【０００１】[0001]
【産業上の利用分野】本発明は、プラント等の異常診断
において、正常と異常を自動的に判定するための異常診
断装置に関する。BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an abnormality diagnosing device for automatically determining normality and abnormality in abnormality diagnosis of a plant or the like.
【０００２】[0002]
【従来の技術】従来、プラント等の異常診断は、センサ
信号のパターンを熟練した運転員が見て、その信号波形
のパターンと過去の経験によって得た異常時のパターン
との類似度から、正常と異常を判定していた。2. Description of the Related Art Conventionally, an abnormality diagnosis of a plant or the like is normally performed by a trained operator who sees a pattern of a sensor signal, and a normality is obtained based on the similarity between the pattern of the signal waveform and the pattern at the time of abnormality obtained from past experience. Was determined to be abnormal.
【０００３】[0003]
【発明が解決しようとする課題】上記のように波形信号
のパターンから異常を判定することは、数値や○×式に
より単純な評価をするのとは違い、熟練した運転員に頼
らざるを得ない。しかし、プラントには多数のセンサが
あり、全てを同時に常時監視するのは不可能である。The determination of abnormality from the pattern of the waveform signal as described above has to rely on a skilled operator, unlike a simple evaluation by a numerical value or a XX formula. Absent. However, there are many sensors in the plant, and it is impossible to monitor all of them simultaneously at the same time.
【０００４】また、仮に運転員に頼らずに、例えば正常
とすべき信号の上下限値を定めて、その範囲内に有るか
無いかを○×式で判定する場合、範囲の境界付近では範
囲内であるか否かで判定が大きく変化し、正常と異常を
判定することができないという問題がある。If, for example, the upper and lower limit values of a signal that should be normal are set and whether or not the signal is within the range is determined by the ○ × formula without depending on the operator, the range near the boundary of the range is determined. There is a problem in that the judgment greatly changes depending on whether or not it is within the range, and normality and abnormality cannot be judged.
【０００５】上記問題を解消するために、各種判定の範
囲を複数設けようとすると、様々な条件の組み合わせに
よりデータ量が膨大となり、データ入力、管理上の問題
が生じる。In order to solve the above problem, if a plurality of ranges for various judgments are to be provided, the amount of data becomes huge due to the combination of various conditions, which causes problems in data input and management.
【０００６】本発明は上記実情に鑑みてなされたもの
で、異常を判定するための運転員を不要とし、信号波形
から異常を正確かつ簡単に判定できる異常診断装置を提
供することを目的とする。The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an abnormality diagnosis device that does not require an operator for determining an abnormality and can accurately and easily determine an abnormality from a signal waveform. .
【０００７】[0007]
【課題を解決するための手段】本発明に係る異常診断装
置は、被診断装置に設置され、該装置の運転状態に応じ
たプロセス信号を出力するセンサと、予め上記被診断装
置の異常時の信号パターンを学習して記憶し、上記セン
サから出力されるプロセス信号のパターンと上記予め記
憶した運転パターンとを比較して、運転状態の異常を判
定するニューラルネットワークと、このニューラルネッ
トワークによる判定結果を表示する表示手段とを具備し
たことを特徴とする。An abnormality diagnosing device according to the present invention is installed in a device to be diagnosed and outputs a process signal according to an operating state of the device, and a device for detecting the abnormality of the device to be diagnosed in advance. A signal pattern is learned and stored, a process signal pattern output from the sensor is compared with the previously stored operation pattern, and a neural network for determining an abnormal operation state and a determination result by this neural network are displayed. And a display means for displaying.
【０００８】[0008]
【作用】被診断装置例えばプラントに設置されたセンサ
は、プラントの運転状態を示すプロセス信号をニューラ
ルネットワークに出力する。このニューラルネットワー
クには、予め異常時の信号波形のパターンを学習により
記憶させておく。このニューラルネットワークは、セン
サから出力されるプロセス信号のパターンと予め記憶し
た信号パターンとを比較して異常を判定し、その結果を
表示手段により表示する。The device to be diagnosed, for example, the sensor installed in the plant outputs a process signal indicating the operating state of the plant to the neural network. In this neural network, the pattern of the signal waveform at the time of abnormality is stored in advance by learning. This neural network compares a process signal pattern output from a sensor with a signal pattern stored in advance to determine an abnormality, and displays the result by a display means.
【０００９】上記の構成とすることにより、運転員は表
示手段により表示された異常と示されたプラントの部分
について、直ちに対策を講じることができる。従って、
異常を判定するための運転員が不要となり、運転員は対
策等の作業に専念できる。With the above arrangement, the operator can immediately take countermeasures for the portion of the plant indicated by the display means as abnormal. Therefore,
An operator is no longer required to determine an abnormality, and the operator can concentrate on work such as countermeasures.
【００１０】[0010]
【実施例】以下、図面を参照して本発明の一あ実施例を
説明する。図１は本発明の一実施例に係る異常診断装置
の構成を示すブロック図である。同図において、１は被
診断装置例えばプラントに設置されたセンサで、プラン
トの運転状態を示すプロセス信号をニューラルネットワ
ーク２へ出力する。一方、ニューラルネットワーク２に
は、予め異常時の信号パターンを学習させ、記憶させて
おく。上記ニューラルネットワーク２は、実際のプラン
ト運転時にセンサ１から出力されるプロセス信号のパタ
ーンと予め記憶している異常時の信号パターンとを比較
して異常を判定し、その結果を表示装置３に出力して表
示画面に表示させる。An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of an abnormality diagnosis device according to an embodiment of the present invention. In the figure, reference numeral 1 denotes a sensor installed in a device to be diagnosed, for example, a plant, which outputs a process signal indicating the operating state of the plant to the neural network 2. On the other hand, in the neural network 2, the signal pattern at the time of abnormality is learned and stored in advance. The neural network 2 compares the process signal pattern output from the sensor 1 at the time of actual plant operation with a previously stored signal pattern at the time of abnormality to determine an abnormality, and outputs the result to the display device 3. And display it on the display screen.
【００１１】上記ニューラルネットワーク２は、図２に
示すように入力層Ａ、中間層（隠れ層）Ｂ、出力層Ｃに
より構成される。入力層Ａは、センサ１から送られてく
るプロセス信号を時間間隔Δｔ毎に複数の振幅信号とし
て取り込み、中間層Ｂに入力する。中間層Ｂは、各入力
値に対して、最初はランダムに初期設定された重み係数
Ｗi を掛け、これらを出力層Ｃへ送る。出力層Ｃは、更
に入力値に対して別の重み係数Ｖi を掛け、出力値Ｚ1
，Ｚ2 を得る。The neural network 2 is composed of an input layer A, an intermediate layer (hidden layer) B, and an output layer C as shown in FIG. The input layer A captures the process signal sent from the sensor 1 as a plurality of amplitude signals at each time interval Δt, and inputs it to the intermediate layer B. The intermediate layer B first multiplies each input value by a weight coefficient Wi initially initialized at random, and sends them to the output layer C. The output layer C further multiplies the input value by another weighting coefficient Vi, and outputs the output value Z1.
, Z2.
【００１２】一方、期待される出力値として、「異
常」、「正常」毎に教師信号Ｄ1 ，Ｄ2を設定してお
く。例えば「異常」を選択すると、教師信号Ｄ1 〜Ｄと
して「Ｄ1＝１」，「Ｄ2 ＝０」が与えられ、また、
「正常」を選択すると、「Ｄ1 ＝０」，「Ｄ2 ＝１」が
与えられる。On the other hand, as expected output values, the teacher signals D1 and D2 are set for each of "abnormal" and "normal". For example, if "abnormal" is selected, "D1 = 1" and "D2 = 0" are given as the teacher signals D1 to D, and
When "Normal" is selected, "D1 = 0" and "D2 = 1" are given.
【００１３】今、「異常」を選択したとすると、ニュー
ラルネットワークは、出力値Ｚ1 ，Ｚ2 と教師信号「Ｄ
1 ＝１」，「Ｄ2 ＝０」との差をそれぞれ取り、いわゆ
る誤差逆伝播法で、つまり出力層Ｃから入力層Ａまでネ
ットワークの逆を辿り、誤差が十分小さくなるように重
み係数を繰り返して修正して、期待される信号波形を形
成していく。If "abnormal" is selected, the neural network outputs the output values Z1 and Z2 and the teacher signal "D".
The difference between "1 = 1" and "D2 = 0" is taken, and the so-called backpropagation method is used, that is, the reverse of the network is traced from the output layer C to the input layer A, and the weighting factors are repeated so that the error becomes sufficiently small. And correct it to form the expected signal waveform.
【００１４】図３は上記ニューラルネットワーク２の具
体的な構成例を、出力ユニットが２つの場合について示
したものである。このニューラルネットワーク２は、１
１個の入力ユニット１１1 〜１１11を持つ入力層Ａ、１
１個の隠れユニット１２1 〜１２11を持つ中間層Ｂ、２
個の出力ユニット１３1 ，１３2 を持つ出力層Ｃ、及び
誤差検出部１４から成っている。FIG. 3 shows a concrete configuration example of the neural network 2 in the case of two output units. This neural network 2 has 1
Input layer A havingone input unit 111 to 1111
Intermediate layer B, 2 withone hidden unit 12 1-1211
The output layer C has a number of output units 131 and 132 and an error detection unit 14.
【００１５】入力層Ａは、センサ１からの入力信号Ｘm
（ｍ：１〜１１）を入力ユニット１１m （ｍ：１〜１
１）により取り込んで中間層Ｂに送る。この中間層Ｂの
隠れユニット１２n （ｎ：１〜１１）は、入力ユニット
１１m からの信号Ｘm に重み係数Ｗnmを掛け、その演算
出力信号Ｙn を出力層Ｃに送る。出力層Ｃの出力ユニッ
ト１３k （ｋ：１，２）は、隠れユニット１２n からの
出力値Ｙn に重み係数Ｖknを掛け、その演算出力信号Ｚ
k を誤差検出部１４へ送る。この誤差検出部１４は、出
力ユニット１３k の出力信号Ｚk と教師信号Ｄk との伝
播誤差δk ，δ′n を求め、この誤差δk ，δ′n が十
分小さくなるように出力層Ｃから入力層Ａ方向へ誤差の
逆伝播を行ない、各層におけるユニットの重み係数を計
算して修正する。The input layer A has an input signal Xm from the sensor 1.
Input unit 11m (m: 1 to 11) (m: 1 to 1)
It is taken in by 1) and sent to the intermediate layer B. Hidden units 12n of the intermediate layer B (n: 1~11) is multiplied by a weighting factor Wnm to signals Xm from the input unit 11m, and sends the calculated output signal Yn in the output layer C. Output unit 13k of the output layer C (k: 1,2) is multiplied by a weighting factor Vkn output value Yn from the hidden units 12n, the operation output signal Z
k is sent to the error detector 14. The error detection unit 14 obtains propagation errors δk and δ ′n between the output signal Zk of the output unit 13k and the teacher signal Dk, and the output layer so that the errors δk and δ ′n are sufficiently small. The error is backpropagated from C to the input layer A, and the weighting coefficient of the unit in each layer is calculated and corrected.
【００１６】図４は、図３における中間層Ｂの１１番目
の隠れユニット１２11の構成例を示したものである。こ
の隠れユニット１２11は、入力層Ａより入力値Ｘ1 〜Ｘ
11が与えられると、それぞれに重み係数Ｗ11,1〜Ｗ
11,11 を掛けたものと、「Ｘ0 ＝１」の入力バイアス
（線形多項式の定数項）に重み係数Ｗ11,0を掛けたもの
との総和「Ｓ11＝Σ（Ｗnm×Ｘm ）」を求め、更に、こ
の総和Ｓ11をシグモイド関数「Ｙ＝（１＋ｅ-s）-1」に
代入して出力値Ｙ11を求める。即ち、ニューロンの伝達
特性を生物の特性に似せる時の近似式を用いて出力値Ｙ
11を求め、これを出力層Ｃに出力する。FIG. 4 shows a configuration example of theeleventh hidden unit 1211 of the intermediate layer B in FIG. The hidden unit 1211 receives input values X1 to X from the input layer A.
When 11 is given, the weighting factors W11,1 to W respectively
The sum of the product of11,11 and the product of the input bias of "X0 = 1" (the constant term of the linear polynomial) multiplied by the weighting factor W11,0 "S11 = Σ (Wnm × Xm ) ”, And the sum S11 is substituted into the sigmoid function“ Y = (1 + e−s )−1 ”, and the output value Y11 is calculated. That is, the output value Y is calculated by using an approximate expression when the transfer characteristic of the neuron is made to resemble the characteristic of the organism.
11 is obtained and is output to the output layer C.
【００１７】なお、中間層Ｂの他の隠れユニット１２1
〜１２10においても、上記隠れユニット１２11と同様に
構成される。次に上記実施例の動作を説明する。Incidentally, another hidden unit 121 of the intermediate layer B1
The same applies to the hidden units 1211 to 1210 as well. Next, the operation of the above embodiment will be described.
【００１８】診断に先立ち、ニューラルネットワーク２
には、予めセンサ信号パターンの異常、正常を学習させ
る。即ち、学習モードを指定し、正常時の信号波形と異
常時の波形を抽出し、そのパターンをニューラルネット
ワーク２に学習させる。この学習に際してはプラント状
態が異常か正常かを選択指定する。Prior to the diagnosis, the neural network 2
In advance, the abnormality or normality of the sensor signal pattern is learned in advance. That is, the learning mode is designated, the signal waveform in the normal state and the waveform in the abnormal state are extracted, and the neural network 2 is made to learn the pattern. In this learning, whether the plant state is abnormal or normal is selected and designated.
【００１９】この判定指定により対応する教師信号Ｄk
（Ｄ1 ，Ｄ2 ）が与えられる。上記学習モードを指定す
ると、ニューラルネットワーク２は、図５に示す逆伝播
学習アルゴリズムに従って運転モデルの学習動作を開始
する。According to this determination designation, the corresponding teacher signal Dk
(D1, D2) is given. When the learning mode is designated, the neural network 2 starts the learning operation of the driving model according to the back propagation learning algorithm shown in FIG.
【００２０】ニューラルネットワーク２は、まず、暫定
的な重み係数の初期値Ｗ0nmとＶ0knを決定する（ステ
ップＡ1 ）。そして、最初に例えば「異常」時のセンサ
信号をニューラルネットワーク２に入力する。ニューラ
ルネットワーク２は、センサ１から出力されるプロセス
信号を時間間隔Δｔ毎に取り込み、その入力値Ｘm に対
して出力値計算（学習）を行ない、各層の出力値Ｙ′
m 、Ｙn 、Ｚk m を求める（ステップＡ2 ）。出力層Ｃ
の出力値Ｚk と教師信号Ｄk から誤差二乗和Ｅを求め、
その値が十分小さいか否かを判定する（ステップＡ3
）。この誤差二乗和Ｅを求めることにより、出力値Ｚk
と教師信号Ｄk との合致度の判定を容易に行なうこと
ができる。The neural network 2 first determines temporary initial values W0nm and V0kn of the weighting factors (step A 1). Then, first, for example, a sensor signal at the time of "abnormal" is input to the neural network 2. The neural network 2 takes in the process signal output from the sensor 1 at each time interval Δt, performs output value calculation (learning) on the input value Xm , and outputs the output value Y ′ of each layer.
m , Yn and Zkm are obtained (step A2). Output layer C
Seek error sum of squares E from the output value Zk and the teacher signal Dk of,
It is determined whether the value is sufficiently small (step A3
). By calculating this error sum of squares E, the output value Zk
It is possible to easily determine the degree of coincidence between the and the teacher signal Dk .
【００２１】上記教師信号Ｄk と出力値Ｚk との誤差二
乗和Ｅを求めるため、次式に示す誤差二乗和計算を行な
う。 Ｅ＝１／２・Σk （Ｚk −Ｄk ）2 …（１） そして、上記の計算により求めた誤差が十分小さいと判
定された場合は、そのまま処理を終了するが、小さくな
ければ、誤差を小さくするために各層、各ユニットの重
み係数を修正する必要があり、出力層Ｃから入力層Ａ方
向へ誤差の逆伝播により、この重み係数を判定するため
の計算を行なう。In order to obtain the error sum of squares E between the teacher signal Dk and the output value Zk , the error sum of squares shown in the following equation is calculated. E = 1 / 2Σk (Zk −Dk )2 (1) Then, when it is determined that the error obtained by the above calculation is sufficiently small, the processing is ended as it is, but if it is not small, In order to reduce the error, it is necessary to modify the weighting coefficient of each layer and each unit, and the calculation for determining this weighting coefficient is performed by the back propagation of the error from the output layer C to the input layer A.
【００２２】上記の重み係数を修正するためには、ま
ず、次式の様な伝播誤差計算を行ない、伝播誤差δk ，
δ′n を求める（ステップＡ4 ）。出力層の伝播誤差δ
k は、 δk ＝（Ｚk −Ｄk ）・Ｚk ・（１−Ｚk ） …（２） により求め、中間層の伝播誤差δ′n は、 δ′n ＝Σk （δk ・Ｖkn）・Ｙn （１−Ｙn ） …（３） により求める。In order to correct the above weighting coefficient, first, the propagation error calculation as shown in the following equation is performed to calculate the propagation error δk ,
Request [delta]'n (step A4). Propagation error δ in the output layer
k is obtained by δk = (Zk −Dk ) · Zk · (1-Zk ) ... (2), and the propagation error δ ′n of the intermediate layer is δ ′n = Σk (δk · Vkn ) · Yn (1-Yn ) ... (3)
【００２３】次に、次式に示す重み係数修正量計算を行
ない、係数修正量ΔＶkn，ΔＷnmを求める（ステップＡ
5 ）。 ΔＶkn＝−δk ・Ｙn …（４） ΔＷnm＝−δ′n ・Ｘm …（５） 即ち、係数修正量ΔＶkn，ΔＷnmは、各ニューロンの出
力値を重みとして算出する。Next, the weight coefficient correction amount shown in the following equation is calculated to obtain the coefficient correction amounts ΔVkn and ΔWnm (step A
Five ). ΔVkn = −δk · Yn (4) ΔWnm = −δ ′n · Xm (5) That is, the coefficient correction amounts ΔVkn and ΔWnm are calculated using the output value of each neuron as a weight.
【００２４】次に、重み係数Ｖkn，Ｗnmを次式により修
正する（ステップＡ6 ）。 Ｖkn＝Ｖkn＋α・ΔＶkn …（６） Ｗnm＝Ｗnm＋α・ΔＷnm …（７） ただし、α＝１（学習係数）とする。Next, the weighting factors Vkn and Wnm are modified by the following equation (step A6). Vkn = Vkn + α · ΔVkn (6) Wnm = Wnm + α · ΔWnm (7) where α = 1 (learning coefficient).
【００２５】そして、上記の計算を、各層、各ユニット
の全てについて終了したか否かを判断し（ステップＡ7
）、終了していなければステップＡ4 に戻って同じ処
理を繰り返して行ない、各層、各ユニット分を終了する
と、ステップＡ2 に戻って、再び出力値を計算する。Then, it is judged whether or not the above calculation is completed for all layers and units (step A7).
If not completed, the process returns to step A4 to repeat the same processing. When each layer and each unit is completed, the process returns to step A2 and the output value is calculated again.
【００２６】以下、同様の処理を繰り返して行ない、ス
テップＡ3 で、誤差二乗和Ｅが十分小さい判断されると
処理を終了する。このようにして各層、各ユニット重み
係数が決定される。The same process is repeated thereafter, and the process ends when it is determined in step A3 that the error sum of squares E is sufficiently small. In this way, each layer and each unit weight coefficient are determined.
【００２７】上記「異常」の学習を終了した後、「正
常」時の信号パターンについても同様にして学習を行な
う。上記のようにして異常と正常の信号パターンがニュ
ーラルネットワーク２で学習され、記憶される。上記ニ
ューラルネットワーク２の学習は、１度行なえば以後は
作業の必要はない。After the learning of the "abnormal" is finished, the signal pattern in the "normal" is similarly learned. As described above, the abnormal and normal signal patterns are learned by the neural network 2 and stored. If the learning of the neural network 2 is performed once, no further work is required.
【００２８】上記学習を終了した後、実際のプラント運
用中のセンサ信号をニューラルネットワーク２へ出力す
る。このニューラルネットワーク２は、ニューラルネッ
トワーク２から出力される信号パターンと予め学習によ
り記憶した信号パターンと比較して異常を判定する。即
ち、ニューラルネットワーク２は、センサ１から出力さ
れるプロセス信号を取り込んで出力値Ｚk を計算し、そ
の出力値Ｚk により「異常」、「正常」の何れであるか
を判断する。例えば出力層Ｃの出力値Ｚ1 ，Ｚ2 が、
「Ｚ1 ＝１，Ｚ2 ＝０」であれば「異常」、「Ｚ1 ＝
０，Ｚ2 ＝１」であれば「正常」であると判断する。そ
して、ニューラルネットワーク２は、上記の判断結果を
確信度と共に表示装置３へ出力する。実際の判定時の出
力値Ｚ1 ，Ｚ2 は，区間［０，１］のアナログ値となる
ので、その値を確信度とみなし、「１」に近いほど高い
確信度としている。After the above learning is completed, the sensor signal during actual plant operation is output to the neural network 2. The neural network 2 determines an abnormality by comparing the signal pattern output from the neural network 2 with the signal pattern stored in advance by learning. That is, the neural network 2 is to calculate the output value Zk fetches the process signal output from the sensor 1, "abnormal" by the output value Zk, determines which of "normal". For example, the output values Z1 and Z2 of the output layer C are
If "Z1 = 1, Z2 = 0", "abnormal", "Z1 =
If 0, Z2 = 1 ", it is determined to be" normal ". Then, the neural network 2 outputs the above determination result to the display device 3 together with the certainty factor. Since the output values Z1 and Z2 at the time of actual determination are analog values in the interval [0, 1], the values are regarded as the certainty factor, and the closer to “1”, the higher the certainty factor.
【００２９】表示装置３は、ニューラルネットワーク２
から送られてきた判断結果を表示画面に表示する。運転
員は、この表示装置３の画面に表示された結果を見てプ
ラントの異常を直ちに知ることができる。The display device 3 is a neural network 2
The judgment result sent from is displayed on the display screen. The operator can immediately know the abnormality of the plant by looking at the result displayed on the screen of the display device 3.
【００３０】[0030]
【発明の効果】以上詳記したように本発明によれば、納
入先のプラントの特性に合わせて一度運転してニューラ
ルネットワークに学習させるだけで、以後は異常を判定
するために運転員は常時注意をはらう必要がなく、異常
時の判定を正確に実施できる。As described in detail above, according to the present invention, the operator always operates in accordance with the characteristics of the plant to which the product is delivered and the neural network learns. It is not necessary to pay attention, and it is possible to make accurate judgments when an abnormality occurs.
【図１】本発明の一実施例に係る異常診断装置の構成
図。FIG. 1 is a configuration diagram of an abnormality diagnosis device according to an embodiment of the present invention.
【図２】同実施例におけるニューラルネットワークの概
念図。FIG. 2 is a conceptual diagram of a neural network in the same embodiment.
【図３】同実施例におけるニューラルネットワークの構
成図。FIG. 3 is a configuration diagram of a neural network in the embodiment.
【図４】図３における隠れユニットの構成図。FIG. 4 is a configuration diagram of a hidden unit in FIG.
【図５】同実施例におけるニューラルネットワークの逆
伝播学習動作を示すフローチャート。FIG. 5 is a flowchart showing a back propagation learning operation of the neural network according to the embodiment.
１ センサ ２ ニューラルネットワーク ３ 表示装置 １１1 〜１１11 入力ユニット １２1 〜１２11 隠れユニット １３1 ，１３2 出力ユニット １４ 誤差検出部1 Sensor 2 Neural Network 3 Display Device 111 to 1111 Input Unit 121 to 1211 Hidden Unit 131 and 132 Output Unit 14 Error Detection Section
───────────────────────────────────────────────────── フロントページの続き (72)発明者 神成 正一郎 神奈川県横浜市中区錦町12番地 三菱重工 業株式会社横浜製作所内 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Shoichiro Kaminari 12 Nishiki-cho, Naka-ku, Yokohama-shi, Kanagawa Mitsubishi Heavy Industries, Ltd. Yokohama Works
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2517994AJPH07234988A (en) | 1994-02-23 | 1994-02-23 | Abnormality diagnostic device |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2517994AJPH07234988A (en) | 1994-02-23 | 1994-02-23 | Abnormality diagnostic device |
| Publication Number | Publication Date |
|---|---|
| JPH07234988Atrue JPH07234988A (en) | 1995-09-05 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| JP2517994AWithdrawnJPH07234988A (en) | 1994-02-23 | 1994-02-23 | Abnormality diagnostic device |
| Country | Link |
|---|---|
| JP (1) | JPH07234988A (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO1998020469A1 (en)* | 1996-11-07 | 1998-05-14 | Rosemount Inc. | Diagnostics for resistance based transmitter |
| US6859755B2 (en) | 2001-05-14 | 2005-02-22 | Rosemount Inc. | Diagnostics for industrial process control and measurement systems |
| US6907383B2 (en) | 1996-03-28 | 2005-06-14 | Rosemount Inc. | Flow diagnostic system |
| US6920799B1 (en) | 2004-04-15 | 2005-07-26 | Rosemount Inc. | Magnetic flow meter with reference electrode |
| US6970003B2 (en) | 2001-03-05 | 2005-11-29 | Rosemount Inc. | Electronics board life prediction of microprocessor-based transmitters |
| US7010459B2 (en) | 1999-06-25 | 2006-03-07 | Rosemount Inc. | Process device diagnostics using process variable sensor signal |
| US7018800B2 (en) | 2003-08-07 | 2006-03-28 | Rosemount Inc. | Process device with quiescent current diagnostics |
| US7046180B2 (en) | 2004-04-21 | 2006-05-16 | Rosemount Inc. | Analog-to-digital converter with range error detection |
| US7085610B2 (en) | 1996-03-28 | 2006-08-01 | Fisher-Rosemount Systems, Inc. | Root cause diagnostics |
| US7206646B2 (en) | 1999-02-22 | 2007-04-17 | Fisher-Rosemount Systems, Inc. | Method and apparatus for performing a function in a plant using process performance monitoring with process equipment monitoring and control |
| US7221988B2 (en) | 2001-03-01 | 2007-05-22 | Rosemount, Inc. | Creation and display of indices within a process plant |
| US7254518B2 (en) | 1996-03-28 | 2007-08-07 | Rosemount Inc. | Pressure transmitter with diagnostics |
| US7272531B2 (en) | 2005-09-20 | 2007-09-18 | Fisher-Rosemount Systems, Inc. | Aggregation of asset use indices within a process plant |
| US7290450B2 (en) | 2003-07-18 | 2007-11-06 | Rosemount Inc. | Process diagnostics |
| US7321846B1 (en) | 2006-10-05 | 2008-01-22 | Rosemount Inc. | Two-wire process control loop diagnostics |
| US7523667B2 (en) | 2003-12-23 | 2009-04-28 | Rosemount Inc. | Diagnostics of impulse piping in an industrial process |
| US7562135B2 (en) | 2000-05-23 | 2009-07-14 | Fisher-Rosemount Systems, Inc. | Enhanced fieldbus device alerts in a process control system |
| US7590511B2 (en) | 2007-09-25 | 2009-09-15 | Rosemount Inc. | Field device for digital process control loop diagnostics |
| US7623932B2 (en) | 1996-03-28 | 2009-11-24 | Fisher-Rosemount Systems, Inc. | Rule set for root cause diagnostics |
| US7627441B2 (en) | 2003-09-30 | 2009-12-01 | Rosemount Inc. | Process device with vibration based diagnostics |
| US7630861B2 (en) | 1996-03-28 | 2009-12-08 | Rosemount Inc. | Dedicated process diagnostic device |
| US8898036B2 (en) | 2007-08-06 | 2014-11-25 | Rosemount Inc. | Process variable transmitter with acceleration sensor |
| US9052240B2 (en) | 2012-06-29 | 2015-06-09 | Rosemount Inc. | Industrial process temperature transmitter with sensor stress diagnostics |
| US9094470B2 (en) | 2002-04-15 | 2015-07-28 | Fisher-Rosemount Systems, Inc. | Web services-based communications for use with process control systems |
| US9201420B2 (en) | 2005-04-08 | 2015-12-01 | Rosemount, Inc. | Method and apparatus for performing a function in a process plant using monitoring data with criticality evaluation data |
| US9207670B2 (en) | 2011-03-21 | 2015-12-08 | Rosemount Inc. | Degrading sensor detection implemented within a transmitter |
| WO2016132468A1 (en)* | 2015-02-18 | 2016-08-25 | 株式会社日立製作所 | Data evaluation method and device, and breakdown diagnosis method and device |
| US9602122B2 (en) | 2012-09-28 | 2017-03-21 | Rosemount Inc. | Process variable measurement noise diagnostic |
| US9927788B2 (en) | 2011-05-19 | 2018-03-27 | Fisher-Rosemount Systems, Inc. | Software lockout coordination between a process control system and an asset management system |
| JP7325695B1 (en)* | 2023-01-23 | 2023-08-14 | 三菱電機株式会社 | DATA PROCESSING DEVICE, DATA PROCESSING METHOD AND DATA PROCESSING PROGRAM |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7085610B2 (en) | 1996-03-28 | 2006-08-01 | Fisher-Rosemount Systems, Inc. | Root cause diagnostics |
| US6907383B2 (en) | 1996-03-28 | 2005-06-14 | Rosemount Inc. | Flow diagnostic system |
| US7623932B2 (en) | 1996-03-28 | 2009-11-24 | Fisher-Rosemount Systems, Inc. | Rule set for root cause diagnostics |
| US7630861B2 (en) | 1996-03-28 | 2009-12-08 | Rosemount Inc. | Dedicated process diagnostic device |
| US7254518B2 (en) | 1996-03-28 | 2007-08-07 | Rosemount Inc. | Pressure transmitter with diagnostics |
| US5828567A (en)* | 1996-11-07 | 1998-10-27 | Rosemount Inc. | Diagnostics for resistance based transmitter |
| WO1998020469A1 (en)* | 1996-11-07 | 1998-05-14 | Rosemount Inc. | Diagnostics for resistance based transmitter |
| US7206646B2 (en) | 1999-02-22 | 2007-04-17 | Fisher-Rosemount Systems, Inc. | Method and apparatus for performing a function in a plant using process performance monitoring with process equipment monitoring and control |
| US7010459B2 (en) | 1999-06-25 | 2006-03-07 | Rosemount Inc. | Process device diagnostics using process variable sensor signal |
| US7562135B2 (en) | 2000-05-23 | 2009-07-14 | Fisher-Rosemount Systems, Inc. | Enhanced fieldbus device alerts in a process control system |
| US7221988B2 (en) | 2001-03-01 | 2007-05-22 | Rosemount, Inc. | Creation and display of indices within a process plant |
| US6970003B2 (en) | 2001-03-05 | 2005-11-29 | Rosemount Inc. | Electronics board life prediction of microprocessor-based transmitters |
| US6859755B2 (en) | 2001-05-14 | 2005-02-22 | Rosemount Inc. | Diagnostics for industrial process control and measurement systems |
| US9760651B2 (en) | 2002-04-15 | 2017-09-12 | Fisher-Rosemount Systems, Inc. | Web services-based communications for use with process control systems |
| US9094470B2 (en) | 2002-04-15 | 2015-07-28 | Fisher-Rosemount Systems, Inc. | Web services-based communications for use with process control systems |
| US7290450B2 (en) | 2003-07-18 | 2007-11-06 | Rosemount Inc. | Process diagnostics |
| US7018800B2 (en) | 2003-08-07 | 2006-03-28 | Rosemount Inc. | Process device with quiescent current diagnostics |
| US7627441B2 (en) | 2003-09-30 | 2009-12-01 | Rosemount Inc. | Process device with vibration based diagnostics |
| US7523667B2 (en) | 2003-12-23 | 2009-04-28 | Rosemount Inc. | Diagnostics of impulse piping in an industrial process |
| US6920799B1 (en) | 2004-04-15 | 2005-07-26 | Rosemount Inc. | Magnetic flow meter with reference electrode |
| US7046180B2 (en) | 2004-04-21 | 2006-05-16 | Rosemount Inc. | Analog-to-digital converter with range error detection |
| US9201420B2 (en) | 2005-04-08 | 2015-12-01 | Rosemount, Inc. | Method and apparatus for performing a function in a process plant using monitoring data with criticality evaluation data |
| US7272531B2 (en) | 2005-09-20 | 2007-09-18 | Fisher-Rosemount Systems, Inc. | Aggregation of asset use indices within a process plant |
| US7321846B1 (en) | 2006-10-05 | 2008-01-22 | Rosemount Inc. | Two-wire process control loop diagnostics |
| US8898036B2 (en) | 2007-08-06 | 2014-11-25 | Rosemount Inc. | Process variable transmitter with acceleration sensor |
| US7590511B2 (en) | 2007-09-25 | 2009-09-15 | Rosemount Inc. | Field device for digital process control loop diagnostics |
| US9207670B2 (en) | 2011-03-21 | 2015-12-08 | Rosemount Inc. | Degrading sensor detection implemented within a transmitter |
| US9927788B2 (en) | 2011-05-19 | 2018-03-27 | Fisher-Rosemount Systems, Inc. | Software lockout coordination between a process control system and an asset management system |
| US9052240B2 (en) | 2012-06-29 | 2015-06-09 | Rosemount Inc. | Industrial process temperature transmitter with sensor stress diagnostics |
| US9602122B2 (en) | 2012-09-28 | 2017-03-21 | Rosemount Inc. | Process variable measurement noise diagnostic |
| WO2016132468A1 (en)* | 2015-02-18 | 2016-08-25 | 株式会社日立製作所 | Data evaluation method and device, and breakdown diagnosis method and device |
| JP7325695B1 (en)* | 2023-01-23 | 2023-08-14 | 三菱電機株式会社 | DATA PROCESSING DEVICE, DATA PROCESSING METHOD AND DATA PROCESSING PROGRAM |
| WO2024157310A1 (en)* | 2023-01-23 | 2024-08-02 | 三菱電機株式会社 | Data processing device, data processing method, and data processing program |
| Publication | Publication Date | Title |
|---|---|---|
| JPH07234988A (en) | Abnormality diagnostic device | |
| KR0172982B1 (en) | Fault diagnosis device and method | |
| US5107454A (en) | Pattern associative memory system | |
| US8209080B2 (en) | System for determining most probable cause of a problem in a plant | |
| US7406451B2 (en) | Method for training neural networks | |
| US20030088381A1 (en) | Sensor fusion using self evaluating process sensors | |
| EP0594168B1 (en) | Fuzzy rule acquisition method and apparatus for fuzzy inference system and fuzzy inference system using the apparatus | |
| CN116432352A (en) | A fault diagnosis method for aeroengine main shaft bearings based on digital twin | |
| CN114386707A (en) | A method and device for predicting track irregularity | |
| JPH05141999A (en) | Fuzzy sensor device | |
| WO2022070656A1 (en) | Abnormality detection model assessment system and assessment method | |
| CN111079348B (en) | Method and device for detecting slowly-varying signal | |
| WO2022019020A1 (en) | Abnormality detection device and abnormality detection program | |
| JPH08117827A (en) | Rolling equipment | |
| KR102601338B1 (en) | Apparatus for predicting equipment failure using learning data based on maintenance history and method thereof | |
| JPH0674876A (en) | Equipment diagnosis method and device | |
| JPH0392795A (en) | Diagnostic method of nuclear power plant | |
| KR960012078B1 (en) | How to diagnose malfunction of automobile parts | |
| JP3192877B2 (en) | Sudden event detection method and sudden event detection device | |
| CN111553055B (en) | A task health degree calculation method oriented to the typical functional structure of satellites | |
| JPH05111474A (en) | Sleep depth determination device | |
| JPH06222710A (en) | Operation training evaluating device | |
| JPH04328669A (en) | Unified neural network and its learning system | |
| Gu et al. | Micro sensor based eye movement detection and neural network based sensor fusion and fault detection and recovery | |
| JP3231860B2 (en) | Fault diagnosis device using neural network |
| Date | Code | Title | Description |
|---|---|---|---|
| A300 | Withdrawal of application because of no request for examination | Free format text:JAPANESE INTERMEDIATE CODE: A300 Effective date:20010508 |