JP2013033459A

JP2013033459A - Abnormality monitoring system and abnormality monitoring method

Info

Publication number: JP2013033459A
Application number: JP2012141812A
Authority: JP
Inventors: Takehide Hirata; 丈英平田; Yoshito Goto; 義人後藤; Takushi Kagawa; 卓士香川
Original assignee: JFE Steel Corp
Current assignee: JFE Steel Corp
Priority date: 2011-06-29
Filing date: 2012-06-25
Publication date: 2013-02-14
Anticipated expiration: 2032-06-25
Also published as: JP5991042B2

Abstract

PROBLEM TO BE SOLVED: To provide an abnormality monitoring system capable of simply detecting an abnormality of equipment which is driven by an electric motor at an early stage with high accuracy, etc.SOLUTION: A statistical monitoring indexing processing part 275 acquires time series data of a torque current performance, a speed performance, and a travel performance of an electric motor in a normal operation of equipment in advance, and calculates an average behavior of correlation among a current performance twice integration value obtained by integrating the torque current value twice, a speed performance once integration value obtained by integrating the speed performance once, and a travel performance value. On the other hand, a statistical monitoring indexing processing part 375 calculates a deviation degree from the average behavior of the correlation in the normal operation of the equipment for the correlation among the current performance twice integration value obtained by integrating the torque current performance twice, the speed performance once integration value obtained by integrating the speed performance once, and the travel performance value to be obtained from the equipment in the operation. A determination processing part 377 determines an abnormality of the equipment based on the calculated deviation degree.

Description

本発明は、プラント設備等の監視対象から得られる時系列データをもとに、監視対象の異常を検知する異常監視システムおよび異常監視方法に関するものである。 The present invention relates to an abnormality monitoring system and an abnormality monitoring method for detecting an abnormality of a monitoring object based on time series data obtained from the monitoring object such as plant equipment.

従来から、鉄鋼製品の製造プラント等で多くを占める、電動機（モータ）で駆動される設備の状態監視では、一般に、その監視対象の設備から得られる信号データに対して適当な上下限を設定することで監視対象の設備の異常を検知している。 Conventionally, in the monitoring of the state of equipment driven by an electric motor (motor) that occupies a large amount in a steel product manufacturing plant or the like, generally appropriate upper and lower limits are set for signal data obtained from the equipment to be monitored. Therefore, the abnormality of the monitored equipment is detected.

たとえば、特許文献１には、電動機の位置指令値と位置実績値、速度指令値と速度実績値の差異信号から確率密度関数を演算し、演算された確率密度関数と予め設定された閾値（上下限）との対比を行うことで電動機の異常を検知して、監視対象の設備の異常を検知することが記載されている。また、特許文献２には、電動機の負荷が一定の定常状態において、電動機の電流値を予め設定された上下限と対比させ、電流値の変動が発生することで電動機と負荷との間の減速機の異常を検知することが記載されている。 For example, in Patent Document 1, a probability density function is calculated from a difference signal between a position command value and actual position value of an electric motor, and a speed command value and actual speed value, and the calculated probability density function and a preset threshold value (above It is described that the abnormality of the motor is detected by comparing with the lower limit), and the abnormality of the equipment to be monitored is detected. Further, in Patent Document 2, in a steady state where the load of the motor is constant, the current value of the motor is compared with upper and lower limits set in advance, and the current value fluctuates, so that deceleration between the motor and the load occurs. It describes the detection of machine abnormalities.

特開２００６−１５８０３１号公報JP 2006-158031 A 特開２００６−１０２８８９号公報JP 2006-102889 A

しかしながら、電動機で駆動される設備では、異常の初期段階では速度や位置の信号データに目立った異常が見られない場合が多い。したがって特許文献１の方法では、初期の異常を捉えにくく、速度や位置の信号データに異常が出現するタイミングでは手遅れになりやすい。また、特許文献２の方法によれば、異常の初期段階でのトルク電流の変化は上下限を設定しても捉えにくい。加えて、鉄鋼製品の製造プラントのように多数の設備で構成される場合に、個々の設備について適切な上下限を設定しなければならず、マンパワーやコストが過大になる。 However, in equipment driven by an electric motor, there are many cases where no noticeable abnormality is observed in the speed and position signal data at the initial stage of abnormality. Therefore, in the method of Patent Document 1, it is difficult to catch the initial abnormality, and it is likely to be too late at the timing when the abnormality appears in the speed and position signal data. Further, according to the method of Patent Document 2, the change in torque current at the initial stage of abnormality is difficult to catch even if upper and lower limits are set. In addition, when it is configured with a large number of facilities such as a steel product manufacturing plant, it is necessary to set appropriate upper and lower limits for each facility, resulting in excessive manpower and cost.

本発明は、上記に鑑みてなされたものであって、電動機で駆動される設備の初期の異常を高精度かつ簡易に検知することができる異常監視システムおよび異常監視方法を提供することを目的とする。 The present invention has been made in view of the above, and it is an object of the present invention to provide an abnormality monitoring system and an abnormality monitoring method that can detect an initial abnormality of equipment driven by an electric motor with high accuracy and ease. To do.

上述した課題を解決し、目的を達成するため、本発明に係る異常監視システムは、電動機で駆動される設備から得られる時系列データをもとに監視対象の前記設備の異常を検知する異常監視システムであって、事前に設備の正常動作時の前記電動機のトルク電流実績と、速度実績と、移動量実績の時系列データを取得して、前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移動量実績値との相関関係の平均的挙動を求める正常時平均的挙動取得手段と、操業時において前記設備から得られる前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移動量実績値との相関関係について、前記設備の正常動作時の相関関係の平均的挙動からの外れ度を算出する操業時外れ度算出手段と、前記操業時外れ度算出手段が算出した前記外れ度をもとに前記設備の異常を判定する判定手段と、を備えることを特徴とする。 In order to solve the above-described problems and achieve the object, the abnormality monitoring system according to the present invention is an abnormality monitoring that detects abnormality of the equipment to be monitored based on time-series data obtained from equipment driven by an electric motor. The system obtains time series data of the torque current results, speed results, and travel results of the motor during normal operation of the equipment in advance, and integrates the torque current results twice to obtain current results twice An average behavior acquisition means for obtaining an average behavior of a correlation between an integral value, a speed history integration value obtained by integrating the speed history once, and a movement amount actual value; and from the equipment during operation Regarding the correlation between the current actual twice integrated value obtained by integrating the obtained torque current actual twice, the speed actual once integrated value obtained by integrating the speed actual once, and the movement amount actual value, the normality of the equipment Phase during operation An out-of-operation degree calculation unit that calculates an out-of-operation degree from an average behavior of the relationship; and a determination unit that determines abnormality of the equipment based on the out-of-operation degree calculated by the out-of-operation degree calculation unit. It is characterized by that.

また、本発明に係る異常監視システムは、上記の発明において、前記正常時平均的挙動取得手段は、前記設備の正常動作時の前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移動量実績値とによって定まる３次元空間内の１点を正常パターンとし、複数の正常パターンに対して主成分分析を行って前記正常パターンの主成分の変換係数を取得し、前記操業時外れ度算出手段は、操業時の前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移動量実績値とによって定まる３次元空間内の点を操業時パターンとし、前記正常パターンの主成分の変換係数をもとに前記操業時パターンの主成分を算出することを特徴とする。 Further, in the abnormality monitoring system according to the present invention, in the above invention, the normal average behavior acquisition means includes a current result twice integrated value obtained by integrating the torque current result twice during normal operation of the facility, and One point in a three-dimensional space determined by a speed result once integrated value obtained by integrating the speed result once and the movement amount actual value is set as a normal pattern, and a principal component analysis is performed on a plurality of normal patterns to perform the principal component analysis. The conversion factor of the main component of the normal pattern is acquired, and the out-of-operation degree calculation means includes a current result twice integrated value obtained by integrating the torque current result during operation twice and a speed obtained by integrating the speed result once. A point in the three-dimensional space determined by the past integration value and the movement amount actual value is used as the operation pattern, and the principal component of the operation pattern is calculated based on the transformation coefficient of the main component of the normal pattern. Specially To.

また、本発明に係る異常監視システムは、上記の発明において、前記正常時平均的挙動取得手段は、前記正常パターンの主成分の寄与率を算出して主要な主成分をさらに決定し、前記判定手段は、前記操業時パターンの主成分のうちの前記主要な主成分以外の外れ成分をもとに前記外れ度を算出し、該外れ度が所定値以上の場合に前記監視対象を異常と判定することを特徴とする。 Further, in the abnormality monitoring system according to the present invention as set forth in the invention described above, the normal average behavior acquisition means further determines a main principal component by calculating a contribution ratio of the principal component of the normal pattern, and the determination The means calculates the degree of deviation based on outliers other than the main principal component among the main components of the operating pattern, and determines that the monitoring target is abnormal when the degree of deviation is a predetermined value or more. It is characterized by doing.

また、本発明に係る異常監視方法は、電動機で駆動される設備から得られる時系列データをもとに監視対象の前記設備の異常を検知する異常監視方法であって、事前に設備の正常動作時の前記電動機のトルク電流実績と、速度実績と、移動量実績の時系列データを取得して、前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移動量実績値との相関関係の平均的挙動を求める正常時平均的挙動取得工程と、操業時において前記設備から得られる前記トルク電流実績を２回積分した電流実績２回積分値と、前記速度実績を１回積分した速度実績１回積分値と、前記移量動実績値との相関関係について、前記設備の正常動作時の相関関係の平均的挙動からの外れ度を算出する操業時外れ度算出工程と、前記操業時外れ度算出工程で算出された前記外れ度をもとに前記設備の異常を判定する判定工程と、を含むことを特徴とする。 The abnormality monitoring method according to the present invention is an abnormality monitoring method for detecting an abnormality of the equipment to be monitored based on time-series data obtained from equipment driven by an electric motor, and normal operation of the equipment in advance. Time series data of the torque current results, speed results, and travel results of the motor at the time is acquired, and the current results are integrated twice, and the speed results are integrated once. The normal speed behavior acquisition step for obtaining the average behavior of the correlation between the speed actual value once integrated value and the movement amount actual value, and the torque current result obtained from the equipment during operation were integrated twice. From the average behavior of the correlation during normal operation of the equipment, the correlation between the current actual integration value, the speed actual integration value obtained by integrating the speed actual performance once, and the displacement movement actual performance value is obtained. Operation to calculate the degree of deviation And out calculation process, characterized in that it comprises a and a determination step of determining an abnormality of the equipment based on the deviated degree calculated by the operating time out calculation process.

本発明によれば、鉄鋼製品の製造プラント等で多くを占める電動機で駆動される設備の初期の異常を高精度かつ簡易に検知することができる。 ADVANTAGE OF THE INVENTION According to this invention, the initial abnormality of the installation driven with the electric motor which occupies most in the steel products manufacturing plant etc. can be detected with high precision and easily.

図１は、本実施の形態の異常監視システムの監視対象設備の構成を説明する模式図である。FIG. 1 is a schematic diagram illustrating the configuration of the monitoring target equipment of the abnormality monitoring system according to the present embodiment. 図２は、本実施の形態の異常監視システムが行う状態監視の原理を説明するための説明図である。FIG. 2 is an explanatory diagram for explaining the principle of state monitoring performed by the abnormality monitoring system of the present embodiment. 図３は、本実施の形態の異常監視システムの全体構成の一例を示すブロック図である。FIG. 3 is a block diagram illustrating an example of the overall configuration of the abnormality monitoring system according to the present embodiment. 図４は、本実施の形態に係る監視指標作成処理手順を示すフローチャートである。FIG. 4 is a flowchart showing a monitoring index creation processing procedure according to the present embodiment. 図５は、本実施の形態に係る異常監視処理手順を示すフローチャートである。FIG. 5 is a flowchart showing an abnormality monitoring processing procedure according to the present embodiment. 図６は、正常時の移動量実績、速度実績、トルク電流実績の各時系列データを例示する図である。FIG. 6 is a diagram illustrating time series data of a movement amount record, a speed record, and a torque current record in a normal state. 図７は、正常時における負荷トルクのばらつき分布を例示する図である。FIG. 7 is a diagram illustrating an example of a distribution of variation in load torque during normal operation. 図８は、正常時の速度実績の１回積分値、トルク電流実績の２回積分値の各データと移動量実績値との関係を例示する図である。FIG. 8 is a diagram exemplifying a relationship between each data of the one-time integral value of the actual speed result and the two-time integral value of the torque current result and the movement amount actual value. 図９は、異常時の移動量実績、速度実績、トルク電流実績の各時系列データを例示する図である。FIG. 9 is a diagram illustrating time series data of the actual movement amount, the actual speed, and the actual torque current at the time of abnormality. 図１０は、異常時の速度実績の１回積分値、トルク電流実績の２回積分値の各データと移動量実績値との関係を例示する図である。FIG. 10 is a diagram exemplifying a relationship between each data of the one-time integral value of the speed record at the time of abnormality and the two-time integral value of the torque current result and the movement amount actual value. 図１１は、本実施の形態に係る異常監視処理により得られる判定結果を例示する図である。FIG. 11 is a diagram illustrating a determination result obtained by the abnormality monitoring process according to the present embodiment.

以下、図面を参照して、本発明の異常監視システムおよび異常監視方法を実施するための一形態について説明する。なお、この実施の形態によって本発明が限定されるものではない。また、図面の記載において、同一部分には同一の符号を付して示している。 Hereinafter, an embodiment for implementing an abnormality monitoring system and an abnormality monitoring method of the present invention will be described with reference to the drawings. Note that the present invention is not limited to the embodiments. Moreover, in description of drawing, the same code | symbol is attached | subjected and shown to the same part.

本実施の形態の異常監視システムは、電動機で駆動される設備、例えば、電動機により鉄鋼製品を所定の位置まで搬送し、製造過程での処理を施すための位置決めする設備を監視対象とし、この監視対象の設備（以下、対象設備）の状態監視を行って異常を検知するものである。 The abnormality monitoring system of the present embodiment targets equipment to be monitored by equipment driven by an electric motor, for example, equipment for conveying steel products to a predetermined position by an electric motor and performing processing in a manufacturing process. It monitors the status of the target equipment (hereinafter, target equipment) and detects an abnormality.

まず、本実施の形態の異常監視システムが行う状態監視の原理について説明する。図１は、電動機による負荷の制御を説明するための説明図である。図１に示すように、電動機１１は、トルク電流によりモータトルクを発生させ、設備１２を所定の位置へ移動させる。その際に電動機１１は、負荷である設備１２の位置を指定する情報が与えられると、位置制御部１３が指定位置へ移動させるための基準となる速度の情報を生成して速度制御部１４に出力し、速度制御部１４がその速度で設備１２を移動させるための基準となるトルク電流の情報を生成して電流制御部１５に出力し、電流制御部１５が所定のトルク電流を生成して電動機１１に出力することで設備１２を駆動する。また、電動機１１にセンサーが配設され、センサーのトルク電流検出部１６が電動機１１のトルク電流の実績値を検出する。センサーの位置検出部１７が電動機１１の位置（移動量）の実績値を検出して位置制御部１３にフィードバックし、センサーの速度検出部１８が電動機１１の速度の実績値を検出して速度制御部１４にフィードバックする。これらセンサーにより検出される位置（移動量）および速度の実績値の時系列データと、電動機１１に入力されるトルク電流の実績値の時系列データとは、データ収集部１９に記録される。 First, the principle of state monitoring performed by the abnormality monitoring system of the present embodiment will be described. FIG. 1 is an explanatory diagram for explaining load control by an electric motor. As shown in FIG. 1, the electric motor 11 generates a motor torque by a torque current and moves the equipment 12 to a predetermined position. At that time, when the information specifying the position of the equipment 12 as a load is given, the electric motor 11 generates information on a speed that serves as a reference for the position control unit 13 to move to the specified position, and sends it to the speed control unit 14 And the speed control unit 14 generates torque current information that serves as a reference for moving the equipment 12 at that speed and outputs it to the current control unit 15. The current control unit 15 generates a predetermined torque current. The equipment 12 is driven by outputting to the electric motor 11. In addition, a sensor is provided in the electric motor 11, and the torque current detection unit 16 of the sensor detects the actual value of the torque current of the electric motor 11. The sensor position detector 17 detects the actual value of the position (movement amount) of the electric motor 11 and feeds it back to the position controller 13, and the sensor speed detector 18 detects the actual value of the speed of the electric motor 11 and controls the speed. Feedback to the unit 14. The time series data of the actual values of the position (movement amount) and speed detected by these sensors and the time series data of the actual values of torque current input to the electric motor 11 are recorded in the data collection unit 19.

このような対象設備の電動機１１と設備１２について、力学上の以下の式（１），（２）が成り立つことが知られている。

It is known that the following equations (1) and (2) in terms of dynamics hold for the electric motor 11 and the equipment 12 of the target equipment.

この式（１），（２）からわかるように、粘性摩擦係数項（式（１）右辺第２項）を無視できる場合、かつ、負荷項（式（１）の右辺第３項）が０の場合には、モータトルクＴ_ｍを２回積分した値は、回転角θ_ｍに相当する移動量に比例する。したがって、モータトルクＴ_ｍに比例するトルク電流を２回積分した値は、移動量に比例する。同様に、速度を１回積分した値は、移動量に比例する。このことから、本発明に係る異常監視システムは、正常時のトルク電流の実績値の２回積分値と、速度の実績値の１回積分値と、移動量の実績値との３値間の相関関係に基づいて、異常の発生を早期に検知するものである。すなわち、正常時の上記３値の相関関係の平均的挙動に基づいて、その外れ度を監視することにより異常を検知する。実際には、粘性摩擦係数項や負荷項は０ではないが、正常時には一定の範囲内に収まるため、外れ度の大きさで異常を検知できる。とくに本実施の形態においては、図２に例示するように、外れ度の監視によく知られた主成分分析の手法を適用する。すなわち、正常時の上記３値を３次元空間内の１点として表すと、各点Ｐ２１は３次元空間において楕円状に分布する。これに対し、異常時には上記３値で表される点Ｐ２２は、正常時の点の主成分（楕円の長軸）に直交する外れ成分が増加して、楕円状の分布から外れる。そこで、平均的挙動として主成分に基づいてＴ^２統計量を算出し、外れ度として主成分に垂直な方向の成分の残差に基づいてＱ統計量を算出し、これらを監視することで実現する。 As can be seen from the equations (1) and (2), when the viscous friction coefficient term (the second term on the right side of the equation (1)) can be ignored, the load term (the third term on the right side of the equation (1)) is 0. in the case of the integrated value of the motor torque T _m 2 times is proportional to the movement amount corresponding to the rotation angle theta _m. Thus, integrated value of the torque current proportional to the motor torque T _m 2 times is proportional to the amount of movement. Similarly, the value obtained by integrating the speed once is proportional to the movement amount. From this, the abnormality monitoring system according to the present invention is between the three values of the two-time integration value of the actual torque current value, the one-time integration value of the actual speed value, and the actual movement amount value. Based on the correlation, the occurrence of abnormality is detected at an early stage. That is, an abnormality is detected by monitoring the degree of deviation based on the average behavior of the above three-value correlation during normal operation. Actually, although the viscous friction coefficient term and the load term are not zero, they are within a certain range at the normal time, so that an abnormality can be detected with the magnitude of the degree of deviation. In particular, in the present embodiment, as illustrated in FIG. 2, a well-known principal component analysis method is applied to monitor the degree of deviation. That is, when the above three values at normal time are expressed as one point in the three-dimensional space, each point P21 is distributed in an elliptical shape in the three-dimensional space. On the other hand, the point P22 represented by the above three values at the time of abnormality deviates from the elliptical distribution due to an increase in the outlier component orthogonal to the main component (ellipse major axis) of the normal point. Therefore, T ² statistic is calculated based on the principal component as the average behavior, Q statistic is calculated based on the residual of the component in the direction perpendicular to the principal component as the degree of deviation, and is realized by monitoring these To do.

なお、図１では電動機１１についての位置および速度を検出しているが、設備１２についての位置および速度にも同様に本発明を適用できる。その場合には、センサーを設備１２に配設する。 In addition, although the position and speed about the electric motor 11 are detected in FIG. 1, the present invention can be similarly applied to the position and speed about the equipment 12. In that case, the sensor is disposed in the facility 12.

図３は、本実施の形態の異常監視システムの全体構成の一例を示すブロック図である。この異常監視システム１は、図３に示すように、オフライン指標作成システム２と、オンライン診断システム３と、時系列ＤＢ４と、監視指標ＤＢ５とを含み、互いにデータの送受が可能に接続されて構成されている。なお、時系列ＤＢ４および監視指標ＤＢ５は、オフライン指標作成システム２またはオンライン診断システム３が備える記憶部２３，３３に保存された構成としてもよい。 FIG. 3 is a block diagram illustrating an example of the overall configuration of the abnormality monitoring system according to the present embodiment. As shown in FIG. 3, the abnormality monitoring system 1 includes an offline index creation system 2, an online diagnostic system 3, a time series DB 4, and a monitoring index DB 5, which are connected to each other so as to be able to transmit and receive data. Has been. The time series DB 4 and the monitoring index DB 5 may be stored in the storage units 23 and 33 provided in the offline index creation system 2 or the online diagnosis system 3.

オフライン指標作成システム２は、例えばワークステーションやパソコン等の汎用コンピュータを用いて実現され、入力部２１と、表示部２２と、記憶部２３と、各部を制御する制御部２５とを含む。 The offline index creation system 2 is realized using a general-purpose computer such as a workstation or a personal computer, and includes an input unit 21, a display unit 22, a storage unit 23, and a control unit 25 that controls each unit.

入力部２１は、例えばキーボードやマウス、タッチパネル、各種スイッチ等の各種入力装置によって実現されるものであり、操作入力に応じた入力信号を制御部２５に出力する。表示部２２は、ＬＣＤやＥＬディスプレイ、ＣＲＴディスプレイ等の表示装置によって実現されるものであり、制御部２５から入力される表示信号をもとに各種画面を表示する。 The input unit 21 is realized by various input devices such as a keyboard, a mouse, a touch panel, and various switches, and outputs an input signal corresponding to an operation input to the control unit 25. The display unit 22 is realized by a display device such as an LCD, an EL display, or a CRT display, and displays various screens based on a display signal input from the control unit 25.

記憶部２３は、更新記憶可能なフラッシュメモリ等のＲＯＭやＲＡＭといった各種ＩＣメモリ、内蔵あるいはデータ通信端子で接続されたハードディスク、ＣＤ−ＲＯＭ等の情報記憶媒体およびその読取装置等によって実現されるものである。この記憶部２３には、オフライン指標作成システム２を動作させ、このオフライン指標作成システム２が備える種々の機能を実現するためのプログラムや、このプログラムの実行中に使用されるデータ等が予め保存され、あるいは処理の都度一時的に保存される。 The storage unit 23 is realized by various IC memories such as ROM and RAM such as flash memory that can be updated and stored, a hard disk built in or connected by a data communication terminal, an information storage medium such as a CD-ROM, and a reading device thereof. It is. In this storage unit 23, a program for operating the offline index creation system 2 and realizing various functions provided in the offline index creation system 2, data used during the execution of this program, and the like are stored in advance. Or, it is temporarily saved for each processing.

制御部２５は、ＣＰＵ等で実現され、入力部２１から入力される入力信号、記憶部２３に保存されるプログラムやデータ等をもとに、オフライン指標作成システム２を構成する各部への指示やデータの転送等を行ってオフライン指標作成システム２の動作を制御する。この制御部２５は、監視指標作成処理部２７を含む。 The control unit 25 is realized by a CPU or the like, based on an input signal input from the input unit 21, a program or data stored in the storage unit 23, and instructions to each unit constituting the offline index creation system 2 Data transfer or the like is performed to control the operation of the offline index creation system 2. The control unit 25 includes a monitoring index creation processing unit 27.

監視指標作成処理部２７は、オンライン診断システム３が行う対象設備の状態監視に用いられる監視指標を作成する処理（監視指標作成処理）を行う機能部であり、過去の操業時に対象設備から得られた時系列データを用いて対象設備の正常な動作状態を指標化する。この監視指標作成処理部２７は、時系列切出処理部２７１と、演算処理部２７３と、正常時平均的挙動取得手段としての統計的監視指標化処理部２７５とを含む。 The monitoring index creation processing unit 27 is a functional unit that performs processing (monitoring index creation processing) for creating a monitoring index used for the status monitoring of the target equipment performed by the online diagnostic system 3 and is obtained from the target equipment during past operations. The normal operating state of the target equipment is indexed using the time series data. The monitoring index creation processing unit 27 includes a time-series extraction processing unit 271, an arithmetic processing unit 273, and a statistical monitoring indexing processing unit 275 as a normal average behavior acquisition unit.

オンライン診断システム３は、オフライン指標作成システム２と同様に、例えばワークステーションやパソコン等の汎用コンピュータを用いて実現され、入力部３１と、表示部３２と、記憶部３３と、各部を制御する制御部３５とを含む。 The online diagnostic system 3 is realized by using a general-purpose computer such as a workstation or a personal computer, as in the offline index creation system 2, and controls the input unit 31, the display unit 32, the storage unit 33, and each unit. Part 35.

入力部３１は、例えばキーボードやマウス、タッチパネル、各種スイッチ等の各種入力装置によって実現されるものであり、操作入力に応じた入力信号を制御部３５に出力する。表示部３２は、ＬＣＤやＥＬディスプレイ、ＣＲＴディスプレイ等の表示装置によって実現されるものであり、制御部３５から入力される表示信号をもとに各種画面を表示する。 The input unit 31 is realized by various input devices such as a keyboard, a mouse, a touch panel, and various switches, for example, and outputs an input signal corresponding to an operation input to the control unit 35. The display unit 32 is realized by a display device such as an LCD, an EL display, or a CRT display, and displays various screens based on a display signal input from the control unit 35.

記憶部３３は、更新記憶可能なフラッシュメモリ等のＲＯＭやＲＡＭといった各種ＩＣメモリ、内蔵あるいはデータ通信端子で接続されたハードディスク、ＣＤ−ＲＯＭ等の情報記憶媒体およびその読取装置等によって実現されるものである。この記憶部３３には、オンライン診断システム３を動作させ、このオンライン診断システム３が備える種々の機能を実現するためのプログラムや、このプログラムの実行中に使用されるデータ等が予め保存され、あるいは処理の都度一時的に保存される。 The storage unit 33 is realized by various IC memories such as ROM and RAM such as flash memory that can be updated and stored, a hard disk connected by a built-in or data communication terminal, an information storage medium such as a CD-ROM, and a reading device thereof. It is. In the storage unit 33, a program for operating the online diagnostic system 3 and realizing various functions provided in the online diagnostic system 3, data used during the execution of the program, and the like are stored in advance, or It is temporarily saved for each processing.

制御部３５は、ＣＰＵ等で実現され、入力部３１から入力される入力信号、記憶部３３に保存されるプログラムやデータ等をもとに、オンライン診断システム３を構成する各部への指示やデータの転送等を行ってオンライン診断システム３の動作を制御する。この制御部３５は、異常監視処理部３７を含む。 The control unit 35 is realized by a CPU or the like, and based on input signals input from the input unit 31 and programs and data stored in the storage unit 33, instructions and data to each unit constituting the online diagnosis system 3 Is transferred to control the operation of the online diagnostic system 3. The control unit 35 includes an abnormality monitoring processing unit 37.

異常監視処理部３７は、対象設備の状態をオンライン（リアルタイム）で監視し、対象設備の異常を検知する処理（異常監視処理）を行う機能部であり、対象設備から得られる時系列データを用いて対象設備の状態監視を行って異常を検知する。この異常監視処理部３７は、時系列切出処理部３７１と、演算処理部３７３と、操業時外れ度算出手段としての統計的監視指標化処理部３７５と、判定手段としての判定処理部３７７とを含む。 The abnormality monitoring processing unit 37 is a functional unit that monitors the state of the target equipment online (in real time) and performs processing for detecting the abnormality of the target equipment (abnormality monitoring processing), and uses time series data obtained from the target equipment. The condition of the target equipment is monitored and an abnormality is detected. The abnormality monitoring processing unit 37 includes a time-series cut-out processing unit 371, an arithmetic processing unit 373, a statistical monitoring indexing processing unit 375 as an out-of-operation degree calculation unit, and a determination processing unit 377 as a determination unit. including.

時系列ＤＢ４には、過去の操業時に対象設備から取得した時系列データが保存される。また、監視指標ＤＢ５には、オフライン指標作成システム２において監視指標作成処理部２７が対象設備の正常な動作状態を指標化した監視指標が保存される。 The time series DB 4 stores time series data acquired from the target equipment during past operations. Also, the monitoring index DB 5 stores a monitoring index in which the monitoring index creation processing unit 27 indexes the normal operation state of the target facility in the offline index creation system 2.

次に、異常監視システム１が行う具体的な処理手順について説明する。図４は、オフライン指標作成システム２が行う監視指標作成処理の処理手順を示すフローチャートである。また、図５は、オンライン診断システム３が行う異常監視処理の処理手順を示すフローチャートである。異常監視システム１は、オフライン指標作成システム２が図４の処理手順に従って監視指標作成処理を行い、オンライン診断システム３が図５の処理手順に従って異常監視処理を行うことで異常監視方法を実施する。なお、ここで説明する処理は、監視指標作成処理を実現するためのプログラムをオフライン指標作成システム２の記憶部２３に保存しておき、オフライン指標作成システム２がこのプログラムを読み出して実行するとともに、異常監視処理を実現するためのプログラムをオンライン診断システム３の記憶部３３に保存しておき、オンライン診断システム３がこのプログラムを読み出して実行することで実現できる。 Next, a specific processing procedure performed by the abnormality monitoring system 1 will be described. FIG. 4 is a flowchart showing a processing procedure of monitoring index creation processing performed by the offline index creation system 2. FIG. 5 is a flowchart showing a processing procedure of abnormality monitoring processing performed by the online diagnosis system 3. In the abnormality monitoring system 1, the offline index creation system 2 performs the monitoring index creation process according to the processing procedure of FIG. 4, and the online diagnostic system 3 performs the abnormality monitoring process according to the processing procedure of FIG. 5. In the process described here, a program for realizing the monitoring index creation process is stored in the storage unit 23 of the offline index creation system 2, and the offline index creation system 2 reads and executes the program. A program for realizing the abnormality monitoring process is stored in the storage unit 33 of the online diagnostic system 3, and the online diagnostic system 3 reads out and executes this program.

オフライン指標作成システム２が行う監視指標作成処理では、図４に示すように、先ず、制御部２５において、監視指標作成処理部２７の時系列切出処理部２７１が、時系列ＤＢ４を参照し、過去の操業時に対象設備から得られた移動量実績、速度実績、トルク電流実績の３種の時系列データのそれぞれについて、正常操業時（対象設備の正常動作時）に得られた時系列データを読み出す（ステップＳ１０１）。そして、時系列切出処理部２７１は、ステップＳ１０１で読み出した正常操業時における３種の時系列データのそれぞれから、移動開始から一定時間のデータを切り出す（ステップＳ１０２）。本実施の形態では、移動開始トリガー信号を基準にして、予め設定した時間長のデータを時系列データのそれぞれから切り出す。 In the monitoring index creation process performed by the offline index creation system 2, as shown in FIG. 4, first, in the control unit 25, the time series extraction processing unit 271 of the monitoring index creation processing unit 27 refers to the time series DB 4, Time series data obtained during normal operation (during normal operation of the target equipment) for each of the three types of time series data obtained from the target equipment during past operations, actual speed, and actual torque current. Read (step S101). Then, the time-series cut-out processing unit 271 cuts out data for a fixed time from the start of movement from each of the three types of time-series data at the normal operation read in step S101 (step S102). In the present embodiment, data having a preset time length is cut out from each of the time series data with reference to the movement start trigger signal.

ここで図６に、正常時の３種の時系列データを例示する。図６では、移動量実績として電動機１１の角度の実績値（以下、電動機角度実績）を示し、速度実績として電動機１１の角速度の実績値（以下、電動機角速度実績）を示し、トルク電流実績に相当する物理量として、トルク電流に比例する（トルク電流から算出される）モータトルクの実績値（以下、トルク実績）を示す。上記３種の実績値のそれぞれを縦軸に、時間を横軸にして、３５の正常事例について移動開始時間を合わせて重畳表示したものである。 Here, FIG. 6 illustrates three types of time-series data at normal times. In FIG. 6, the actual value of the angle of the motor 11 (hereinafter referred to as “motor angle actual result”) is shown as the movement amount result, the actual value of the angular velocity of the motor 11 (hereinafter referred to as “motor angular speed result”) is shown as the speed result, and it corresponds to the torque current result. As a physical quantity to be performed, an actual value of motor torque (hereinafter referred to as torque actual) proportional to the torque current (calculated from the torque current) is shown. Each of the above three kinds of actual values is displayed on the vertical axis and the time is plotted on the horizontal axis, and the movement start times are superimposed and displayed for 35 normal cases.

この図６によれば、電動機角度実績の時系列データと、電動機角速度実績の時系列データは、それぞれ３５例についてパターンがほぼ一致していることがわかる。一方、トルク実績の時系列データは、３５例のパターンにバラつきがあることがわかる。 According to FIG. 6, it can be seen that the pattern of the time series data of the actual motor angle and the time series data of the actual motor angular velocity are substantially the same for each of 35 cases. On the other hand, it can be seen that the time-series data of the torque results has variations in the patterns of 35 examples.

このトルク実績のばらつきは負荷トルクのばらつきによる。負荷トルクは負荷を動かすためのトルクである。図７に、正常時における負荷トルクのばらつき分布を例示する。図７からわかるように、正常時の負荷トルクにはばらつきがあるため、モータトルク、すなわちトルク電流実績にもばらつきが生じる。このことから、トルク電流実績のみの監視による異常の検知は困難であることがわかる。 This variation in torque results is due to variations in load torque. The load torque is a torque for moving the load. FIG. 7 illustrates an example of a distribution of variation in load torque in a normal state. As can be seen from FIG. 7, since the load torque at the normal time varies, the motor torque, that is, the torque current results also vary. This indicates that it is difficult to detect an abnormality by monitoring only the torque current results.

次に、ステップＳ１０３の処理では、演算処理部２７３が、ステップＳ１０２で切り出した速度実績の時系列データに対して１回積分する処理を施し、トルク電流実績の時系列データに対して２回積分する処理を施す。その後、移動量実績、速度実績の１回積分、トルク電流実績の２回積分の各時系列データのそれぞれに対し、以後の監視の前処理として正規化処理を行う。なお、ステップＳ１０３の正規化処理は、以後の処理を３種の時系列データそれぞれの最大値の大小に影響されることなく行うためのものである。 Next, in the process of step S103, the arithmetic processing unit 273 performs a process of integrating the time series data of the speed record cut out in step S102 once and integrates the time series data of the torque current result twice. Apply the process. After that, normalization processing is performed as preprocessing for subsequent monitoring for each time series data of the movement amount record, the speed record one-time integration, and the torque current record two-time integration. Note that the normalization processing in step S103 is for performing the subsequent processing without being affected by the maximum value of each of the three types of time-series data.

ここで図８に、正常時の速度実績の１回積分値、トルク電流実績の２回積分値の各データと移動量実績値との関係を例示する。図８の各データは図６の３５の正常事例の各時系列データに基づく移動量実績（電動機角度実績）値、速度実績（電動機角速度実績）の１回積分実績値、トルク電流実績（トルク実績）の２回積分値のそれぞれを縦軸に、移動量実績（角度実績）値を横軸にして、正常時の３５例について重畳表示したものである。この図８によれば、電動機角度実績値、電動機角速度実績の１回積分値、トルク電流実績（トルク実績）の２回積分値は、互いに相関があることがわかる。 Here, FIG. 8 illustrates the relationship between each data of the normal speed actual integration value and the torque current actual twice integration value and the movement amount actual value. Each data in FIG. 8 is a movement amount record (motor angle record) value, speed record (motor angular speed record) one-time integration record, torque current record (torque record) based on each time-series data of 35 normal cases in FIG. ) With the vertical axis representing the two-time integral value and the horizontal axis representing the actual movement amount (angular result) value, 35 cases in the normal state are superimposed and displayed. According to FIG. 8, it can be seen that the actual motor angle value, the one-time integrated value of the actual motor angular velocity, and the two-time integrated value of the actual torque current (torque actual) have a correlation with each other.

続いて、ステップＳ１０４の処理では、統計的監視指標化処理部２７５が、正常時平均的挙動取得工程として、ステップＳ１０３での正規化処理後の３種の時系列データをそれぞれ所定数サンプリングし、対応する３種のデータを取得する。そして、ステップＳ１０５の処理では、統計的監視指標化処理部２７５は、ステップＳ１０４で取得した３種のデータを３次元空間内の点として表した複数の正常パターンの主成分分析を行う。 Subsequently, in the process of step S104, the statistical monitoring indexing processing unit 275 samples a predetermined number of each of the three types of time series data after the normalization process in step S103 as a normal average behavior acquisition step, Acquire the corresponding three types of data. In the process of step S105, the statistical monitoring indexing processing unit 275 performs principal component analysis of a plurality of normal patterns representing the three types of data acquired in step S104 as points in the three-dimensional space.

例えば、本実施の形態では、図２に示した３次元空間において楕円状に分布する各点Ｐ２１によって表される複数の正常パターンについて公知の主成分分析を行う。この主成分分析では、各正常パターンのサンプリングデータを用い、主成分毎の固有ベクトルを求めて各主成分の変換係数を取得し、主成分の式を取得する。次いで、累積寄与率が予め設定される閾値（例えば０．８）以上となる主要な主成分である上位の主成分の成分数Ｒを決定する。なお、成分数Ｒを決定する閾値は、固定値としてもよいし、ユーザ操作に従って可変に設定することとしてもよく、適宜設定してよい。ここで決定した上位Ｒ個の主成分（第１主成分〜第Ｒ主成分）によって、正常パターンを特徴付ける主要な特性が決定される。一方、第Ｒ＋１主成分より下位の主成分である外れ成分（残差）は、正常パターンの特性を決定する際の寄与度が低い。 For example, in the present embodiment, a known principal component analysis is performed on a plurality of normal patterns represented by the points P21 distributed elliptically in the three-dimensional space shown in FIG. In this principal component analysis, sampling data of each normal pattern is used to obtain eigenvectors for each principal component, to obtain transformation coefficients for each principal component, and to obtain principal component expressions. Next, the component number R of the upper principal component, which is the main principal component, whose cumulative contribution rate is equal to or greater than a preset threshold value (for example, 0.8) is determined. The threshold for determining the number of components R may be a fixed value, may be set variably according to a user operation, and may be set as appropriate. The main characteristics that characterize the normal pattern are determined by the top R principal components (first to R-th principal components) determined here. On the other hand, an outlier component (residual), which is a principal component lower than the (R + 1) th principal component, has a low contribution when determining the characteristics of the normal pattern.

このような主成分分析は以下のように数式で表現できる。分析対象のデータ行列を次式（３）とする。

ただし、Ｎはサンプリングデータ数、Ｐはデータ項目数（本実施の形態では３）である。また、採用する主成分数をＲ、ＵとＶを直交行列として、データ行列Ｘの特異値分解を次式（４）のように表すこととする。

このとき、主成分Ｖ_Ｒの主成分得点Ｔ_Ｒは、次式（５）のように表せる。

また、主成分と直交する残差は、次式（６）のように表せる。

これらの式（５）（６）に基づいて、平均的挙動を示す指標としてのＴ^２統計量は、次式（７）のように表される。

また、外れ度を示す指標としてのＱ統計量は、次式（８）のように表せる。

Such principal component analysis can be expressed by a mathematical formula as follows. The data matrix to be analyzed is represented by the following equation (3).

However, N is the number of sampling data, and P is the number of data items (3 in this embodiment). Further, the singular value decomposition of the data matrix X is expressed as the following equation (4), with the number of principal components to be adopted being R, U and V being orthogonal matrices.

In this case, the principal component score _{T R} of the main component _{V R} can be expressed as the following equation (5).

Further, the residual orthogonal to the main component can be expressed as the following equation (6).

Based on these formulas (5) and (6), the T ² statistic as an index indicating the average behavior is expressed as the following formula (7).

Further, the Q statistic as an index indicating the degree of detachment can be expressed as the following equation (8).

以上のようにして主成分分析を行い、正常パターンの主成分の変換係数によって表される主成分の式（主成分Ｖ_Ｒ）および上位の主成分の成分数Ｒを取得したならば、ステップＳ１０６の処理では、統計的監視指標化処理部２７５は、取得した主成分の式（主成分Ｖ_Ｒ）および主成分の成分数Ｒを対象設備の正常な動作状態を指標化した監視指標として監視指標ＤＢ５に保存する。 After performing the principal component analysis as described above and obtaining the principal component formula (principal component V _R ) and the number R of the higher principal components, which are represented by the transformation coefficients of the principal components of the normal pattern, step S106. In the process of, the statistical monitoring indexing processing unit 275 uses the acquired principal component formula (principal component V _R ) and the number R of the principal component components as monitoring indexes that index the normal operating state of the target facility. Save to DB5.

また、オンライン診断システム３が行う異常監視処理では、図５に示すように、先ず、制御部３５の異常監視処理部３７において、時系列切出処理部３７１が、操業中に対象設備から得られた移動量実績、速度実績、トルク電流実績の３種の時系列データを入力し（ステップＳ２０１）、入力された３種の時系列データのそれぞれから、移動開始から一定時間のデータを切り出す（ステップＳ２０２）。ここでの処理は図４のステップＳ１０１と同様に、移動開始トリガー信号を基準にして、予め設定した時間長のデータを時系列データのそれぞれから切り出す。切り出された３種のデータは、記憶部３３に一時保存する。その後、ステップＳ２０３の処理では、演算処理部３７３が、ステップＳ２０２で切り出した速度実績の時系列データに対して１回積分する処理を施し、トルク電流実績の時系列データに対して２回積分する処理を施す。その後、移動量実績、速度実績の１回積分、トルク電流実績の２回積分の各時系列データのそれぞれに対し、前処理として正規化処理を行う。 Further, in the abnormality monitoring process performed by the online diagnosis system 3, as shown in FIG. 5, first, in the abnormality monitoring processing part 37 of the control part 35, a time series cut-out processing part 371 is obtained from the target equipment during operation. The three types of time series data, that is, the movement amount record, the speed record, and the torque current record are input (step S201), and the data for a certain period of time from the start of movement is cut out from each of the input three types of time series data (step S202). In this process, similarly to step S101 in FIG. 4, data of a preset time length is cut out from each of the time-series data with reference to the movement start trigger signal. The extracted three types of data are temporarily stored in the storage unit 33. Thereafter, in the process of step S203, the arithmetic processing unit 373 performs a process of integrating once with respect to the time series data of the speed record cut out in step S202, and integrates twice with respect to the time series data of the torque current result. Apply processing. After that, normalization processing is performed as preprocessing for each time series data of the movement amount record, speed record one-time integration, and torque current record two-time integration.

ここで、図９に操業時に異常が発生した場合（異常時）の移動量実績、速度実績、トルク電流実績の３種の時系列データを例示する。図９は、３の異常事例について、移動量実績（電動機角度実績）、速度実績（電動機角速度実績）、トルク電流実績（トルク実績）の時系列データを、図６の３５の正常事例の各時系列データに移動開始時間を合わせて重畳表示したものである。この図９によれば、電動機角度実績の時系列データと、電動機角速度実績の時系列データは、異常時のパターンも正常時のパターンとほとんど差異が認められない。一方、トルク実績の時系列データは、正常時と異常時のＳ／Ｎ比は１．６であり、トルク実績、すなわちトルク電流の監視だけでは異常初期段階での検出は困難であることがわかる。 Here, FIG. 9 exemplifies three types of time-series data of movement amount results, speed results, and torque current results when an abnormality occurs during operation (at the time of abnormality). FIG. 9 shows the time series data of the movement amount results (motor angle results), the speed results (motor angular speed results), and the torque current results (torque results) for each of the three normal cases. The series data is superimposed and displayed with the movement start time. According to FIG. 9, the time series data of the motor angle results and the time series data of the motor angular speed results show almost no difference between the abnormal pattern and the normal pattern. On the other hand, the time-series data of the actual torque shows that the S / N ratio at normal time and abnormal time is 1.6, and it is difficult to detect at the initial stage of abnormality only by monitoring the actual torque, that is, torque current. .

また、図１０に異常時の速度実績の１回積分値、トルク実績の２回積分値の各データと移動量実績値との関係を例示する。図１０の各データは図９の３の異常事例の各時系列データに基づく移動量実績（電動機角度実績）値、速度実績（電動機角速度実績）の１回積分値、トルク電流実績（トルク実績）の２回積分値のそれぞれのデータを、図８の３５の正常事例のデータに重畳表示したものである。この図１０によれば、正常時、異常時ともに電動機角度実績値、電動機角速度実績の１回積分値、トルク実績の２回積分値は、互いに相関があること、また図１０の点線で囲まれる部分に示されるように、異常時にはトルク実績の２回積分値が正常時のそれより大きく外れることがわかる。 Moreover, FIG. 10 illustrates the relationship between each data of the once integrated value of the actual speed and the twice integrated value of the actual torque and the movement amount actual value at the time of abnormality. Each data in FIG. 10 is a movement amount actual (motor angle actual) value, a speed actual result (motor angular speed actual) one-time integral value, torque current actual (torque actual) based on each time series data of 3 abnormal cases in FIG. The data of the two-time integral values are superimposed and displayed on the data of 35 normal cases in FIG. According to FIG. 10, the motor angle actual value, the motor angular velocity actual one-time integral value, and the torque actual two-time integral value are correlated with each other in both normal and abnormal states, and are surrounded by a dotted line in FIG. 10. As shown in the part, it can be seen that the two-time integrated value of the actual torque is significantly different from that in the normal state when there is an abnormality.

続いて、ステップＳ２０４の処理では、統計的監視指標化処理部３７５が、監視指標ＤＢ５を参照し、対象設備の監視指標、すなわち、正常パターンの主成分の変換係数によって表される主成分の式（主成分Ｖ_Ｒ）および主成分の成分数Ｒを読み出す。続いて、ステップＳ２０５の処理では、統計的監視指標化処理部３７５は、操業時外れ度算出工程として、ステップＳ２０３での正規化処理後の３種の時系列データを所定数ずつサンプリングし、対応する３種のデータを取得する。そして、ステップＳ２０６の処理では、統計的監視指標化処理部３７５は、ステップＳ２０５で取得した３種のデータを３次元空間内の点として表した操業時パターンにステップＳ２０４で読み出した主成分の式（主成分Ｖ_Ｒ）を適用することで、操業時パターン、すなわちステップＳ２０５で取得したサンプリングデータのＲ個の主成分およびこのＲ個の主成分より下位の主成分である外れ成分（残差）を算出する。 Subsequently, in the process of step S204, the statistical monitoring indexing processing unit 375 refers to the monitoring index DB 5, and represents the principal component expression represented by the monitoring index of the target facility, that is, the conversion coefficient of the principal component of the normal pattern. Read out (principal component V _R ) and the number of components R of the main component. Subsequently, in the process of step S205, the statistical monitoring indexing processing unit 375 samples a predetermined number of the three types of time series data after the normalization process in step S203 as the operation outlier calculation step. Three types of data are acquired. In the process of step S206, the statistical monitoring indexing processing unit 375 calculates the principal component formula read in step S204 to the operation pattern representing the three types of data acquired in step S205 as points in the three-dimensional space. By applying (principal component V _R ), an operating pattern, that is, R principal components of the sampling data acquired in step S205 and an outlier component (residual) which is a principal component lower than the R principal components. Is calculated.

その後、ステップＳ２０７の処理では、判定処理部３７７が、判定工程として、主成分のＴ^２統計量を算出して平均的挙動を示す指標とするとともに、外れ成分のＱ統計量を算出して主成分の外れ度を示す指標とし、この平均的挙動と外れ度とをもとに対象設備の異常を判定する。例えば、判定処理部３７７は、予め設定される閾値を用いて外れ度を閾値処理することで外れ度の大小を判定する。そして、判定処理部３７７は、外れ度が大きい場合に、対象設備の動作状態を異常と判定する。 Thereafter, in the process of step S207, the determination processing unit 377 calculates the main component T ² statistic as an index indicating an average behavior and calculates the Q statistic of the outlier component as the determination step. An index indicating the degree of component detachment is used, and the abnormality of the target equipment is determined based on the average behavior and the degree of detachment. For example, the determination processing unit 377 determines the magnitude of the detachment degree by thresholding the detachment degree using a preset threshold value. The determination processing unit 377 determines that the operation state of the target facility is abnormal when the degree of detachment is large.

図１１に、本実施の形態の異常監視処理により得られる判定結果を例示する。図１１の横軸は時系列データのサンプリング回数、縦軸はそれぞれ平均的挙動を示す指標（Ｔ^２統計量）、外れ度を示す指標（Ｑ統計量）を示す。図中の各ピークが１回の移動を示す。図１１の点線で囲まれる部分に示されるように、外れ度が所定の閾値を越えてトルク電流実績（トルク実績）が異常と判定された場合には、平均的挙動も正常時より大きくなることがわかる。ただし、外れ度の方がより大きく外れ、平均的挙動だけで異常を判定することは難しいことがわかる。 FIG. 11 illustrates a determination result obtained by the abnormality monitoring process of the present embodiment. The horizontal axis the number of times of sampling the time-series data in FIG. 11, an index indicating a respective vertical axis the average behavior (T ² statistic) indicating the index of the out-degree (Q statistic). Each peak in the figure shows one movement. As shown in the portion surrounded by the dotted line in FIG. 11, when the degree of detachment exceeds a predetermined threshold value and the torque current performance (torque performance) is determined to be abnormal, the average behavior should be larger than normal. I understand. However, it is understood that the degree of detachment deviates more greatly, and it is difficult to determine an abnormality only by average behavior.

以上説明したように、本実施の形態では、正常時のトルク電流の実績値の２回積分値と、速度の実績値の１回積分値と、移動量の実績値との３値間の相関関係について、正常時の上記３値の相関関係の平均的挙動に基づいて、その外れ度を監視することにより異常を検知することとした。外れ度の監視には、よく知られた主成分分析の手法を適用し、平均的挙動として主成分に基づいてＴ^２統計量を算出し、外れ度として主成分に垂直な方向の成分の残差に基づいてＱ統計量を算出することで異常を判定することとした。 As described above, in the present embodiment, the correlation between the three values of the two-time integrated value of the actual torque current value, the one-time integrated value of the actual speed value, and the actual value of the movement amount is obtained. Regarding the relationship, the abnormality was detected by monitoring the degree of deviation based on the average behavior of the above three-value correlation during normal operation. For monitoring the degree of deviation, a well-known principal component analysis method is applied to calculate the T ² statistic based on the principal component as an average behavior, and as the degree of deviation, the residual component in the direction perpendicular to the principal component is calculated. Anomalies were determined by calculating the Q statistic based on the difference.

本実施の形態によれば、異常の初期段階でも、電動機で駆動される設備の異常を高精度かつ簡易に検知することができる。 According to the present embodiment, it is possible to detect an abnormality of equipment driven by an electric motor with high accuracy and easily even at an initial stage of abnormality.

なお、本実施の形態では、鉄鋼製品の製造プラントのプラント設備を監視対象として異常を検知する場合について説明したが、鉄鋼製品の製造プラントに限らず、各種製品の製造プラント設備等の所望の監視対象から得られる時系列データをもとに、その異常を検知する場合に同様に適用できる。 In addition, although this Embodiment demonstrated the case where the abnormality was detected for the plant equipment of the steel product manufacturing plant as a monitoring object, it is not limited to the steel product manufacturing plant, but desired monitoring of various product manufacturing plant equipment, etc. The present invention can be similarly applied when detecting an abnormality based on time-series data obtained from an object.

以上のように、本発明の異常監視システムおよび異常監視方法は、監視対象の種類等を意識することなく汎用的に適用し、かつその異常を高精度に検知するのに適している。 As described above, the abnormality monitoring system and abnormality monitoring method of the present invention are suitable for general-purpose application without being aware of the type of monitoring target and the like and detecting the abnormality with high accuracy.

１異常監視システム
２オフライン指標作成システム
２１入力部
２２表示部
２３記憶部
２５制御部
２７監視指標作成処理部
２７１時系列切出処理部
２７３演算処理部
２７５統計的監視指標化処理部
３オンライン診断システム
３１入力部
３２表示部
３３記憶部
３５制御部
３７異常監視処理部
３７１時系列切出処理部
３７３演算処理部
３７５統計的監視指標化処理部
３７７判定処理部
４時系列ＤＢ
５監視指標ＤＢ
１１電動機
１２設備 DESCRIPTION OF SYMBOLS 1 Abnormality monitoring system 2 Offline index production system 21 Input part 22 Display part 23 Storage part 25 Control part 27 Monitoring index production process part 271 Time series extraction process part 273 Operation process part 275 Statistical monitoring indexization process part 3 Online diagnostic system 31 Input Unit 32 Display Unit 33 Storage Unit 35 Control Unit 37 Abnormality Monitoring Processing Unit 371 Time Series Extraction Processing Unit 373 Arithmetic Processing Unit 375 Statistical Monitoring Indexing Processing Unit 377 Judgment Processing Unit 4 Time Series DB
5 Monitoring index DB
11 Electric motor 12 Equipment

Claims

An abnormality monitoring system for detecting an abnormality of the equipment to be monitored based on time-series data obtained from equipment driven by an electric motor,
Time series data of the torque current record, speed record, and travel record of the motor during normal operation of the facility in advance is acquired, and the current record twice integrated value obtained by integrating the torque current record twice; A normal average behavior acquisition means for obtaining an average behavior of a correlation between a speed actual one-time integrated value obtained by integrating a speed actual once and the movement amount actual value;
Correlation between the current actual value twice integrated value obtained by integrating the torque current results obtained from the equipment during operation, the speed actual value integrated value obtained by integrating the speed actual value once, and the movement amount actual value With respect to the operational deviation degree calculating means for calculating the deviation degree from the average behavior of the correlation during normal operation of the equipment,
Determining means for determining an abnormality of the equipment based on the degree of outage calculated by the out-of-operation degree-of-operation calculating means;
An abnormality monitoring system comprising:

The normal average behavior acquisition means includes a current result two-time integration value obtained by integrating the torque current result during normal operation of the facility twice, a speed result one-time integration value obtained by integrating the speed result once, One point in the three-dimensional space determined by the movement amount actual value is set as a normal pattern, a principal component analysis is performed on a plurality of normal patterns to obtain conversion coefficients of the main components of the normal pattern,
The out-of-operation degree calculation means includes a current result twice integrated value obtained by integrating the torque current result during operation twice, a speed result integrated value obtained by integrating the speed result once, and a movement amount actual value. 2. The abnormality according to claim 1, wherein a point in a three-dimensional space determined by is used as an operation pattern, and the principal component of the operation pattern is calculated based on a conversion coefficient of the principal component of the normal pattern. Monitoring system.

The normal average behavior acquisition means further determines a main principal component by calculating a contribution ratio of the main component of the normal pattern,
The determination means calculates the degree of deviation based on outliers other than the main principal component of the main components of the operating pattern, and the monitoring target is abnormal when the degree of deviation is a predetermined value or more. The abnormality monitoring system according to claim 1, wherein the abnormality monitoring system is determined.

An abnormality monitoring method for detecting an abnormality of the equipment to be monitored based on time-series data obtained from equipment driven by an electric motor,
Time series data of the torque current record, speed record, and travel record of the motor during normal operation of the facility in advance is acquired, and the current record twice integrated value obtained by integrating the torque current record twice; A normal average behavior acquisition step for obtaining an average behavior of a correlation between a speed actual one-time integrated value obtained by integrating a speed actual once and the movement amount actual value;
Correlation between the current actual integrated value obtained by integrating the torque current actual obtained twice from the equipment at the time of operation, the speed actual integrated value obtained by integrating the speed actual once, and the displacement movement actual value For the relationship, an out-of-operation degree calculation step for calculating a degree of deviation from the average behavior of the correlation during normal operation of the equipment;
A determination step of determining an abnormality of the equipment based on the degree of outage calculated in the out-of-operation degree-of-operation calculation step;
An abnormality monitoring method comprising: