DE19907454A1 - Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters - Google Patents

Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters

Info

Publication number
DE19907454A1
DE19907454A1 DE1999107454 DE19907454A DE19907454A1 DE 19907454 A1 DE19907454 A1 DE 19907454A1 DE 1999107454 DE1999107454 DE 1999107454 DE 19907454 A DE19907454 A DE 19907454A DE 19907454 A1 DE19907454 A1 DE 19907454A1
Authority
DE
Germany
Prior art keywords
vibration
operating parameters
quantities
operating
parameters
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.)
Withdrawn
Application number
DE1999107454
Other languages
German (de)
Inventor
Georg Schneider
Rolf Zoeller
Manfred Weigel
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.)
Brueel and Kjaer Vibro GmbH
Original Assignee
Carl Schenck AG
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 Carl Schenck AG filed Critical Carl Schenck AG
Priority to DE1999107454 priority Critical patent/DE19907454A1/en
Publication of DE19907454A1 publication Critical patent/DE19907454A1/en
Withdrawn legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01HMEASUREMENT OF MECHANICAL VIBRATIONS OR ULTRASONIC, SONIC OR INFRASONIC WAVES
    • G01H1/00Measuring characteristics of vibrations in solids by using direct conduction to the detector
    • G01H1/003Measuring characteristics of vibrations in solids by using direct conduction to the detector of rotating machines

Abstract

With the method for model-based vibration diagnostic monitoring of rotating machines, it is possible to more precisely determine the relationships between the machine's vibration behavior and operating parameters while reducing the effort. This is achieved by carrying out the monitoring and analysis in the operational phase. The dynamics of the operating parameters are removed from the monitoring. It is thus possible to distinguish between limit violations caused by fluctuations in the operating parameters and violations caused by the actual change in the state of the machine.

Description

The invention relates to a method for model-based Vibration diagnostic monitoring of rotating machines.

The main task of a procedure for vibration diagnostic There is monitoring of the condition of rotating machines including an assessment, if possible without business interruption the current machine condition, the load on the machine and to allow any changes in the machine condition chen. Machine condition is understood as the evaluation the technical condition of the machine based on the Ge totality of the current values of all vibration quantities and Be drive parameters. Vibration sizes are all from the Vibration signal time functions derivable parameters, at for example, RMS value of the vibration speed or Spit Zen value of the rotational frequency vibration path component. Operating par meters are, for example, speed, power, excitation current, Temperatures and pressures.

An assessment of the machine condition, in which a qualita tive statement about the technical condition of the machine fen is done by analyzing the measured vibration sizes taking into account the operating parameters.

Such a method is known from the publication "VIBROCAM 5000, The Turboma Diagnostic Monitoring System schinen, C081, from Carl Schenck AG ", especially for use on steam turbine sets, gas turbines, turbopumps, Turbocompressors and hydroelectric machines is suitable.  

The rotors of the machines mentioned together with the Bearings and the foundation of a complex spring-mass system. The Vibration behavior depends heavily on the operating regime Operating mode, the operating status and the installation conditions gene from the machine, so that for each individual measuring point Machine operating mode, operating regime and operating status dependent, individual vibration quantities determined and for Assessment must be used.

Operating mode is basically to be distinguished end modes of the machine, such as. B. startup, normal operation and Spout. An operating regime differentiates within a Be possible different ways of working such. B. Turbine operation, pump operation and phase shifter operation at Pump storage sets in hydropower plants. The operating status is determined by the values of the significant operating parameters in characterized the operating regimes.

Changes in the vibration behavior can e.g. B. by Abnut tongues and damage, overloads and deformations due to interference in normal operation and due to influences the electrical network. The causes of Vibrations are essentially based on their appearance characterized. The highest information ge the frequencies of the dominant signal component hold le in the oscillation spectrum and the frequencies of the signal component le where changes occur.

In the known method of vibration diagnostic monitoring First, vibration quantities are recorded and the operating parameters in the respective operating regime and Operating state, as well as a frequency analysis and formation of Characteristics, the vibration behavior and its change characterize. Then in a learning phase the Normal range and the normal behavior of the selective characteristic  in the fluctuation range of the operating parameters for all loads drive regime and operating conditions determined. In the following The next step is a limit value comparison of the current ones len selective parameters with the corresponding parameters the normal state, so that alarms may be triggered or impending critical machine conditions in good time can be signaled.

From DE 37 25 123 is still a method for Schwin diagnostic diagnostic monitoring of rotating machines, ins special thermal turbomachinery known. With this ver are driving as vibration quantities har to the rotational frequency monical signal components in different states and stored in a pointer memory. Then in one Reference value store the arithmetic mean for each signal share filed. The Dif Reference value between the current state and the average reference condition determined, which is then compared with the normal range becomes. In addition, the associated measurement point operating parameters. With the help of this measurement data a function is then provided in a regressor, which is the reference value depending on the operating parameters can predict.

From the methods described above for vibration diagnosis The interrelationship is the monitoring of rotating machines between machine vibration behavior and operation parameters cannot be determined adequately. Furthermore are for the vibration levels of the different operating regimes and Operating conditions to specify a variety of limit values what to a large amount of data and a considerable amount of work wall leads.

Due to the known state of the art ing invention the task, the relationships between  vibration behavior of the machine and operating parameters when reducing the effort more precisely to determine the over to improve the monitoring and assessment of the machine condition.

This object is achieved by those specified in claim 1 Features resolved.

With the inventive method for model-based Vibration diagnostic monitoring of rotating machines it becomes possible to automatically dependencies of the vibrations to determine and display operating parameters. this leads to not only to significantly reduce the amount of data involved in the previous monitoring had to be saved, but also gives a better insight into the causes of the Vibrations. Changes in the machine condition are recognized. By optimizing the operating parameters The machine can operate with less vibration be enough.

Since the forward selection as well as the modeling based on the vibration magnitude currently recorded in the operating phase and operating parameters takes place, is the largest possible Agreement of the complex obtained in the modeling Model and the actual current operating state of the Machine reached. The com plexe model is adapted to the constant changes. Here through is a more current version compared to the prior art evaluation of the machine condition possible.

Furthermore, in the method according to the invention te reference operating parameters. The vibration test keeping the rotating machine is made up of a portion which results from known operating parameters, and given otherwise from parts resulting from changes in the machine result, together. Based on the current  tracked model and the inclusion of the previously determined the reference operating parameters, the fluctuations in the Vibration behavior resulting from the current operating parameters result, are removed so that the proportions resulting from Ver Changes in the machine condition result, who is extracted the.

The monitoring of the machine status becomes significant simplified because only the extracted portions that come from changes machine condition result, are visible. There on the one hand, it is then possible to immediately get information about to maintain the machine condition of the machine as well as one create long-term trend about the condition of the machine len.

Furthermore, Dar is free from dynamic influences position of the vibration behavior possible. One by Be changes in drive parameters hidden change of the machine condition is thus visible and a further assessment accessible.

In a further development of the inventive concept is in front seen that when evaluating the relative deviation of the Vibration quantities for all operating regimes and operating states a few or a single characteristic value is specified. Hereby will drastically reduce the number of Limit values for monitoring the machine status enough.

The present invention is based on the vibration diagnosis tical monitoring of a pump storage set explained in more detail.

Show it:

Figure 1 is a schematic representation of the shaft train of a pump memory set with the measuring points and the data processing unit.

Fig. 2 is a diagram of the learning phase in a block diagram;

Fig. 3 is an illustration of the model formation in the operating phase Be in a block diagram;

Fig. 4 is a schematic representation of the monitoring process in the operating phase.

In Fig. 1, the shaft train 1 of a pump storage replacement with the measuring points for vibration measurement is shown schematically. The bearings and the measuring planes for wave vibration measurements are each assigned to transducers 2 , 3 for detecting vibration signals. The vibration signals determined at the measuring points via the sensors 2 , 3 are forwarded to a data processing unit 4 (represented by arrows 5 ). At the same time, a reference signal 3 ″ (one pulse per machine revolution) is derived from the shaft train 1 by means of reference transducer 3 ′ and fed to the data processing unit 4. In addition, a large number of transducers are provided which show the different operating parameters, such as, for example, power, excitation current, pressures and temperatures, the measurement signals for the operating parameters are likewise forwarded to the data processing unit 4 (represented by arrows 6 ). In the data processing unit 4 , vibration values 5 'are determined from the vibration signals 5 and, if appropriate, from the reference signals 3, and stored or buffered. At the same time, the measured values for the operating parameters 6 'are also stored.

In the model-based method for vibration diagnostic monitoring according to the invention, a model is first of all carried out in a learning phase, which is shown schematically in FIG. 2, using a simple model.

The first goal of modeling is to determine whether all operating parameters are measured that have a decisive influence on the vibration behavior of the machine. This can be checked by predicting the vibration quantities from the operating parameters alone. It is first checked whether it is possible to draw a sufficiently precise figure from the parameter values to determine the vibration magnitudes. If this is possible, the information sought is represented in the data. This is the approach

i = (x i )

made, denoting the model function. A possible approach for is a linear combination of generally nonlinear basic functions. The model for a data point of the target variable y i = y ( its ) (t s : sampling time) for any state vector x i is then given by

where x j is a basic function of the model.

In the learning phase, all the measured operating parameters 6 ′ and stored in the data processing unit are first transmitted to a forward selection unit 7 . The vibration quantities 5 'are also transmitted. It is first assumed that a linear prediction model that the TO state vector x i = (x i 1, x i 2,... X i d) T combined operation parameter 6 '(the various operating parameters be designated by the high indices) by Linear combination of its components on the vibration quantities y i 1 , y i 2,. . . y 1 k i = 1, 2,. . . N maps (the high indices denoting the different vibration quantities). The forward selection process, which is described in detail below, is then used to assess the relevance of the vibration-determining operating parameters.

The selection of the vibration-determining operating parameters is thus traced back to a model structure determination problem, because the individual operating parameters can be understood as terms of a model and the terms that lead to an optimal model can be selected by means of term selection methods. The relevant operating parameters 8 selected in this way are the output variables of the forward selection unit 7 .

The model structure is determined using the forward selection method described below, which is carried out in the forward selection unit 7 in the learning phase. For an initially empty set of operating parameters, the size that reduces the quadratic error χ 2 most is gradually added. This results in a ranking that indicates which operating parameters have the greatest influence on the vibration variables. The more operating parameters are taken into account, the smaller this quadratic error χ 2 . However, it only refers to the data of the learning phase (training data) and does not allow any statement about how the model reacts to unknown data (test data). The quadratic error χ 2 is too unsuitable for the selection of relevant terms.

The so-called estimate provides a necessary statement prediction error with the test data. This indicates how exactly the trained model for future, unknown data predicts. If there is sufficient data from the learning phase this can be done by dividing the data into a Trai quantity of data and test data.  

Another option is to use a very much more efficient method that is known in statistics and is called "cross-validation" (B. Efron and R. J. Tibshirani "An Introduction to the Bootstrap", Chapman and Hall, 1993). With this method, several divisions are made Training and test data sets made. An extreme Va riante of it is to convert the N data points into one training data set of size N-1 and a test data set of size 1 to share. This procedure is called "Leave-One-Out (LOO) Cross-Vali dation ". The selection criterion σ2 then results as Average of the square errors when predicting the left test records.

With i (x i ) the prediction of the i-th data set after the model has been trained with the other N-1 data sets, the test data error σ 2 results in:

The advantage of this method is on the one hand that there is no interference solution of the mean by the division into training and Test data volume arises and secondly, that the entire Training and test data set can be used for training can.

In contrast to the quadratic error χ 2 , which indicates how well the model prediction matches the training data, σ 2 gives a measure of the match for unknown data sets. The previously introduced error function zuvor 2 decreases monotonically with the increase in new basic functions and is therefore unsuitable for selecting relevant terms. However, the LOO error σ 2 initially decreases with the addition of new basic functions and increases again from a critical number, since the error between the data sets of the training volume increases (overfit). This property is used to select relevant terms.

The calculation of the optimal model structure according to the relationships given above

requires knowledge of N data records.

In Fig. 3, the further modeling is shown schematically, which he follows in the adaptive modeling unit 20 . The further modeling takes place in the operating phase, taking into account the currently measured vibration quantities 5 'and operating parameters 6 '.

The operating parameters, which are known in the previously described internal phase as relevant for the prediction of the vibration quantities, are fed as input variables 8 to a polynomial generator 9 . A complex res and thus more powerful model is determined in the polynomial generator 9 . Complex models can be realized by adding power terms and product terms. These models are called polynomial models. The selection of the optimal model terms from a given superset is again a form of the model structure determination and is carried out by forward selection, as described below in the forward selection unit 11 .

In the operating phase, it is also desirable to receive updated values of sizes χ 2 and σ 2 for each newly measured data set. The analog quantities are J (n) and j 2 (n) and are calculated according to the following formulas in the forward selection unit 11 :

These variables correspond to the variables χ 2 and σ 2 calculated in the learning phase with the difference that an exponential weighting is carried out using the so-called memory factor λ. Here n represents the current point in time (n = nt s ; t s : sampling time). In addition, the values are updated with the receipt of each new data record. The memory factor λ determines how quickly the errors J (n) and j 2 (n) adapt to the current situation.

In the forward selection unit 11 , the structure of the model is determined from the vibration quantities 5 and the basic function 10 provided by the polynomial generator 9 . The optimal parameters a j are then determined by minimizing the weighted quadratic model error J (n):

in the subsequent recursive least squares parameter estimation unit 13 on the basis of the supplied vibration quantities 5 ′ and the selected basic function 12 . This leads to a linear system of equations, the solution of which provides the sought model parameter 14 .

To solve this system of equations during loading Updating the drive phase is a recursive solution driving necessary. Recursive means that with each new measured his data set (vibration size and operating parameters) Update of the solutions of the system of equations in the sense of least squares. Such a procedure is as "recursive least squares" from Simon Haykin, Adaptive Fil ter Theory, Prentice Hall, Chapter 13, pages 477-486, 1991 known.

The model structure is determined using the forward selection method described below in the forward selection unit 11 . This method is analogous to the method described for the forward selection unit 7 .

If a mere adaptation of the model parameters is insufficient to determine a sufficiently precise model of the machine, a forward selection in the forward selection unit 11 is carried out automatically. For this purpose, it is determined which operating parameter reduces the weighted quadratic error J (n) the most. This determines the order in which the various operating parameters are added to the model. At the same time it is considered how the exponentially weighted test data error

changed. If this error increases after adding a Be drive parameters so this last added Be  drive parameters removed from the model and another Hin increase in operating parameters in the model stopped. The new model structure is determined.

The result is a compact formulaic relationship (hereinafter referred to as optimized model 12 ) between vibration quantities and operating parameters. If in the further course of the monitoring an adaptation of the model parameters is insufficient to determine a sufficiently precise model, then this is an indication that the machine state has changed structurally and a new forward selection in FIG. 11 is necessary. This is carried out automatically. In this way, there is always an up-to-date formula. Until this new correlation is not yet complete, the previously used one is transferred to the computing unit 15 . If the new relationship exists, it is now transferred to the computing unit 15 as an optimized model 12 . The values for the model parameters aj 14 are determined in the least square parameter estimation unit 13 , taking into account the optimized model 12 . In the arithmetic unit 15 , the prediction of the vibration variables 16 is then carried out according to the optimized model 12 , taking into account the optimal model parameters 14 , which are likewise fed to the arithmetic unit 15 . The predicted vibration variables 16 are available for further processing as output variables of the computing unit 15 .

In Fig. 4, the monitoring process running in the operating phase is shown schematically.

The preselected operating parameters 8 represent the input variables both for the shaft train 1 of the pump storage unit, which is shown schematically, and for the adaptive model formation unit 20. The arrow 19 indicated in the adaptive model formation unit 20 indicates the adaptation of the model parameters to the currently measured ones Vibration sizes 5 '. In the difference image 18 , the difference between the measured vibration quantities 5 ′ and the vibration quantities 16 predicted in the adaptive model formation unit 20 by the model 20 a is formed and as an error to be minimized in the adaptive model formation unit 20 by the model 20 a according to the mathematical relationships

considered. The model 20 a formed in the adaptive model formation unit 20 , consisting of an optimized model 12 and the associated model parameters 14 , which contains the functional relationship between vibration quantities and operating parameters, is displayed in the display unit 21 .

An always current copy of the model 20 a formed in the adaptive model formation unit is referred to as 20 b and is transferred to an adaptive model formation unit 20 '. This copied adaptive model 20 b receives as input variables not fixed time-varying operational parameter, referred to as a reference operating parameters 17th One or more preferred parameter settings typical of the operation (for example low, medium and peak load) can be selected as suitable values. The temporal change in both the model parameters and the model structure results in time-variable predictions of the vibration quantities 23 , although the reference operating parameters 17 are constant. These changes in the vibration quantities are directly related to the change in the machine state and are displayed in the display device 22 .

The vibration quantities 23 determined in this way, which are cleared of the vibration changes resulting from the changes in the operating parameters, can be monitored in a subsequent conventional monitoring system against predetermined limit values.

Claims (1)

  1. Method for model-based vibration diagnosis Monitoring of rotating machines, in which in a Learning phase first of all vibration quantities and company savings meters are recorded and saved and then a Modeling takes place in several steps, first using a simple, e.g. linear model the operation combined to the state vector x (i) parameters by linear combination of its components are mapped to the vibration quantities y (i) and by comparing the measured and the predicted Vibration quantities first checked using the model is whether all vibration-relevant quantities are recorded and a Be evaluation of the ranking of the operating parameters for relevance done, then current in the operating phase Vibration quantities and operating parameters are recorded then using selected relevant operating parameters modeling with a complex model takes place and then another with the forward selection procedure Evaluation of the ranking of the operating parameters on rele vanz takes place, so that a functional connection Basis of a complex model between selected rele vant operating parameters and vibration variables can be derived is, with both the forward selection and the Mo dell formation based on the currently recorded Schwin size and operating parameters and from the current functional context information on Evaluation of the rotating machine can be obtained and  additionally with previously defined reference operating parameters meters and the current functional relationship The machine condition is assessed.
DE1999107454 1999-02-22 1999-02-22 Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters Withdrawn DE19907454A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
DE1999107454 DE19907454A1 (en) 1999-02-22 1999-02-22 Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
DE1999107454 DE19907454A1 (en) 1999-02-22 1999-02-22 Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters
AU25491/00A AU2549100A (en) 1999-02-22 2000-02-18 Method for conducting model-based, vibration diagnostic monitoring of rotating machines
PCT/EP2000/001307 WO2000050857A1 (en) 1999-02-22 2000-02-18 Method for conducting model-based, vibration diagnostic monitoring of rotating machines
DK200101240A DK200101240A (en) 1999-02-22 2001-08-21 Method of model-based oscillatory diagnostic monitoring of rotating machines

Publications (1)

Publication Number Publication Date
DE19907454A1 true DE19907454A1 (en) 2000-08-24

Family

ID=7898368

Family Applications (1)

Application Number Title Priority Date Filing Date
DE1999107454 Withdrawn DE19907454A1 (en) 1999-02-22 1999-02-22 Method for model based vibration diagnostic monitor of rotating machines by judging machine condition being related to current function and previously established reference operating parameters

Country Status (4)

Country Link
AU (1) AU2549100A (en)
DE (1) DE19907454A1 (en)
DK (1) DK200101240A (en)
WO (1) WO2000050857A1 (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7010445B2 (en) 2002-01-23 2006-03-07 Csi Technology, Inc. Automated fault diagnosis device and method
DE102009038011A1 (en) 2009-08-20 2011-03-10 Schenck Rotec Gmbh Method for automatic detection and detection of errors on a balancing machine
WO2015158343A3 (en) * 2014-04-16 2015-12-10 Schaeffler Technologies AG & Co. KG Method for reducing low-frequency vibrations in the drive train of a motor vehicle
DE102018115354A1 (en) * 2018-06-26 2020-01-02 Rolls-Royce Deutschland Ltd & Co Kg Device and method for determining at least one rotation parameter of a rotating device

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3725123A1 (en) * 1986-10-27 1988-04-28 Inst Energieversorgung Method for selective vibration monitoring of rotating machines
DE4243882C1 (en) * 1992-12-23 1994-01-05 Baleanu Michael Alin Method and device for monitoring a technical process
DE4406723A1 (en) * 1993-02-27 1994-09-01 Boetius & Partner Informations Method for monitoring the operational state of a machine or installation
DE19754918A1 (en) * 1996-12-10 1998-06-18 Caterpillar Inc Improvements in component machine testing using vibration data analysis processed by a neural network

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5602761A (en) * 1993-12-30 1997-02-11 Caterpillar Inc. Machine performance monitoring and fault classification using an exponentially weighted moving average scheme

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3725123A1 (en) * 1986-10-27 1988-04-28 Inst Energieversorgung Method for selective vibration monitoring of rotating machines
DE4243882C1 (en) * 1992-12-23 1994-01-05 Baleanu Michael Alin Method and device for monitoring a technical process
DE4406723A1 (en) * 1993-02-27 1994-09-01 Boetius & Partner Informations Method for monitoring the operational state of a machine or installation
DE19754918A1 (en) * 1996-12-10 1998-06-18 Caterpillar Inc Improvements in component machine testing using vibration data analysis processed by a neural network

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
HAMSCHER,Walter (Hrsg.), et.al.: Model-based Diagnosis, Morgan Kaufmann Publishers, Inc., San Mateo, CA, USA, 1992, S.249-254 *
ISERMANN,Rolf: Modellgestützte Überwachung und Fehlerdiagnose Technischer Systeme (Teil 1). In: atp - Automatisierungstechnische Praxis 38, 1996, 5, S.9,10,12-14,16-20 *
REMBOLD,Ulrich: Methoden der Künstlichen Intelligenz in der industriellen Fertigung. In: TR Technische Rundschau, H.13, 1992, S.38-49 *
VIBROCAM 5000 der Fa. Schenck Vibro GmbH, Darmstadt, 1998, S.3-11 *

Cited By (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7010445B2 (en) 2002-01-23 2006-03-07 Csi Technology, Inc. Automated fault diagnosis device and method
DE102009038011A1 (en) 2009-08-20 2011-03-10 Schenck Rotec Gmbh Method for automatic detection and detection of errors on a balancing machine
US8561463B2 (en) 2009-08-20 2013-10-22 Schenck Rotec Gmbh Method for the automatic detection and identification of errors in a balancing machine
DE102009038011B4 (en) 2009-08-20 2018-03-08 Schenck Rotec Gmbh Method for automatic detection and detection of errors on a balancing machine
DE102009038011B9 (en) * 2009-08-20 2018-04-12 Schenck Rotec Gmbh Method for automatic detection and detection of errors on a balancing machine
WO2015158343A3 (en) * 2014-04-16 2015-12-10 Schaeffler Technologies AG & Co. KG Method for reducing low-frequency vibrations in the drive train of a motor vehicle
CN106233024A (en) * 2014-04-16 2016-12-14 舍弗勒技术股份两合公司 The method of the low-frequency vibration in the PWTN reducing motor vehicles
US10215240B2 (en) 2014-04-16 2019-02-26 Schaeffler Technologies AG & Co. KG Method for reducing low-frequency vibrations in the drive train of a motor vehicle
CN106233024B (en) * 2014-04-16 2019-06-07 舍弗勒技术股份两合公司 The method of the low-frequency vibration in powertrain for reducing motor vehicle
DE102018115354A1 (en) * 2018-06-26 2020-01-02 Rolls-Royce Deutschland Ltd & Co Kg Device and method for determining at least one rotation parameter of a rotating device

Also Published As

Publication number Publication date
DK200101240A (en) 2001-08-21
AU2549100A (en) 2000-09-14
WO2000050857A1 (en) 2000-08-31

Similar Documents

Publication Publication Date Title
US10527520B2 (en) Operating wind motors and determining their remaining useful life
US20170293862A1 (en) Machine learning device and machine learning method for learning fault prediction of main shaft or motor which drives main shaft, and fault prediction device and fault prediction system including machine learning device
US4352293A (en) Method of diagnosing vibration of rotary machine
JP4649497B2 (en) Method for diagnosing lifetime of member and method and apparatus for diagnosing the lifetime
US10809153B2 (en) Detecting apparatus, detection method, and program
ES2540859T3 (en) Inferential signal generator for instrumented equipment and procedures
US5486995A (en) System for real time optimization
US5608845A (en) Method for diagnosing a remaining lifetime, apparatus for diagnosing a remaining lifetime, method for displaying remaining lifetime data, display apparatus and expert system
KR0157052B1 (en) Improved plant operating system employing a deterministic, probabilistic and subjective modeling system
US8332164B2 (en) Method for determining fatigue damage in a power train of a wind turbine
US6651012B1 (en) Method and apparatus for trending and predicting the health of a component
US5511004A (en) Diagnostic method for an evolutionary process
DE60219111T2 (en) Method and system for the diagnosis of turbine engines
AU618111B2 (en) Method and apparatus for diagnosing machines
US6820222B2 (en) Apparatus and method for processor power measurement in a digital signal processor using trace data and simulation techniques
Engel et al. Prognostics, the real issues involved with predicting life remaining
US6711523B2 (en) Method and apparatus for determining a condition indicator for use in evaluating the health of a component
US7395188B1 (en) System and method for equipment life estimation
CA2438056C (en) Process unit monitoring program
US8116990B2 (en) Method and system for real-time prognosis analysis and usage based residual life assessment of turbine engine components and display
DE102011102034A1 (en) Online adjustment of a process-analytical model with effective process operation
RU2512610C2 (en) Method and monitoring system of vibrating phenomena appearing in gas-turbine engine of aircraft in operating time
US5922963A (en) Determining narrowband envelope alarm limit based on machine vibration spectra
US8170835B2 (en) Data analysis applications
US4060716A (en) Method and apparatus for automatic abnormal events monitor in operating plants

Legal Events

Date Code Title Description
OM8 Search report available as to paragraph 43 lit. 1 sentence 1 patent law
8127 New person/name/address of the applicant

Owner name: BRUEEL & KJAER VIBRO GMBH, 64293 DARMSTADT, DE CAR

8110 Request for examination paragraph 44
8139 Disposal/non-payment of the annual fee