TW202311879A - Device management system, cause of failure estimation method for device and non-transitory programmable memory medium wherein the device management system includes a log data acquisition mechanism, a cluster information extraction mechanism, an abnormality calculation mechanism, and a cause of failure estimation mechanism - Google Patents

Device management system, cause of failure estimation method for device and non-transitory programmable memory medium wherein the device management system includes a log data acquisition mechanism, a cluster information extraction mechanism, an abnormality calculation mechanism, and a cause of failure estimation mechanism Download PDF

Info

Publication number
TW202311879A
TW202311879A TW111132172A TW111132172A TW202311879A TW 202311879 A TW202311879 A TW 202311879A TW 111132172 A TW111132172 A TW 111132172A TW 111132172 A TW111132172 A TW 111132172A TW 202311879 A TW202311879 A TW 202311879A
Authority
TW
Taiwan
Prior art keywords
information
cluster
inter
management system
log
Prior art date
Application number
TW111132172A
Other languages
Chinese (zh)
Other versions
TWI824681B (en
Inventor
大西貴子
Original Assignee
日商歐姆龍股份有限公司
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 日商歐姆龍股份有限公司 filed Critical 日商歐姆龍股份有限公司
Publication of TW202311879A publication Critical patent/TW202311879A/en
Application granted granted Critical
Publication of TWI824681B publication Critical patent/TWI824681B/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
    • G06F11/079—Root cause analysis, i.e. error or fault diagnosis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment
    • G06F11/3466—Performance evaluation by tracing or monitoring
    • G06F11/3476—Data logging

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Debugging And Monitoring (AREA)
  • Exchange Systems With Centralized Control (AREA)

Abstract

The present invention provides a technology capable of accurately estimating the cause of a failure of a device used in combination with a computer. The device management system of the present invention is a management system of a device. The device management system is characterized in that it includes a log data acquisition mechanism, a cluster information extraction mechanism, an abnormality calculation mechanism, and a cause of failure estimation mechanism. The log data acquisition mechanism is configured to obtain a log, wherein the log is a record of software actions related to the control of the device. The cluster information extraction mechanism is configured to obtain cluster information and inter-cluster transition information from the obtained log. The cluster information is the information representing the contents of each step in operations of the device. The inter-cluster transition information is the information in transition between a step and another step. The abnormality calculation mechanism is configured to calculate an abnormality between each of the obtained inter-cluster transition information. The cause of failure estimation mechanism is configured to estimate a cause of failure of the device based on the abnormality calculated by the abnormality calculation mechanism.

Description

裝置管理系統、裝置的障礙原因推測方法以及非暫時性地記憶程式的記憶媒體Device management system, device failure cause estimation method, and storage medium for non-temporarily storing programs

本發明是有關於一種裝置管理系統、裝置的障礙原因推測方法以及程式。The present invention relates to a device management system, a device failure cause estimation method, and a program.

關於產業用途的裝置（或設備）的保養業務，於產生了裝置動作的停止、性能的降低等障礙的情形時，進行其原因分析。另外，此種障礙原因的分析通常是由操作員（operator）一方面相互參照軟體（software）的動作記錄（日誌）、裝置的機械零件的運轉狀態（各種感測器測量值、馬達轉速等）、控制裝置（電腦）的運轉狀態（中央處理單元（Central Processing Unit，CPU）使用率、記憶體使用量、網路收發量、基板溫度等）等多樣的資訊，一方面進行分析。Regarding the maintenance work of equipment (or equipment) for industrial use, when troubles such as stoppage of equipment operation and performance degradation occur, the cause analysis is performed. In addition, the analysis of the causes of such failures is usually performed by the operator (operator) cross-referencing the software (software) action records (logs), the operating status of the mechanical parts of the device (various sensor measurements, motor speed, etc.) , the operating status of the control device (computer) (Central Processing Unit (Central Processing Unit, CPU) usage rate, memory usage, network sending and receiving volume, substrate temperature, etc.), and other information, on the one hand to analyze.

然而，根據如此般對照多樣的資訊進行分析的方法，有操作員的負擔大，分析結果大幅度地受到個人的經驗、知識影響等問題。However, according to the method of analyzing various pieces of information in this way, the burden on the operator is heavy, and there are problems such as that the analysis result is largely influenced by personal experience and knowledge.

針對此種問題，近年來提出了包含自動化在內的和保養業務的效率化有關的各種方法。例如，正致力於蓄積與裝置的狀態有關的資料，以有助於故障對策的自動化。尤其於反覆進行一定的簡單動作般的裝置中，有效的是藉由學習自感測器獲得的訊號資料從而檢測異常值或變化點（所謂偏離值），提出有使用該些資訊來進行障礙原因的推測或故障的預測（例如非專利文獻1等）。In response to such a problem, in recent years, various methods related to efficiency of maintenance work including automation have been proposed. For example, efforts are being made to accumulate data on the status of devices to contribute to the automation of failure countermeasures. Especially in devices that repeatedly perform certain simple actions, it is effective to detect abnormal values or change points (so-called deviation values) by learning the signal data obtained from the sensor, and to propose the cause of the failure using this information. speculation or failure prediction (for example, Non-Patent Document 1, etc.).

然而，對於檢查裝置或加工機器等與控制裝置（電腦）組合而進行複雜動作般的裝置而言，使用簡單資料的先前技術難以獲得充分的結果。鑒於所述方面，近年來亦進行了下述研究，即：使用深層學習等方法來靈活運用自多數的感測器獲取的大量的資料（例如非專利文獻2等）。However, it is difficult to obtain sufficient results using conventional techniques using simple data for devices that perform complicated operations in combination with a control device (computer), such as an inspection device or a processing machine. In view of the above, in recent years, studies have also been conducted on using methods such as deep learning to utilize a large amount of data acquired from many sensors (for example, Non-Patent Document 2, etc.).

而且，亦提出有代替感測器資料而靈活運用控制裝置的軟體日誌或維護記錄等文本資料，以這些作為對象進行學習，推測維護的最適時機（例如非專利文獻3等）。 [現有技術文獻] [非專利文獻] In addition, it has also been proposed to use text data such as software logs and maintenance records of the control device instead of sensor data, and use these as objects to learn and estimate the optimal timing of maintenance (for example, non-patent document 3, etc.). [Prior art literature] [Non-patent literature]

[非專利文獻1]費雷羅.S（Ferreiro, S.）、孔德.E（Konde, E.）、費爾南德.S（Fernandez, S.）及皮拉多.A（Prado, A.），2016，工業4.0：生產設備的預測性智能維護（Industry 4.0 : predictive intelligent maintenance for production equipment），歐洲預後與健康管理學會會議（European Conference of the Prognostics and Health Management Society），no（pp.1-8），researchgate.net. [非專利文獻2]埃德姆吉米.T.T（Ademujimi, T.T.）、布倫代奇.M.P（Brundage, M.P.）及帕布.V.V（Prabhu,V.V.），2017，當前機器學習技術於製造診斷中的應用綜述（A Review of Current Machine Learning Techniques Used in Manufacturing Diagnosis），生產管理系統的進步（Advances in Production Management Systems），智能、協作及可持續製造之路（The Path to Intelligent, Collaborative and Sustainable Manufacturing）（pp.407-415），施普林格國際出版公司（Springer International Publishing.） [非專利文獻3]帕蒂爾.R.B（Patil,R.B.）、帕蒂爾.M.A（Patil, M.A.）、拉維.V.（Ravi, V.）及內克.S（Naik, S.），2017，用於自機器日誌對成像設備進行糾正性維護的預測建模（Predictive modeling for corrective maintenance of imaging devices from machine logs），會議論文集：...IEEE醫學和生物學工程年度國際會議（Conference proceedings : ...Annual International Conference of the IEEE Engineering in Medicine and Biology Society），IEEE醫學和生物學工程學會（IEEE Engineering in Medicine and Biology Society），會議（Conference），2017，1676-1679，ieeexplore.ieee.org. [Non-Patent Document 1] Ferreiro. S (Ferreiro, S.), Conde. E (Konde, E.), Fernandez. S (Fernandez, S.) and Pilardo. A (Prado, A.), 2016, Industry 4.0: Predictive Intelligent Maintenance for Production Equipment, European Conference of the Prognostics and Health Management Society, no (pp .1-8), researchgate.net. [Non-Patent Document 2] Ademujimi, T.T (Ademujimi, T.T.), Brundage, M.P (Brundage, M.P.) and Prabhu.V.V (Prabhu, V.V.), 2017, Current Machine Learning Technology in Manufacturing Diagnosis A Review of Current Machine Learning Techniques Used in Manufacturing Diagnosis, Advances in Production Management Systems, The Path to Intelligent, Collaborative and Sustainable Manufacturing ( pp.407-415), Springer International Publishing. [Non-Patent Document 3] Patil. R.B (Patil, R.B.), Patil. M.A (Patil, M.A.), Ravi. V. (Ravi, V.) and Naik. S (Naik, S.) , 2017, Predictive modeling for corrective maintenance of imaging devices from machine logs (Predictive modeling for corrective maintenance of imaging devices from machine logs), Conference Proceedings:...IEEE Annual International Conference on Medical and Biological Engineering ( Conference proceedings : ...Annual International Conference of the IEEE Engineering in Medicine and Biology Society), IEEE Engineering in Medicine and Biology Society (IEEE Engineering in Medicine and Biology Society), Conference (Conference), 2017, 1676-1679, ieeexplore. ieee.org.

[發明所欲解決之課題] 然而，為了推測與電腦組合的相對較複雜的裝置的障礙原因，僅使用感測器資料的先前方法有下述問題，即：無法應對與軟體的動作連動地產生的不良狀況等。而且，即便為學習軟體日誌或維護記錄等文本資料進行分析的方法，亦有下述問題，即：雖可記錄、學習已知的障礙或劣化的狀態來進行分析，但關於未知或預料之外的障礙則難以學習，無法應對此種障礙等。 [Problem to be Solved by the Invention] However, conventional methods using only sensor data in order to estimate failure causes of relatively complex devices combined with a computer have a problem in that they cannot cope with malfunctions that occur in conjunction with software operations. Furthermore, even in the method of analyzing text data such as software logs and maintenance records, there is a problem that although it is possible to record and learn known failures or deterioration states for analysis, there is a problem regarding unknown or unexpected It is difficult to learn, unable to deal with such obstacles, etc.

本發明是鑒於所述般的實際情況而成，其目的在於提供一種可高精度地進行與電腦組合使用的裝置的障礙原因的推測的技術。 [解決課題之手段] The present invention is made in view of the above-mentioned actual situation, and an object of the present invention is to provide a technology capable of accurately estimating the cause of a failure of a device used in combination with a computer. [Means to solve the problem]

為了達成所述目的，本發明採用以下的結構。即， 一種裝置管理系統，為裝置的管理系統，其特徵在於包括： 日誌資料獲取機構，獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作記錄； 叢集資訊提取機構，自所述獲取的日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示所述裝置的運轉的各步驟的內容的資訊，所述叢集間過渡資訊為和一個所述步驟與另一個所述步驟之間的過渡有關的資訊； 異常度計算機構，算出所述提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測機構，基於由所述異常度計算機構所算出的所述異常度來推測所述裝置的障礙原因。 In order to achieve the object, the present invention employs the following structures. Right now, A device management system is a device management system, characterized in that it comprises: The log data acquisition mechanism acquires a log, and the log is a record of software actions related to the control of the device; a cluster information extracting unit for extracting cluster information and inter-cluster transition information from the set of acquired logs, the cluster information being information representing the contents of each step in the operation of the device, and the inter-cluster transition information being and information relating to transitions between one said step and another said step; an abnormality calculation mechanism, which calculates the abnormality of the extracted transition information between clusters; and The failure cause estimation means estimates a failure cause of the device based on the abnormality degree calculated by the abnormality degree calculation means.

根據此種結構，關於產生了障礙的裝置，可針對和裝置的運轉有關的各個細小行為算出異常度，並基於所述異常度來進行障礙原因的推測，因而即便對於未知（或預料之外的）障礙原因，亦可推測其為障礙的原因。According to this structure, the degree of abnormality can be calculated for each small behavior related to the operation of the device in which a failure has occurred, and the cause of the failure can be estimated based on the degree of abnormality. Therefore, even for unknown (or unexpected) ) the cause of the disorder, it can also be speculated that it is the cause of the disorder.

而且，所述異常度計算機構亦可基於所述裝置的正常時的所述叢集間過渡資訊，來算出所述提取的各個所述叢集間過渡資訊的異常度。若為此種結構，則可不使用裝置有障礙時的學習資料，而僅基於裝置的正常運轉時的資料來算出障礙產生時的異常度，因而可自簡單結構的裝置至複雜的裝置而對各種對象適用。Moreover, the abnormality calculation unit may also calculate the abnormality of each of the extracted inter-cluster transition information based on the inter-cluster transition information when the device is normal. With such a structure, it is possible to calculate the degree of abnormality when a failure occurs based only on the data of the normal operation of the device without using the learning data when the device has a failure. Object applies.

而且，亦可於所述叢集間過渡資訊中，包含和所述裝置的多個所述步驟間的過渡的產生頻率有關的資訊，且所述裝置管理系統更包括：叢集間過渡資訊評價機構，基於所述產生頻率來進行所述提取的各個所述叢集間過渡資訊的加權，所述異常度計算機構使用進行了所述加權的資訊來算出所述異常度。藉由如此般包括基於步驟間的過渡的產生頻率來進行加權的機構，從而可有效率且高精度地算出異常度。Moreover, the inter-cluster transition information may also include information related to the frequency of occurrence of transitions between multiple steps of the device, and the device management system further includes: an inter-cluster transition information evaluation mechanism, Weighting of the extracted inter-cluster transition information is performed based on the generation frequency, and the abnormality calculation means calculates the abnormality using the weighted information. With such a mechanism including weighting based on the occurrence frequency of transitions between steps, the degree of abnormality can be calculated efficiently and with high precision.

而且，亦可更包括：硬體資訊獲取機構，獲取和所述裝置的硬體的狀態有關的硬體資訊，所述叢集間過渡資訊評價機構基於所述硬體資訊獲取機構獲取的所述硬體資訊，進一步進行所述提取的各個所述叢集間過渡資訊的加權。Moreover, it may further include: a hardware information acquisition unit, which acquires hardware information related to the state of the hardware of the device, and the inter-cluster transition information evaluation unit is based on the hardware information acquired by the hardware information acquisition unit. Volume information, further weighting the extracted inter-cluster transition information.

此處，所謂硬體資訊，為各種感測器資料及自所述感測器資料獲得的和裝置的硬體方面的動作、狀態有關的資訊。如此，藉由使用硬體資訊進一步進行加權，從而可更高精度地算出異常度。Here, the so-called hardware information refers to various sensor data and information related to the operation and state of the hardware of the device obtained from the sensor data. In this way, by further performing weighting using hardware information, the degree of abnormality can be calculated with higher accuracy.

而且，所述障礙原因推測機構亦可推測為，於由所述異常度計算機構所算出的所述異常度滿足既定條件的所述叢集間過渡資訊所指定的所述步驟中，存在所述裝置的障礙原因。所謂既定條件，具體而言，例如可設為超過既定的臨限值的情形等。於該情形時，臨限值可由用戶預先設定，亦可藉由根據裝置的運轉實績進行學習從而自動設定。藉由如此設定，從而可有效率地推測裝置的障礙原因。Furthermore, the failure cause estimation means may infer that the device exists in the step specified by the inter-cluster transition information in which the abnormality degree calculated by the abnormality degree calculation means satisfies a predetermined condition. cause of the obstacle. Specifically, the predetermined condition may be, for example, a case where a predetermined threshold value is exceeded. In this case, the threshold value may be set in advance by the user, or it may be set automatically by learning from the actual operation performance of the device. By setting in this way, it is possible to efficiently estimate the failure cause of the device.

而且，所述裝置管理系統亦可更包括：顯示機構，可顯示表示由所述異常度計算機構所算出的所述異常度、及/或由所述障礙原因推測機構所推測的所述障礙原因的資訊。根據此種結構，用戶可容易地確認所推測出的障礙原因。Furthermore, the device management system may further include: a display unit capable of displaying the abnormality degree calculated by the abnormality degree calculation unit and/or the failure cause estimated by the failure cause estimation unit. information. According to such a configuration, the user can easily confirm the presumed cause of the failure.

而且，所述裝置管理系統亦可更包括：有向圖生成機構，以所述叢集資訊作為節點，以所述叢集間過渡資訊作為邊緣，生成表示所述叢集資訊及所述叢集間過渡資訊的關係的有向圖，所述顯示機構可顯示所述有向圖。Moreover, the device management system may further include: a directed graph generating mechanism, which uses the cluster information as a node and the inter-cluster transition information as an edge to generate a graph representing the cluster information and the inter-cluster transition information. A directed graph of relationships, the display mechanism can display the directed graph.

根據此種結構，用戶能以有向圖的態樣來確認和裝置的控制有關的軟體的動作，可將所述資訊靈活運用於裝置的管理、保養。According to this structure, the user can confirm the operation of the software related to the control of the device in the form of a directed graph, and can utilize the information for management and maintenance of the device.

而且，所述叢集間過渡資訊亦可藉由既定的方法進行了增加重要度的評價的加權，所述有向圖生成機構生成可視認各個所述叢集間過渡資訊的所述加權的、所述有向圖。再者，此處的加權的方法並無特別限定，例如可如上所述，設為基於叢集間過渡的產生頻率、對應的硬體資訊（感測器資料）等的加權。根據此種結構，用戶可確認反映出加權的有向圖，故而可自有向圖獲取更詳細的資訊。Furthermore, the inter-cluster transition information may be weighted by a predetermined method to increase the evaluation of importance, and the directed graph generating means generates the weighted, the directed graph. Furthermore, the weighting method here is not particularly limited, for example, as mentioned above, it can be set as weighting based on the generation frequency of inter-cluster transitions, corresponding hardware information (sensor data), and the like. According to this structure, the user can confirm the directed graph that reflects the weight, and thus can obtain more detailed information from the directed graph.

而且，所述有向圖生成機構亦可藉由將表示所述叢集間過渡資訊的所述加權的數值顯示於所述邊緣的附近，從而生成以可視認的方式表現所述加權的有向圖。Furthermore, the directed graph generating means may also generate a directed graph that expresses the weighted value in a visually recognizable manner by displaying the weighted value representing the inter-cluster transition information near the edge. .

而且，所述有向圖生成機構亦可藉由對表示各個所述叢集間過渡資訊的所述邊緣的清晰度設置差異來進行顯示，從而生成以可視認的方式表現所述加權的有向圖。此處，所謂對清晰度設置差異，例如可想到將邊緣的線的粗細度對應權重而增粗，或者將邊緣的線的亮度或明度對應權重而提高等。Moreover, the directed graph generation mechanism may also display by setting a difference in the sharpness of the edge representing the transition information between each of the clusters, thereby generating a directed graph that expresses the weighted in a visually recognizable manner. . Here, setting a difference in sharpness is conceivable, for example, to increase the thickness of the edge line according to the weight, or to increase the brightness or lightness of the edge line according to the weight.

而且，亦可於所述叢集資訊中包含作為自所述日誌提取的文本資訊的單詞，所述有向圖生成機構將各個所述叢集資訊所含的所述單詞以出現次數由多到少的順序提取既定數，並且生成使用所述提取的單詞作為表示所述叢集資訊的內容的資訊的、所述有向圖。若為此種結構，則用戶可基於單詞容易地掌握有向圖的各節點的內容。In addition, words that are text information extracted from the log may also be included in the cluster information, and the directed graph generation mechanism ranks the words included in each cluster information in descending order of occurrence times A predetermined number is sequentially extracted, and the directed graph using the extracted words as information representing the contents of the cluster information is generated. With such a configuration, the user can easily grasp the contents of each node of the directed graph based on words.

而且，所述裝置管理系統亦可更包括：提取日誌顯示圖像生成機構，自所述日誌的集合中，提取與滿足既定條件的所述叢集間過渡資訊對應的日誌，作為表示所述滿足既定條件的所述叢集間過渡資訊的內容的資訊，並且生成表示所述提取的日誌的內容的、提取日誌顯示圖像，所述顯示機構可顯示所述提取日誌顯示圖像。Moreover, the device management system may further include: an extracting log display image generation mechanism, extracting from the set of logs the log corresponding to the inter-cluster transition information satisfying the predetermined condition, as a representation of the satisfying the predetermined The information of the content of the inter-cluster transition information of the conditions, and an extracted log display image representing the content of the extracted log are generated, and the display means can display the extracted log display image.

再者，所謂此處提及的「滿足既定條件」，可設為異常度超過既定值的情形、用戶選擇了所述有向圖的與所述叢集間過渡資訊對應的邊緣的情形等。若為此種結構，則用戶可迅速確認與所述叢集間過渡資訊對應的日誌。Furthermore, the so-called "satisfying the predetermined condition" mentioned here may be a case where the degree of abnormality exceeds a predetermined value, a case where the user selects an edge of the directed graph corresponding to the inter-cluster transition information, and the like. With such a configuration, the user can quickly confirm the log corresponding to the inter-cluster transition information.

而且，所述提取日誌顯示圖像亦可彈出顯示於表示與所述顯示圖像所示的提取日誌對應的所述叢集間過渡資訊的、所述邊緣的附近。若為此種顯示，則可容易地掌握彈出顯示的所述提取日誌顯示圖像、與表示對應的所述叢集間過渡資訊的所述邊緣的關係。再者，提取日誌顯示圖像的顯示部位並無特別限定，亦可無關乎所述彈出顯示而設置特定的顯示區域。Moreover, the extracted log display image may also be pop-up displayed near the edge representing the inter-cluster transition information corresponding to the extracted log shown in the displayed image. Such a display makes it possible to easily grasp the relationship between the extracted log display image displayed as a popup and the edge indicating the corresponding inter-cluster transition information. Furthermore, the display area of the extracted log display image is not particularly limited, and a specific display area may be provided regardless of the pop-up display.

而且，本發明亦可作為裝置的障礙原因推測方法而適用，所述裝置的障礙原因推測方法為推測裝置的障礙原因的方法，且包含： 日誌資料獲取步驟，獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作歷程資訊； 叢集資訊提取步驟，自所獲取的所述日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示由所述裝置進行的處理的各步驟的內容的資訊，所述叢集間過渡資訊為和所述裝置的多個所述步驟間的過渡有關的資訊； 異常度計算步驟，算出所述提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測步驟，基於所述異常度計算步驟中算出的所述異常度來推測所述裝置的障礙原因。 Furthermore, the present invention can also be applied as a method for estimating a cause of a failure of a device, which is a method for estimating a cause of a failure of a device, and includes: The log data obtaining step is to obtain a log, and the log is related to the control of the device, and the information of the action history of the software; a cluster information extraction step of extracting cluster information and inter-cluster transition information from the acquired set of logs, the cluster information being information indicating the contents of each step of processing performed by the device, and the inter-cluster transition information information is information relating to transitions between a plurality of said steps of said device; an abnormality calculation step, calculating the abnormality of the extracted transition information between the clusters; and The failure cause estimation step estimates a failure cause of the device based on the abnormality degree calculated in the abnormality degree calculation step.

而且，本發明亦可理解為用以使電腦執行所述方法的程式、及非暫時性地記錄有此種程式的電腦可讀取的記錄媒體。Furthermore, the present invention can also be understood as a program for causing a computer to execute the above method, and a computer-readable recording medium on which such a program is recorded non-temporarily.

再者，所述結構及處理各自只要不產生技術上的矛盾，則可相互組合而構成本發明。 [發明的效果] In addition, each of the said structure and process can be combined with each other as long as no technical contradiction arises, and this invention is comprised. [Effect of the invention]

根據本發明，可提供一種可高精度地進行與資訊處理裝置組合使用的裝置的障礙原因的推測的技術。According to the present invention, it is possible to provide a technology capable of accurately estimating the cause of a failure of a device used in combination with an information processing device.

以下，基於圖式對本發明的實施例加以說明。然而，以下的各例所記載的結構要素的尺寸、材質、形狀、其相對配置等只要無特別記載，則並非意指將本發明的範圍僅限定於該些。Hereinafter, embodiments of the present invention will be described based on the drawings. However, unless otherwise specified, dimensions, materials, shapes, relative arrangements, and the like of constituent elements described in the following examples are not intended to limit the scope of the present invention to these only.

＜適用例＞ （適用例的結構） 本發明例如可適用作外觀檢查裝置的管理系統，所述外觀檢查裝置藉由對利用攝像機構拍攝檢查對象物所得的圖像進行處理，從而對檢查對象物進行檢查。圖1為表示本適用例的裝置管理系統1的概略的示意圖。 ＜Application example＞ (Structure of application example) The present invention is applicable, for example, as a management system of an appearance inspection device that inspects an inspection object by processing an image of the inspection object captured by an imaging mechanism. FIG. 1 is a schematic diagram showing an outline of a device management system 1 of this application example.

裝置管理系統1包含資訊處理終端100及外觀檢查裝置120。資訊處理終端100可與外觀檢查裝置120一體地構成，亦可為與外觀檢查裝置120可通訊地連接的分立的裝置，例如可包含通用的電腦。再者，資訊處理終端100可包含單一的電腦，亦可包含互相協作的多台電腦。外觀檢查裝置120例如為藉由拍攝零件搭載基板等檢查對象物並進行圖像處理從而自動進行檢查對象物的檢查的裝置。The device management system 1 includes an information processing terminal 100 and an appearance inspection device 120 . The information processing terminal 100 may be integrally formed with the appearance inspection device 120 , or may be a separate device communicably connected to the appearance inspection device 120 , and may include, for example, a general-purpose computer. Furthermore, the information processing terminal 100 may include a single computer, or may include multiple computers cooperating with each other. The appearance inspection device 120 is, for example, a device that automatically inspects the inspection object by taking an image of the inspection object such as a component mounting board and performing image processing.

資訊處理終端100包括日誌資料獲取部101、叢集資訊提取部102、叢集間過渡資訊評價部103、有向圖生成部104、基準資料生成部105、異常度計算部106、障礙原因推測部107、顯示部108、記憶部109的各功能部。除此以外，雖未圖示，但亦可包括滑鼠或鍵盤等各種輸入機構、通訊機構等。The information processing terminal 100 includes a log data acquisition unit 101, a cluster information extraction unit 102, an inter-cluster transition information evaluation unit 103, a directed graph generation unit 104, a reference data generation unit 105, an abnormality calculation unit 106, a failure cause estimation unit 107, Functional units of the display unit 108 and the memory unit 109 . In addition, although not shown, various input mechanisms such as a mouse or a keyboard, communication mechanisms, and the like may be included.

外觀檢查裝置120成為包括下述部分的結構：輸送機124，將檢查對象物O搬送至拍攝位置；相機121，拍攝檢查對象物O；以及X平台122及Y平台123，使相機121沿水平方向移動。而且，外觀檢查裝置120雖未圖示，但包括對所拍攝的圖像進行處理的圖像處理部、基於圖像進行檢查的檢查處理部、及輸出檢查結果的輸出處理部等。The appearance inspection device 120 has a structure including the following parts: the conveyor 124 transports the inspection object O to the photographing position; the camera 121 photographs the inspection object O; and the X platform 122 and the Y platform 123 make the camera 121 horizontally move. Furthermore, although not shown, the appearance inspection device 120 includes an image processing unit that processes captured images, an inspection processing unit that performs inspection based on images, an output processing unit that outputs inspection results, and the like.

（障礙原因推測的方法） 本適用例的裝置管理系統1事先準備使用外觀檢查裝置120的正常動作時的多個資料進行學習（模型化）而得的基準資料，於外觀檢查裝置120產生了障礙的情形時，基於所述基準資料來推測障礙原因。 (Method of inferring the cause of the obstacle) The device management system 1 of this application example prepares in advance reference data obtained by learning (modelling) using a plurality of data during normal operation of the visual inspection device 120, and when a failure occurs in the visual inspection device 120, based on the Benchmark data to infer the cause of the obstacle.

具體而言，首先由日誌資料獲取部101獲取和外觀檢查裝置120的正常動作時的控制有關的軟體日誌（以下亦簡稱為日誌）。日誌如圖2所示以文本資訊的形式構成，藉由利用叢集資訊提取部102對所述文本資訊進行處理，從而提取表示外觀檢查裝置120的運轉的各步驟的內容的、叢集資訊。進而，提取叢集間過渡資訊，此叢集間過渡資訊為和外觀檢查裝置120的運轉的一個步驟與另一個步驟之間的過渡有關的資訊。Specifically, first, the log data acquisition unit 101 acquires a software log (hereinafter also simply referred to as a log) related to the control during normal operation of the appearance inspection device 120 . The log is constituted as text information as shown in FIG. 2 , and by processing the text information by the cluster information extraction unit 102 , cluster information representing the contents of each step in the operation of the appearance inspection device 120 is extracted. Furthermore, inter-cluster transition information is extracted, which is information related to the transition between one step and another step in the operation of the visual inspection device 120 .

進而，由有向圖生成部104製作表示所提取的各個叢集資訊及叢集間過渡資訊的關係的有向圖。將所述處理以製作基準資料所需要的程度重覆多次，獲取多個有向圖。進而，由基準資料生成部105將所獲取的多個有向圖替換為矩陣表現，並且對矩陣的各要素算出平均、分散，將其作為基準資料進行保存。Furthermore, the directed graph generating unit 104 creates a directed graph representing the relationship between the extracted cluster information and the inter-cluster transition information. This process is repeated as many times as necessary to create the reference data to obtain a plurality of directed graphs. Furthermore, the reference data generation unit 105 replaces the acquired directed graphs with a matrix representation, and calculates the average and dispersion for each element of the matrix, and stores them as reference data.

另外，於外觀檢查裝置120產生了障礙的情形時，獲取和障礙產生時的控制有關的日誌，藉由與基準資料的製作時相同的處理來製作有向圖，將其變換為矩陣表現。繼而，由異常度計算部106將所獲取的障礙產生時的矩陣資料的各要素與基準資料的矩陣的各要素進行比對，針對各要素算出表示與基準資料的偏離的大小的異常度。繼而，障礙原因推測部107認為與異常度為既定的臨限值以上的要素對應的步驟（或步驟間的過渡）為障礙的原因的可能性高，將所述步驟推測為障礙原因。In addition, when a failure occurs in the visual inspection device 120 , logs related to control at the time of failure are obtained, and a directed graph is created by the same process as when creating reference data, and converted into a matrix representation. Next, the abnormality degree calculating unit 106 compares each element of the acquired matrix data at the occurrence of a failure with each element of the reference data matrix, and calculates an abnormality degree indicating the magnitude of the deviation from the reference data for each element. Next, the failure cause estimation unit 107 considers that a step (or a transition between steps) corresponding to an element whose degree of abnormality is equal to or greater than a predetermined threshold value is highly likely to be a failure cause, and estimates the step as a failure cause.

如以上般，本適用例的裝置管理系統1可僅基於正常動作時的資料來製作基準資料，藉由進行障礙產生時的資料與基準資料的比對從而推測障礙的產生原因。藉此，對於未知的障礙原因亦可高精度地進行推測。As above, the device management system 1 of this application example can create reference data based only on the data during normal operation, and compare the data when a fault occurs with the reference data to estimate the cause of the fault. Thereby, it is possible to estimate with high precision even unknown failure causes.

＜實施形態1＞ 繼而，基於圖1至圖9對本發明的實施形態加以更詳細說明。首先，對本實施形態的裝置管理系統1的資訊處理終端100包括的功能部加以說明。 ＜Embodiment 1＞ Next, an embodiment of the present invention will be described in more detail based on FIGS. 1 to 9 . First, the functional units included in the information processing terminal 100 of the device management system 1 of the present embodiment will be described.

（資訊處理終端的功能） 日誌資料獲取部101獲取日誌，所述日誌為和外觀檢查裝置120的控制有關的、軟體的動作記錄。叢集資訊提取部102自所獲取的日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示外觀檢查裝置120的運轉的各步驟的內容的資訊，所述叢集間過渡資訊為和一個步驟與另一步驟之間的過渡有關的資訊。 (Function of information processing terminal) The log data acquisition unit 101 acquires a log, which is a record of software operations related to the control of the appearance inspection device 120 . The cluster information extracting unit 102 extracts the cluster information and the inter-cluster transition information, which are information indicating the contents of each step in the operation of the visual inspection device 120, and the inter-cluster transition information, from the acquired log collection. Information about the transition between one step and another.

再者，所述叢集間過渡資訊中，包含和外觀檢查裝置120的多個步驟間的過渡的產生頻率有關的資訊，叢集間過渡資訊評價部103至少使用所述產生頻率的資訊來進行所提取的各個叢集間過渡資訊的加權。Furthermore, the inter-cluster transition information includes information related to the frequency of occurrence of transitions between multiple steps of the visual inspection device 120, and the inter-cluster transition information evaluation unit 103 uses at least the information of the occurrence frequency to extract The weighting of transition information between clusters of .

而且，有向圖生成部104以所提取的叢集資訊作為節點，以叢集間過渡資訊作為邊緣，生成表示各個叢集資訊及叢集間過渡資訊的關係的有向圖。基準資料生成部105生成基準資料，此基準資料成為用以推測障礙原因的基準。具體而言，將對外觀檢查裝置120的正常運轉時的多個日誌資料進行採樣所獲取的多個有向圖替換為矩陣表現，並且對矩陣的各要素算出平均、分散，將其作為基準資料保存於記憶部109。Then, the directed graph generation unit 104 generates a directed graph representing the relationship between each cluster information and the inter-cluster transition information using the extracted cluster information as nodes and the inter-cluster transition information as edges. The reference data generating unit 105 generates reference data which is used as a reference for estimating the cause of the failure. Specifically, a plurality of directed graphs obtained by sampling a plurality of log data during normal operation of the visual inspection device 120 are replaced with a matrix representation, and the average and distribution of each element of the matrix are calculated and used as reference data Stored in the memory unit 109 .

異常度計算部106將根據障礙產生時的日誌資料所生成的有向圖替換為矩陣表現，並且將矩陣的各要素與所述基準資料進行比對，藉此針對各要素算出表示與基準資料的偏離的大小的異常度。矩陣的各要素分別與自日誌提取的各個叢集間過渡資訊對應，故而矩陣的各要素的異常度成為對應的各個叢集間過渡資訊的異常度。The abnormality calculation unit 106 replaces the directed graph generated from the log data at the time of failure with a matrix representation, and compares each element of the matrix with the reference data, thereby calculating the relationship between the representation and the reference data for each element. The degree of abnormality in the magnitude of the deviation. Each element of the matrix corresponds to each inter-cluster transition information extracted from the log, so the anomaly degree of each element of the matrix becomes the anomaly degree of each corresponding inter-cluster transition information.

障礙原因推測部107認為與所算出的異常度為既定的臨限值以上的叢集間過渡資訊對應的軟體日誌所示的步驟為障礙的原因的可能性高，將所述步驟推測為障礙原因。The failure cause estimation unit 107 considers that the steps indicated in the software logs corresponding to the inter-cluster transition information whose calculated abnormality degree is equal to or greater than a predetermined threshold value are highly likely to be the cause of the failure, and estimate the steps as the cause of the failure.

顯示部108為液晶顯示器等圖像顯示裝置，顯示包含所述有向圖、所推測的障礙原因、叢集間過渡資訊的異常度等的各種資訊。記憶部109例如可包含讀入專用記憶體（Read Only Memory，ROM）、隨機存取記憶體（Random Access Memory，RAM）等主記憶部與可抹除可程式唯讀記憶體（Erasable Programmable Read Only Memory，EPROM）、硬碟驅動器（Hard Disc Drive，HDD）、可移動媒體（removable media）等輔助記憶部。The display unit 108 is an image display device such as a liquid crystal display, and displays various information including the directed graph, the estimated cause of failure, the degree of abnormality of transition information between clusters, and the like. Memory unit 109 can include, for example, main memory units such as read-only memory (Read Only Memory, ROM), random access memory (Random Access Memory, RAM), and erasable programmable read-only memory (Erasable Programmable Read Only Memory). Memory, EPROM), hard disk drive (Hard Disc Drive, HDD), removable media (removable media) and other auxiliary memory units.

於輔助記憶部，除了操作系統（Operating System，OS）、各種程式以外，可保存所述基準資料、管理對象裝置的運轉實績或維護記錄等各種資訊。再者，藉由將保存於輔助記憶部的程式下載至主記憶部的作業區域並執行，藉由程式的執行來控制各結構部等，從而可實現達成所述般的既定目的之功能部。再者，一部分或全部的功能部亦可藉由特殊應用積體電路（Application Specific Integrated Circuit，ASIC）或現場可程式閘陣列（Field Programmable Gate Array，FPGA）般的硬體電路來實現。In the auxiliary memory unit, besides the operating system (Operating System, OS) and various programs, various information such as the above-mentioned reference data, actual operation results and maintenance records of the managed device can be stored. Furthermore, by downloading and executing the program stored in the auxiliary memory to the work area of the main memory, the execution of the program controls each structural part, thereby realizing the functional part that achieves the above-mentioned intended purpose. Furthermore, a part or all of the functional parts may also be realized by hardware circuits such as Application Specific Integrated Circuit (ASIC) or Field Programmable Gate Array (Field Programmable Gate Array, FPGA).

（障礙原因推測處理的流程） 繼而，對本實施形態的裝置管理系統1的外觀檢查裝置120的障礙原因推測處理的流程加以說明。圖3為表示裝置管理系統1的障礙原因推測處理的一例的流程圖。如圖3所示，裝置管理系統1首先基於自正常運轉時的外觀檢查裝置120獲取的資料來生成基準資料（S101）。 (Flow of fault cause estimation processing) Next, the flow of the failure cause estimation process of the visual inspection device 120 in the device management system 1 of the present embodiment will be described. FIG. 3 is a flowchart showing an example of failure cause estimation processing of the device management system 1 . As shown in FIG. 3 , the device management system 1 first generates reference data based on the data acquired from the visual inspection device 120 during normal operation ( S101 ).

此處，參照圖4，對步驟S101的基準資料生成處理加以詳細說明。圖4為表示本實施形態的基準資料生成處理的子路徑的流程的流程圖。如圖4所示，首先日誌資料獲取部101獲取正常運轉時的日誌資料（S201）。繼而，叢集資訊提取部102進行基於既定的規則將日誌資訊分離的處理（S202）。圖5中表示與此種日誌資訊的分離處理有關的說明圖。如圖5所示，日誌資料的各行包含表示時刻的時刻資料部分及消息字符串，叢集資訊提取部102將日誌以1行為單位分解為時刻資料與消息字符串。此處，對於消息字符串，進行進一步刪除數字及符號而分解為單詞單位的處理。Here, referring to FIG. 4 , the reference data creation process in step S101 will be described in detail. Fig. 4 is a flow chart showing the flow of sub-paths in the reference data creation process of the present embodiment. As shown in FIG. 4 , first, the log data acquisition unit 101 acquires log data during normal operation ( S201 ). Next, the cluster information extraction unit 102 performs a process of separating log information based on a predetermined rule ( S202 ). FIG. 5 shows an explanatory diagram related to such log information separation processing. As shown in FIG. 5 , each row of the log data includes a time data part and a message character string representing the time, and the cluster information extraction unit 102 decomposes the log into time data and message character strings in units of one row. Here, the message character string is further decomposed into word units by deleting numerals and symbols.

繼而，叢集資訊提取部102例如進行使用TF-IDF的方法將日誌各行的單詞集合進行向量化的處理。TF-IDF為公知的方法，故而省略詳細的說明，為基於單詞的出現頻率（Term Frequency，TF）及逆文檔頻率（Inverse Document Frequency，IDF）此兩個指標來求出單詞的重要度的方法。Next, the cluster information extraction unit 102 performs, for example, a process of vectorizing the word set of each line of the log using the method of TF-IDF. TF-IDF is a well-known method, so detailed explanations are omitted. It is a method for calculating the importance of words based on two indicators: word frequency (Term Frequency, TF) and inverse document frequency (Inverse Document Frequency, IDF). .

叢集資訊提取部102進而使用例如K-平均法，如圖6所示，將所有行的向量集合叢集化為例如200個叢集（S203）。圖6為對叢集化的日誌線加以說明的說明圖。再者，K-平均法為廣為人知的叢集化方法，故而省略詳細的說明。The cluster information extraction unit 102 further uses, for example, the K-mean method, as shown in FIG. 6 , to cluster the vector sets of all rows into, for example, 200 clusters ( S203 ). FIG. 6 is an explanatory diagram illustrating clustered log lines. Furthermore, the K-average method is a well-known clustering method, so detailed description is omitted.

繼而，叢集資訊提取部102如圖7所示，基於輸出各日誌線的時刻而生成使叢集編號連號的日誌叢集序列。藉由如此以時序排列叢集編號，從而可獲取和叢集間的過渡有關的資訊。即，可提取如此自日誌的文本消息（單詞）獲得的表示外觀檢查裝置120的運轉的各步驟的內容的叢集資訊、及作為和一個叢集（步驟）與另一個叢集的過渡有關的資訊的叢集間過渡資訊。即，本實施形態中，步驟S202、步驟S203的處理相當於叢集資訊提取步驟。Next, as shown in FIG. 7 , the cluster information extraction unit 102 generates a log cluster sequence in which the cluster numbers are consecutively numbered based on the time at which each log line is output. By arranging the cluster numbers in time series in this way, information on transitions between clusters can be obtained. That is, it is possible to extract cluster information representing the contents of each step of the operation of the visual inspection apparatus 120 obtained from the text message (word) of the log in this way, and a cluster as information related to the transition from one cluster (step) to another cluster transition information. That is, in this embodiment, the processing of step S202 and step S203 corresponds to the cluster information extraction step.

繼而，有向圖生成部104製作以各叢集資訊作為節點且以叢集間過渡資訊作為邊緣的有向圖（S204）。此時，亦可於各節點顯示對應的叢集的叢集編號。Next, the directed graph generating unit 104 creates a directed graph having each cluster information as a node and inter-cluster transition information as an edge ( S204 ). At this time, the cluster number of the corresponding cluster may also be displayed on each node.

繼而，叢集間過渡資訊評價部103基於有向圖的節點間（即對應的叢集間）的過渡頻率，進行有向圖的各邊緣的加權（S205）。圖8A及圖8B中，表示對S205中進行的加權加以說明的說明圖。圖8A為基於輸出日誌線的時刻將與各日誌線對應的叢集的叢集編號以時序自左向右排列的圖。圖8B為表示反映出加權的有向圖的圖。於圖8B的有向圖的邊緣附近，記載有數值，此數字表示所述邊緣（即叢集間的過渡）的產生頻率。若參照圖8A，則自叢集編號2向叢集編號2的過渡產生兩次，自叢集編號2向叢集編號0的過渡產生兩次，自叢集編號0向叢集編號6的過渡產生三次，自叢集編號6向叢集編號0的過渡產生一次，自叢集編號6向叢集編號2的過渡產生一次。另外，於圖8B的有向圖中，將所述過渡的次數顯示於邊緣附近，並且將邊緣的粗細度對應過渡的產生頻率而顯示得粗。Next, the inter-cluster transition information evaluation unit 103 performs weighting of each edge of the directed graph based on the transition frequency between nodes of the directed graph (that is, between corresponding clusters) ( S205 ). FIG. 8A and FIG. 8B show explanatory diagrams for explaining the weighting performed in S205. FIG. 8A is a diagram in which the cluster numbers of the clusters corresponding to each log line are arranged in time series from left to right based on the time when the log line is output. Fig. 8B is a diagram showing a directed graph reflecting weighting. Numerical values are recorded near the edges of the directed graph in FIG. 8B , and the numbers indicate the frequency of occurrence of the edges (ie transitions between clusters). Referring to Figure 8A, the transition from cluster number 2 to cluster number 2 occurs twice, the transition from cluster number 2 to cluster number 0 occurs twice, the transition from cluster number 0 to cluster number 6 occurs three times, and the transition from cluster number 6 occurs three times. The transition from cluster number 6 to cluster number 0 occurs once, and the transition from cluster number 6 to cluster number 2 occurs once. In addition, in the directed graph of FIG. 8B , the number of transitions is displayed near the edge, and the thickness of the edge is displayed thicker corresponding to the frequency of the transition.

藉由如此般進行步驟S201至步驟S205的處理，從而對一個正常運轉時的日誌資料的一系列處理結束。圖9中表示對一個正常運轉時的日誌資料的一系列處理結束時生成的有向圖的一例。By performing the processing from step S201 to step S205 in this way, a series of processing on one log data during normal operation ends. FIG. 9 shows an example of a directed graph generated at the end of a series of processing for one log data during normal operation.

繼而，基準資料生成部105進行下述處理，即：判定是否以生成基準資料所需要的既定數（例如100件）獲取了如所述般經加權的有向圖（S206）。此處，於未獲取既定數的有向圖資料的情形時，回到步驟S201，獲取新的正常運轉時的日誌資料，重覆以後的處理。Next, the reference data generating unit 105 performs a process of determining whether or not the weighted directed graphs as described above have been acquired by the predetermined number (for example, 100) required for generating the reference data ( S206 ). Here, when the predetermined number of directed graph data has not been acquired, return to step S201 to acquire new log data during normal operation, and repeat subsequent processing.

另一方面，於步驟S206中判定為既定數的有向圖資料獲取完畢的情形時，進入步驟S207，基準資料生成部105進行將所獲取的所有的有向圖變換為矩陣表現的處理。具體而言，如下述式（1）所示，進行變換為邊緣權重矩陣W的處理，所述邊緣權重矩陣W將有向圖的自一個節點向另一節點過渡的邊緣的權重設為矩陣的各要素。如所述示例般將叢集的個數設為200個的情形時，邊緣權重矩陣W成為200×200的矩陣。 [數1]

Figure 02_image001
On the other hand, when it is judged in step S206 that the predetermined number of directed graph data has been acquired, the process proceeds to step S207, and the reference data generation unit 105 converts all the acquired directed graphs into matrix representations. Specifically, as shown in the following formula (1), the process of converting to an edge weight matrix W is performed, and the edge weight matrix W sets the weight of the edge transitioning from one node to another node of the directed graph as the weight of the matrix elements. When the number of clusters is set to 200 as in the above example, the edge weight matrix W becomes a 200×200 matrix. [number 1]
Figure 02_image001

此處，矩陣的要素W 00表示自表示叢集編號0的節點0（以下亦以簡稱為節點0的方式記載）向節點0過渡的邊緣（即叢集間過渡資訊）的權重，W n0表示自節點n向節點0過渡的邊緣的權重。即，W ij表示自節點i向節點j過渡的邊緣的權重。 Here, element W 00 of the matrix represents the weight of the edge (that is, inter-cluster transition information) from node 0 indicating cluster number 0 (hereinafter referred to as node 0 for short) to node 0, and W n0 represents the self-node The weight of the edge where n transitions to node 0. That is, W ij represents the weight of an edge transitioning from node i to node j.

基準資料生成部105若對所有的有向圖完成替換為所述矩陣表現的處理，則進行對矩陣的各要素算出平均及分散的處理（S208）。例如，於使用100件的正常運轉時的資料的情形時，作為將100件矩陣資料綜合的結果，分別算出表示100件的平均的一個平均權重矩陣、及表示100件的分散的一個分散權重矩陣作為基準資料（S209）。再者，以下關於平均權重矩陣的各要素，記載為表示100件的W ij的平均的μ ij，關於分散權重矩陣的各要素，記載為表示100件的W ij的分散的σ ij。 When the process of replacing all the directed graphs with the above-mentioned matrix expression is completed, the reference data generation unit 105 performs the process of calculating the average and distribution of each element of the matrix ( S208 ). For example, when using the data of 100 cases of normal operation, as a result of integrating the matrix data of 100 cases, one average weight matrix representing the average of 100 cases and one distribution weight matrix representing the dispersion of 100 cases are respectively calculated. as reference data (S209). Hereinafter, each element of the average weight matrix is described as μ ij representing the average of 100 W ij , and each element of the dispersed weight matrix is described as σ ij representing the distribution of 100 W ij .

如此，若步驟S209的處理結束，則基準資料生成處理（S101）的一系列子路徑結束。使說明回到圖3的表示障礙原因推測處理的流程圖，若步驟S101的處理結束，則基準資料生成部105將所生成的基準資料保存於記憶部109（S102）。In this way, when the process of step S209 ends, a series of sub-routes of the reference material generation process ( S101 ) ends. Referring back to the flowchart showing the failure cause estimation process in FIG. 3 , when the process of step S101 is completed, the reference data generation unit 105 stores the generated reference data in the storage unit 109 ( S102 ).

繼而，於外觀檢查裝置120產生了障礙時，日誌資料獲取部101獲取所述障礙產生時的日誌資料（S103）。繼而，執行下述一系列處理，即：自障礙產生時的日誌資料中進行叢集資訊的提取，基於其而生成經加權的有向圖，獲得將有向圖進行矩陣變換而得的資料（S104）。步驟S104中進行的具體的處理內容與所述步驟S202至步驟S205及步驟S207中進行的處理相同。因此，省略此處的說明。Next, when a failure occurs in the visual inspection device 120 , the log data acquisition unit 101 acquires the log data at the time of the failure ( S103 ). Then, perform the following series of processing, that is: extract cluster information from the log data when the obstacle occurs, generate a weighted directed graph based on it, and obtain the data obtained by matrix transformation of the directed graph (S104 ). The specific processing content performed in step S104 is the same as the processing performed in step S202 to step S205 and step S207. Therefore, description here is omitted.

繼而，異常度計算部106將步驟S104中獲取的障礙產生時的矩陣資料與基準資料進行比對，藉此針對障礙產生時的各矩陣要素算出異常度a ij（S105）。具體而言，基於下述式（2）算出異常度。 [數2]

Figure 02_image003
Next, the abnormality degree calculation unit 106 compares the matrix data obtained in step S104 at the occurrence of the failure with the reference data, thereby calculating the abnormality a ij for each matrix element at the time of the failure ( S105 ). Specifically, the degree of abnormality is calculated based on the following formula (2). [number 2]
Figure 02_image003

繼而，障礙原因推測部107將滿足既定條件（例如超過臨限值等）的異常度的矩陣要素推測為障礙的原因（S106）。繼而，例如於推測為矩陣要素W ij為障礙的原因的情形時，由於矩陣要素W ij具有節點i的叢集資訊及節點j的叢集資訊，故而將該些叢集資訊（或者對應的日誌線）作為表示所推測的障礙原因的資訊而顯示於顯示部108（S107）。 Next, the failure cause estimation unit 107 estimates a matrix element of an abnormality degree satisfying a predetermined condition (for example, exceeding a threshold value) as a cause of failure ( S106 ). Then, for example, when it is inferred that the matrix element W ij is the cause of the obstacle, since the matrix element W ij has the cluster information of node i and the cluster information of node j, these cluster information (or the corresponding log line) are used as Information indicating the estimated failure cause is displayed on the display unit 108 ( S107 ).

如以上般，障礙原因推測處理的一系列流程結束。再者，步驟S101及步驟S102的處理無需於每次進行障礙原因的推測時執行，亦可一次製作基準資料並保存後，自步驟S103的處理開始障礙原因推測處理。當然，亦可適當執行步驟S101及步驟S102的處理，視需要更新基準資料。As above, the series of flow of the failure cause estimation process ends. Furthermore, the processes of step S101 and step S102 do not need to be performed every time the cause of failure is estimated, and the reference data may be created once and stored, and then the cause of failure estimation process may be started from the process of step S103. Of course, the processing of step S101 and step S102 can also be properly performed, and the reference data can be updated as needed.

根據以上般的裝置管理系統1，僅基於正常動作時的資料來製作基準資料，藉由進行障礙產生時的資料與基準資料的比對，從而可算出表示各步驟的要素的異常度，推測障礙的產生原因。藉此，對於未知的障礙原因亦可高精度地推測。而且，藉由利用叢集間的過渡的產生頻率進行加權，從而對於重要度高的事項可適當算出異常度。According to the device management system 1 as above, reference data are created based only on the data during normal operation, and by comparing the data when a fault occurs with the reference data, the abnormality of the elements representing each step can be calculated and the fault can be estimated. the cause of. Thereby, it is possible to estimate with high precision even unknown failure causes. Furthermore, by performing weighting using the frequency of occurrence of transitions between clusters, the degree of abnormality can be appropriately calculated for items with high importance.

（變形例1） 再者，所述實施形態1中，說明了將作為表示所推測的障礙原因的資訊的叢集資訊顯示於顯示部108，但可於顯示部108顯示各種資訊。圖10A及圖10B為表示作為顯示於顯示部108的資訊的一例的有向圖的圖。圖10A表示於各節點顯示有對應的叢集編號的、通常的有向圖。圖10B為表示有向圖的變形顯示例的圖。有向圖生成部104亦可將與各節點對應的叢集資訊所含的單詞按出現次數由多到少的順序提取，並且生成使用所述提取的單詞作為表示與各個節點對應的叢集的內容的資訊的、有向圖（參照圖10B），將其顯示於顯示部108。由此，用戶可基於單詞容易地掌握有向圖的各節點的內容。 (Modification 1) In addition, in the above-mentioned first embodiment, it was described that the cluster information, which is information indicating the estimated failure cause, is displayed on the display unit 108 , but various kinds of information may be displayed on the display unit 108 . 10A and 10B are diagrams showing a directed graph as an example of information displayed on the display unit 108 . FIG. 10A shows a general directed graph in which each node shows a corresponding cluster number. FIG. 10B is a diagram showing a modified display example of a directed graph. The directed graph generation unit 104 may also extract the words included in the cluster information corresponding to each node in descending order of appearance times, and generate an image using the extracted words as the content representing the cluster corresponding to each node. A directed graph of information (see FIG. 10B ) is displayed on the display unit 108 . Thereby, the user can easily grasp the content of each node of the directed graph based on words.

（變形例2） 圖11為表示本實施形態1的進而另一變形例的裝置管理系統2的概略結構的概略圖。再者，以下對於與所述實施例中所說明相同的結構、處理，標註相同的符號，省略重覆的說明。如圖11所示，本變形例的裝置管理系統2於資訊處理終端200中包括提取日誌顯示圖像生成部201，僅此方面與實施形態1的裝置管理系統1不同，其他方面相同。 (Modification 2) FIG. 11 is a schematic diagram showing a schematic configuration of a device management system 2 according to still another modified example of the first embodiment. In the following, the same configurations and processes as those described in the above-mentioned embodiments are assigned the same reference numerals, and redundant descriptions are omitted. As shown in FIG. 11 , the device management system 2 of this modification includes an extracted log display image generator 201 in the information processing terminal 200 , and differs from the device management system 1 of the first embodiment only in this respect, and is the same in other respects.

提取日誌顯示圖像生成部201自日誌的集合中，提取與滿足既定條件的邊緣對應的日誌，作為表示滿足所述既定條件的邊緣（叢集間過渡資訊）的內容的資訊，並且生成表示所述提取的日誌的內容的、提取日誌顯示圖像，顯示於顯示部108。圖12中表示使和所述邊緣的內容有關的提取日誌顯示圖像於有向圖的邊緣附近彈出顯示的狀態的顯示畫面例。The extracted log display image generator 201 extracts, from the set of logs, a log corresponding to an edge satisfying a predetermined condition as information representing the content of an edge (inter-cluster transition information) satisfying the predetermined condition, and generates the The extracted log display image of the content of the extracted log is displayed on the display unit 108 . FIG. 12 shows an example of a display screen in a state where the extracted log display image related to the content of the edge is pop-up displayed near the edge of the directed graph.

再者，所謂此處提及的「滿足既定條件」，可設為異常度超過既定值的情形、用戶藉由滑鼠操作等選擇了所述有向圖的與所述叢集間過渡資訊對應的邊緣的情形等。而且，提取日誌顯示圖像不限於有向圖的邊緣附近，能以任意的態樣顯示，亦可設為於畫面上具有專用的顯示區域的用戶介面。若為此種本變形例的結構，則用戶可迅速確認與叢集間過渡資訊對應的日誌。Furthermore, the so-called "satisfies the predetermined conditions" mentioned here can be set as the situation where the degree of abnormality exceeds the predetermined value, and the user selects the transition information of the directed graph corresponding to the inter-cluster transition information by mouse operation, etc. edge cases, etc. Furthermore, the extracted log display image is not limited to the vicinity of the edge of the directed graph, and can be displayed in any form, and can also be used as a user interface having a dedicated display area on the screen. According to the configuration of this modified example, the user can quickly confirm the log corresponding to the inter-cluster transition information.

＜實施形態2＞ 繼而，基於圖13至圖17對本發明的另一實施形態加以說明。圖13為表示本實施形態的裝置管理系統3的概略結構的概略圖。如圖13所示，本實施形態的裝置管理系統3與裝置管理系統1相比，於資訊處理終端300中包括感測器資料獲取部301的方面不同。而且，本實施形態的叢集間過渡資訊評價部303如後述，於進行一部分不同處理的方面與實施形態1的叢集間過渡資訊評價部103不同。其他方面與實施形態1的裝置管理系統1相同。 ＜Embodiment 2＞ Next, another embodiment of the present invention will be described based on FIGS. 13 to 17 . FIG. 13 is a schematic diagram showing a schematic configuration of the device management system 3 of the present embodiment. As shown in FIG. 13 , the device management system 3 of this embodiment differs from the device management system 1 in that the information processing terminal 300 includes a sensor data acquisition unit 301 . Furthermore, the inter-cluster transition information evaluation unit 303 of the present embodiment differs from the inter-cluster transition information evaluation unit 103 of the first embodiment in that it performs a part of different processing, as will be described later. Other points are the same as the device management system 1 of the first embodiment.

感測器資料獲取部301獲取探測外觀檢查裝置120的硬體（例如輸送機124、相機121、X平台122、Y平台123、輸出裝置等）的狀態的資訊的、感測器資料。感測器資料為記錄硬體的狀態的、時序的數值資料，自外觀檢查裝置120所包括的各種感測器、馬達、位置控制系統等器件獲取。感測器資料以文本形式、二進制形式的哪一種記錄取決於裝置的規格，則只要可獲取時間與數值的對應關係，則可為任何形式。再者，本實施形態中，各種感測器、感測器資料獲取部301等相當於硬體資訊獲取機構。The sensor data acquiring unit 301 acquires sensor data for detecting the state information of the hardware (such as the conveyor 124 , camera 121 , X platform 122 , Y platform 123 , output device, etc.) of the appearance inspection device 120 . The sensor data is time-series numerical data that records the state of the hardware, and is obtained from various sensors, motors, position control systems and other devices included in the visual inspection device 120 . Whether the sensor data is recorded in text form or binary form depends on the specifications of the device, and any form may be used as long as the correspondence relationship between time and value can be obtained. Furthermore, in this embodiment, various sensors, the sensor data acquiring unit 301, etc. are equivalent to hardware information acquiring means.

關於由外觀檢查裝置120進行的自動檢查處理，針對一個檢查對象的處理大致包含六個進程。具體而言，進程分為檢查對象物O的搬入、檢查對象物O的拍攝、圖像處理、良好與否判定、檢查結果輸出、檢查對象物O的搬出。另外，各進程中硬體的動作不同，各進程的長度或重覆次數亦不一定。因此，記錄硬體的狀態的感測器資料亦視對象物而大幅度地變動。即，可謂藉由對硬體的動作（表示所述動作的感測器資料）進行學習從而推測裝置的故障原因並不容易。Regarding the automatic inspection process performed by the visual inspection device 120, the processing for one inspection object roughly includes six processes. Specifically, the process is divided into carrying in of the inspection object O, imaging of the inspection object O, image processing, good or bad judgment, output of inspection results, and export of the inspection object O. In addition, the actions of the hardware in each process are different, and the length or number of repetitions of each process is also not fixed. Therefore, the sensor data that records the state of the hardware also varies greatly depending on the object. That is, it can be said that it is not easy to estimate the failure cause of the device by learning the operation of the hardware (sensor data indicating the operation).

繼而，對本實施形態的裝置管理系統3的外觀檢查裝置120的障礙原因推測處理的流程加以說明。圖14為表示裝置管理系統3的障礙原因推測處理的一例的流程圖。如圖14所示，作為總體的流程，與實施形態1的情形的處理大致相同。Next, the flow of the failure cause estimation process of the visual inspection device 120 in the device management system 3 of the present embodiment will be described. FIG. 14 is a flowchart showing an example of failure cause estimation processing in the device management system 3 . As shown in FIG. 14, the overall flow is almost the same as the processing in the case of the first embodiment.

本實施形態的裝置管理系統3首先基於自正常運轉時的外觀檢查裝置120獲取的資料來生成基準資料（S301）。此處，基於圖15對步驟S301的子路徑加以說明。圖15為表示本實施形態的基準資料生成處理的子路徑的流程的流程圖。The device management system 3 of the present embodiment first generates reference data based on data acquired from the visual inspection device 120 during normal operation ( S301 ). Here, the subroutine of step S301 will be described based on FIG. 15 . Fig. 15 is a flow chart showing the flow of sub-routes in the reference data creation process of the present embodiment.

如圖15所示，本實施形態的基準資料生成時的子路徑亦大致與實施形態1相同，自步驟S201至步驟S205為止進行與實施形態1相同的處理。即，獲取正常運轉時的日誌資料（S201），進行將所述日誌資訊分離的處理（S202），進行所分離的日誌資訊的叢集化（S203），使用叢集化的資料生成有向圖（S204），進行基於叢集間的過渡的產生頻率的、邊緣的加權（S205）。As shown in FIG. 15, the sub-paths in the reference data generation of this embodiment are also substantially the same as in the first embodiment, and the same processing as that in the first embodiment is performed from step S201 to step S205. That is, acquire log data during normal operation (S201), perform processing for separating the log information (S202), perform clustering of the separated log information (S203), and generate a directed graph using the clustered data (S204 ), edge weighting is performed based on the generation frequency of transitions between clusters ( S205 ).

本實施形態中，作為其後續步驟，感測器資料獲取部301獲取外觀檢查裝置120的正常運轉時的感測器資料（S401）。進而，叢集間過渡資訊評價部303進行基於步驟S401中獲取的感測器資料將有向圖的邊緣進一步加權的處理（S402）。In the present embodiment, as a subsequent step, the sensor data acquisition unit 301 acquires sensor data during normal operation of the appearance inspection device 120 ( S401 ). Furthermore, the inter-cluster transition information evaluation unit 303 performs a process of further weighting the edges of the directed graph based on the sensor data acquired in step S401 ( S402 ).

具體而言，叢集間過渡資訊評價部303使用變化點檢測（Change-Finder）算法，將自硬體獲取的感測器資料（時序數值資料）變換為以時序表示各時刻的變化的大小的資料（變化得分）。圖16中表示說明感測器資料與變化得分的關係的說明圖。再者，關於Change-Finder算法，為公知的方法，故而省略詳細的說明。Specifically, the inter-cluster transition information evaluation unit 303 uses a Change-Finder algorithm to convert sensor data (time-series numerical data) acquired from the hardware into data representing the magnitude of change at each time in time series. (change score). FIG. 16 is an explanatory diagram illustrating the relationship between the sensor data and the change score. In addition, since the Change-Finder algorithm is a known method, a detailed description thereof will be omitted.

此時，如圖16所示，成為變化得分的基礎的感測器資料的數值範圍視成為對象的硬體等而不同，故而變化得分亦反映出此種差異，數值範圍不均一。因此，將所有的變化得分於0～1之間歸一化。At this time, as shown in FIG. 16 , the numerical range of the sensor data used as the basis of the change score differs depending on the target hardware, etc., so the change score also reflects this difference, and the numerical range is not uniform. Therefore, all variance scores were normalized between 0 and 1.

繼而，叢集間過渡資訊評價部303對叢集資訊提取部102所生成的日誌叢集序列（參照圖7）映射變化得分，使哪個叢集間的過渡時變化得分大進行對應。圖17中表示映射有變化得分的日誌叢集序列的示例。Next, the inter-cluster transition information evaluation unit 303 maps change scores to the log cluster sequence (see FIG. 7 ) generated by the cluster information extraction unit 102 , and associates which cluster has a larger transition time change score. An example of a sequence of log clusters mapped with change scores is shown in FIG. 17 .

繼而，叢集間過渡資訊評價部303藉由反映出有向圖的各邊緣（各叢集間過渡）的、變化得分的大小，從而對有向圖的各邊緣進行基於硬體資訊的加權。Next, the inter-cluster transition information evaluation unit 303 performs weighting based on hardware information for each edge of the directed graph (inter-cluster transition) by reflecting the magnitude of the change score.

如此，若步驟S402的處理結束，則進入步驟S206，基準資料生成處理（S301）的子路徑的以後的處理與實施形態1中所說明相同，故而省略此處的說明。In this way, when the processing of step S402 is completed, the process proceeds to step S206, and subsequent processing in the sub-path of the reference data generation processing (S301) is the same as that described in Embodiment 1, and thus the description here is omitted.

使說明回到圖14的表示障礙原因推測處理的流程圖，若步驟S301的處理結束，則基準資料生成部105將所生成的基準資料保存於記憶部109（S102）。繼而，於外觀檢查裝置120產生障礙時，日誌資料獲取部101獲取障礙產生時的日誌資料（S103），感測器資料獲取部301獲取表示障礙產生時的硬體狀態的感測器資料（S302）。Referring back to the flowchart showing the failure cause estimation process in FIG. 14 , when the processing in step S301 is completed, the reference data generation unit 105 stores the generated reference data in the storage unit 109 ( S102 ). Next, when a failure occurs in the appearance inspection device 120, the log data acquisition unit 101 acquires the log data when the failure occurs (S103), and the sensor data acquisition unit 301 acquires sensor data indicating the hardware state when the failure occurs (S302 ).

繼而，於資訊處理終端300中執行下述一系列處理，即：自障礙產生時的日誌資料中進行叢集資訊的提取，基於其而生成經加權的有向圖，進而對所述有向圖進行基於感測器資料的加權，獲取將有向圖進行矩陣變換而得的資料（S303）。再者，步驟S303中進行的具體的處理內容與所述步驟S301的子路徑的步驟S202至步驟S205、步驟S402及步驟S207中進行的處理相同。因此，省略重覆的說明。Then, the following series of processes are executed in the information processing terminal 300, that is, cluster information is extracted from the log data when an obstacle occurs, a weighted directed graph is generated based on it, and then the directed graph is processed Based on the weighting of the sensor data, obtain data obtained by performing matrix transformation on the directed graph ( S303 ). Furthermore, the specific processing content performed in step S303 is the same as the processing performed in steps S202 to S205, step S402 and step S207 of the sub-path of step S301. Therefore, repeated descriptions are omitted.

而且，本實施形態的步驟S105以後的處理亦與實施形態1的內容相同，故而省略此處的說明。In addition, the processing after step S105 in this embodiment is also the same as that in the first embodiment, so the description here is omitted.

根據本實施形態的裝置管理系統3，可使用表示外觀檢查裝置120的硬體狀態的感測器資料，進一步進行有向圖的邊緣（即叢集間過渡資訊）的加權。僅以叢集間的過渡的產生頻率進行了加權的有向圖有下述可能性，即：將與外觀檢查裝置120的進程的切換、或硬體狀態的變化對應的（頻率低的）重要的叢集間過渡資訊評價得過低。相對於此，本實施形態的裝置管理系統3中，檢測感測器資料的重要的變化點，基於其來進一步進行邊緣的加權，故而可抑制將重要的叢集間過渡資訊評價得過低，獲得更高精度的障礙原因的推測結果。According to the device management system 3 of this embodiment, the sensor data representing the hardware state of the visual inspection device 120 can be used to further perform weighting on the edges of the directed graph (ie, inter-cluster transition information). A directed graph weighted only by the occurrence frequency of transitions between clusters may include (low-frequency) important Inter-cluster transition information is undervalued. On the other hand, in the device management system 3 of the present embodiment, important change points of sensor data are detected, and edge weighting is further performed based on them, so it is possible to suppress the overestimation of important inter-cluster transition information and obtain The result of guessing the cause of the failure with higher precision.

＜其他＞ 所述各實施形態僅對本發明進行例示性說明，本發明不限定於所述具體形態。本發明可於其技術思想的範圍內進行各種變形及組合。例如，所述實施形態中，說明了以外觀檢查裝置作為對象的管理系統，但裝置管理系統的管理對象的裝置不限於此。如上文所述，不使用障礙產生時的異常的資料，而使用僅利用正常運作時的資料進行學習而得的基準資料來進行障礙的產生原因的推測，故而可僅使用裝置的實際運用線可收集的資料來進行運用，因而可將本發明適用於各種裝置。 ＜Other＞ Each of the above-mentioned embodiments is merely an illustration of the present invention, and the present invention is not limited to the above-mentioned specific embodiments. The present invention can be variously modified and combined within the scope of the technical idea. For example, in the above-mentioned embodiments, the management system targeted at the visual inspection device was described, but the device managed by the device management system is not limited to this. As mentioned above, the cause of the failure can be estimated using only the reference data obtained by learning only the data during normal operation without using the abnormal data when the failure occurs, so that only the actual operating line of the device can be used. The collected data can be used, so the present invention can be applied to various devices.

而且，作為正常運轉時的資料，可不將裝置所處理的對象物（檢查或加工的對象）限定於一個，而收集對多個對象物進行處理時的資料後，將該些混合而生成基準資料。於該情形時，亦不需要對象物於哪個時機如何變更等資料，可僅由軟體日誌及感測器資料來生成基準資料。In addition, as the data during normal operation, the object processed by the device (the object of inspection or processing) is not limited to one, but the data of processing multiple objects can be collected, and these can be mixed to generate reference data . In this case, data such as when and how the object changes are not required, and reference data can be generated only from software logs and sensor data.

而且，所述實施形態中設為包含管理對象的裝置的系統，但亦可僅將所述實施形態的資訊處理終端理解為本發明的管理系統。即，本發明亦可理解為包含與管理對象的裝置分立地構成的資訊處理終端的、裝置管理終端。Furthermore, in the above-mentioned embodiment, the system including the devices to be managed can be considered, but only the information processing terminal of the above-mentioned embodiment can be understood as the management system of the present invention. That is, the present invention can also be understood as a device management terminal including an information processing terminal configured separately from a managed device.

＜附註1＞ 一種裝置管理系統，為裝置的管理系統（1、2、3），其特徵在於包括： 日誌資料獲取機構（101），獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作記錄； 叢集資訊提取機構（102），自所述獲取的日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示所述裝置的運轉的各步驟的內容的資訊，所述叢集間過渡資訊為和一個所述步驟與另一個所述步驟之間的過渡有關的資訊； 異常度計算機構（106），算出所述提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測機構（107），基於由所述異常度計算機構所算出的所述異常度來推測所述裝置的障礙原因。 ＜Note 1＞ A device management system is a device management system (1, 2, 3), characterized by comprising: A log data acquisition mechanism (101), which acquires a log, the log is a record of software actions related to the control of the device; A cluster information extracting unit (102), extracting cluster information and inter-cluster transition information from the acquired log collection, the cluster information is information indicating the contents of each step in the operation of the device, and the inter-cluster transition information is information relating to a transition between one said step and another said step; An abnormality calculation mechanism (106), which calculates the abnormality of the extracted transition information between clusters; and A failure cause estimation means (107) estimates a failure cause of the device based on the abnormality degree calculated by the abnormality degree calculation means.

＜附註2＞ 一種裝置的障礙原因推測方法，推測裝置的障礙原因，且包含： 日誌資料獲取步驟（S201、S103），獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作歷程資訊； 叢集資訊提取步驟（S202、S203），自所獲取的所述日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示由所述裝置進行的處理的各步驟的內容的資訊，所述叢集間過渡資訊為和所述裝置的多個所述步驟間的過渡有關的資訊； 異常度計算步驟（S105），算出所述提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測步驟（S106），基於由所述異常度計算機構所算出的所述異常度來推測所述裝置的障礙原因。 ＜Note 2＞ A method for estimating the cause of a failure of a device, estimating the cause of a failure of the device, and comprising: The step of obtaining log data (S201, S103), obtaining a log, the log is information about the operation history of the software related to the control of the device; a cluster information extraction step (S202, S203), extracting cluster information and inter-cluster transition information from the acquired set of logs, the cluster information being information representing the contents of each step of processing performed by the device, said inter-cluster transition information is information relating to transitions between said steps of said device; An abnormality calculation step (S105), calculating the abnormality of the extracted transition information between clusters; and The failure cause estimation step ( S106 ) estimates a failure cause of the device based on the abnormality degree calculated by the abnormality degree calculation means.

1、2、3:裝置管理系統 100、200、300:資訊處理終端 101:日誌資料獲取部 102:叢集資訊提取部 103:叢集間過渡資訊評價部 104:有向圖生成部 105:基準資料生成部 106:異常度計算部 107:障礙原因推測部 108:顯示部 109:記憶部 120:外觀檢查裝置 121:相機 122:X平台 123:Y平台 124:輸送機 201:提取日誌顯示圖像生成部 301:感測器資料獲取部 O:檢查對象物 S101～S107、S201～S209、S301～S303、S401、S402:步驟 1, 2, 3: Device management system 100, 200, 300: information processing terminal 101: Log Data Acquisition Department 102:Cluster Information Extraction Department 103: Inter-cluster Transition Information Evaluation Department 104:Directed graph generation unit 105: Benchmark data generation department 106: Abnormal Degree Calculation Department 107: Obstacle cause speculation department 108: display part 109: memory department 120: Appearance inspection device 121: camera 122: X platform 123: Y platform 124: Conveyor 201: Extraction log display image generation part 301: Sensor Data Acquisition Department O: check object S101～S107, S201～S209, S301～S303, S401, S402: steps

圖1為表示實施形態1的裝置管理系統的概略的示意圖。 圖2為表示軟體日誌的示例的說明圖。 圖3為表示由實施形態1的裝置管理系統進行的處理的流程的流程圖。 圖4為表示實施形態1的裝置管理系統的處理的子路徑的流程圖。 圖5為對軟體日誌的分離處理加以說明的說明圖。 圖6為對叢集化的日誌線（logline）加以說明的說明圖。 圖7為對由實施形態1的裝置管理系統所生成的日誌叢集序列加以說明的說明圖。 圖8A為對由實施形態1的裝置管理系統所生成的有向圖加以說明的第一圖。圖8B為對由實施形態1的裝置管理系統所生成的有向圖加以說明的第二圖。 圖9為表示由實施形態1的裝置管理系統所生成的有向圖的一例的圖。 圖10A為表示實施形態1的變形例中畫面顯示的有向圖的一例的圖。圖10B為表示實施形態1的變形例中畫面顯示的有向圖的另一例的圖。 圖11為表示實施形態1的另一變形例的裝置管理系統的概略的示意圖。 圖12為對由實施形態1的另一變形例的裝置管理系統所顯示的畫面的一例加以說明的圖。 圖13為表示實施形態2的裝置管理系統的概略的示意圖 圖14為表示由實施形態2的裝置管理系統所進行的處理的流程的流程圖。 圖15為表示實施形態2的裝置管理系統的處理的子路徑的流程圖。 圖16為表示感測器資料與變化得分的關係的說明圖。 圖17為表示由實施形態2的裝置管理系統所生成的、映射有變化得分的日誌叢集序列的示例的說明圖。 FIG. 1 is a schematic diagram showing an outline of a device management system according to Embodiment 1. As shown in FIG. FIG. 2 is an explanatory diagram showing an example of a software log. Fig. 3 is a flowchart showing the flow of processing performed by the device management system according to the first embodiment. Fig. 4 is a flowchart showing sub-routes of processing in the device management system according to the first embodiment. FIG. 5 is an explanatory diagram explaining software log separation processing. FIG. 6 is an explanatory diagram illustrating clustered loglines. Fig. 7 is an explanatory diagram illustrating a log cluster sequence generated by the device management system according to the first embodiment. Fig. 8A is a first diagram illustrating a directed graph generated by the device management system according to the first embodiment. Fig. 8B is a second diagram illustrating a directed graph generated by the device management system according to the first embodiment. Fig. 9 is a diagram showing an example of a directed graph generated by the device management system according to the first embodiment. Fig. 10A is a diagram showing an example of a directed graph displayed on a screen in a modification of the first embodiment. Fig. 10B is a diagram showing another example of the directed graph displayed on the screen in the modified example of the first embodiment. Fig. 11 is a schematic diagram showing an outline of a device management system according to another modification of the first embodiment. Fig. 12 is a diagram illustrating an example of a screen displayed by the device management system according to another modification of the first embodiment. Fig. 13 is a schematic diagram showing an outline of a device management system according to Embodiment 2 Fig. 14 is a flowchart showing the flow of processing performed by the device management system according to the second embodiment. Fig. 15 is a flowchart showing sub-routes of processing in the device management system according to the second embodiment. FIG. 16 is an explanatory diagram showing the relationship between sensor data and change scores. Fig. 17 is an explanatory diagram showing an example of a log cluster sequence to which a change score is mapped, generated by the device management system according to the second embodiment.

1:裝置管理系統 1: Device management system

100:資訊處理終端 100: Information processing terminal

101:日誌資料獲取部 101: Log Data Acquisition Department

102:叢集資訊提取部 102:Cluster Information Extraction Department

103:叢集間過渡資訊評價部 103: Inter-cluster Transition Information Evaluation Department

104:有向圖生成部 104:Directed graph generation unit

105:基準資料生成部 105: Benchmark data generation department

106:異常度計算部 106: Abnormal Degree Calculation Department

107:障礙原因推測部 107: Obstacle cause speculation department

108:顯示部 108: display part

109:記憶部 109: memory department

120:外觀檢查裝置 120: Appearance inspection device

121:相機 121: camera

122:X平台 122: X platform

123:Y平台 123: Y platform

124:輸送機 124: Conveyor

O:檢查對象物 O: check object

Claims (15)

一種裝置管理系統，為裝置的管理系統，其特徵在於包括： 日誌資料獲取機構，獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作記錄； 叢集資訊提取機構，自獲取的所述日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示所述裝置的運轉的各步驟的內容的資訊，所述叢集間過渡資訊為和一個所述步驟與另一個所述步驟之間的過渡有關的資訊； 異常度計算機構，算出提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測機構，基於由所述異常度計算機構所算出的所述異常度來推測所述裝置的障礙原因。 A device management system is a device management system, characterized in that it comprises: The log data acquisition mechanism acquires a log, and the log is a record of software actions related to the control of the device; a cluster information extracting unit extracts cluster information and inter-cluster transition information from the acquired collection of logs, the cluster information is information indicating the contents of each step in the operation of the device, and the inter-cluster transition information is and information relating to transitions between one said step and another said step; an abnormality calculation mechanism, which calculates the abnormality of the extracted transition information between the clusters; and The failure cause estimation means estimates a failure cause of the device based on the abnormality degree calculated by the abnormality degree calculation means. 如請求項1所述的裝置管理系統，其中所述異常度計算機構基於所述裝置的正常時的所述叢集間過渡資訊，來算出提取的各個所述叢集間過渡資訊的異常度。The device management system according to claim 1, wherein the abnormality calculation unit calculates the abnormality of each of the extracted inter-cluster transition information based on the inter-cluster transition information when the device is normal. 如請求項1或請求項2所述的裝置管理系統，其中所述叢集間過渡資訊中，包含和所述裝置的多個所述步驟間的過渡的產生頻率有關的資訊， 所述裝置管理系統更包括：叢集間過渡資訊評價機構，基於所述產生頻率來進行提取的各個所述叢集間過渡資訊的加權， 所述異常度計算機構使用進行了所述加權的資訊來算出所述異常度。 The device management system according to claim 1 or claim 2, wherein the inter-cluster transition information includes information related to the frequency of occurrence of transitions between multiple steps of the device, The device management system further includes: an inter-cluster transition information evaluation mechanism that weights each of the inter-cluster transition information extracted based on the generation frequency, The abnormality degree calculating means calculates the abnormality degree using the weighted information. 如請求項3所述的裝置管理系統，更包括： 硬體資訊獲取機構，獲取和所述裝置的硬體的狀態有關的硬體資訊， 所述叢集間過渡資訊評價機構基於所述硬體資訊獲取機構所獲取的所述硬體資訊，進一步進行提取的各個所述叢集間過渡資訊的加權。 The device management system as described in claim 3, further comprising: a hardware information acquisition mechanism for acquiring hardware information related to the state of the hardware of the device, The inter-cluster transition information evaluation unit further weights the extracted inter-cluster transition information based on the hardware information acquired by the hardware information acquisition unit. 如請求項1或請求項2所述的裝置管理系統，其中所述障礙原因推測機構推測為，於由所述異常度計算機構所算出的所述異常度滿足既定條件的所述叢集間過渡資訊所指定的步驟中，存在所述裝置的障礙原因。The device management system according to claim 1 or claim 2, wherein the failure cause estimation unit infers that the inter-cluster transition information when the abnormality degree calculated by the abnormality degree calculation unit satisfies a predetermined condition In the specified step, there is a cause of failure of the device. 如請求項1或請求項2所述的裝置管理系統，更包括： 顯示機構，能夠顯示表示由所述異常度計算機構所算出的所述異常度、及/或由所述障礙原因推測機構所推測的所述障礙原因的資訊。 The device management system as described in claim 1 or claim 2, further comprising: The display means can display information indicating the abnormality degree calculated by the abnormality degree calculation means and/or the failure cause estimated by the failure cause estimation means. 如請求項6所述的裝置管理系統，更包括： 有向圖生成機構，以所述叢集資訊作為節點，以所述叢集間過渡資訊作為邊緣，生成表示各個所述叢集資訊及所述叢集間過渡資訊的關係的有向圖， 所述顯示機構能夠顯示所述有向圖。 The device management system as described in claim 6, further comprising: a directed graph generation mechanism, using the cluster information as a node and the inter-cluster transition information as an edge, to generate a directed graph representing the relationship between each of the cluster information and the inter-cluster transition information, The display mechanism is capable of displaying the directed graph. 如請求項7所述的裝置管理系統，其中所述叢集間過渡資訊藉由既定的方法進行了增加重要度的評價的加權， 所述有向圖生成機構生成能夠視認各個所述叢集間過渡資訊的所述加權的、所述有向圖。 The device management system according to claim 7, wherein the inter-cluster transition information is weighted to increase the importance of evaluation by a predetermined method, The directed graph generating means generates the weighted directed graph capable of visually recognizing each of the inter-cluster transition information. 如請求項8所述的裝置管理系統，其中所述有向圖生成機構藉由將表示所述叢集間過渡資訊的所述加權的數值顯示於所述邊緣的附近，從而生成以能夠視認的方式表現所述加權的有向圖。The device management system according to claim 8, wherein the directed graph generation mechanism displays the weighted value representing the inter-cluster transition information near the edge, thereby generating the graph in a visually recognizable manner. A directed graph representing the weights. 如請求項8所述的裝置管理系統，其中所述有向圖生成機構藉由對表示各個所述叢集間過渡資訊的所述邊緣的清晰度設置差異進行顯示，從而生成以能夠視認的方式表現所述加權的有向圖。The device management system according to claim 8, wherein the directed graph generation mechanism generates a visually recognizable representation by displaying the difference in definition of the edges representing the transition information between each of the clusters. The weighted directed graph. 如請求項7所述的裝置管理系統，其中所述叢集資訊中包含作為自所述日誌提取的文本資訊的單詞， 所述有向圖生成機構將各個所述叢集資訊所含的所述單詞按出現次數由多到少的順序進行提取，並且生成使用提取的所述單詞作為表示所述叢集資訊的內容的資訊的、所述有向圖。 The device management system according to claim 7, wherein the cluster information includes words as text information extracted from the log, The directed graph generation mechanism extracts the words contained in each of the cluster information in descending order of occurrence times, and generates a graph using the extracted words as information representing the content of the cluster information. , the directed graph. 如請求項7所述的裝置管理系統，更包括： 提取日誌顯示圖像生成機構，自所述日誌的集合中，提取與滿足既定條件的所述叢集間過渡資訊對應的日誌，作為表示滿足所述既定條件的所述叢集間過渡資訊的內容的資訊，並且生成表示提取的所述日誌的內容的、提取日誌顯示圖像， 所述顯示機構能夠顯示所述提取日誌顯示圖像。 The device management system as described in claim 7, further comprising: extracting log display image generation means extracting, from the set of logs, a log corresponding to the inter-cluster transition information satisfying a predetermined condition as information representing the content of the inter-cluster transition information satisfying the predetermined condition , and generate an extracted log display image representing the content of the extracted log, The display means can display the extraction log display image. 如請求項12所述的裝置管理系統，其中所述提取日誌顯示圖像彈出顯示於表示與所述提取日誌顯示圖像所示的提取日誌對應的所述叢集間過渡資訊的、所述邊緣的附近。The device management system according to claim 12, wherein the extraction log display image is pop-up displayed on the edge of the edge representing the inter-cluster transition information corresponding to the extraction log shown in the extraction log display image nearby. 一種裝置的障礙原因推測方法，推測裝置的障礙原因，且包括： 日誌資料獲取步驟，獲取日誌，所述日誌為和所述裝置的控制有關的、軟體的動作歷程資訊； 叢集資訊提取步驟，自所獲取的所述日誌的集合中提取叢集資訊及叢集間過渡資訊，所述叢集資訊為表示由所述裝置進行的處理的各步驟的內容的資訊，所述叢集間過渡資訊為和所述裝置的多個所述步驟間的過渡有關的資訊； 異常度計算步驟，算出提取的各個所述叢集間過渡資訊的異常度；以及 障礙原因推測步驟，基於所述異常度計算步驟中算出的所述異常度來推測所述裝置的障礙原因。 A method for estimating the cause of a device failure, which includes: The log data obtaining step is to obtain a log, and the log is related to the control of the device, and the information of the action history of the software; a cluster information extraction step of extracting cluster information and inter-cluster transition information from the acquired set of logs, the cluster information being information indicating the contents of each step of processing performed by the device, and the inter-cluster transition information information is information relating to transitions between a plurality of said steps of said device; An abnormality calculation step, calculating the abnormality of the extracted transition information between the clusters; and The failure cause estimation step estimates a failure cause of the device based on the abnormality degree calculated in the abnormality degree calculation step. 一種記憶媒體，非暫時性地記憶用以使電腦執行如請求項14所述的方法的各步驟的程式。A memory medium that non-temporarily stores the program used to make the computer execute the steps of the method described in Claim 14.
TW111132172A 2021-09-06 2022-08-26 Device management system, device failure cause estimation method, and memory medium for non-temporarily storing programs TWI824681B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2021145011A JP2023038095A (en) 2021-09-06 2021-09-06 Device management system, method of estimating failure cause of device, and program
JP2021-145011 2021-09-06

Publications (2)

Publication Number Publication Date
TW202311879A true TW202311879A (en) 2023-03-16
TWI824681B TWI824681B (en) 2023-12-01

Family

ID=85226567

Family Applications (1)

Application Number Title Priority Date Filing Date
TW111132172A TWI824681B (en) 2021-09-06 2022-08-26 Device management system, device failure cause estimation method, and memory medium for non-temporarily storing programs

Country Status (4)

Country Link
JP (1) JP2023038095A (en)
CN (1) CN115774656A (en)
DE (1) DE102022208921A1 (en)
TW (1) TWI824681B (en)

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7650109B2 (en) * 2002-12-26 2010-01-19 Japan Tobacco Inc. Analyzing system, analyzing method using the same, and system for collecting survey results for use in analysis
TWI684139B (en) * 2018-10-16 2020-02-01 中華電信股份有限公司 System and method of learning-based prediction for anomalies within a base station
JP7467168B2 (en) 2020-03-11 2024-04-15 キヤノン株式会社 Image pickup device and imaging device

Also Published As

Publication number Publication date
TWI824681B (en) 2023-12-01
CN115774656A (en) 2023-03-10
DE102022208921A1 (en) 2023-03-09
JP2023038095A (en) 2023-03-16

Similar Documents

Publication Publication Date Title
JP6723669B2 (en) Abnormality detection program, abnormality detection method, and abnormality detection device
US10373065B2 (en) Generating database cluster health alerts using machine learning
JP5431235B2 (en) Equipment condition monitoring method and apparatus
CN102265227B (en) Method and apparatus for creating state estimation models in machine condition monitoring
US20170193372A1 (en) Health Management Using Distances for Segmented Time Series
CN111459700A (en) Method and apparatus for diagnosing device failure, diagnostic device, and storage medium
JP6521096B2 (en) Display method, display device, and program
JP2009064407A (en) Process analysis apparatus, process analysis method, and process analysis program
JP6489235B2 (en) System analysis method, system analysis apparatus, and program
JP2019016039A (en) Method for diagnosing abnormal state of process and abnormal state diagnosis apparatus
JP2009086896A (en) Failure prediction system and failure prediction method for computer
JP6950504B2 (en) Abnormal candidate extraction program, abnormal candidate extraction method and abnormal candidate extraction device
JP6917805B2 (en) Data filtering device and method
JP6918735B2 (en) Monitoring device, monitoring method and monitoring program
JP2009110262A (en) Filter device, filtering method for filter device and filtering program
TWI824681B (en) Device management system, device failure cause estimation method, and memory medium for non-temporarily storing programs
US20220222402A1 (en) Information processing device, information processing method, and information processing program
JP7019364B2 (en) Monitoring device, monitoring method, monitoring program, display device, display method and display program
JP6247777B2 (en) Abnormality diagnosis apparatus and abnormality diagnosis method
JP2018014093A (en) Process state diagnosis method and state diagnosis device
JP6627258B2 (en) System model generation support device, system model generation support method, and program
JP6787873B2 (en) Abnormal type judgment device, abnormal type judgment method and program
JP2010102462A (en) Apparatus, method and program for estimating trouble
JP6797789B2 (en) State estimation device, state estimation method and program
JP6973445B2 (en) Display method, display device, and program