過資訊輸出介面103，從故障風險指標推定裝置1輸出故障風險指標的推定值。 Through the information input interface 102, the information A for statistical evaluation is input to the failure risk index estimation device 1. Through the information output interface 103, the estimated value of the failure risk index is output from the failure risk index estimation device 1.

第1記憶部5，可以設置在具有資料庫100的記憶裝置中，亦可設置在第4A圖所示的處理電路104的內部記憶體。另外，第1記憶部5亦可設置在第4B圖所示的記憶體105。 The first memory unit 5 may be provided in a memory device having the database 100, or may be provided in the internal memory of the processing circuit 104 shown in FIG. 4A. In addition, the first memory unit 5 may be provided in the memory 105 shown in FIG. 4B.

故障風險指標推定裝置1中的模型式構築部4a及參數推定部4b的各功能由處理電路實現。 The functions of the model construction unit 4a and the parameter estimation unit 4b in the failure risk index estimation device 1 are realized by the processing circuit.

亦即，故障風險指標推定裝置1具備處理電路，其基於從 FMEA結果DB2讀取的FMEA結果和統計評估用資訊A，構築表示故障風險指標的推移之模型式，基於從FMEA結果DB2和既定作業間隔DB3讀取的資訊，統計推定模型式的參數值，使得從模型式算出的保養作業間隔和既定的保養作業間隔的差分為最小。 That is, the failure risk index estimation device 1 includes a processing circuit, which is based on FMEA result DB2 reads the FMEA result and statistical evaluation information A, constructs a model formula indicating the progress of the failure risk indicator, based on the information read from the FMEA result DB2 and the predetermined operation interval DB3, statistically estimates the model-type parameter values, making The difference between the maintenance work interval calculated from the model formula and the predetermined maintenance work interval is the smallest.

處理電路，可以為專用的硬體，亦可為執行記憶在記憶體的程式之CPU(Central Processing Unit)。 The processing circuit may be dedicated hardware or a CPU (Central Processing Unit) that executes a program stored in the memory.

處理電路為如第4A圖所示的專用的硬體的情況下，處理電路104為例如單一電路、複合電路、程式化處理器、並列程式化處理器、ASIC(Application Specific Integrated Circuit)、FPGA(Field-Programmable Gate Array)、或者其組合。 When the processing circuit is dedicated hardware as shown in FIG. 4A, the processing circuit 104 is, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA ( Field-Programmable Gate Array), or a combination thereof.

模型式構築部4a及參數推定部4b個別的功能可以用個別的處理電路實現，亦可將其功能整合而由1個處理電路實現。 The individual functions of the model building unit 4a and the parameter estimation unit 4b may be realized by individual processing circuits, or their functions may be integrated and realized by one processing circuit.

處理電路為如第4B圖所示的處理器106的情況下，模型式構築部4a及參數推定部4b的各功能由軟體、韌體、或者由軟體和韌體的組合實現。 When the processing circuit is the processor 106 shown in FIG. 4B, each function of the model construction unit 4a and the parameter estimation unit 4b is realized by software, firmware, or a combination of software and firmware.

軟體或韌體係記載為程式，並記憶在記憶體105中。處理器106，讀取並執行記憶在記憶體105中的程式，藉此實現各部的功能。 The software or firmware system is recorded as a program, and is memorized in the memory 105. The processor 106 reads and executes the program stored in the memory 105, thereby realizing the functions of each part.

亦即，故障風險指標推定裝置1具有記憶體105，用以記憶程式，其由處理器106實行時，執行第5圖所示之步驟ST1、步驟ST2。另外，這些程式使得電腦執行模型式構築部4a及參數推定部4b的程序或方法。 That is, the failure risk index estimation device 1 has a memory 105 for memorizing programs, and when executed by the processor 106, it executes steps ST1 and ST2 shown in FIG. In addition, these programs cause the computer to execute the program or method of the model construction unit 4a and the parameter estimation unit 4b.

記憶體105為例如RAM(Random Access Memory)、ROM(Read Only Memory)、快閃記憶體、EPROM(Erasable Programmable Read Only Memory)、EEPROM(Electrically-EPROM)等的非揮發性或揮發性的半導體記憶體、或磁碟、軟碟(flexible disk)、光磁碟、CD(compact disk，光碟)、迷你光碟(minidisc)、DVD(Digital Versatile Disc)等。 The memory 105 is, for example, RAM (Random Access Memory), ROM (Read Only Memory), flash memory, EPROM (Erasable Programmable Read Only Memory), EEPROM (Electrically-EPROM) and other non-volatile or volatile semiconductor memory, or disk, flexible disk (flexible disk), optical disk, CD (compact disk), mini-disc (minidisc), DVD (Digital Versatile Disc), etc.

再者，模型式構築部4a及參數推定部4b的各功能，亦可一部分用專用的硬體實現，一部分用軟體或者韌體實現。例如，模型式構築部4a，用作為專用的硬體的處理電路實現其功能，參數推定部4b，則由處理器106讀取並執行記憶在記憶體105的程式來實現其功能亦可。 In addition, the functions of the model building unit 4a and the parameter estimation unit 4b may be partially realized by dedicated hardware and partially realized by software or firmware. For example, the model construction unit 4a may realize its function by a dedicated hardware processing circuit, and the parameter estimation unit 4b may read and execute a program stored in the memory 105 by the processor 106 to realize its function.

像這樣，處理電路，能夠藉由硬體、軟體、韌體或其組合，實現上記之各個功能。 In this way, the processing circuit can realize the above functions by hardware, software, firmware, or a combination thereof.

繼之說明動作。 Followed by the description of the action.

第5圖為表示實施形態1的故障風險指標推定裝置的動作的流程圖，表示從求取故障風險指標的推定值到記憶在第1記憶部5為止的一連串處理。第6圖為表示故障風險指標R(t)和中間評估指標S,W,M之關係的圖，表示保養作業後的經過時間中的故障風險指標R(t)之推移。第7圖為表示FMEA結果的評估項目C,E,D和中間評估指標S,W,M的關聯性的圖。 FIG. 5 is a flowchart showing the operation of the failure risk index estimation device according to the first embodiment, and shows a series of processes from obtaining the estimated value of the failure risk index to storing in the first storage unit 5. Figure 6 is a diagram showing the relationship between the failure risk index R (t) and the intermediate evaluation indexes S, W, and M, and shows the transition of the failure risk index R (t) in the elapsed time after the maintenance operation. Fig. 7 is a diagram showing the correlation between the evaluation items C, E, and D of the FMEA results and the intermediate evaluation indexes S, W, and M.

以下，參照第6圖及第7圖，順著第5圖說明故障風險指標推定裝置1的動作。 Hereinafter, referring to FIGS. 6 and 7, the operation of the failure risk index estimation device 1 will be described along FIG. 5.

首先，指標推定部4，基於從FMEA結果DB2讀取的FMEA結果和統計評估用資訊A，構築故障風險指標的模型式(步驟ST1)。 First, the index estimation unit 4 builds a model formula of the failure risk index based on the FMEA result read from the FMEA result DB2 and the information A for statistical evaluation (step ST1).

故障風險指標R(t)，如下記式(1)所示，按照統計分布f(t) 而推移(時間變化)。在下記式(1)中，時間規模係數S為用以調整R(t)的時間方向之變化的係數，為關於故障風險增加速度的參數。風險重要性係數W為用以調整R(t)的大小的係數，為關於故障風險的重要性程度的參數。 The failure risk index R (t), as shown in the following formula (1), according to the statistical distribution f (t) And the passage (time change). In the following formula (1), the time scale coefficient S is a coefficient for adjusting the change in the time direction of R (t), and is a parameter regarding the rate of increase in the risk of failure. The risk importance coefficient W is a coefficient for adjusting the magnitude of R (t), and is a parameter regarding the importance degree of the failure risk.

R(t)=W×f(S×t)‧‧‧(1) R (t) = W × f (S × t) ‧‧‧‧ (1)

指標推定部4之R(t)的統計推定中，如第6圖所示，以R(t)超過作為容許上限閾值的“1.0”之作業間隔作為下一次進行保養作業的適當間隔TE時，求出為了安全而將保養作業提前進行的作業間隔。例如、指標推定部4，設定安全邊際M，由下記式(2)算出由模型式估計的作業間隔T。安全邊際M為從適當間隔TE的經過時間回溯的時間間隔。 In the statistical estimation of R (t) of the index estimation unit 4, as shown in FIG. 6, when the operation interval at which R (t) exceeds "1.0" as the allowable upper limit threshold is used as the appropriate interval T E for the next maintenance operation , To find the work interval that advances the maintenance work for safety. For example, the index estimation unit 4 sets the safety margin M, and calculates the working interval T estimated by the model formula from the following formula (2). M margin of safety for the time elapsed from the time interval back to an appropriate interval T E.

T=TE-M‧‧‧(2) T = T E -M‧‧‧ (2)

指標推定部4之R(t)的統計推定中，時間規模係數S、風險重要性係數W、及安全邊際M，是在得到故障風險指標的推定值之前被指定了數值的參數。在此將其稱之為中間評估指標。 In the statistical estimation of R (t) of the index estimation unit 4, the time scale coefficient S, the risk importance coefficient W, and the safety margin M are parameters that are assigned numerical values before the estimated value of the failure risk index is obtained. This is called an intermediate evaluation index here.

統計分布f(t)按照韋伯分布的情況下，統計分布f(t)可以用下記式(3)所示的韋伯分布的累積密度函數表示。在此，形狀參數γ和尺度參數為韋伯分布的累積密度函數中的統計分布參數。 When the statistical distribution f (t) follows the Weber distribution, the statistical distribution f (t) can be expressed by the cumulative density function of the Weber distribution shown in the following formula (3). Here, the shape parameter γ and the scale parameter It is the statistical distribution parameter in the cumulative density function of the Weber distribution.

另外，上記式(1)，可以使用形狀參數γ和尺度參數，以下記式(4)表示。下記式(4)中，尺度參數只有出現作為時間規模係數S的積，作為=1時僅推定S的參數。 In addition, in equation (1) above, shape parameter γ and scale parameter can be used , Represented by the following formula (4). In the following formula (4), the scale parameter Only the product of the time scale factor S appears as When = 1, only the parameters of S are estimated.

f(t)=1-exp(-(t/)γ)‧‧‧(3) f (t) = 1-exp (-(t / ) γ ) ‧‧‧ (3)

R(t)=W×(1-exp(-(S×t)γ))‧‧‧(4) R (t) = W × (1-exp (-(S × t) γ )) ‧‧‧‧ (4)

經過時間t為適當間隔TE時，故障風險指標R(TE)=1.0，所以，適當間隔TE可以用上記式(4)，以下記式(5)表示。下記式(5)中，ln表示自然對數。 T is the elapsed time interval T E proper, failure risk index R (T E) = 1.0, therefore, can be used for an appropriate interval T E note of formula (4), referred to the following formula (5). In the following formula (5), ln represents the natural logarithm.

TE=(1/S)×(-ln(1-(1/W)))1/γ‧‧‧(5) T E = (1 / S) × (-ln (1- (1 / W))) 1 / γ ‧‧‧ (5)

如第7圖所示，FMEA結果的評估項目中，有故障頻度層級、影響大小層級、檢出可能性層級，以下，以C為故障頻度層級、以E為影響大小層級、以D為檢出可能性層級。FMEA結果的評估項目和中間評估指標之間，具有如第7圖所示的關聯性。 As shown in Figure 7, the FMEA result evaluation items include failure frequency level, impact size level, and detection possibility level. Below, C is the failure frequency level, E is the impact size level, and D is the detection level. Possibility level. There is a correlation as shown in Figure 7 between the evaluation items of the FMEA results and the intermediate evaluation indicators.

故障頻度層級C，可以作為評估對象的零件發生故障的頻度之指標，所以和關於故障風險的增加速度的時間規模係數S具有關聯性。 The failure frequency level C can be used as an indicator of the frequency of occurrence of the failure of the part to be evaluated, so it is related to the time scale factor S regarding the rate of increase of the failure risk.

影響大小層級E，可以作為故障對於評估對象的零件之影響大小之指標，所以和關於故障風險的重要性程度的風險重要性係數W、關於將保養作業提前的程度的安全邊際M兩方都有關聯性。 The impact size level E can be used as an indicator of the impact size of the fault on the parts to be evaluated, so both the risk importance coefficient W about the importance of the fault risk and the safety margin M about the degree of advancement of maintenance operations are both Relevance.

檢出可能性層級D，可以作為評估對象的零件發生的故障之檢出容易度之指標，所以和關於保養作業提早的程度之安全邊際M有關聯性。 The detection possibility level D can be used as an indicator of the ease of detection of the failure of the parts to be evaluated, so it is related to the safety margin M regarding the degree of early maintenance work.

指標推定部4，統計推定模型式的參數S,W,M,γ,值，使得從模型式估計的保養作業間隔T和既定保養作業間隔TS之差分為最小(步驟ST2)。尺度參數固定為“1”，所以，指標推定部4就各零件及各檢查項目算出模型式的參數S,W,M,γ的值，並使其與零件及檢查項目對應地記憶在第1記憶部5。 Index estimation unit 4, statistically estimate model-type parameters S, W, M, γ, Value so that the difference between the maintenance operation interval T estimated from the model expression and the predetermined maintenance operation interval T S is minimized (step ST2). Scale parameter Since it is fixed at "1", the index estimation unit 4 calculates the values of the model parameters S, W, M, and γ for each part and each inspection item, and stores it in the first memory unit corresponding to the part and the inspection item 5.

繼之，說明指標推定部4的詳細動作。 Next, the detailed operation of the index estimation unit 4 will be described.

第8圖為表示模型式構築部及參數推定部的動作的流程圖。以下，假設統計分布為韋伯分布的情況。 FIG. 8 is a flowchart showing the operation of the model-based construction unit and the parameter estimation unit. In the following, it is assumed that the statistical distribution is the Weber distribution.

模型式構築部4a，將FMEA結果DB2和既定作業間隔DB3的各表格資料合併(步驟ST1a)。表格資料為，如第2A圖及第2B圖所示的由各項目的資訊構成的資料，表格資料當中，零件ID、檢查項目ID及接下來橫向排列的資訊為記錄資料。 The model building unit 4a merges the table data of the FMEA result DB2 and the predetermined work interval DB3 (step ST1a). The table data is, as shown in FIGS. 2A and 2B, data composed of items of information. Among the table data, the part ID, inspection item ID, and the next horizontally arranged information are recorded data.

例如，第2B圖所示的既定作業間隔DB3的資訊，為由零件ID及檢查項目ID及接下來的既定作業間隔月數構成表格資料的資訊。 For example, the information of the predetermined operation interval DB3 shown in FIG. 2B is information that constitutes the table data from the part ID and the inspection item ID and the number of months of the next predetermined operation interval.

模型式構築部4a，從既定作業間隔DB3讀取表格資料，基於此表格資料中的零件ID及檢查項目ID，檢索具有零件ID及檢查項目ID的FMEA結果DB2的表格資料。 The model construction part 4a reads the table data from the predetermined work interval DB3, and retrieves the table data of the FMEA result DB2 having the part ID and the inspection item ID based on the part ID and the inspection item ID in the table data.

模型式構築部4a，執行合併，亦即將此檢索所特定的表格資料的FMEA結果與從既定作業間隔DB3讀取的表格資料組合。 The model construction unit 4a performs the merge, that is, combines the FMEA result of the specific table data retrieved with the table data read from the predetermined operation interval DB3.

模型式構築部4a，對於既定作業間隔DB3中的所有的表格資料執行上記處理，產生既定的保養作業間隔和FMEA結果組合而成的資訊。此組合資訊如第9圖所示。第9圖中，用作業間隔的月數表示既定的保養作業間隔TS。 The model construction part 4a performs the above-mentioned processing on all the table data in the predetermined operation interval DB3, and generates information combining the predetermined maintenance operation interval and the FMEA result. This combined information is shown in Figure 9. In FIG. 9, the predetermined maintenance work interval T S is represented by the number of months of the work interval.

FMEA結果DB2的表格資料當中，對於不具有與既定作業間隔DB3的表格資料相同的零件ID及檢查項目ID的表格資料，直接將其追加到第9圖所示的組合資訊。 Among the FMEA result DB2 table data, for the table data that does not have the same part ID and inspection item ID as the predetermined operation interval DB3 table data, it is directly added to the combination information shown in FIG. 9.

模型式構築部4a，就各零件ID及各檢查項目ID，對應FMEA結果的評估項目及評估層級，分配與此FMEA結果關聯的中間評估指標的參數(步驟ST2a)。 The model building part 4a assigns the parameters of the intermediate evaluation index associated with the FMEA result for each part ID and each inspection item ID, corresponding to the evaluation item and evaluation level of the FMEA result (step ST2a).

模型式構築部4a，上記組合資訊的FMEA結果中，對於FMEA的故障相關評估項目和此評估項目的評估層級相同者，分配共通的中間評估指標之參數，以構築模型式。 The model construction part 4a, in the FMEA result of the above-mentioned combination information, assigns the parameters of the common intermediate assessment index to the model of the FMEA failure-related assessment item and the assessment level of this assessment item at the same level.

在零件ID及檢查項目ID相異的資料間的上記模型式中，也在共通的參數設定相同值。藉此，能夠在相異的零件或檢查項目中評估共通作用的故障之影響。 In the above model formula between data with different part IDs and inspection item IDs, the same value is also set for common parameters. In this way, it is possible to evaluate the effects of common failures on different parts or inspection items.

第10圖為表示將中間評估指標的參數分配給FMEA結果的評估項目之結果的圖。例如，在第9圖中，作為對應於(零件ID，檢查項目ID)=(EQ001,MT001)的FMEA結果之評估項目的故障頻度層級C，評估層級為“2”。第9圖所示組合資訊的FMEA結果當中，故障頻度層級C的評估層級為“2”的，還有(零件ID，檢查項目ID)=(EQ002,MT002)。模型式構築部4a，如第10圖所示，將共通的參數S2分配給與故障頻度層級C有關聯性的時間規模係數S的參數。 Figure 10 is a diagram showing the results of assigning the parameters of the intermediate evaluation index to the evaluation items of the FMEA results. For example, in Fig. 9, as the failure frequency level C of the evaluation item corresponding to the FMEA result of (part ID, inspection item ID) = (EQ001, MT001), the evaluation level is "2". Among the FMEA results of the combined information shown in Figure 9, the evaluation level of the failure frequency level C is "2", and (Part ID, Inspection Item ID) = (EQ002, MT002). Model constructing unit of formula 4a, as shown in FIG. 10, the common parameters assigned to fault S 2 C with a frequency level parameter time-scale coefficient of correlation of S.

作為對應於(零件ID，檢查項目ID)=(EQ001,MT001)的FMEA結果的評估項目之影響大小層級E，評估層級為“3”。 The impact level level E, which is the evaluation item corresponding to the FMEA result of (part ID, inspection item ID) = (EQ001, MT001), the evaluation level is "3".

上記的組合資訊之FMEA結果中，影響大小層級E的評估層級為“3”的，還有(零件ID，檢查項目ID)=(EQ001,MT002)。模型式構築部4a，如第10圖所示，將共通的參數W3分配給與影響大小層級E有關聯性的風險重要性係數W的參數。 Among the FMEA results of the combination information mentioned above, the evaluation level of the impact size level E is "3", and (Part ID, Inspection Item ID) = (EQ001, MT002). As shown in FIG. 10, the model construction part 4a assigns the common parameter W 3 to the parameter of the risk importance coefficient W related to the influence level E.

(EQ001,MT001)的資料和(EQ001,MT002)的資料中，作為FMEA結果的評估項目的檢出可能性層級D之評估層級都是“3”。因此，模型式構築部4a，如第10圖所示，將共通的參數M3,3分配給與檢出可能性層級D有關聯性的安全邊際M的參數。 In the data of (EQ001, MT001) and (EQ001, MT002), the evaluation level of the detection possibility level D as the evaluation item of the FMEA result is "3". Therefore, as shown in FIG. 10, the model construction part 4a assigns the common parameters M 3, 3 to the parameters of the safety margin M related to the detection possibility level D.

模型式構築部4a，當完成了(零件ID，檢查項目ID)=(EQ001,MT001)之資料的中間評估指標之參數的分配時，基於上記式(2)到上記式(5)，構築作業間隔T、故障風險指標R(t)及適當間隔TE以作為故障風險指標的模型式。 The model construction part 4a, when the parameter assignment of the intermediate evaluation index of the data of (Part ID, inspection item ID) = (EQ001, MT001) is completed, based on the above formula (2) to the above formula (5), constructing work The interval T, the failure risk index R (t) and the appropriate interval T E are used as the model formula of the failure risk index.

(EQ001,MT001)的資料之中間評估指標的參數為S2、W3、M3,3，因此其模型式如下記。 (EQ001, MT001) The parameters of the intermediate evaluation indicators are S 2 , W 3 , and M 3,3 , so the model formula is as follows.

T=TE-M3,3 T = T E -M 3,3

R(t)=W3×{1-exp(-(S2×t)γ)} R (t) = W 3 × {1-exp (-(S 2 × t) γ )}

TE=(1/S2)×(-ln(1-(1/W3)))1/γ T E = (1 / S 2 ) × (-ln (1- (1 / W 3 ))) 1 / γ

可以用影響大小層級E的係數因子ME和檢出可能性層級D的係數因子MD的積ME×MD來表現安全邊際M。藉此，安全邊際M為固定的值，能夠減少推定對象的參數的個數。 Level can affect the size of the coefficient factor E M E D levels and the possibility of detecting the coefficient of volume factor M D M E × M D to show safety margin M. With this, the safety margin M is a fixed value, and the number of parameters of the estimation target can be reduced.

繼之，參數推定部4b，基於模型式構築部4a所分配的參數從上記模型式算出作業間隔T，統計推定作業間隔T和既定的作業間隔TS之差分為最小的參數(步驟ST3a)。 Next, the parameter estimating unit 4b calculates the working interval T from the above-mentioned model formula based on the parameters assigned by the model formula constructing unit 4a, and statistically estimates the parameter where the difference between the estimated working interval T and the predetermined working interval T S is the smallest (step ST3a).

由於統計分布f(t)依據韋伯分布，作業間隔T，如第11圖所示，係以依據中間評估指標的參數決定的變數之形式求出。在第11圖中，Zi為以下記式(6)表示的參數。 Since the statistical distribution f (t) is based on the Weber distribution, the operating interval T, as shown in Figure 11, is obtained in the form of variables determined according to the parameters of the intermediate evaluation index. In Fig. 11, Z i is a parameter represented by the following formula referred to (6).

Zi=(-ln(1-(1/Wi)))1/γ‧‧‧(6) Z i = (-ln (1- (1 / W i ))) 1 / γ ‧‧‧ (6)

參數推定部4b，針對上記組合資訊中所有的零件ID及檢查項目ID，推定中間評估指標的參數S,W,M及形狀參數γ，使得作業間隔T和既定的作業間隔TS之差分的平方和為最小。再者，如前述，尺度參數為“1”。 The parameter estimation unit 4b estimates the parameters S, W, M and the shape parameter γ of the intermediate evaluation index for all the part IDs and inspection item IDs in the above combination information so that the difference between the operation interval T and the predetermined operation interval T S is squared The sum is minimal. Furthermore, as mentioned above, the scale parameter Is "1".

這些參數的推定方法可以為例如共軛梯度法。不過，不限 定於共軛梯度法，只要是能夠推定既定的作業間隔TS和誤差為最小的參數的方法即可。 The estimation method of these parameters may be, for example, a conjugate gradient method. However, it is not limited to the conjugate gradient method, as long as it can estimate a predetermined working interval T S and a parameter with the smallest error.

參數推定部4b，將如前述般推定的結果，就各零件ID及各檢查項目ID分類，與零件ID及檢查項目ID對應地記憶在第1記憶部5中(步驟ST4a)。因為統計分布f(t)依據韋伯分布，所以在第3圖所示之“統計分布”的項目設定“韋伯分布”，在“統計分布參數”的項目設定形狀參數γ的值。 The parameter estimation unit 4b classifies the results of the estimation as described above into each component ID and each inspection item ID, and stores them in the first storage unit 5 corresponding to the component ID and the inspection item ID (step ST4a). Since the statistical distribution f (t) is based on the Weber distribution, the “Weber distribution” is set in the item of “Statistical distribution” shown in FIG. 3, and the value of the shape parameter γ is set in the item of “Statistical distribution parameter”.

參數推定部4b，可以將彼此類似的檢查項目間的其中一方的參數值之統計推定所使用的資訊，用於檢查項目間的另一方的參數值的統計推定。例如，通電檢查和絕緣檢查都是檢查電氣的導通狀態的，因此可稱之為類似的檢查項目。因此，參數推定部4b，將通電檢查的參數值的統計推定所使用的資訊，用於絕緣檢查的參數值的統計推定中。藉此，能夠將用於統計推定的資訊再利用，能夠減少推定所需的處理負荷。 The parameter estimation unit 4b may use information used for statistical estimation of the parameter value of one of the inspection items similar to each other for statistical estimation of the parameter value of the other of the inspection items. For example, both the power-on inspection and the insulation inspection check the electrical conduction state, so it can be called a similar inspection item. Therefore, the parameter estimation unit 4b uses the information used for the statistical estimation of the parameter value of the energization inspection for the statistical estimation of the parameter value of the insulation inspection. With this, the information used for statistical estimation can be reused, and the processing load required for estimation can be reduced.

如上述，此實施形態1的故障風險指標推定裝置1中，模型式構築部4a，基於從FMEA結果DB2讀取的FMEA結果和統計評估用資訊A，構築表示按照統計分布的故障風險指標的推移之模型式。參數推定部4b，基於從FMEA結果DB2和既定作業間隔DB3讀取的資訊，統計推定模型式的參數值，使得從模型式算出的保養作業間隔和既定的保養作業間隔的差分為最小。藉由此構成，即使在沒有或缺少設備的保養實績資料的情況下，也能夠適當推定設備發生故障的風險之指標。 As described above, in the failure risk index estimation device 1 of the first embodiment, the model construction unit 4a constructs a transition indicating the failure risk index according to the statistical distribution based on the FMEA result read from the FMEA result DB2 and the information A for statistical evaluation Model. The parameter estimation unit 4b statistically estimates the parameter values of the model formula based on the information read from the FMEA result DB2 and the predetermined operation interval DB3 so that the difference between the maintenance operation interval calculated from the model expression and the predetermined maintenance operation interval is minimized. With this configuration, even if there is no or lack of maintenance performance data of the equipment, the index of the risk of equipment failure can be appropriately estimated.

實施形態1的故障風險指標推定裝置1中，模型式構築部4a，當FMEA的故障相關評估項目和該評估項目的 評估層級相同時，將共通的參數分配給和評估項目關聯的模型式的參數。藉此，能夠在相異的零件或檢查項目中評估共通作用的故障之影響。 In the failure risk index estimation device 1 of the first embodiment, the model construction unit 4a is used to evaluate the failure-related evaluation items of the FMEA and the When the evaluation levels are the same, the common parameters are assigned to the model-type parameters associated with the evaluation project. In this way, it is possible to evaluate the effects of common failures on different parts or inspection items.

另外，實施形態1的故障風險指標推定裝置1中，參數推定部4b，將彼此類似的檢查項目間的其中一方的參數值之統計推定所使用的資訊，用於檢查項目間的另一方的參數值的統計推定。 In addition, in the failure risk index estimation device 1 of the first embodiment, the parameter estimation unit 4b uses information used for statistical estimation of the parameter value of one of the inspection items that are similar to each other to check the parameter of the other side Statistical presumption of value.

藉此，能夠將用於統計推定的資訊再利用，能夠減少推定所需的處理負荷。 With this, the information used for statistical estimation can be reused, and the processing load required for estimation can be reduced.

實施形態2. Embodiment 2.

第12圖為表示本發明之實施形態2的故障風險指標推定裝置1A的功能構成的方塊圖。在第12圖中，與第1圖相同的構成要素係標示以相同符號並省略其說明。故障風險指標推定裝置1A，除了實施形態1所示的構成之外，還包括設備資訊DB6、保養實績DB7、故障實績DB8、限縮部9、限縮資料記憶部10、指標推定部11、第2記憶部12及合併部13。 FIG. 12 is a block diagram showing the functional configuration of the failure risk index estimation device 1A according to Embodiment 2 of the present invention. In FIG. 12, the same constituent elements as those in FIG. 1 are marked with the same symbols and their description is omitted. The failure risk index estimation device 1A includes, in addition to the configuration shown in the first embodiment, equipment information DB6, maintenance performance DB7, failure performance DB8, limitation unit 9, limitation data storage unit 10, index estimation unit 11, and 2Memory section 12 and merge section 13.

設備資訊DB6為，記憶包含設備、構成設備的零件、及各零件的檢查項目的設備資訊之DB。設備資訊DB6中記憶了例如第13A圖所示的項目資訊。“設備ID”的項目中設定了設備的識別資訊。“零件ID”及“檢查項目ID”則與第2A圖所示者相同。 The equipment information DB6 is a DB that stores equipment information including equipment, parts constituting the equipment, and inspection items of each part. The device information DB6 stores, for example, the item information shown in FIG. 13A. The identification information of the device is set in the item of "device ID". "Part ID" and "inspection item ID" are the same as those shown in Fig. 2A.

“保養開始日時”的項目中，設定了零件的各檢查項目的保養契約開始的日時。再者，保養契約中，個別執行的保養作業的實施日時為使用第13B圖的後述保養作業實施日時。 In the item "maintenance start date and time", the date and time when the maintenance contract for each inspection item of the parts is set is set. In addition, in the maintenance contract, the implementation date and time of the maintenance operation performed individually is the implementation date and time of the maintenance operation described later using FIG. 13B.

保養實績DB7為，記憶構成設備的零件的各檢查 項目的保養作業實績資料之DB。保養實績DB7中記憶了如第13B圖所示的項目資訊。“保養實績ID”的項目中，設定了保養作業實績資料的識別資訊。 The maintenance record DB7 is for memorizing each inspection of parts constituting the equipment DB of actual performance data of project maintenance operations. The maintenance performance DB7 stores the item information shown in Figure 13B. In the item of "maintenance performance ID", identification information of maintenance performance data is set.

“設備ID”、“零件ID”及“檢查項目ID”，則與第13A圖與第2A圖所示者相同。“保養作業實施日時”的項目中，設定了各零件的保養作業的實施日時。 "Equipment ID", "Part ID" and "Inspection Item ID" are the same as those shown in Fig. 13A and Fig. 2A. In the item "Maintenance operation date and time", the execution date and time of the maintenance operation of each part is set.

故障實績DB8為，記憶構成設備的各零件的故障實績資料的DB。故障實績DB8中記憶了如第13C圖所示的項目資訊。“故障實績ID”的項目中，設定了故障實績資料的識別資訊。“設備ID”和“零件ID”，則與第13B圖所示者相同。“故障發生日時”的項目中，設定了零件發生故障的日時。“關聯檢查項目ID”的項目中，設定了已發生故障關聯的檢查項目之識別資訊。 The failure performance DB 8 is a DB that stores failure performance data of each component constituting the device. The fault information DB8 stores the item information shown in Figure 13C. In the item of "failure performance ID", identification information of failure performance data is set. "Equipment ID" and "Part ID" are the same as those shown in Figure 13B. In the item "date and time of failure", the date and time of failure of the parts are set. In the item of "associated inspection item ID", the identification information of the inspection item related to the failure has been set.

限縮部9，將設備資訊DB6、保養實績DB7及故障實績DB8中所記憶的資訊，分類為與FMEA結果DB2的各FMEA結果對應的資訊。 The reduction unit 9 classifies the information stored in the equipment information DB6, the maintenance performance DB7, and the failure performance DB8 as information corresponding to the FMEA results of the FMEA result DB2.

例如，限縮部9，就各保養實績資料集計保養作業實施後的無故障之間隔，並將表示集計的無故障之間隔的資訊就各FMEA結果進行分類，並記憶在限縮資料記憶部10中。 For example, the reduction unit 9 aggregates the maintenance-free performance data after the maintenance operation is performed, and classifies the information indicating the aggregated failure-free interval according to each FMEA result, and stores it in the reduction data storage unit 10 in.

限縮資料記憶部10為，記憶由限縮部9所分類的資訊之DB。限縮資料記憶部10中記憶了如第14A圖所示的項目的資訊。 The reduction data storage unit 10 is a DB that stores information classified by the reduction unit 9. The information of the items shown in FIG. 14A is stored in the limited data storage unit 10.

在第14A圖中，“故障頻度層級”、“影響大小層級”、“檢出可能性層級”為FMEA的評估項目，與第2A圖所示者相同。 In Fig. 14A, "fault frequency hierarchy", "impact magnitude hierarchy", and "detection possibility hierarchy" are FMEA evaluation items, which are the same as those shown in Fig. 2A.

“零件ID”及“檢查項目ID”與第2A圖所示者相同。“保 養作業實施日時”與第13B圖所示者相同。 "Part ID" and "inspection item ID" are the same as those shown in Fig. 2A. "Bao The maintenance operation date and time are the same as those shown in Fig. 13B.

“無故障持續月數”的項目中，設定了下列任一者：保養作業後維持無故障直到下次保養作業的月數、保養作業後直到發生故障為止的月數及保養作業後維持無故障直到現在的月數。“故障發生旗標”的項目中，設定了表示零件是否發生故障之值。例如，當零件發生故障時設定為“1”、未發生故障則設定為“0”。 In the item "months without failure", any one of the following is set: the number of months to maintain trouble-free after maintenance operation until the next maintenance operation, the number of months to failure after maintenance operation and the maintenance without failure after maintenance operation The number of months until now. In the item of "Fault Occurrence Flag", a value indicating whether a part has failed is set. For example, it is set to "1" when a part fails, and set to "0" if no failure occurs.

指標推定部11，基於由限縮部9所分類的資訊，推定與保養作業實績資料和故障實績資料的關係對應的故障風險指標的第2推定值。 The index estimation unit 11 estimates the second estimated value of the failure risk index corresponding to the relationship between the maintenance work actual performance data and the failure actual performance data based on the information classified by the restriction unit 9.

例如，指標推定部11，從限縮資料記憶部10讀取與處理對象的FMEA結果對應的資訊，基於所讀取的資訊，就各零件及各檢查項目統計推定故障風險指標。藉由此推定所得到的故障風險指標的推定值，和推定所使用的實資料數一起記憶在第2記憶部12。 For example, the index estimation unit 11 reads the information corresponding to the FMEA result of the processing target from the reduced data storage unit 10, and statistically estimates the failure risk index for each part and each inspection item based on the read information. The estimated value of the failure risk index obtained by the estimation is stored in the second storage unit 12 together with the actual data number used for the estimation.

第2記憶部12，將指標推定部11所推定的故障風險指標之推定值，就各零件及各檢查項目記憶之。第2記憶部12中記憶如第14B圖所示的項目的資訊。第14B圖中，“零件ID”及“檢查項目ID”與第2A圖所示者相同。“統計分布”及“統計分布參數”與第3圖所示者相同。“時間規模係數”、“風險重要性係數”、“安全邊際”為故障風險指標的第2推定值之參數，與第3圖所示者相同。“實資料數”的項目中，設定了指標推定部11進行的故障風險指標的統計推定所使用的資料的數量。 The second memory unit 12 memorizes the estimated value of the failure risk index estimated by the index estimating unit 11 for each part and each inspection item. The second memory section 12 memorizes the information of the items shown in FIG. 14B. In Fig. 14B, "Part ID" and "Inspection Item ID" are the same as those shown in Fig. 2A. "Statistical distribution" and "statistical distribution parameters" are the same as those shown in Figure 3. "Time scale factor", "risk importance factor", and "safety margin" are the parameters of the second estimated value of the failure risk index, and are the same as those shown in Figure 3. In the item of “number of real data”, the number of data used for statistical estimation of the failure risk index by the index estimating unit 11 is set.

合併部13，將參數推定部4b所推定的第1推定值和指標推定部11所推定的第2推定值按比例分配，算出最終 的故障風險指標的推定值。 The merging unit 13 distributes the first estimated value estimated by the parameter estimating unit 4b and the second estimated value estimated by the index estimating unit 11 in proportion to calculate the final The estimated value of the failure risk index.

例如，合併部13，對應於第1推定值的推定中估計的資料數和第2推定值的推定中使用的實資料數，將第1推定值和第2推定值按比例分配，算出最終的故障風險指標的推定值。從合併部13輸出表示故障風險指標的推定值的資訊B。 For example, the merging unit 13 allocates the first estimated value and the second estimated value in proportion to the estimated number of data used in the estimation of the first estimated value and the number of real data used in the estimation of the second estimated value, and calculates the final The estimated value of the failure risk index. Information B indicating the estimated value of the failure risk index is output from the merging unit 13.

表示故障風險指標的推定值的資訊B，由第14C圖所示項目的資訊構成。在第14C圖中，“零件ID”及“檢查項目ID”與第2A圖所示者相同。“統計分布”及“統計分布參數”與第3圖所示者相同。“時間規模係數”、“風險重要性係”、“安全邊際”為故障風險指標的第2推定值的參數，與第3圖所示者相同。 The information B representing the estimated value of the failure risk index is composed of the information of the items shown in Fig. 14C. In Fig. 14C, "Part ID" and "Inspection Item ID" are the same as those shown in Fig. 2A. "Statistical distribution" and "statistical distribution parameters" are the same as those shown in Figure 3. "Time scale factor", "risk importance system", and "safety margin" are the parameters of the second estimated value of the failure risk index, which are the same as those shown in Figure 3.

故障風險指標推定裝置1A中的FMEA結果DB2、既定作業間隔DB3、設備資訊DB6、保養實績DB7及故障實績DB8，為第4A圖及第4B圖所示的資料庫100。分別記憶在FMEA結果DB2及既定作業間隔DB3中的資訊，透過DB輸出入介面101，輸入到指標推定部4。 The FMEA result DB2, the predetermined operation interval DB3, the equipment information DB6, the maintenance performance DB7, and the failure performance DB8 in the failure risk index estimation device 1A are the databases 100 shown in FIGS. 4A and 4B. The information stored in the FMEA result DB2 and the predetermined operation interval DB3, respectively, is input to the index estimation unit 4 through the DB input / output interface 101.

分別記憶在設備資訊DB6、保養實績DB7及故障實績DB8的資訊，透過DB輸出入介面101，被輸入到限縮部9。 The information stored in the equipment information DB6, the maintenance performance DB7, and the failure performance DB8, respectively, are input and output to the interface 101 through the DB, and are input to the reduction unit 9.

統計評估用資訊A，透過資訊輸入介面102，被輸入到故障風險指標推定裝置1A。透過資訊輸出介面103，從合併部13輸出表示最終故障風險指標的推定值的資訊B。 The information A for statistical evaluation is input to the failure risk index estimation device 1A through the information input interface 102. Through the information output interface 103, information B representing the estimated value of the final failure risk index is output from the merging unit 13.

在此設想第1記憶部5、限縮資料記憶部10及第2記憶部12設置在具有資料庫100的記憶裝置中，但也可以設置在第4A圖所示的處理電路104的內部記憶體。另外，第1記憶部5、限縮資料記憶部10及第2記憶部12亦可設置在第4B圖所示 的記憶體105中。 It is assumed here that the first memory unit 5, the limited data memory unit 10, and the second memory unit 12 are provided in a memory device having a database 100, but they may also be provided in the internal memory of the processing circuit 104 shown in FIG. 4A . In addition, the first memory unit 5, the limited data storage unit 10, and the second memory unit 12 may also be provided as shown in FIG. 4B Memory 105.

故障風險指標推定裝置1A中的指標推定部4、限縮部9、指標推定部11及合併部13的各功能，係由處理電路實現。 The functions of the index estimating unit 4, the narrowing unit 9, the index estimating unit 11, and the merging unit 13 in the failure risk index estimating device 1A are realized by the processing circuit.

亦即，故障風險指標推定裝置1A具有用以執行前述各部之功能的處理之處理電路。處理電路可以為專用的硬體，也可以是執行記憶在記憶體中之程式的CPU。 That is, the failure risk index estimation device 1A has a processing circuit for performing the processing of the functions of the aforementioned various parts. The processing circuit can be dedicated hardware or a CPU that executes a program stored in memory.

處理電路為如第4A圖所示的專用的硬體的情況下，處理電路104為例如單一電路、複合電路、程式化處理器、並列程式化處理器、ASIC(Application Specific Integrated Circuit)、FPGA(Field-Programmable Gate Array)、或者其組合。 When the processing circuit is dedicated hardware as shown in FIG. 4A, the processing circuit 104 is, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC (Application Specific Integrated Circuit), an FPGA ( Field-Programmable Gate Array), or a combination thereof.

指標推定部4、限縮部9、指標推定部11及合併部13個別的功能可以用個別的處理電路實現，亦可將其功能整合而由1個處理電路實現。 The individual functions of the index estimating unit 4, the narrowing unit 9, the index estimating unit 11, and the merging unit 13 can be realized by individual processing circuits, or their functions can be integrated and realized by one processing circuit.

處理電路為如第4B圖所示的處理器106的情況下，指標推定部4、限縮部9、指標推定部11及合併部13的各功能由軟體、韌體、或者由軟體和韌體的組合實現。 In the case where the processing circuit is the processor 106 shown in FIG. 4B, the functions of the index estimation unit 4, the reduction unit 9, the index estimation unit 11, and the merging unit 13 are software, firmware, or software and firmware Of the combination.

軟體或韌體係記載為程式，並記憶在記憶體105中。處理器106，讀取並執行記憶在記憶體105中的程式，藉此實現各部的功能。 The software or firmware system is recorded as a program, and is memorized in the memory 105. The processor 106 reads and executes the program stored in the memory 105, thereby realizing the functions of each part.

亦即、故障風險指標推定裝置1A具有記憶體105，用以記憶程式，其由處理器106執行時，執行第15圖所示的步驟ST1b到步驟ST8b為止的處理。 That is, the failure risk index estimation device 1A has a memory 105 for storing a program, and when executed by the processor 106, the processing from step ST1b to step ST8b shown in FIG. 15 is executed.

另外，這些程式使得電腦執行指標推定部4、限縮部9、指標推定部11、及合併部13的程序或方法。 In addition, these programs cause the computer to execute the procedures or methods of the index estimating unit 4, the narrowing unit 9, the index estimating unit 11, and the merging unit 13.

指標推定部4、限縮部9、指標推定部11及合併 部13的各功能，亦可一部分用專用的硬體實現，一部分用軟體或者韌體實現。例如，指標推定部4及限縮部9用作為專用的硬體的處理電路實現其功能，指標推定部11及合併部13，則由處理器106讀取並執行記憶在記憶體105的程式來實現其功能亦可。像這樣，處理電路，能夠藉由硬體、軟體、韌體或其組合，實現上記之各個功能。 Index estimation unit 4, reduction unit 9, index estimation unit 11 and merger Each function of the unit 13 may also be partially realized by dedicated hardware and partially realized by software or firmware. For example, the index estimation unit 4 and the reduction unit 9 use dedicated processing circuits to implement their functions. The index estimation unit 11 and the merge unit 13 are read by the processor 106 and executed by the program stored in the memory 105. Realize its function. In this way, the processing circuit can realize the above functions by hardware, software, firmware, or a combination thereof.

繼之說明動作。 Followed by the description of the action.

第15圖為表示故障風險指標推定裝置1A的動作的流程圖，表示從求出故障風險指標的第1推定值和第2推定值到輸出最終的推定值為止的一連串處理。 FIG. 15 is a flowchart showing the operation of the failure risk index estimation device 1A, and shows a series of processes from obtaining the first estimated value and the second estimated value of the failure risk index to outputting the final estimated value.

限縮部9，將設備資訊DB6、保養實績DB7及故障實績DB8中所記憶的資訊，分類為與各FMEA結果對應的資訊(步驟ST1b)。 The constriction unit 9 classifies the information stored in the equipment information DB6, the maintenance performance DB7, and the failure performance DB8 as information corresponding to the results of each FMEA (step ST1b).

由限縮部9分類的資訊被記憶在限縮資料記憶部10中。 The information classified by the narrowing-down section 9 is stored in the narrowing-down data storage section 10.

指標推定部11，從限縮資料記憶部10讀取與處理對象的FMEA結果對應的資訊，並基於已讀取的資訊，就各零件及各檢查項目統計推定故障風險指標(步驟ST2b)。 The index estimation unit 11 reads the information corresponding to the FMEA result of the processing target from the reduction data storage unit 10, and statistically estimates the failure risk index for each part and each inspection item based on the read information (step ST2b).

指標推定部11，將上記推定所得到的第2推定值和實資料數，與零件ID及檢查項目ID對應地記憶在第2記憶部12中(步驟ST3b)。 The index estimating unit 11 stores the second estimated value and the actual data number obtained by the above estimation in the second storage unit 12 in correspondence with the part ID and the inspection item ID (step ST3b).

尚未將FMEA結果DB2中記憶的所有的FMEA結果予以處理的情況下(步驟ST4b：否)，重複從步驟ST1b起的處理。 When all the FMEA results stored in the FMEA result DB2 have not been processed (step ST4b: No), the processing from step ST1b is repeated.

已將FMEA結果DB2中記憶的所有的FMEA結果都予以處理的情況下(步驟ST4b：是)，移行到步驟ST5b的處理。 When all the FMEA results memorized in the FMEA result DB2 have been processed (step ST4b: Yes), the process moves to step ST5b.

在步驟ST5b中，指標推定部4，和實施形態1一樣，推定作為第1推定值的參數值。 In step ST5b, the index estimating unit 4 estimates the parameter value as the first estimated value, as in the first embodiment.

指標推定部4，將已推定的第1推定值，與零件ID及檢查項目ID對應地記憶在第1記憶部5中(步驟ST6b)。 The index estimating unit 4 stores the estimated first estimated value in the first storage unit 5 in correspondence with the component ID and the inspection item ID (step ST6b).

合併部13，讀取第1記憶部5中記憶的第1推定值、以及第2記憶部12中記憶的第2推定值，將已讀取的第1推定值和第2推定值按比例分配，算出最終的故障風險指標的推定值(步驟ST7b)。之後，合併部13，輸出已算出的表示故障風險指標的推定值的資訊B(步驟ST8b)。 The merging unit 13 reads the first estimated value stored in the first memory unit 5 and the second estimated value stored in the second memory unit 12, and distributes the read first estimated value and the second estimated value in proportion To calculate the final estimated value of the failure risk index (step ST7b). After that, the merging unit 13 outputs the calculated information B indicating the estimated value of the failure risk index (step ST8b).

再者，第15圖中，表示在指標推定部4進行的第1推定值的推定之前，先進行由指標推定部11進行的第2推定值的推定的情況，但並不限定於此。 In addition, FIG. 15 shows a case where the second estimated value is estimated by the index estimating unit 11 before the first estimated value is estimated by the index estimating unit 4, but it is not limited to this.

例如，也可以在指標推定部11進行的第2推定值的推定之前，先進行指標推定部4的第1推定值的推定。另外，亦可將指標推定部4的第1推定值的推定和指標推定部11進行的第2推定值的推定同時並行。 For example, the first estimation value of the index estimation unit 4 may be estimated before the estimation of the second estimation value by the index estimation unit 11. In addition, the estimation of the first estimated value by the index estimation unit 4 and the estimation of the second estimated value by the index estimation unit 11 may be performed in parallel.

繼之，詳細說明限縮部9的動作。 Next, the operation of the constriction unit 9 will be described in detail.

第16圖為表示限縮部9的動作的流程圖，表示從分類設備資訊DB6、保養實績DB7及故障實績DB8中記憶的資訊到記憶在限縮資料記憶部10為止的一連串處理。 FIG. 16 is a flowchart showing the operation of the reduction unit 9, and shows a series of processes from the information stored in the classification device information DB6, the maintenance record DB7, and the failure record DB8 to the reduction data storage unit 10.

首先，限縮部9，將FMEA結果DB2、設備資訊DB6及保養實績DB7的各表格資料合併(步驟ST1c)。 First, the reduction unit 9 merges the table data of the FMEA result DB2, the device information DB6, and the maintenance performance DB7 (step ST1c).

例如，限縮部9，基於從設備資訊DB6讀取的表格資料的設備ID、零件ID及檢查項目ID，檢索具有相同設備ID、零件ID及檢查項目ID的保養實績DB7的表格資料。限縮部9，將設定於此檢索所特定的表格資料的“保養作業實施日時”的 項目中的資訊，合併到從設備資訊DB6讀取的表格資料中。 For example, the narrowing unit 9 searches the table data of the maintenance performance DB 7 having the same device ID, part ID, and inspection item ID based on the device ID, part ID, and inspection item ID of the table data read from the device information DB6. The constriction unit 9 sets the "maintenance operation implementation date and time" of the specified form data searched here The information in the project is merged into the table data read from the device information DB6.

再者，限縮部9，在沒有與設備資訊相同的設備ID、零件ID及檢查項目ID的保養實績資料的情況下，不刪除檢索中使用的設備ID、零件ID及檢查項目ID的記錄資料而讓其留在合併後的表格資料中。 In addition, the reduction unit 9 does not delete the record data of the equipment ID, part ID, and inspection item ID used in the search if there is no maintenance performance data of the same equipment ID, part ID, and inspection item ID as the equipment information. Instead, leave it in the combined form.

繼之，限縮部9，基於合併後的表格資料中的設備ID、零件ID及檢查項目ID，檢索具有相同設備ID、零件ID及檢查項目ID的FMEA結果DB2的表格資料。限縮部9，將此檢索所特定的表格資料中的FMEA結果合併於上記合併後的表格資料。 Next, the restriction unit 9 searches the FMEA result DB2 table data having the same device ID, part ID, and inspection item ID based on the device ID, part ID, and inspection item ID in the merged table data. The constriction unit 9 merges the FMEA results in the form data specified in this search into the merged form data.

限縮部9，對設備資訊DB6中所有的表格資料執行上記處理，產生將設備資訊和保養實績資料就各FMEA結果分類的資訊。 The constriction section 9 performs the above-mentioned processing on all the table data in the equipment information DB6, and generates information that classifies equipment information and maintenance performance data according to the results of each FMEA.

繼之，限縮部9，在如前述般合併的表格資料當中，僅限縮於對應於評估對象的FMEA結果的記錄資料(步驟ST2c)。 Next, the limiter 9 is limited to the record data corresponding to the FMEA result of the evaluation object among the table data merged as described above (step ST2c).

限縮部9，對於上記合併後的表格資料的記錄資料，追加“無故障持續月數”及“故障發生旗標”的項目。 The constriction unit 9 adds items of "non-failure duration months" and "failure occurrence flag" to the recorded data of the merged form data.

步驟ST3c中，限縮部9，基於評估對象的FMEA結果所對應的設備ID及零件ID，從故障實績DB8中進行檢索，找出執行了關於此零件ID所對應的零件的保養作業之後直到下一次保養作業為止的期間中最早發生的故障實績資料。再者，若沒有執行保養作業，則檢索保養開始日時之後直到下一次保養作業為止的期間中，最早發生的故障實績資料。 In step ST3c, the narrowing section 9 searches the failure performance DB 8 based on the equipment ID and part ID corresponding to the FMEA result of the evaluation object, and finds that the maintenance work on the part corresponding to this part ID is performed until Actual performance data of the earliest failures in the period up to one maintenance operation. In addition, if no maintenance operation is performed, the failure performance data that occurred the earliest in the period from the maintenance start date to the next maintenance operation is searched.

限縮部9，基於檢索結果的故障實績資料，算出零件沒有發生故障的無故障持續月數，將已算出的月數設定在“無故障持續月數”的項目中，將作為表示故障已發生的值之“1”設定 在“故障發生旗標”的項目中。 The constriction unit 9 calculates the number of months of failure-free duration based on the failure performance data of the search results, and sets the calculated number of months in the item of "months of failure-free duration" as the indication that the failure has occurred "1" setting of the value In the "Fault Occurrence Flag" item.

另一方面，若上記檢索中沒有找到故障實績，則限縮部9算出值到下一次保養作業日時為止的時間間隔，將該月數設定在“無故障持續月數”的項目中，將作為表示故障未發生的值之“0”設定在“故障發生旗標”的項目中。 On the other hand, if the failure record is not found in the above search, the reduction unit 9 calculates the time interval from the value to the next maintenance operation date, and sets the number of months in the item of "months without failure" as The value "0" indicating that the failure has not occurred is set in the "Fault Occurrence Flag" item.

再者，若尚未決定下一次保養作業的日時，則將到現時點為止的間隔設定在“無故障持續月數”的項目中。 In addition, if the date and time of the next maintenance operation has not yet been decided, the interval up to the present point is set in the item of "months without failure".

限縮部9，對於步驟ST1c中已合併的表格資料中的所有的記錄資料都執行上記處理。藉此，產生將設備資訊、保養實績資料和故障實績資料就各FMEA結果分類後的資訊。 The constriction unit 9 executes the above-mentioned processing for all the record data in the table data merged in step ST1c. In this way, information is generated that classifies equipment information, maintenance performance data, and failure performance data based on the results of each FMEA.

限縮部9，將處理結果的表格資料記憶在限縮資料記憶部10中(步驟ST4c)。如第14A圖所示，上記表格資料為由“故障頻度層級”、“影響大小層級”、“檢出可能性層級”、“設備ID”、“零件ID”、“檢查項目ID”、“保養作業實施日時”、“無故障持續月數”及“故障發生旗標”之項目構成的資料。 The reduction unit 9 stores the table data of the processing result in the reduction data storage unit 10 (step ST4c). As shown in Figure 14A, the above table data is composed of "fault frequency level", "influence size level", "detection possibility level", "equipment ID", "part ID", "inspection item ID", "maintenance Information on the items of "operation time and date", "non-failure duration months" and "fault occurrence flag".

繼之，詳細說明指標推定部11的動作。 Next, the operation of the index estimation unit 11 will be described in detail.

第17圖為表示指標推定部11的動作之流程圖，表示推定第2推定值之後到記憶在第2記憶部12為止的一連串處理。 FIG. 17 is a flowchart showing the operation of the index estimating unit 11, and shows a series of processes until the second estimated value is estimated and stored in the second memory unit 12.

在步驟ST1d中，指標推定部11，從記憶在限縮資料記憶部10的表格資料中，檢索處理對象的零件ID及檢查項目ID相同的記錄資料。指標推定部11，算出在檢索出的記錄資料中，設定於“無故障持續月數”的項目之值為保養作業後的經過時間t(t=1,2,‧‧‧)以下的記錄資料之數量。像這樣算出的記錄資料的數量，為保養作業後的經過月數所對應的零件的累計台數。 In step ST1d, the index estimation unit 11 retrieves the record data having the same part ID and inspection item ID as the processing target from the table data stored in the reduction data storage unit 10. The index estimating unit 11 calculates the value of the item set in the "non-failure continuous months" in the retrieved log data as the log data of the elapsed time t (t = 1,2, ‧‧‧‧) after the maintenance operation The number. The number of recorded data calculated in this way is the cumulative number of parts corresponding to the elapsed months after maintenance work.

指標推定部11，從限縮資料記憶部10中所記憶的表格資料，檢索“故障發生旗標”的項目中設定為“1”的記錄資料。 The index estimation unit 11 retrieves the record data set to “1” in the item of “failure occurrence flag” from the table data stored in the reduction data storage unit 10.

指標推定部11，基於檢索到的記錄資料，算出各零件ID及各檢查項目ID之保養作業後的經過月數所對應的故障件數(步驟ST2d)。 The index estimation unit 11 calculates the number of failures corresponding to the elapsed months after the maintenance operation for each part ID and each inspection item ID based on the retrieved log data (step ST2d).

繼之，指標推定部11算出實績故障率，其係為將上記保養作業後的經過月數對應的故障件數除以上記累計台數所得之值(步驟ST3d)。 Next, the index estimation unit 11 calculates the actual performance failure rate, which is a value obtained by dividing the number of failures corresponding to the elapsed months after the maintenance operation described above by the cumulative number of units (step ST3d).

接著，指標推定部11，特定近似在步驟ST3d算出的實績故障率的推移的統計分布，統計推定依據此統計分布的模型式的參數值(步驟ST4d)。例如，推定出上記式(4)中風險重要性係數W為1時的故障風險指標R(t)所近似的實績故障率的推移之時間規模係數S及形狀參數γ。 Next, the index estimation unit 11 specifies a statistical distribution that approximates the transition of the actual failure rate calculated in step ST3d, and statistically estimates the parameter value of the model formula based on this statistical distribution (step ST4d). For example, the time scale factor S and shape parameter γ of the transition of the actual failure rate approximated by the failure risk index R (t) when the risk importance coefficient W in the above equation (4) is 1, are estimated.

在此考慮參數推定使用共軛梯度法，也可以用推定韋伯分布之參數的已知的方法。 Here, it is considered that the conjugate gradient method is used for parameter estimation, and a known method for estimating the parameters of the Weber distribution may also be used.

風險重要性係數W的值係就各FMEA結果設定。例如，可以設定從使用者接收之值，亦可設定從實績資料求出的故障發生時的損失額×故障率之差分為最小的係數之值。安全邊際M的值，可以設定為從使用者接受之值，亦可設定為M=0。 The value of risk importance coefficient W is set for each FMEA result. For example, the value received from the user may be set, and the value of the coefficient at which the difference between the amount of loss at the time of failure obtained by the actual performance data and the failure rate is the smallest may be set. The value of the safety margin M can be set to a value accepted from the user or M = 0.

再者，在此使用韋伯分布作為統計分布，但也可以用γ分布、對數常態分布等地統計分布，只要是和實績資料的誤差為最小的統計分布即可。 In addition, the Weber distribution is used as the statistical distribution here, but it is also possible to use a geographical distribution such as a gamma distribution and a lognormal distribution, as long as it is the statistical distribution with the smallest error from the actual performance data.

指標推定部11，確認在限縮資料記憶部10所記憶的表格資料的零件ID及檢查項目ID的組合當中，是否有未處理的組合(步驟ST5d)。 The index estimation unit 11 confirms whether there is an unprocessed combination among the combination of the part ID and the inspection item ID of the table data stored in the reduction data storage unit 10 (step ST5d).

限縮資料記憶部10中記憶的表格資料的零件ID及檢查項目ID的組合當中，有未處理的組合的情況下(步驟ST5d：否)，回到步驟ST1d，重複執行前述的處理。 When there is an unprocessed combination among the combination of the part ID and the inspection item ID of the table data stored in the limited data storage unit 10 (step ST5d: No), the process returns to step ST1d and the aforementioned processing is repeatedly executed.

在零件ID及檢查項目ID的所有組合都已處理的情況下(步驟ST5d：是)，指標推定部11，將作為已推定之參數值的第2推定值與零件ID及檢查項目ID對應地記憶在第2記憶部12中(步驟ST6d)。 When all combinations of the part ID and the inspection item ID have been processed (step ST5d: Yes), the index estimation unit 11 memorizes the second estimated value, which is the estimated parameter value, in association with the part ID and the inspection item ID In the second storage unit 12 (step ST6d).

因為統計分布是按照韋伯分布，所以在第14B圖所示的“統計分布”的項目中設定為“韋伯分布”，在“統計分布參數”的項目中設定形狀參數γ的值。 Since the statistical distribution is based on the Weber distribution, the item “Statistical distribution” shown in FIG. 14B is set to “Weber distribution”, and the value of the shape parameter γ is set to the item “Statistical distribution parameter”.

繼之，詳細說明合併部13的動作。 Next, the operation of the merging unit 13 will be described in detail.

第18圖為表示合併部13的動作之流程圖，表示由第1推定值和第2推定值算出最終的故障風險指標之推定值到輸出為止的一連串處理。合併部13，從第2記憶部12讀取處理對象的零件ID及檢查項目ID、其所對應的第2推定值及實資料數(步驟ST1e)。 FIG. 18 is a flowchart showing the operation of the merging unit 13 and shows a series of processes from the first estimated value and the second estimated value to the final estimated value of the failure risk index to the output. The merging unit 13 reads the part ID and inspection item ID of the processing target, the corresponding second estimated value and the actual data number from the second storage unit 12 (step ST1e).

繼之，合併部13，從第1記憶部5讀取處理對象的零件ID及檢查項目ID及其所對應的第1推定值(步驟ST2e)。 Next, the merging unit 13 reads the part ID and inspection item ID of the processing target and their corresponding first estimated values from the first storage unit 5 (step ST2e).

對於第1記憶部5所記憶的第1推定值當中，不具有和從第2記憶部12讀取的第2推定值對應的零件ID及檢查項目ID相同的零件ID及檢查項目ID的記錄資料，直接從第1記憶部5讀取之。 Among the first estimated values stored in the first storage unit 5, there is no record data of the same part ID and inspection item ID as the part ID and inspection item ID corresponding to the second estimated value read from the second storage unit 12 , Read it directly from the first memory section 5.

假設從第2記憶部12讀取的記錄資料中所設定的實資料數為NA、保養作業後的各經過時間的零件之累計台數持續為NA、故障件數為NA×RA(t)。合併部13，在從第1記憶部5讀取的具有相同零件ID及檢查項目ID的記錄資料中，求出作 為以第1推定值的統計推定估計的資料數量的NI(步驟ST3e)。 It is assumed that the number of real data set in the log data read from the second memory unit 12 is N A , the cumulative number of parts of each elapsed time after the maintenance operation continues to be N A , and the number of defective parts is N A × R A (t). The merging unit 13 obtains N I which is the number of data estimated by the statistical estimation of the first estimated value from the recorded data having the same part ID and inspection item ID read from the first storage unit 5 (step ST3e).

例如，合併部13，使用統計評估用資訊A中的容許誤差△及信賴率α，從下記式(7)算出NI。下記式(7)中，z(α)表示標準正規分布的上側100α%。當△=0.1、α=0.99(99%)時，z(α)=2.326、NI=2168。 For example, the merging unit 13 calculates N I from the following equation (7) using the allowable error Δ and the reliability rate α in the information A for statistical evaluation. In the following formula (7), z (α) represents the upper 100α% of the standard normal distribution. When △ = 0.1, α = 0.99 (99%), z (α) = 2.326, N I = 2168.

NI=(((4/△2)+(1/2))×z(α)2)‧‧‧(7) N I = (((4 / △ 2 ) + (1/2)) × z (α) 2 ) ‧‧‧ (7)

假設在從第1記憶部5讀取的具有相同零件ID及檢查項目ID的記錄資料中，保養作業後的各經過時間之零件的累計台數持續為依據上記式(7)算出的NI，故障件數為NI×RI(t)。 Suppose that in the record data having the same part ID and inspection item ID read from the first storage unit 5, the cumulative number of parts of each elapsed time after the maintenance operation continues to be N I calculated according to the above formula (7), The number of faults is N I × R I (t).

合併部13，將表示從第2記憶部12讀取的記錄資料的故障風險指標RA(t)、和表示從第1記憶部5讀取的具有相同零件ID及檢查項目ID的記錄資料的故障風險指標RI(t)按比例分配(步驟ST4e)。 The merging unit 13 combines the failure risk index R A (t) indicating the recorded data read from the second memory unit 12 and the recorded data indicating the same part ID and inspection item ID read from the first memory unit 5 The failure risk index R I (t) is distributed proportionally (step ST4e).

合併部13，以NA+NI為保養作業後的各經過時間的零件之累計台數、以NA×RA(t)+NI×RI(t)為故障件數，重新統計推定故障風險指標RF(t)。推定方法可使用例如共軛梯度法。到此為止的處理惟按比例分配處理。如此一來，合併部13，將作為第1推定值的統計推定所估計的資料數的NI用於其和第2推定值的按比例分配。藉此，能夠適當地將第1推定值和第2推定值按比例分配。 The merging unit 13 takes N A + N I as the cumulative number of parts with each elapsed time after the maintenance operation, and takes N A × R A (t) + N I × R I (t) as the number of failures, and re-counts Estimated failure risk index R F (t). For the estimation method, for example, a conjugate gradient method can be used. The processing up to this point only distributes the processing in proportion. In this way, the merging unit 13 uses N I, which is the number of data estimated by the statistical estimation of the first estimated value, for its proportional distribution with the second estimated value. With this, the first estimated value and the second estimated value can be appropriately distributed in proportion.

合併部13，確認在分別記憶於第1記憶部5及第2記憶部12的資訊當中，是否有尚未按比例分配的零件ID及檢查項目ID的組合(步驟ST5e)。 The merging unit 13 confirms whether there is a combination of the part ID and the inspection item ID that have not been proportionally allocated among the information stored in the first memory unit 5 and the second memory unit 12 (step ST5e).

分別繼易於第1記憶部5及第2記憶部12的資訊當中，有尚未按比例分配的零件ID及檢查項目ID的組合時(步驟 ST5e：否)，對於未處理的組合重複執行從步驟ST1e起的處理。 When there is a combination of the part ID and the inspection item ID that have not been proportionally distributed among the information in the first storage section 5 and the second storage section 12, respectively (step ST5e: No), the processing from step ST1e is repeatedly executed for the unprocessed combination.

所有的零件ID及檢查項目ID的組合都已按比例分配時(步驟ST5e：是)，合併部13，從故障風險指標RF(t)的參數值產生表示最終的故障風險指標的推定值的資訊B並輸出之(步驟ST6e)。 When all combinations of part IDs and inspection item IDs have been allocated in proportion (step ST5e: Yes), the merging unit 13 generates an estimated value indicating the final failure risk index from the parameter value of the failure risk index R F (t) Information B is output (step ST6e).

如上述，實施形態2之故障風險指標推定裝置1A中，限縮部9將分別記憶在、設備資訊DB6、保養實績DB7、及故障實績DB8的資訊分類為對應於各FMEA結果的資訊。指標推定部11，基於限縮部9所分類的資訊，推定出對應於保養作業實績資料和故障實績資料的關係的故障風險指標的推定值。合併部13，將指標推定部4所推定的第1推定值和指標推定部11所推定的第2推定值按比例分配，算出最終的故障風險指標的推定值。藉由此構成，即使在沒有或缺少設備的保養實績資料的情況下，也能夠適當推定設備發生故障的風險之指標。 As described above, in the failure risk index estimation device 1A of Embodiment 2, the reduction unit 9 classifies the information stored in the equipment information DB6, the maintenance performance DB7, and the failure performance DB8 as information corresponding to the results of each FMEA. The index estimation unit 11 estimates the estimated value of the failure risk index corresponding to the relationship between the maintenance work actual performance data and the failure actual performance data based on the information classified by the limitation unit 9. The merging unit 13 allocates the first estimated value estimated by the index estimating unit 4 and the second estimated value estimated by the index estimating unit 11 in proportion to calculate the final estimated value of the failure risk index. With this configuration, even if there is no or lack of maintenance performance data of the equipment, the index of the risk of equipment failure can be appropriately estimated.

另外，實施形態2的故障風險指標推定裝置1A中，合併部13，將第1推定值的統計推定所估計的資料數用於其和第2推定值的按比例分配。藉由此構成，能夠適當地將第1推定值和第2推定值按比例分配。 In addition, in the failure risk index estimation device 1A of the second embodiment, the merging unit 13 uses the number of data estimated by the statistical estimation of the first estimated value for its proportional distribution with the second estimated value. With this configuration, the first estimated value and the second estimated value can be appropriately distributed in proportion.

再者，本發明在其發明範圍內，可以進行各實施形態的自由組合或各實施形態的任意構成要素的變形、或者各實施形態中的任意構成要素的省略。 In addition, within the scope of the invention, the present invention can be freely combined with each embodiment, modified with any constituent element of each embodiment, or omitted with any constituent element in each embodiment.

【產業上的利用可能性】 【Industrial application possibilities】

本發明的故障風險指標推定裝置，即使在沒有或缺少設備的保養實績資料的情況下，也能夠適當推定設備發生故障的風險之指標，因此可以適用於例如各種的機械系統。 The failure risk index estimation device of the present invention can appropriately estimate the risk index of equipment failure even when there is no or lack of maintenance performance data of the equipment, so it can be applied to various mechanical systems, for example.

Claims (5)

一種故障風險指標推定裝置，其包括：模型式構築部，基於表示構成設備之零件的各檢查項目的故障模式影響解析結果的資訊及表示故障風險指標的推定所使用的統計分布的資訊，構築表示依據前記統計分布的前記故障風險指標之推移的模型式；參數推定部，基於表示構成設備的零件的各檢查項目的故障模式影響解析結果的資訊及表示零件的各檢查項目之既定的保養作業間隔的資訊，統計推定由前記模型式算出的保養作業間隔和前記既定的保養作業間隔之差分為最小的前記模型式的參數值；限縮部，將包含設備、構成設備的零件及各零件的檢查項目的設備資訊、構成設備的零件的各檢查項目之保養作業實績資料及構成設備的各零件的故障實績資料，分類為對應於各個故障模式影響解析結果的資訊；指標推定部，基於前記限縮部所分類的資訊，推定出前記保養作業實績資料和前記故障實績資料之關係所對應的前記故障風險指標的推定值；合併部，將作為前記參數推定部所推定之參數值的第1推定值和作為前記指標推定部所推定的推定值之第2推定值按比例分配，算出最終的前記故障風險指標的推定值。A failure risk index estimation device, including: a model-type construction unit, constructing a representation based on information indicating the analysis result of the failure mode of each inspection item constituting a part of equipment and information indicating the statistical distribution used to estimate the failure risk index The model formula for the transition of the predecessor failure risk indicators based on the statistical distribution of the predecessor; the parameter estimation unit, based on the information indicating the analysis result of the failure mode of each inspection item constituting the component of the equipment and the predetermined maintenance operation interval indicating each inspection item of the component Information, statistically estimate that the difference between the maintenance operation interval calculated by the preceding model formula and the previously established maintenance operation interval is the smallest parameter value of the previous model model; the constriction section will include equipment, parts constituting the equipment, and inspection of each part Item equipment information, maintenance operation performance data for each inspection item that constitutes equipment parts, and failure performance data for each component that constitutes equipment, are classified as information corresponding to the analysis results of the impact of each failure mode; index estimation unit, based on the previous limit Information classified by the Ministry, presumed The estimated value of the previous failure risk index corresponding to the relationship between the previous maintenance performance data and the previous failure performance data; the merging unit will use the first estimated value as the parameter value estimated by the previous parameter estimation unit and the estimated value as the previous index estimation unit The second estimated value of the estimated value is distributed in proportion to calculate the final estimated value of the previous failure risk indicator. 如申請專利範圍第1項所記載的故障風險指標推定裝置，前記合併部，將前記第1推定值的統計推定所估計的資料數用於其與前記第2推定值的按比例分配。As in the failure risk index estimation device described in item 1 of the patent application scope, the preamble merger unit uses the estimated number of data estimated by the statistical presumption of the preamble 1 to the proportional distribution of the presumed second presumption. 如申請專利範圍第1項所記載的故障風險指標推定裝置，前記模型式構築部，故障模式影響解析的故障相關評估項目和該評估項目的評估層級相同的情況下，將共通的參數分配為前記評估項目關聯的前記模型式的參數。If the failure risk index estimation device described in item 1 of the patent application scope, the predecessor model-type construction unit, and the failure-related evaluation items for the analysis of the failure mode effect are the same as the evaluation level of the evaluation item, the common parameters are assigned as the preface Evaluate the pre-modeled parameters associated with the project. 如申請專利範圍第1項所記載的故障風險指標推定裝置，前記參數推定部，將彼此類似的檢查項目間的其中一方的參數值之統計推定所使用的資訊，用於檢查項目間的另一方的參數值的統計推定。The failure risk index estimation device described in item 1 of the patent application scope, the pre-parameter estimation section, uses information used for statistical estimation of the parameter value of one of the inspection items that are similar to each other for the inspection of the other Statistical estimation of parameter values. 一種故障風險指標推定方法，其包括：模型式構築步驟，模型式構築部基於表示構成設備之零件的各檢查項目的故障模式影響解析結果的資訊及表示故障風險指標的推定所使用的統計分布的資訊，構築表示依據前記統計分布的前記故障風險指標之推移的模型式；參數推定步驟，參數推定部基於表示構成設備的零件的各檢查項目的故障模式影響解析結果的資訊及表示零件的各檢查項目之既定的保養作業間隔的資訊，統計推定由前記模型式算出的保養作業間隔和前記既定的保養作業間隔之差分為最小的前記模型式的參數值；限定步驟，限縮部將包含設備、構成設備的零件及各零件的檢查項目的設備資訊、構成設備的零件的各檢查項目之保養作業實績資料及構成設備的各零件的故障實績資料，分類為對應於各個故障模式影響解析結果的資訊；指標推定步驟，指標推定部基於前記限縮部所分類的資訊，推定出前記保養作業實績資料和前記故障實績資料之關係所對應的前記故障風險指標的推定值；合併步驟，合併部將作為前記參數推定部所推定之參數值的第1推定值和作為前記指標推定部所推定的推定值之第2推定值按比例分配，算出最終的前記故障風險指標的推定值。A method for estimating a failure risk index, including a model-building step, the model-building part is based on information indicating the analysis result of the failure mode of each inspection item constituting a part of equipment and the statistical distribution used to estimate the failure risk index Information, constructing a model formula representing the transition of the predecessor failure risk index based on the statistical distribution of the predecessor; parameter estimation step, the parameter estimation section is based on information indicating the analysis result of the failure mode of each inspection item constituting the equipment component and each inspection of the component The information of the scheduled maintenance operation interval of the project, statistically predicts that the difference between the maintenance operation interval calculated by the previous model and the previous predetermined maintenance operation interval is the smallest parameter value of the previous model; the limited steps, the limit part will include equipment, The equipment information of the parts constituting the equipment and the inspection items of each part, the maintenance performance data of each inspection item constituting the parts of the equipment, and the failure performance data of each part constituting the equipment are classified as information corresponding to the impact analysis results of each failure mode ; Index estimation step , The index estimation unit estimates the estimated value of the previous failure risk index corresponding to the relationship between the previous maintenance performance data and the previous failure performance data based on the information classified in the previous limit and shrinkage department; the merge step will be used as the previous parameter estimation unit The first estimated value of the estimated parameter value and the second estimated value as the estimated value estimated by the previous index estimating unit are distributed in proportion to calculate the final estimated value of the previous fault risk index.
TW106114251A 2016-12-09 2017-04-28 Fault risk index estimation device and fault risk index estimation method TWI646476B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
PCT/JP2016/086738 WO2018105104A1 (en) 2016-12-09 2016-12-09 Failure risk index estimation device and failure risk index estimation method
??PCT/JP2016/086738 2016-12-09

Publications (2)

Publication Number Publication Date
TW201822080A TW201822080A (en) 2018-06-16
TWI646476B true TWI646476B (en) 2019-01-01

Family

ID=62490824

Family Applications (1)

Application Number Title Priority Date Filing Date
TW106114251A TWI646476B (en) 2016-12-09 2017-04-28 Fault risk index estimation device and fault risk index estimation method

Country Status (5)

Country Link
JP (1) JP6501982B2 (en)
KR (1) KR102097953B1 (en)
CN (1) CN110023967B (en)
TW (1) TWI646476B (en)
WO (1) WO2018105104A1 (en)

Families Citing this family (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP7213637B2 (en) * 2018-08-07 2023-01-27 日鉄テックスエンジ株式会社 Maintenance management device, maintenance management method and program
JP7265128B2 (en) * 2019-03-04 2023-04-26 日本製鉄株式会社 Facility management support device, facility management support method, program, and computer-readable recording medium
WO2020188636A1 (en) * 2019-03-15 2020-09-24 三菱電機株式会社 Slippage estimation device and slippage estimation method
CN110376982B (en) * 2019-07-11 2021-06-29 江南大学 Control analysis method based on improved FMEA
CN111582751B (en) * 2020-05-19 2022-06-14 国网吉林省电力有限公司 Time-weighted electricity purchasing risk early warning method
CN111651933B (en) * 2020-05-22 2023-09-26 宁波诺丁汉新材料研究院有限公司 Industrial boiler fault early warning method and system based on statistical inference
JP2023012668A (en) * 2021-07-14 2023-01-26 三菱重工業株式会社 Failure predicting device, failure predicting method, and program
KR102489119B1 (en) 2021-10-21 2023-01-18 김지예 Smart FMEA system and process management system and method using thereof
JP7511797B2 (en) 2022-07-01 2024-07-05 三菱電機株式会社 Maintenance support system, maintenance support method, and maintenance support program
CN115982960B (en) * 2022-12-08 2023-09-01 国家管网集团北方管道有限责任公司 Intelligent risk prevention and control capability evaluation method for pipeline oil delivery station

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6223143B1 (en) * 1998-08-31 2001-04-24 The United States Government As Represented By The Administrator Of The National Aeronautics And Space Administration Quantitative risk assessment system (QRAS)
JP2005011327A (en) * 2003-05-29 2005-01-13 Tokyo Electric Power Co Inc:The Repair plan making support device and method
JP2005085178A (en) * 2003-09-11 2005-03-31 Mitsubishi Electric Corp System for preparing equipment operation plan
US20050149570A1 (en) * 2003-12-19 2005-07-07 Kabushiki Kaisha Toshiba Maintenance support method, storage medium, and maintenance support apparatus
JP2008009990A (en) * 2006-06-29 2008-01-17 Toshiba Corp Maintenance planning system and maintenance planning method
JP2009251822A (en) * 2008-04-03 2009-10-29 Toshiba Corp Complex diagnosis maintenance plan supporting system and supporting method for same
US20120221190A1 (en) * 2011-02-24 2012-08-30 Bae Systems Plc Reliability centered maintenance
TW201250507A (en) * 2011-06-02 2012-12-16 Hon Hai Prec Ind Co Ltd System and method for validating design of electronic products
CN104076809A (en) * 2013-03-26 2014-10-01 三菱电机株式会社 Data processing device and data processing method
CN104166800A (en) * 2014-08-11 2014-11-26 工业和信息化部电子第五研究所 Component FMEA analysis method and system based on failure mechanisms
EP2837984A2 (en) * 2013-08-05 2015-02-18 Uptime Engineering GmbH Process to optimize the maintenance of technical systems
CN104463421A (en) * 2014-11-06 2015-03-25 朱秋实 Big data modeling equipment dynamic optimization maintenance method based on real-time status
US20160077164A1 (en) * 2014-09-17 2016-03-17 Kabushiki Kaisha Toshiba Failure sign diagnosis system of electrical power grid and method thereof

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2004252549A (en) 2003-02-18 2004-09-09 Hokkaido Sekiyu Kyodo Bichiku Kk Rated operation management method and rated operation management device
JP4967430B2 (en) * 2006-04-11 2012-07-04 オムロン株式会社 Defect management device, defect management program, and recording medium recording the same
JP4907507B2 (en) 2007-12-03 2012-03-28 新日本製鐵株式会社 Equipment maintenance plan creation support apparatus, method, program, and computer-readable recording medium
CN102208028B (en) * 2011-05-31 2013-06-19 北京航空航天大学 Fault predicting and diagnosing method suitable for dynamic complex system
FR2989500B1 (en) * 2012-04-12 2014-05-23 Airbus Operations Sas METHOD, DEVICES AND COMPUTER PROGRAM FOR AIDING THE TROUBLE TOLERANCE ANALYSIS OF AN AIRCRAFT SYSTEM USING REDUCED EVENT GRAPHICS
JP2016004279A (en) * 2014-06-13 2016-01-12 株式会社日立製作所 Maintenance system and maintenance method
JP6251201B2 (en) 2015-01-08 2017-12-20 三菱重工業株式会社 Occupancy rate prediction device and availability rate prediction method
CN105956789A (en) * 2016-05-24 2016-09-21 国网四川省电力公司 Quantitative risk evaluation method for power equipment based on state evaluation

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6223143B1 (en) * 1998-08-31 2001-04-24 The United States Government As Represented By The Administrator Of The National Aeronautics And Space Administration Quantitative risk assessment system (QRAS)
JP2005011327A (en) * 2003-05-29 2005-01-13 Tokyo Electric Power Co Inc:The Repair plan making support device and method
JP2005085178A (en) * 2003-09-11 2005-03-31 Mitsubishi Electric Corp System for preparing equipment operation plan
US20050149570A1 (en) * 2003-12-19 2005-07-07 Kabushiki Kaisha Toshiba Maintenance support method, storage medium, and maintenance support apparatus
JP2008009990A (en) * 2006-06-29 2008-01-17 Toshiba Corp Maintenance planning system and maintenance planning method
JP2009251822A (en) * 2008-04-03 2009-10-29 Toshiba Corp Complex diagnosis maintenance plan supporting system and supporting method for same
US20120221190A1 (en) * 2011-02-24 2012-08-30 Bae Systems Plc Reliability centered maintenance
TW201250507A (en) * 2011-06-02 2012-12-16 Hon Hai Prec Ind Co Ltd System and method for validating design of electronic products
CN104076809A (en) * 2013-03-26 2014-10-01 三菱电机株式会社 Data processing device and data processing method
EP2837984A2 (en) * 2013-08-05 2015-02-18 Uptime Engineering GmbH Process to optimize the maintenance of technical systems
CN104166800A (en) * 2014-08-11 2014-11-26 工业和信息化部电子第五研究所 Component FMEA analysis method and system based on failure mechanisms
US20160077164A1 (en) * 2014-09-17 2016-03-17 Kabushiki Kaisha Toshiba Failure sign diagnosis system of electrical power grid and method thereof
CN104463421A (en) * 2014-11-06 2015-03-25 朱秋实 Big data modeling equipment dynamic optimization maintenance method based on real-time status

Also Published As

Publication number Publication date
WO2018105104A1 (en) 2018-06-14
KR20190071812A (en) 2019-06-24
TW201822080A (en) 2018-06-16
JPWO2018105104A1 (en) 2019-03-28
KR102097953B1 (en) 2020-04-07
CN110023967B (en) 2022-12-23
CN110023967A (en) 2019-07-16
JP6501982B2 (en) 2019-04-17

Similar Documents

Publication Publication Date Title
TWI646476B (en) Fault risk index estimation device and fault risk index estimation method
US8707268B2 (en) Testing operations of software
US9183067B2 (en) Data preserving apparatus, method and system therefor
US11095728B2 (en) Techniques for automatically interpreting metric values to evaluate the health of a computer-based service
US10467132B1 (en) Variability system and analytics for continuous reliability in cloud-based workflows
US20090271235A1 (en) Apparatus and method for generating survival curve used to calculate failure probability
CN107992410B (en) Software quality monitoring method and device, computer equipment and storage medium
US20150025872A1 (en) System, method, and apparatus for modeling project reliability
JPWO2008020468A1 (en) Program analysis method and apparatus
US20140095934A1 (en) Test case production utilizing problem reports
JP2019032671A (en) Cause estimation method and program
JP6975086B2 (en) Quality evaluation method and quality evaluation equipment
JP5745561B2 (en) Prediction error evaluation apparatus and method
US9348733B1 (en) Method and system for coverage determination
JP6545397B2 (en) Causal relationship evaluation device, causal relationship evaluation system and causal relationship evaluation method
JP2021015321A (en) Procedure identification device, calculation model generation device, procedure identification method, procedure identification program, calculation model generation method, calculation model generation program, learning data generation device and calculation program
KR102426581B1 (en) Method and apparatus for analyzing safety of automotive software
CN106547690B (en) Based on the Software Reliability Modeling method and device under new fitting criterion
Gupta et al. Developing software bug prediction models using various software metrics as the bug indicators
US10437710B2 (en) Code coverage testing utilizing test-to-file maps
Krasich Modeling of SW reliability in early design with planning and measurement of its reliability growth
Triwijoyo et al. Software reliability measurement base on failure intensity
Fedotova et al. Applicability of extreme value theory to the execution time prediction of programs on socs
JP2014203330A (en) Product quality evaluation device and method
JP7428190B2 (en) Resource allocation calculation device, resource allocation calculation method, and resource allocation calculation program

Legal Events

Date Code Title Description
MM4A Annulment or lapse of patent due to non-payment of fees