JPS59188575A - Analyzing device for life distribution - Google Patents
Analyzing device for life distributionInfo
- Publication number
- JPS59188575A JPS59188575A JP58062158A JP6215883A JPS59188575A JP S59188575 A JPS59188575 A JP S59188575A JP 58062158 A JP58062158 A JP 58062158A JP 6215883 A JP6215883 A JP 6215883A JP S59188575 A JPS59188575 A JP S59188575A
- Authority
- JP
- Japan
- Prior art keywords
- data
- distribution
- elapsed time
- regression analysis
- life distribution
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Landscapes
- Tests Of Electronic Circuits (AREA)
- Testing Of Individual Semiconductor Devices (AREA)
Abstract
Description
【発明の詳細な説明】 〔発明のオＵ用分野〕 本発明は、電子部品の信頼性評価におい工。[Detailed description of the invention] [Field of invention] The present invention is an odor technology for evaluating the reliability of electronic components.
とくにその寿命分布を％足する装置に関するものである
。In particular, it relates to a device that adds up the life distribution as a percentage.
従来１次のような方法乃至手段を用いて電子部品の寿命
分布を特定していた。まず、経時的に多数個の被測定電
子部品の故障判定を行い。Conventionally, the life distribution of electronic components has been specified using the following first-order method or means. First, a large number of electronic components to be measured are determined to be faulty over time.
試験経過時間と累積故障率データを求める。次に被測定
電子部品の寿命分布に近いと思われる第１の確率紙に上
記データをプロットし目視でその直線性をチェックし、
直線的であれば第１の確率紙の定義する分布形と与なし
、直緋から読みとれる母数をもって寿命分布ビ特定して
いた。しかし、第１の確率紙にプロットしたとき直憩的
でないときは、第２の確率紙に再びプロットし直し、第
１の確率紙と同様な方法で寿命分布を特定していた。更
に第２の確率紙にプロットしたときでも直線的でないと
きは、第３゜第４．・・・とプロットする確率紙を変え
、プロットが直線的と思われる確率紙で纂１の確率紙で
述べた手順により寿命分布を特定していた。Obtain test elapsed time and cumulative failure rate data. Next, plot the above data on a first probability paper that is considered to be close to the life distribution of the electronic component to be measured, and visually check its linearity.
If it was linear, it would be the same as the distribution shape defined by the first probability paper, and the lifespan distribution would be specified using the parameter that could be read from the straight line. However, if the plotting on the first probability paper was not direct, it was plotted again on the second probability paper, and the life distribution was specified in the same manner as the first probability paper. Furthermore, if it is not linear even when plotted on the second probability paper, the 3rd degree, 4th degree, etc. ...I changed the probability paper on which I plotted, and specified the life distribution using the probability paper whose plot seemed to be linear, using the procedure described in Volume 1 of the probability paper.
しかしこのような方法では、データの１０ツトの都度確
率紙を選び、目視で直線性をチェツりをする必要があり
、寿命分布の特定に解析者の経験が要求されたり、人手
による試行錯誤的解析のため、寿命分布の特定に時間が
かかったりするような欠点がありた。However, with this method, it is necessary to select a probability paper for each 10 points of data and visually check the linearity, requiring the experience of the analyst to identify the life distribution, and requiring manual trial and error. Due to the analysis, it has the disadvantage that it takes time to identify the life distribution.
本発明の目的は、上ｆｚＧした諸欠点を解決し。 The object of the present invention is to solve the above-mentioned drawbacks.
自動的に電子部品の寿命分布を特定する解析装置を提供
することＫある。An object of the present invention is to provide an analysis device that automatically identifies the life distribution of electronic components.
本発明は、解析者の経＊が要求されるデータ１０ツトの
直紡性のチェックを、蛾形回帰分析による重相関係数の
比較で行い、寿命分布の特定を自動的に行なうようにな
したことを％徴とする。The present invention checks the directness of 10 pieces of data that require the analysis* by comparing multiple correlation coefficients using moth-shaped regression analysis, and automatically identifies the lifespan distribution. What you did is expressed as a percentage.
以下１本発明の一実施例を第１図乃至第３図を引用し説
明する。第１図は本発明による寿命分布解析装置の各部
の構成を示すブロック図であり、経時的に多数の板側定
電子部品７の故障判定を行い、その判定結果を試験経過
時間ｔと累積故障率Ｆと忙対応付けてデータ収集する故
障判定部２と、これらのデータ及びその他必要なデータ
を記憶しておくデータ記憶部３と、さらに記憶されてい
る試験経過時間ｔと累積故障率Ｆとを線形回帰分析する
前に分布形状に合うよう変数変換するデータ変換部４と
、変数変換された上記データを線型回帰分析する回帰分
析部５と、解析結果を表示及びハードコピーする出力部
６及びこれら各部２〜６を目的に沿って制御する制御部
１とから成っている。An embodiment of the present invention will be described below with reference to FIGS. 1 to 3. FIG. 1 is a block diagram showing the configuration of each part of the life distribution analysis device according to the present invention, in which a large number of plate-side constant electronic components 7 are judged for failure over time, and the judgment results are calculated based on the test elapsed time t and cumulative failure. A failure determination section 2 collects data in association with the failure rate F, a data storage section 3 that stores these data and other necessary data, and a stored test elapsed time t and cumulative failure rate F. a data conversion unit 4 that converts variables to match the distribution shape before linear regression analysis; a regression analysis unit 5 that performs linear regression analysis on the variable-converted data; an output unit 6 that displays and hard copies the analysis results; It consists of a control section 1 that controls each of these sections 2 to 6 according to the purpose.
以下１２１１ｃ示したフローチャートを用いて。Using the flowchart shown below 1211c.
第１図の動作を説明する。The operation shown in FIG. 1 will be explained.
なおデータ変換部４では寿命分布特定のために、ワイプ
ル分布、正規分布、対数正規分布。Note that the data converter 4 uses Weipul distribution, normal distribution, and lognormal distribution to specify the life distribution.
極値分布（指数型）及び極値分布（コーシー型）の５種
類の分布形に対し直線関係にするため。To create a linear relationship for five types of distribution types: extreme value distribution (exponential type) and extreme value distribution (Cauchy type).
Ｗ、１表に示す変数変換式忙従って試験経過時間を及び
累積故障率Ｆの各データを変数変換する。W, the variable conversion formula shown in Table 1. Accordingly, each data of test elapsed time and cumulative failure rate F is converted into variables.
さて制御部１は、まず寿命分布を特定したい被測定電子
部品７の故障判定と経時的にその判・ ３ ・
足結果を試験経過時間ｔと累積故障率Ｆとに対応づけて
データを収集するよう故障判定部２・に指令する。故障
判定部２はこの指令を受け、第２図のステップ１１を実
行する。この操作”は制御部１にあらかじめ設定されて
いる試験終了時間Ｋ ｒｘるまで行なわれ、収集された
データはステップ１２に示すようにデータ記憶部３に記
憶される。Now, the control unit 1 first collects data by determining the failure of the electronic component 7 to be measured whose life distribution is to be specified, and by associating the results with the test elapsed time t and the cumulative failure rate F. The fault determination unit 2 is instructed to do so. The failure determination section 2 receives this command and executes step 11 in FIG. 2. This operation is continued until the test end time Krx, which is preset in the control unit 1, is reached, and the collected data is stored in the data storage unit 3 as shown in step 12.
なお第２表は集積（ロ）路部品を被測定電子部品７とし
、これを高温高ｇ＆漂境に放置したときの試験経過時間
ｔと累積故障率Ｆのオリジナルデータ例であり、上記デ
ータ記憶部３に記憶されている。Table 2 is an example of the original data of the test elapsed time t and the cumulative failure rate F when the electronic component to be measured 7 is an integrated circuit component and is left in a high temperature, high gravity, and floating environment. It is stored in section 3.
次に制御１部１は、線形回帰分析の前処理としてＷｃ１
表に示した変数変換式に従ってオリジナルデータを変俣
丁べくデータ変換ｆＭＳ４に指令しステップ１３を実行
する。本例で示したオリジナルデータの場合は、第３表
に示すような値にデータ変換され、に換後のデータもデ
ータ記憶部３に記憶される。Next, the control unit 1 performs Wc1 as preprocessing for linear regression analysis.
Step 13 is executed by instructing the data conversion fMS 4 to transform the original data according to the variable conversion formula shown in the table. In the case of the original data shown in this example, the data is converted into the values shown in Table 3, and the converted data is also stored in the data storage section 3.
・ ４ ・
さらに制御部１は１回帰分析部５にＲＬｉ １表に示し
た各分布形に対応した変数変換後のデータを選択し、説
明変数として試験経過時間ｔまたはそのデータ変換値、
目的変数として累積故障率Ｆのデータ変換値を選び、ワ
イブル分布から極値分布（コーンーｍ）までの６種類の
分布形に対応する線型回帰分析を指令し１回帰分析部５
はステップ１４及び１４′にてこれを実行する。・ 4 ・ Furthermore, the control unit 1 selects the data after variable conversion corresponding to each distribution form shown in the RLi 1 table to the regression analysis unit 5, and sets the test elapsed time t or its data conversion value as an explanatory variable.
The data conversion value of the cumulative failure rate F is selected as the objective variable, and the linear regression analysis corresponding to six types of distribution shapes from the Weibull distribution to the extreme value distribution (cone-m) is commanded.1 Regression analysis unit 5
does this in steps 14 and 14'.
回帰分析結果は第３図に示すようになる。しかし、制御
部１はこの中で一番直腺性のよいものを得るために、ス
テップ１５にて回帰分析による重相関係数Ｒを最大にす
る変数変換の組に対応する分布形を選ぶ。本例では第６
図かられかるように同図ＣＣ） Ｋ示す対数正規分布が
選ばれ、この分布の母数が第１表に示す式によって計算
される。The regression analysis results are shown in Figure 3. However, in order to obtain the one with the best directness among them, the control unit 1 selects in step 15 the distribution form corresponding to the set of variable transformations that maximizes the multiple correlation coefficient R obtained by regression analysis. In this example, the 6th
As can be seen from the figure, the lognormal distribution shown in Figure CC) K is selected, and the parameters of this distribution are calculated by the formula shown in Table 1.
第１表の推定母数の計算式において、 ｂ、 、 ｂ。In the formula for calculating estimated parameters in Table 1, b, , b.
はそれぞれ線形回帰分析で求められた回帰式の零次及び
１次の偏回帰係数である。本例の場合分布形は対数正規
分布と推定され、その母数は。are the zero-order and first-order partial regression coefficients of the regression equation obtained by linear regression analysis, respectively. In this example, the distribution shape is estimated to be lognormal distribution, and its parameter is.
ト特足でき、ステップ１６にて出力＠６で表示される。The special foot is executed, and the output @6 is displayed in step 16.
なお、解析結果の表示及びノ１−トコビーは。In addition, the display of analysis results and No. 1 Tocobee.
重相関係数Ｒを求人とする分布形のグラフ及びその抽足
母数のみならず、他の分布形に河しても任Ａｒこ出力で
きるよう虻なされる。It is designed to be able to output not only the graph of the distribution type and its extraction parameter using the multiple correlation coefficient R, but also the graph of other distribution types.
（以下余白） 、 ７ 。(Margin below) , 7.
、 ８゜
〔発明の効果〕
本発明による寿命分布解析装置を用いることによって、
先見的に分布形の分らない寿命分布形状及び母数の特足
を自動的に行うことができるので、将来の寿命予測、信
頼性設計、保守計画等の迅速化及び精度同上に寄与する
などの効果がある。, 8゜[Effects of the invention] By using the life distribution analysis device according to the present invention,
Since it is possible to automatically calculate the life distribution shape and parameters of which the distribution shape is unknown in advance, it contributes to speeding up and accuracy of future life prediction, reliability design, maintenance planning, etc. effective.
第１図は本発明による寿命分布解析ＩＪ：、置のプａツ
ク図、＄２図は第１肉の動作説明用のフローチャート、
第３図は各分布形に対応した線形回帰分析結果を示す線
図である。
１・・・制御部 ２・・・故障判定部３・・・
データ記憶部 ４・・・データ変換部５・・・回帰分
析部 ６・・・出力部７・・・被測定電子部品
代理人弁理士 高 槁 明 夫
第２図
第 ｊFigure 1 is a diagram of the life distribution analysis IJ:, position according to the present invention, Figure 2 is a flowchart for explaining the operation of the first meat,
FIG. 3 is a diagram showing the results of linear regression analysis corresponding to each distribution type. 1...Control unit 2...Failure determination unit 3...
Data storage section 4... Data conversion section 5... Regression analysis section 6... Output section 7... Patent attorney representing the electronic component to be measured Akio Ko Figure 2 j
Claims (1)
果を試験経過時間と累積故障率と忙対応付けてデータ収
集する故障判定部と、該故障判定部で収集された上記デ
ータ及びその他必要なデータ′ｆｊｒ：記憶するデータ
記憶部と、該データ記憶部に記憶されている試験経過時
間データ及び累積故障率データを複数個の変換式で変数
変換するデータに換部と、変数変換された上記試験経過
時間データと累積故障率データを回帰分析する１ｇＩ帰
分析部と、該回帰分析結果を表示記録する出力部及びこ
れら各部の動作を制御する制（ＩＱ１部とからなる寿命
分布解析装置。A failure determination unit that determines the failure of a large number of electronic components over time and collects data by correlating the results with the test elapsed time and cumulative failure rate, and the above-mentioned data and other data collected by the failure determination unit. Necessary data 'fjr: A data storage unit to store, a conversion unit that converts the test elapsed time data and cumulative failure rate data stored in the data storage unit into variables using a plurality of conversion formulas; A life distribution analysis device consisting of a 1gI regression analysis section that performs regression analysis on the test elapsed time data and cumulative failure rate data, an output section that displays and records the regression analysis results, and a system (IQ1 section) that controls the operation of each of these sections. .
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP58062158A JPS59188575A (en) | 1983-04-11 | 1983-04-11 | Analyzing device for life distribution |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP58062158A JPS59188575A (en) | 1983-04-11 | 1983-04-11 | Analyzing device for life distribution |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| JPS59188575A true JPS59188575A (en) | 1984-10-25 |
Family
ID=13192021
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP58062158A Pending JPS59188575A (en) | 1983-04-11 | 1983-04-11 | Analyzing device for life distribution |
Country Status (1)
| Country | Link |
|---|---|
| JP (1) | JPS59188575A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0655573U (en) * | 1993-01-21 | 1994-08-02 | 株式会社ワコール | Drop sewing tools for sewing machines |
-
1983
- 1983-04-11 JP JP58062158A patent/JPS59188575A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPH0655573U (en) * | 1993-01-21 | 1994-08-02 | 株式会社ワコール | Drop sewing tools for sewing machines |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10718542B2 (en) | Building management system with system identification using multi-step ahead error prediction | |
| CN109061341A (en) | Kalman filtering transformer fault prediction technique and system neural network based | |
| US5373457A (en) | Method for deriving a piecewise linear model | |
| Balasooriya et al. | Progressively censored reliability sampling plans for the Weibull distribution | |
| CN111537841B (en) | Optimization method and system suitable for ground fault type identification | |
| JP2012528392A5 (en) | ||
| CN108663582A (en) | A kind of fault diagnosis method and system of transformer | |
| KR20090001148A (en) | Virtual metrology system and virtual metrology method | |
| CN108830417B (en) | ARMA (autoregressive moving average) and regression analysis based life energy consumption prediction method and system | |
| JP7214884B2 (en) | Information processing device, information processing method, computer program and information processing system | |
| CN111143981B (en) | Virtual test model verification system and method | |
| CN106021838B (en) | A kind of Complex Electronic Systems Based method for predicting residual useful life | |
| CN116258273A (en) | Hydraulic prediction method and system for wet double-clutch transmission, vehicle and storage medium | |
| JP2001273006A (en) | Operation evaluation system of energy plant | |
| JPS59188575A (en) | Analyzing device for life distribution | |
| Gorbiychuk et al. | Evaluation of dynamic properties of gas pumping units according to the results of experimental researches | |
| CN117607691A (en) | Tracking detection method and device for charging working temperature of lithium ion battery | |
| EP2738701A2 (en) | Method and apparatus for statistical electronic circuit simulation | |
| KR20220135246A (en) | Estimation apparatus and method for estimating error factors | |
| Sharp | Simple metrics for evaluating and conveying prognostic model performance to users with varied backgrounds | |
| WO2009055967A1 (en) | Real-time model validation | |
| CN105656453B (en) | A kind of optical fiber current mutual inductor random noise Real-Time Filtering method based on time series | |
| WO2012048734A1 (en) | Method for controlling an industrial process | |
| JP2020190994A (en) | Facility maintenance management system | |
| CN115526137B (en) | Simulation verification method for disconnecting link switch operation ticket |