JPS608905A - Abnormality diagnosing method of rotary machine of composite cycle plant - Google Patents

Abnormality diagnosing method of rotary machine of composite cycle plant

Info

Publication number
JPS608905A
JPS608905A JP58115974A JP11597483A JPS608905A JP S608905 A JPS608905 A JP S608905A JP 58115974 A JP58115974 A JP 58115974A JP 11597483 A JP11597483 A JP 11597483A JP S608905 A JPS608905 A JP S608905A
Authority
JP
Japan
Prior art keywords
abnormality
rotating
waveform
same
machines
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP58115974A
Other languages
Japanese (ja)
Inventor
Masayuki Toubou
昌幸 当房
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Toshiba Corp
Original Assignee
Toshiba Corp
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 Toshiba Corp filed Critical Toshiba Corp
Priority to JP58115974A priority Critical patent/JPS608905A/en
Publication of JPS608905A publication Critical patent/JPS608905A/en
Pending legal-status Critical Current

Links

Classifications

    • HELECTRICITY
    • H02GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02HEMERGENCY PROTECTIVE CIRCUIT ARRANGEMENTS
    • H02H7/00Emergency protective circuit arrangements specially adapted for specific types of electric machines or apparatus or for sectionalised protection of cable or line systems, and effecting automatic switching in the event of an undesired change from normal working conditions
    • H02H7/06Emergency protective circuit arrangements specially adapted for specific types of electric machines or apparatus or for sectionalised protection of cable or line systems, and effecting automatic switching in the event of an undesired change from normal working conditions for dynamo-electric generators; for synchronous capacitors

Landscapes

  • Engineering & Computer Science (AREA)
  • Power Engineering (AREA)
  • Testing And Monitoring For Control Systems (AREA)
  • Engine Equipment That Uses Special Cycles (AREA)
  • Control Of Velocity Or Acceleration (AREA)

Abstract

PURPOSE:To simplify a diagnostic algorithm, and to diagnose an on-line abnormality by executing a pattern comparison of a power spectrum waveform between each rotary machine installed on each unit, when diagnosing an abnormality of the rotary machine of the same design. CONSTITUTION:In a composite cycle plant having generating units 8-11 of (n) shafts in the whole plant, rotary machines of the same design of total (n) sets are installed. As for these rotary machines, when the operating condition is the same, a power spectrum waveform shows the almost same waveform under the operating state of a normal time. Accordingly, an abnormality of one specified set can be detected by comparing it with a waveform obtained by averaging other (n)-1 sets, and taking a deviation of two waveforms. It is detected by a fact that an average spectrum waveform has a peak peculiar to its rotary machine even in case of a normal state, by a disturbance coming from other than the cause of an abnormality, a large deviation is not generated when a difference is taken between the machines of the same kind, which are not abnormal, and only a peak due to an abnormality shows a large deviation. It is displayed by a CRT2 and informed to an operator.

Description

【発明の詳細な説明】 〔発明の技術外野〕 本発明は、ガスタービンど蒸気タービンとを組み合わせ
、1ユニツトが1軸又は2軸で構成される複合サイクル
発電ユニットが複数ユニット設置される複合ザイクルプ
ラントの各ユニットの同−設計の回転イ我の回転に伴な
う異常を、その振動波形の周波数解析手法によシ、計算
機を用いてオンライン処理で診断する複合ザイクルプラ
ントの回転機異常診断方法VC関する。
Detailed Description of the Invention [Technical Field of the Invention] The present invention relates to a combined cycle power generation unit in which a plurality of combined cycle power generation units each consisting of one or two shafts are installed in combination with a steam turbine such as a gas turbine. A rotating machine abnormality diagnosis method for a complex cycle plant in which abnormalities associated with the rotation of each unit of the plant with the same design are diagnosed through online processing using a computer using a frequency analysis method of the vibration waveform. Related to VC.

〔発明の技術的背景とその間z1点〕 一般の回転柚における磯ピiの異常診断において、振動
波形の計測がおこなわれてきた。機器の異常は、その原
因に応じて特有の振動波形を生み出す。
[Technical background of the invention and z1 point] In the abnormality diagnosis of Isopi i in a general rotating yuzu, vibration waveforms have been measured. Equipment abnormalities produce unique vibration waveforms depending on the cause.

その振動波形をみることで異常診断を下すことができる
。ここで問題になるのは、波形に含まれているノイズ成
分が波形の情報を読の−と力にくくすることである。明
らかに振動波形を目視して異常の識別可能なものは、機
器の運斬継続ができないほどの重大な損害を受けた場合
でちり、波形を見る以前VC振動が大きいこと、あるい
は回転音によって認知され、然るべく処理きれなければ
ならない。そこで機器異常診断としては、そこに至る以
前のノイズに埋もれた振動波形から、有効な情報を早期
に正確に読みとることが必要とされる。この信号処理技
術としては、数学的に種々の手法があシ、フーリエ変換
を用いた周波数解析法は、得られた波形データを離散的
に高速で処理するFFT(ファーストフーリエトランス
フオーム)アルゴリズム及び近年のマイクロプロセッサ
の発展に伴ナウハードウエア化されたパッケージプログ
ラム等により広く工学的に応用されている。
Abnormalities can be diagnosed by looking at the vibration waveform. The problem here is that noise components contained in the waveform make the information in the waveform difficult to read. Abnormalities that can be clearly identified by visually observing the vibration waveform are likely to be caused by severe damage to the equipment that makes it impossible to continue operating. and must be able to process it accordingly. Therefore, in diagnosing equipment abnormalities, it is necessary to quickly and accurately read valid information from the vibration waveforms buried in the noise before that point. There are various mathematical methods for this signal processing technology, including the frequency analysis method using Fourier transform, the FFT (Fast Fourier Transform) algorithm that processes the obtained waveform data discretely at high speed, and the recent With the development of microprocessors, it has now been widely applied in engineering through packaged programs converted into hardware.

よく知られているように、回転異常が起った場合、その
異常原因によplある周波数に於けるパワーが増大し、
周波数分析を行った場合、それはパワースペクトラムの
波形上にピークとなって現われる。よって周波数分析手
法による異常診断は、診断対象スペクトラム波形を正常
時の波形や異常時の典型的な波形とパターン比較し、そ
の異常に伴なうピークを検知し、逆に異常原因を探求す
ることが基本的な構成である。しかしここで問題になる
のはパターンの比較やパターンの認識に伴なう困難さが
介在することである。すなわち通常の回転機のスペクト
ラム波形には、回転数と同期あ(3) るいはその整数倍のピークが現われ、その他に、軸受等
の工作精度上の限界、あるいは運転条件(例えばタービ
ンでは負荷のとり方、発電機では系統側からの影響、ポ
ンプ等では吐出流量変化)等、BA器異常以外の外乱的
要因によるピークが波形読取を困難にし、かつ先に述べ
たように、明らかにピークが現われた場合は、通常機器
運転の続行が不能となる場合が多く、異常診断として要
求されるその微妙で曖昧なピーク徴候を捕えるだめのパ
ターン認識Vζは膨大なデータの蓄積をベースにした、
非常に微妙なものになる。またあるピーク徴候が検知さ
れたとして、どれくらいのピーク値を示せば異常と診断
を下すのがというL2きい値判定の問題もあり、いずれ
ンこせよそれは人間がチャート紙上のパワースペクトラ
ム波形のパターン認識細に比較検討していくという方法
であシ、計算機によるオンラインでの波形のパターン認
へ及び診断のアルゴリズムは大規模かつ複雑になシ、事
実上非常な困難が伴なう。
As is well known, when a rotation abnormality occurs, the power at a certain frequency increases due to the cause of the abnormality,
When frequency analysis is performed, it appears as a peak on the power spectrum waveform. Therefore, abnormality diagnosis using frequency analysis methods involves pattern comparison of the spectrum waveform to be diagnosed with normal waveforms and typical abnormal waveforms, detecting the peak associated with the abnormality, and conversely searching for the cause of the abnormality. is the basic configuration. However, the problem here is that there are difficulties associated with pattern comparison and pattern recognition. In other words, in the spectrum waveform of a normal rotating machine, a peak that is synchronized with the rotation speed (3) or an integer multiple thereof appears, and there are also peaks that are caused by limitations in the machining accuracy of bearings, or operating conditions (for example, in the case of a turbine, the load Peaks caused by disturbance factors other than BA equipment abnormalities, such as the influence from the grid side for generators, and changes in discharge flow rate for pumps, etc., make it difficult to read the waveform, and as mentioned earlier, peaks clearly appear. In such cases, it is often impossible to continue normal equipment operation, and pattern recognition Vζ, which is required to detect subtle and ambiguous peak symptoms required for abnormality diagnosis, is based on the accumulation of a huge amount of data.
It becomes very subtle. There is also the problem of determining the L2 threshold, which is the peak value required to diagnose an abnormality when a certain peak symptom is detected. Although this method requires detailed comparative studies, the algorithms for on-line waveform pattern recognition and diagnosis using a computer are large-scale and complex, and are extremely difficult in practice.

次ニ複合サイクル発電ユニットが複数ユニット−(4) 設置される小によ多構成される複合ザイクルプラントの
回転機相互間…iに伴う問題点を述べる。従来の火力プ
ラントと異なって各回転軸に同一般計の回転機が設置さ
れる為、発電プラント全体の診断を考えると多くの機器
を診断しなければならない。つiシ各回転軸を構成する
回転機を順々に診断する手間と、パターン認識の為のデ
ータの蓄積を考照しなければならず、従来の火力プラン
ト以上の困難さが伴い、計算機による珍重アルゴリズム
が必要となってくる。
Next, a plurality of combined cycle power generation units (4) We will discuss the problems associated with the inter-rotating machines of a combined cycle plant that is installed in a large number of small units. Unlike conventional thermal power plants, each rotary shaft is equipped with a rotating machine of the same general size, so in order to diagnose the entire power plant, many pieces of equipment must be diagnosed. However, it is more difficult than conventional thermal power plants, as it requires consideration of the trouble of sequentially diagnosing the rotating machines that make up each rotating shaft and the accumulation of data for pattern recognition. A highly valued algorithm is required.

〔発明の目的〕[Purpose of the invention]

本発明は、複合サイクル発電ユニットが複数ユニットで
構成される複合サイクルプラントに於ける各ユニットの
同一般計の回転機の異常を、その振動波形の周波数解析
手法によシ診断する際、各ユニットニ設置される回転機
の相互間でのパワースペクトラム波形のパターン比較を
行うことにより、診断アルゴリズムを簡素化し、それよ
シ複数の回転機に対して計算機によるオンライン処理で
の異常診断を可能にする被合サイクルプラントの回転機
異常診断方法を得る事を目的とするものである。
The present invention provides a method for diagnosing an abnormality in a rotating machine of the same general type in each unit in a combined cycle power generation unit consisting of a plurality of units using a frequency analysis method of its vibration waveform. By comparing patterns of power spectrum waveforms between installed rotating machines, the diagnostic algorithm is simplified, and it is also possible to diagnose abnormalities of multiple rotating machines using online computer processing. The purpose of this study is to obtain a method for diagnosing abnormalities in rotating machines of a combined cycle plant.

〔発明の概要〕[Summary of the invention]

上記目的を達成するため本発明に於ては、夫々の回転軸
系に含まれる同一種類の回転機が同一般計のガスグーピ
ンと発電機、蒸気タービンと発電機又はガスタービンと
発電機と蒸気タービン及び回転補機とから構成される複
数の回転軸系を有する複合サイクルプラントの前記ガス
タービン、発電機、蒸気タービン、回転補機の中の少な
くとも一種類の複数の回転機の回転に伴なう異常診断を
前記回転機の生ずる振動波形の周波数分析により行なう
際に、前記各回転軸系に含まれる同一種類の回転機相互
間でのパワースペクトラム波形のパターン比較を計x哉
を用いて行なうようにするものである。
In order to achieve the above object, in the present invention, the same type of rotary machine included in each rotating shaft system is the same general gas turbine and generator, steam turbine and generator, or gas turbine, generator and steam turbine. Accompanying the rotation of at least one type of multiple rotating machines among the gas turbine, generator, steam turbine, and rotating auxiliary machines of a combined cycle plant having a plurality of rotating shaft systems consisting of When abnormality diagnosis is performed by frequency analysis of vibration waveforms generated by the rotating machines, power spectrum waveform patterns are compared between rotating machines of the same type included in each of the rotating shaft systems using a total of It is something to do.

〔発明の実施例〕[Embodiments of the invention]

以下図面を参照して本発明を説明する。第1図は本発明
の一実施例を示すブロック図で、1は計算機、2はCR
T表示装置、3は周波数分析器、4〜7は振動センサ、
8〜11は複合サイクル発電ユニットの回転軸系(回転
補機を含む)である。
The present invention will be explained below with reference to the drawings. FIG. 1 is a block diagram showing one embodiment of the present invention, where 1 is a computer, 2 is a CR
T display device, 3 is a frequency analyzer, 4 to 7 are vibration sensors,
8 to 11 are rotating shaft systems (including rotating auxiliary machines) of the combined cycle power generation unit.

以下の説明では、複合サイクル発電ユニットはlユニッ
ト中に回転側1系とし、てガスターピント蒸気タービン
と発電・礪とを有する1軸型を例にとり、又回転補機と
しては各種ポンプとこれを1駆動するモータを有するも
のとするが、これに限定されるものでtよない。
In the following explanation, the combined cycle power generation unit has one system on the rotating side in the unit, and a single-shaft type with a Gastarpinto steam turbine and a power generation unit is taken as an example, and various pumps and this are used as rotating auxiliary equipment. Although the present invention is assumed to have one driving motor, the present invention is not limited to this.

一般に同一プラント全体置される各ユニットの回転機種
は、設計と運転の便宜上回−設計とされる。上記の如く
各ユニットには異なる機種の複数の回転機が設置される
が、図面で(1)回転軸系全体を単1(C1個の円とし
て示す。尚振動センサ4〜7も各ユニットニつV・て1
個のみ示したが、通常各回転機の軸受に設けられる。
Generally, the rotary machines of each unit located throughout the same plant are designed for convenience in design and operation. As mentioned above, a plurality of rotating machines of different models are installed in each unit, but in the drawing (1) the entire rotating shaft system is shown as a single (C1 circle).The vibration sensors 4 to 7 are also TsuV・te1
Although only one is shown, it is usually provided in the bearing of each rotating machine.

このようにして各発電ユニットの回転軸系或は回転補機
8〜11の回転に伴なう振動データは、各振動センサ4
〜7によって検知され、周波数分析器3Vこ送られる。
In this way, vibration data accompanying the rotation of the rotating shaft system of each power generating unit or rotating auxiliary machines 8 to 11 is collected by each vibration sensor 4.
~7 and sent to the frequency analyzer 3V.

この周波数分析器3は、FFT(ファースト フーリエ
 トランスフオーム)アルゴリズムのプログラムを有す
るマイクロプロセッサであシ、リアルタイム処理でパワ
ースペクトラムを生成する。周波数分析器3より得られ
たパワースペクトラムは割算・f;il VC送られ、
以下に述べる診断アルゴリズムにより診断されて、その
結果は(3RT表示装置2に表示され運転員(C知らさ
れる。尚診断結果VC応じてベル或はブザー等により警
報を与えるのが好捷しい。
The frequency analyzer 3 is a microprocessor having an FFT (Fast Fourier Transform) algorithm program, and generates a power spectrum through real-time processing. The power spectrum obtained from the frequency analyzer 3 is divided by f;il and sent to the VC.
Diagnosis is performed using the diagnostic algorithm described below, and the results are displayed on the 3RT display device 2 and notified to the operator (C).It is preferable that a warning is given by a bell, buzzer, etc. in accordance with the diagnosis result VC.

次に計算機1が有する診断°アルゴリズムを1軸型発′
、にユニットに適用した場合について説明する。
Next, develop the diagnostic algorithm of computer 1 into a single-axis type.
, the case where it is applied to the unit will be explained.

R+ ’!ユニット8〜11かプラント全体でn°軸1
らる複合サイクルプラントニは、全体でn台の同−設計
の回転機種が設置される。こルらの回転・、、)種は運
転条件が同じであれば、正常時の通常;運転状態下では
、パワースペクトラム波形はほぼ同じ波形を示す。よっ
て或特定の1台の異常は、他の(n−1)台の平均をと
った波形と比較し、2つの波形の偏差をとることによつ
′C明確に浮び上ってくる。
R+'! n° axis 1 for units 8 to 11 or the entire plant
A total of n rotary machines of the same design will be installed in the combined cycle plant. If the operating conditions are the same, these rotations are normal during normal operation; under operating conditions, the power spectrum waveforms show almost the same waveform. Therefore, an abnormality in one specific device becomes clearly apparent by comparing the waveform with the average of the other (n-1) devices and taking the deviation of the two waveforms.

何故ならば、この平均スペクトラム波形は、先に述べた
異常原因μ外からぐる外乱により、正常状態でもその回
転機種固有のピークを有しておシ、異常でない同一機種
間で差をとれば、その固有のピークVCは大きな偏差は
出す、異常によるピークのみが大きな偏Mk示すからで
ある。通電0白中の同一機種の回転機器が2台以上同時
VC同じ異常を発生することは極めてまれであることを
考慮すルば、上記波形の相互比較による判尾は充分に合
理性を有する。
This is because this average spectrum waveform has a peak unique to the rotating machine even under normal conditions due to disturbances outside the abnormality cause μ mentioned above, and if you take the difference between the same models that are not abnormal, This is because the unique peak VC shows a large deviation, and only the peak due to an abnormality shows a large deviation Mk. Considering that it is extremely rare for two or more of the same type of rotating equipment to simultaneously generate the same VC abnormality when energized (0 or 0), the conclusion based on mutual comparison of the waveforms is sufficiently reasonable.

以下の説明に関し、下記の//ポル及び記号を使用ター
る。
In the following description, the following symbols are used:

回転機種固有 :P n 1lilIl中iqu:6る2機種 :Pi(1≦
i≦0)Piの14散型パワースペクトラA : 5′
CP + ) :(X I 1 + X + 2 +(
IXmの行ベクトル) ”’+”im)Pi診1す7時
の平均スペクトラム:−7(at)台」−〔4G〕−シ
ー[Pi):]−1 (IXmO行ベクトル) パワースペクトラム偏差 : 2(pH) ==7(p
i)−[相]Gi)(ixmO行ベクトル) ” [d
t1dt2−− +dtm)=太のパワースペクトラム
偏差:dlσ−eMax (d r kInIn軸中短
軸るP機種Piに異常が生ずると、Plのパワースペク
トラム、7(pI)には、その原因に応じである周波数
σにピークを与える。一般にこれを検知することは、先
に述べたようKP機種固有のピークとの弁別が困姪とな
るが、n軸中i軸以外にある機種PよフPi診断時の平
均スペクトラム、7(C11)をめてパワースペクトラ
ムa M ’CP + )を計算し、最大のパワースペ
クトラム偏M d rσ=Max(d+k) をめれば
、周波数σでのピーク形成をもってPiのパワースペク
トラム、7(pBが有する異常像候を代表していると悶
えられる。それよシ周波θσでピーク形成を与える異常
原因がn軸中i MにあるP機種Piにあることが診断
できる。
Rotating model specific: P n 1liilIl medium iqu: 6ru 2 models: Pi (1≦
i≦0) Pi's 14-dispersion power spectrum A: 5'
CP + ): (X I 1 + X + 2 + (
IXm row vector) ``'+''im) Pi diagnosis 17 o'clock average spectrum: -7 (at) range'' - [4G] - Sea [Pi): ] -1 (IXmO row vector) Power spectrum deviation: 2(pH) ==7(p
i) − [phase] Gi) (ixmO row vector) ” [d
t1dt2-- +dtm) = Thick power spectrum deviation: dlσ-eMax (d r kInIn axis medium short axis P model Pi has an abnormality, the power spectrum of Pl, 7 (pI) will vary according to the cause. A peak is given at a certain frequency σ.Generally, it is difficult to detect this and distinguish it from the peak specific to the KP model as mentioned earlier, but it is difficult to detect this when distinguishing it from the peak specific to the KP model. Calculate the power spectrum a M 'CP + ) by taking the average spectrum, 7 (C11) at The power spectrum of 7 (pB) is representative of the abnormal image characteristics that it has.It can be diagnosed that the cause of the abnormality that gives rise to a peak at the frequency θσ is in the P model Pi located at iM in the n axis. .

次に具体的な診断アルゴリズムの一例各第2図に示すフ
ローチャートを用いて説明する。
Next, an example of a specific diagnostic algorithm will be explained using the flowchart shown in FIG. 2.

診断が起動すると、n台の回転機機種Pの夫々について
パワースペクトラム”(PH) 、”(P2) 、 =
・、f(Pn)をめる。この時振動の時間的なバラツキ
を吸収するだめ何回かの平均をめる。次にn個のパワー
スペクトラムを加算してP全体のバヮースペクトラム、
、7[G) を計算する。n軸中i軸にあるP機種Pi
の診断の為に、Piのパワースペクトラム−7CPi]
とP全体のパワースペクトラムス〔)〕とからP’ 珍
に’r時の平均スペクトラムy1:Gi)の平均スペク
トラム波形パターンを計鴇する請求められたPi診断時
の平均スペクトラム=7(Gl)とPlのパワースペク
トラム、7(pI)よりパワースペクトラム偏差Jpi
)を計算し、最大のパワースペクトラム(i % d 
Iσ=Max(dilclをめて、ピーク値周波孜σi
t認知する。
When the diagnosis is started, the power spectrum "(PH), "(P2), = for each of the n rotating machine types P
・, calculate f(Pn). At this time, in order to absorb the temporal variation in vibration, an average is calculated several times. Next, add the n power spectra to get the power spectrum of the entire P,
, 7[G). P model Pi on the i-axis of the n-axis
For diagnosis, power spectrum of Pi - 7CPi]
Calculate the average spectrum waveform pattern of the average spectrum y1:Gi) from P' and the power spectrum of the entire P ().In rare cases, calculate the average spectrum waveform pattern of the average spectrum y1:Gi) at the time of 'r.The average spectrum at the time of Pi diagnosis requested = 7 (Gl) From the power spectrum of and Pl, 7 (pI), the power spectrum deviation Jpi
) and calculate the maximum power spectrum (i % d
Iσ=Max(with dicl, peak value frequency σi
tRecognize.

これよりn軸中i軸(CあるP機種PIを診断するPi
のパワースペクトラム−7(Pi) 、 pi診断時の
平均スペクトラム−71:Gl)及びピーク値周波数σ
iの3つのPi診断要素を計算機1が記憶する。計算機
1はこれら3つの要素をCRT表示装行2に表示し、P
iのパワースペクトラム、7(pt)が示す異常徴候を
ピーク値固波赦σIに於けるピーク形成をもって運転員
に的確に知らぜると共(C必要・tて応じて警報を与え
る。i軸にあるり1以外のJ l’71fにちるPJに
ついても同様にして診断す石ことができる。
From this, the i-axis of the n-axes (Pi to diagnose a certain P model PI)
Power spectrum -7 (Pi), average spectrum at the time of pi diagnosis -71:Gl) and peak value frequency σ
The computer 1 stores the three Pi diagnostic elements of i. Computer 1 displays these three elements on CRT display row 2, and
The power spectrum of i, 7 (pt) accurately informs the operator of the abnormality sign shown by the peak formation at the peak value σI (C necessary/t, and gives an alarm accordingly. It is possible to diagnose PJs other than J1'71f in the same way.

以上の説明から明らかなように、計算m1t7C上記診
断アルゴリズムを実現するためのプログラムを与えれば
、回転機器の異常診断を計算機1によりオンライン処理
で実施することが可能となる。
As is clear from the above description, if a program for implementing the calculation m1t7C diagnostic algorithm described above is given, abnormality diagnosis of rotating equipment can be performed by the computer 1 through online processing.

以上診断アルゴリズムの一例について説明したが、本発
明は以下に示す種々の異なる態様で実施することができ
る。
Although an example of a diagnostic algorithm has been described above, the present invention can be implemented in various different embodiments as shown below.

(1)上記の例では、1つの回転機種を診断の対象とし
たが、周波数外析器3のチャンネルを切替えること1よ
シ、複数の機種にわたる診断を1台の周波数汁析器によ
シ実施することができる。
(1) In the above example, one rotating model was targeted for diagnosis, but it is also possible to diagnose multiple models with one frequency analyzer by switching the channel of the external frequency analyzer 3. It can be implemented.

(2)複合サイクルプラントにあっては、その運用上清
に全軸を運転するとは限らない。よってこのような場合
には、通常運転中の機器だけを選択して上述のアルゴリ
ズムを適用する。更に運転台数が2台以下と少ない場合
には、相互比較が意味を持たなくなるから、それ以前の
最新の平均スペクトラムを計算機IK記憶させておき、
この最新の平均スペクトラムとの比較Vこよって異常診
断アルゴリスムを適用していく。
(2) In a combined cycle plant, all the shafts are not necessarily operated for its operational supernatant. Therefore, in such a case, only the devices in normal operation are selected and the above algorithm is applied. Furthermore, if the number of operating vehicles is small, such as 2 or less, mutual comparison becomes meaningless, so the latest average spectrum before that is stored in the computer IK,
Based on this comparison with the latest average spectrum, an abnormality diagnosis algorithm is applied.

(3JFFT(ファーストフーリエトランスフオーム)
ニよるパワースペクトラムは、有限個のデータ打切によ
るフーリエ変換でちる。よって波形の歪み庖生ずるため
、この歪を矯正する信号処理技術がなされCI/′、る
。上記の例では単VC2つの波形の偏差金求めただけで
あるが、使用される信号処理技術によつ−Cは、全周波
数帯の偏差を一律に比較するのではなく、周波数による
補正全入れた偏差により異常診断アルゴリズムを適用し
ていく。
(3JFFT (Fast Fourier Transform)
The power spectrum due to 2 is obtained by Fourier transform with a finite number of data truncation. As a result, waveform distortion occurs, and signal processing techniques have been developed to correct this distortion. In the above example, only the deviation of the waveforms of two single VCs was calculated, but depending on the signal processing technology used, -C does not uniformly compare the deviations of all frequency bands, but instead calculates the deviation by frequency. An abnormality diagnosis algorithm is applied based on the detected deviation.

以上記載の実21tli態様は、1゛μ」れも計算機l
のプログラムの簡単な拡張及び変杉により容易に実現で
きることは言うまでもない。
The actual 21tli mode described above is calculated using a calculator l
It goes without saying that this can be easily realized by a simple extension of the program and by Hensugi.

〔発明の効果〕〔Effect of the invention〕

一般に回転機器の回転に伴なう異常診断を、その振動波
形の周波数解析手法によシ来行する際に、パワースペク
トラム波形のパターン比較が問題になるが、本発明によ
れば、複合サイクルプラントにあっては、同−設計機種
が複数台設置されるという特徴をとらえることによシ、
同−設計j涜指間での相互比較を行うことで診断アルゴ
リズムが非常に簡素化される利点が得られるのみならず
、実機試験等による膨大なデータを蓄積することなしに
比較的容易に、しかも充分に合理性をもった異常診断が
可能となる。またこの簡素化された診断アルゴリズムは
、オンラインでの計算機処理を可能とし、更に繁雑で曖
昧、かつ微妙な判定を要する波形のパターン認はという
困難な作栗をも回避させる利点がある。
Generally, when diagnosing abnormalities associated with the rotation of rotating equipment using a frequency analysis method of its vibration waveform, a problem arises in comparing patterns of power spectrum waveforms, but according to the present invention, it is possible to By taking into account the characteristic that multiple models of the same design are installed,
Not only can you obtain the advantage of greatly simplifying the diagnostic algorithm by performing mutual comparisons between the same designs, but you can also compare them relatively easily without accumulating huge amounts of data from actual machine tests, etc. Furthermore, it becomes possible to diagnose abnormalities with sufficient rationality. Furthermore, this simplified diagnostic algorithm enables online computer processing, and also has the advantage of avoiding the difficult task of recognizing waveform patterns that are complex, ambiguous, and require delicate judgments.

斯くして周波数解析手法により、計算機によるオンライ
ン処理を可能とする複合サイクルジラントの回転機異常
診断方法を得ることができる。
In this way, by using the frequency analysis method, it is possible to obtain a method for diagnosing a rotating machine abnormality in a combined cycle gillant, which enables online processing by a computer.

【図面の簡単な説明】[Brief explanation of drawings]

第1図は本発明の一実施例を示すブロック図、第2図は
本発明に使用する診断アルゴリズムの一例を示すフロー
チャートである。 1・・・計n機 2・・・OR,T表示装置3・・周波
数分析器 4〜7・・振動センサ8〜11・・・回転軸
系或は回転補機 代理人 弁理士 則 近 憲 佑 (ほか1名)第1図 / 第2図
FIG. 1 is a block diagram showing an embodiment of the present invention, and FIG. 2 is a flowchart showing an example of a diagnostic algorithm used in the present invention. 1...N devices in total 2...OR, T display device 3...Frequency analyzer 4-7...Vibration sensors 8-11...Rotating shaft system or rotating auxiliary machine agent Patent attorney Nori Chika Ken Yu (and 1 other person) Figure 1/ Figure 2

Claims (1)

【特許請求の範囲】[Claims] 大々の回転軸系に含まれる同一種類の回転機が同−設計
のガスタービンと発電機、蒸気タービンと発′IIL機
又はガスタービンと発電機と蒸気タービン及び回転補機
とから構成される複数の回転軸系を有する複合サイクル
プラントの前記ガスタービン、発電機、蒸気ターピベ回
転補機の中の少なくとも一種類の複数の回転機の回転に
伴なう異常診゛断を前記回転機の生ずる振動波形の周波
数分析により行なう際に、前記各回転軸系に含まれる同
一種類の回転機相互間でのパワースペクトラム波形のパ
ターン比較を計算機を用いて行なうようにした複合サイ
クルプラントの回転機異常診断方法。
A rotating machine of the same type included in a large rotating shaft system is composed of a gas turbine and a generator, a steam turbine and a generator, or a gas turbine, a generator, a steam turbine, and a rotating auxiliary machine of the same design. Diagnosis of abnormalities associated with the rotation of at least one type of a plurality of rotating machines among the gas turbine, generator, and steam turbine rotating auxiliary equipment of a combined cycle plant having a plurality of rotating shaft systems is performed. A rotary machine abnormality diagnosis for a combined cycle plant in which a computer is used to compare patterns of power spectrum waveforms between rotating machines of the same type included in each rotating shaft system when frequency analysis of vibration waveforms is performed. Method.
JP58115974A 1983-06-29 1983-06-29 Abnormality diagnosing method of rotary machine of composite cycle plant Pending JPS608905A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP58115974A JPS608905A (en) 1983-06-29 1983-06-29 Abnormality diagnosing method of rotary machine of composite cycle plant

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP58115974A JPS608905A (en) 1983-06-29 1983-06-29 Abnormality diagnosing method of rotary machine of composite cycle plant

Publications (1)

Publication Number Publication Date
JPS608905A true JPS608905A (en) 1985-01-17

Family

ID=14675733

Family Applications (1)

Application Number Title Priority Date Filing Date
JP58115974A Pending JPS608905A (en) 1983-06-29 1983-06-29 Abnormality diagnosing method of rotary machine of composite cycle plant

Country Status (1)

Country Link
JP (1) JPS608905A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10883895B2 (en) 2016-12-15 2021-01-05 Mitsubishi Electric Corporation Abnormality diagnostic device for power transmission mechanism and abnormality diagnostic method for power transmission mechanism

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10883895B2 (en) 2016-12-15 2021-01-05 Mitsubishi Electric Corporation Abnormality diagnostic device for power transmission mechanism and abnormality diagnostic method for power transmission mechanism

Similar Documents

Publication Publication Date Title
Su et al. Induction machine condition monitoring using neural network modeling
US3731070A (en) Gas turbine engine analyzer
Wegerich Similarity based modeling of time synchronous averaged vibration signals for machinery health monitoring
JPS6312243B2 (en)
WO2005071620A2 (en) Method and apparatus to diagnose mechanical problems in machinery
EP3339830B1 (en) Modeling and visualization of vibration mechanics in residual space
CN114035140B (en) Fault detection and fault tolerance control method for current sensor of induction motor
CN115410729A (en) Online monitoring method, system and medium for vibration of shaft seal pump of nuclear reactor
EP3974646A1 (en) Method and system for wind speed determination using vibration data
JPH01101418A (en) Diagnosing device for rotary machine
JP3103193B2 (en) Diagnostic equipment for rotating machinery
JPS608905A (en) Abnormality diagnosing method of rotary machine of composite cycle plant
Tseng et al. A Diagnostic System for Speed‐Varying Motor Rotary Faults
JPH04204021A (en) Apparatus for diagnosing vibration and sound of rotating machine
JPH01270623A (en) Apparatus for diagnosing vibration of rotary machine
JPS63150633A (en) Shaft vibration diagnosing device
Lim et al. Motor fault detection method for vibration signal using FFT residuals
JPH02232529A (en) Method and apparatus for diagnosing vibration of rotary machine
JPS58100734A (en) Method and device for detecting fault of rotary equipment
JPH0399234A (en) Diagnosing method for abnormality of rotary machine
JPH01216218A (en) Diagnostic device for shaft vibration abnormality
JPS5857008A (en) Diagnosis of anomaly of water feeding pump and driving turbine thereof
JPH10111713A (en) Rotor diagnosis device
JPS628023A (en) Diagnosis of rotary machine
Su et al. A neural network method for induction machine fault detection with vibration signal