WO2019043600A1 - Remaining useful life estimator - Google Patents
Remaining useful life estimator Download PDFInfo
- Publication number
- WO2019043600A1 WO2019043600A1 PCT/IB2018/056597 IB2018056597W WO2019043600A1 WO 2019043600 A1 WO2019043600 A1 WO 2019043600A1 IB 2018056597 W IB2018056597 W IB 2018056597W WO 2019043600 A1 WO2019043600 A1 WO 2019043600A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- rul
- features
- data
- fault
- estimating
- 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.)
- Ceased
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/14—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object using acoustic emission techniques
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M13/00—Testing of machine parts
- G01M13/04—Bearings
- G01M13/045—Acoustic or vibration analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/36—Detecting the response signal, e.g. electronic circuits specially adapted therefor
- G01N29/42—Detecting the response signal, e.g. electronic circuits specially adapted therefor by frequency filtering or by tuning to resonant frequency
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4409—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by comparison
- G01N29/4418—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by comparison with a model, e.g. best-fit, regression analysis
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4472—Mathematical theories or simulation
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/46—Processing the detected response signal, e.g. electronic circuits specially adapted therefor by spectral analysis, e.g. Fourier analysis or wavelet analysis
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0221—Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0259—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the response to fault detection
- G05B23/0283—Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL]
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Signal Processing (AREA)
- Acoustics & Sound (AREA)
- Automation & Control Theory (AREA)
- Mathematical Physics (AREA)
- Algebra (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
Abstract
A method and a system for estimating the remaining useful life, which estimates the life cycle of rotatory equipment in real-time, the method comprising the steps of: receiving high frequency vibration signals; pre-processing the received signals, comprising cleaning the data received from the sensors; extracting features, comprising applying various transformations on the raw data and extracting statistical variables or features, wherein waveforms are processed; selecting features, comprising capturing all relevant information in a minim number of variables; classifying the information, allowing information creation based on technics that automatically classify the data according to the features included, wherein the classification step is used for fault detection, identification of the component involved in the fault and of the fault type, and to estimate the severity of the fault; and estimating the Remaining Useful Life (RUL), wherein a specific model is associated to a specific class, inferring the status of the machinery and estimating RUL by means of non-linear multivariate regression techniques.
Description
REMAINING USEFUL LIFE ESTIMATOR
SPECIFICATION
The present text describes a technical solution named "Remaining Useful Life Estimator" or "RUL Estimator", directed to estimate the life cycle of rotatory equipment in real-time. The solution relates to all industries using rotatory equipment, especially in Mining, Refining, Power, Pulp and Paper.
Rotating mechanical components are widely used in all industries, their failure are common and can lead to safety issues, production losses, and unplanned downtime costs.
Background
Prior art solutions might implement hardware allowing for vibration data acquisition, using said information for estimating the lifespan of rotatory equipment based either on algorithms or in condition monitoring. Current developments focus on the hardware for data collection, or on the software for estimating the health status of mechanical components. Besides, the analysis and detection is made by human analysts.
The claimed solution is unique by including special data processing methods in seamlessly relationship with a data collection system operating in real time, combining real time measurements with lifespan predictions based on the behavior of the monitored equipment. Description of the technical solution
As indicated above, the proposed technical solution is directed estimate the life cycle of rotatory equipment in real-time. The solution aims at:
1. Automatic fault detection and identification in rotating mechanical components,
2. Automatic estimation of the severity level of faults,
3. Automatic estimation of the remaining useful life at any time of the rotating mechanical components.
Current monitoring technics allow for fault detection and identification relying on human expert or analyst expertise. Our solution goes beyond that, i.e. the automatic detection and identification of failures and the prediction of the Remaining Useful Life (RUL) of the rotating mechanical components after that failure.
The solution involves a system and a computer implemented method that work together to give the best estimation of the RUL, as depicted in Figure 1 (Conceptual Representation of the Invention).
The system comprises a vibration data collection equipment, which alternative implements a method for capturing and processing data, called PeakVue. Said method of capturing and processing data is embedded in the system, particularly in the data collection equipment. The data collection equipment combines prediction and protection capabilities to provide a complete online machinery monitoring solution, integrating protection, prediction, real-time performance monitoring and process automation. The data collection equipment can even be deployed in a prediction-only configuration to overlay on an existing protection system. The data collection equipment is field-proven, providing API 670 protection on the industry's most critical equipment: steam turbine generators, gas turbines, boiler feed pumps, offshore compressors, pipeline turbo compressors, chemical industry compressors, turbo exhausters, blowers, and boosters.
Besides, the system comprises at least one receiver for receiving high frequency vibration signals which can be complemented by other measurement technics such as temperature, acoustics, or oil analysis, using proper sensing means for said purposes.
On the other hand, the computer implemented method comprises a set of steps that interpret and transform the raw signals from the sensors into actionable information.
Once the signals are acquired, data processing is applied for calculating the RUL. Figure 2 (Data processing process for RUL Estimation) shows the data processing process implemented by the method and system of the proposed solution.
Pre-processing: This first step comprises cleaning the data received from the sensors. This includes removing offsets, noises, biases, and trends, smoothing, filtering high frequency signals where no relevant information is present, checking data integrity, replacing and normalizing for comparing a wide range of signals from various data sources. Some approaches for this step are the implementation of normalization, filtering and whitening of the received data.
Features Extraction (figure 3): This second step is of the most important of the process. It consists of applying various transformations on the raw data and extracting statistical variables or features. In this case, waveforms are processed, not trends. Some transformations are Fast Fourier Transformation, Wavelets, Root Mean Square (RMS) and Kurtosis, among others.
Wavelets are mathematical functions that cut up data into different frequency components, and then study each component with a resolution matched to its scale. This approach would result in obtaining information from time and frequency domain and performing a multiscale analysis. Some energy- wavelets channels show tendencies that can be useful for RUL estimation, as can be seen in figure 4 (Wavelet vibration analysis over an example bearing).
Feature Selection: Once features are generated/extracted, the selection of those is a fundamental step to capture all relevant information in a minim number of variables. The features can be selected by Principal Components Analysis (PCA), Sequential Feature Selection (SFS), based on Mutual Information techniques or by manifold learning algorithms as T distributed Stochastic Neighbor Embedding (tSNE).
The implementation of the feature selection generally the strategy of figure 5 (feature selection). Classification: The classification step allows information creation based on technics that automatically classify the data according to the features included. In this case, the classification step is used for fault detection, identification of the component involved in the fault and of the fault type, and to estimate the severity of the fault.
RUL Estimation: The estimation of the Remaining Useful Life is based on all the previous steps. The estimation is based on a multivariate regression using models based on the class detected in
the previous step. A specific model is associated to a specific class. Using the information form previous step (Classification) the system can infer the status of the machinery. Using non-linear multivariate regression techniques as Random Forest or Support Vector Regression, the system could estimate the RUL. These techniques can be combined with other statistical models as Kalman Filters and Markov models.
The RUL of an asset or component is the remaining time left from the measuring time to the end of the useful life. The RUL can be measured through the time, so a curve can be constructed based on predictions. The "real RUL" (ground truth) or the time the piece actually lived is always a straight line (see figure 6: RUL prediction and Ground Truth (GT) of an example bearing).
In this example, mainly, acceleration signals are processed (velocity and displacement can also be computed), wherein the Waveform is of a high frequency signal (> 5kHz) recorded for a short time (~ 0.1 to 10 s), the trend is time series of long periods (days, weeks, etc.), usually taken as a "snapshot" (RMS value) every second and the Spectrum is the distribution of power into frequency components composing that signal. Usually computed using FFT (Fast Fourier Transform). In figure 7 an example of waveform, trend and spectrum of a vibration measurement is shown.
The system and method integration performs a complete automated estimation of the health of mechanical components.
In a particular embodiment of the solution the features extracted from the signal are captured and processed using an special method called PeakVue, technology that provides earlier, more accurate indications of developing faults in rolling element bearing machines, providing a simple, reliable indication of equipment health via a single trend. PeakVue filters out traditional vibration signals to focus on impacting, a much better indicator of overall asset health on any type of rolling element bearing machine, for instance. As a measure of impacting, PeakVue readings are much easier to interpret. As an example, a good machine, properly
installed and well lubricated, should normally not have any impacting on it. This establishes the zero principle: The PeakVue measurement on a good machine should be at or close to zero. As common machinery faults begin to appear on rotating equipment (e.g. rolling element bearing defects, gear defects, insufficient lubrication, or pump cavitation), the PeakVue reading typically can be evaluated using the Rule of lO's: A PeakVue value of 10 would implies abnormal situation developing (monitoring should be made more closely), PeakVue value of 20 would implies serious abnormal situation (an action plan should be developed), PeakVue value of 40 would implies critical abnormal situation (action plan should be implemented). With these simple principles, PeakVue is a powerful tool to bring reliability to the control room. Operators with no special training in machinery diagnostics can use PeakVue measurements quickly and easily to determine both when a piece of rotating equipment is healthy and when an abnormal situation is present. Once an abnormal situation has been identified using the PeakVue overall, detailed diagnostic information can be extracted from the PeakVue waveform or spectrum to determine the exact nature of the defect. PeakVue can visualize distress signals on a machine that are simply not visible with other vibration measurements. Earlier indication of developing defects facilitates optimum maintenance planning and minimizes the impact on production.
Further information is detailed in the following examples:
Example 1
In the example, 3 sets of 4 coupled bearings are tested (see figure 8 scheme, Arrangement of bearings, example). Waveforms of 1 sec length every 10 minutes are obtained. ~ 1 month per test.
SET 1
4 bearings / 2 channels (horizontal/vertical)
At the end of the test-to-failure experiment, inner race defect occurred in bearing 3 and roller element defect in bearing 4. Vibration analysis in four bearings, set 1, is shown in figure 9.
SET 2
4 bearings / 1 channel
At the end of the test-to-failure experiment, outer race failure occurred in bearing 1. Vibration analysis in four bearings, set 2, is shown in figure 10.
SET 3
4 bearings / 1 cannel
At the end of the test-to-failure experiment, outer race failure occurred in bearing 3. Vibration analysis in four bearings, set 3, is shown in figure 11.
Test Set Up. Set 1
2 artificial signals were generated from random samples of three original bearings signals:
• One (1) training data generated with 468 'records' (balanced training)
o 156 records for each class (3 classes)
• One (1) test data generated with 180 'records'
o 60 records for each class (3 classes)
• Feature extraction: RMS, Kurtosis, WPD (5 levels) and wavelets (energy)
• Random Forest Classifier (300 trees).
• Four (4) classification tests were performed (by random permutation of the records). Results of set 1 are shown in figure 12.
Test Set Up. Set 2
2 artificial signals were generated from random samples of three original bearings signals:
• One (1) training data generated with 468 'records' (balanced training)
o 156 records for each class (3 classes)
• One (1) test data generated with 3220 'records'
o 1200 records (class 1), 1200 records (class 2) and 820 records (class 3)
• Feature extraction: RMS, Kurtosis, WPD (5 levels) and wavelets (energy)
• A Random Forest classifier (300 trees) was used.
• Four (4) classification tests were performed (by random permutation of the records)
Results of set 2 are shown in figure 13.
Test Set Up. Set 3
2 artificial signals were generated from random samples of three original bearings signals:
• One (1) training data generated with 468 'records' (balanced training)
o 156 records for each class (3 classes)
• One (1) test data generated with 20692 'records'
o 8984 records (class 1), 8256 records (class 2) and 3452 records (class 3)
• Feature extraction: RMS, Kurtosis, WPD (5 levels) and wavelets (energy)
• A random forest classifier (300 trees) was used.
• One (1) classification test was performed.
Resume Tests
Table 1: Resume tests sets 1 to 3.
In order to improve classification results, a temporal window was tested on database. Using a window of 9 records the result was compared with the classifier without window. Results reached -99,6 on all tests. Between parentheses the classification rates using temporal window are shown.
Table 2: Resume tests sets 1 to 3 using temporal window.
Example 2: RUL
A training step is performed with data given to the classifier and regression model to train the system. Training samples are shown in figure 14.
RUL estimation window is presented as known data and the task is predict the future RUL, as shown in Figure 15.
Using the outputs of the classifier, a RUL regression model per class was trained with a database. Multiple test and simulation to find best parameters were performed. Post processing filtering or weighting were implemented to smooth the curves. Normalization of the RUL estimation between [0, 1] enable incorporate different length waveforms to training set. RUL results are shown in figure 16.
After multiple test using different set up for Classification and post processing best results were obtained (until now) using regression per class and RUL normalization. Random Forest has performed better than SVM for regression
Claims
1. A method for estimating the remaining useful life (RUL), which estimates the life cycle of rotatory equipment in real-time, the method comprising the steps of:
receiving high frequency vibration signals which can be complemented by other measurement technics such as temperature, acoustics, or oil analysis, using proper sensing means for said purposes;
pre-processing the received signals, comprising cleaning the data received from the sensors, including at least one of removing offsets, noises, biases, and trends, smoothing, filtering high frequency signals where no relevant information is present, checking data integrity, replacing and normalizing for comparing a wide range of signals from various data sources; extracting features, comprising applying various transformations on the raw data and extracting statistical variables or features, wherein waveforms are processed, and wherein the transformations are selected from the group of at least one of Fast Fourier Transformation, Wavelets, Root Mean Square (RMS) and Kurtosis;
selecting features, comprising capturing all relevant information in a minim number of variables, wherein the features can be selected by at least one of Principal Components Analysis (PCA), Sequential Feature Selection (SFS), based on Mutual Information techniques or by manifold learning algorithms as T distributed Stochastic Neighbor Embedding (tSNE);
classifying the information, allowing information creation based on technics that automatically classify the data according to the features included, wherein the classification step is used for fault detection, identification of the component involved in the fault and of the fault type, and to estimate the severity of the fault; and
estimating the Remaining Useful Life (RUL), which is based on all the previous steps, mainly on a multivariate regression using models based on the class detected in the previous step, wherein a specific model is associated to a specific class, inferring the status of the
machinery and estimating RUL by means of non-linear multivariate regression techniques as Random Forest or Support Vector Regression, which can be combined with other statistical models as Kalman Filters and Markov models.
2. A remaining useful life (RUL) estimator, which estimates the life cycle of rotatory equipment in real-time by implementing the method of claim 1, the system comprising:
a vibration data collection equipment, having a method of capturing and processing data embedded in the data collection equipment;
at least one receiver for receiving high frequency vibration signals which can be complemented by other measurement technics such as temperature, acoustics, or oil analysis, using proper sensing means for said purposes.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US201762551285P | 2017-08-29 | 2017-08-29 | |
| US62/551,285 | 2017-08-29 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2019043600A1 true WO2019043600A1 (en) | 2019-03-07 |
Family
ID=65525130
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/IB2018/056597 Ceased WO2019043600A1 (en) | 2017-08-29 | 2018-08-29 | Remaining useful life estimator |
Country Status (1)
| Country | Link |
|---|---|
| WO (1) | WO2019043600A1 (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032174A (en) * | 2019-04-26 | 2019-07-19 | 南京航空航天大学 | A kind of hierarchical fault diagnosis model and method |
| CN110888059A (en) * | 2019-12-03 | 2020-03-17 | 西安科技大学 | Algorithm based on improved random forest combined cubature Kalman power battery state of charge estimation |
| CN110991055A (en) * | 2019-12-08 | 2020-04-10 | 中国航空综合技术研究所 | Residual life prediction system for rotary mechanical products |
| CN112327810A (en) * | 2020-11-17 | 2021-02-05 | 山东科技大学 | Fault estimation method of dynamic event triggered transmission Markov jump system |
| CN112749453A (en) * | 2020-12-16 | 2021-05-04 | 安徽三禾一信息科技有限公司 | Complex equipment residual service life prediction based on improved SVR |
| CN113093712A (en) * | 2021-04-08 | 2021-07-09 | 重庆理工大学 | Active vehicle-mounted running state monitoring and fault forecasting system for automobile transmission system |
| CN113487086A (en) * | 2021-07-06 | 2021-10-08 | 新智数字科技有限公司 | Method and device for predicting remaining service life of equipment, computer equipment and medium |
| CN115774852A (en) * | 2022-11-17 | 2023-03-10 | 东方电气集团东方电机有限公司 | Classification model training method for water turbine cavitation recognition |
| CN115905820A (en) * | 2022-09-30 | 2023-04-04 | 杭州电力设备制造有限公司 | A Disturbance Identification Method for Power Quality in High Noise Environment |
| CN115908861A (en) * | 2022-11-07 | 2023-04-04 | 中国华能集团清洁能源技术研究院有限公司 | A bearing remaining service life prediction method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070198215A1 (en) * | 2006-02-22 | 2007-08-23 | Bonanni Pierino G | Method, system, and computer program product for performing prognosis and asset management services |
| US8442778B2 (en) * | 2009-11-16 | 2013-05-14 | Nrg Systems, Inc. | Data acquisition system for condition-based maintenance |
| US8781982B1 (en) * | 2011-09-23 | 2014-07-15 | Lockheed Martin Corporation | System and method for estimating remaining useful life |
-
2018
- 2018-08-29 WO PCT/IB2018/056597 patent/WO2019043600A1/en not_active Ceased
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070198215A1 (en) * | 2006-02-22 | 2007-08-23 | Bonanni Pierino G | Method, system, and computer program product for performing prognosis and asset management services |
| US8442778B2 (en) * | 2009-11-16 | 2013-05-14 | Nrg Systems, Inc. | Data acquisition system for condition-based maintenance |
| US8781982B1 (en) * | 2011-09-23 | 2014-07-15 | Lockheed Martin Corporation | System and method for estimating remaining useful life |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110032174B (en) * | 2019-04-26 | 2020-08-11 | 南京航空航天大学 | Layered fault diagnosis model and method |
| CN110032174A (en) * | 2019-04-26 | 2019-07-19 | 南京航空航天大学 | A kind of hierarchical fault diagnosis model and method |
| CN110888059A (en) * | 2019-12-03 | 2020-03-17 | 西安科技大学 | Algorithm based on improved random forest combined cubature Kalman power battery state of charge estimation |
| CN110888059B (en) * | 2019-12-03 | 2022-04-19 | 西安科技大学 | Charge state estimation algorithm based on improved random forest combined volume Kalman |
| CN110991055A (en) * | 2019-12-08 | 2020-04-10 | 中国航空综合技术研究所 | Residual life prediction system for rotary mechanical products |
| CN110991055B (en) * | 2019-12-08 | 2022-09-16 | 中国航空综合技术研究所 | Residual life prediction system for rotary mechanical products |
| CN112327810A (en) * | 2020-11-17 | 2021-02-05 | 山东科技大学 | Fault estimation method of dynamic event triggered transmission Markov jump system |
| CN112327810B (en) * | 2020-11-17 | 2021-07-30 | 山东科技大学 | A fault estimation method for dynamic event-triggered transmission Markov hopping systems |
| CN112749453A (en) * | 2020-12-16 | 2021-05-04 | 安徽三禾一信息科技有限公司 | Complex equipment residual service life prediction based on improved SVR |
| CN112749453B (en) * | 2020-12-16 | 2023-10-13 | 安徽三禾一信息科技有限公司 | Remaining service life prediction method of complex equipment based on improved SVR |
| CN113093712B (en) * | 2021-04-08 | 2023-08-15 | 重庆理工大学 | Active vehicle transmission system on-board operation status monitoring and fault prediction system |
| CN113093712A (en) * | 2021-04-08 | 2021-07-09 | 重庆理工大学 | Active vehicle-mounted running state monitoring and fault forecasting system for automobile transmission system |
| CN113487086A (en) * | 2021-07-06 | 2021-10-08 | 新智数字科技有限公司 | Method and device for predicting remaining service life of equipment, computer equipment and medium |
| CN113487086B (en) * | 2021-07-06 | 2024-04-26 | 新奥新智科技有限公司 | Method, device, computer equipment and medium for predicting residual service life of equipment |
| CN115905820A (en) * | 2022-09-30 | 2023-04-04 | 杭州电力设备制造有限公司 | A Disturbance Identification Method for Power Quality in High Noise Environment |
| CN115908861A (en) * | 2022-11-07 | 2023-04-04 | 中国华能集团清洁能源技术研究院有限公司 | A bearing remaining service life prediction method and device |
| CN115774852A (en) * | 2022-11-17 | 2023-03-10 | 东方电气集团东方电机有限公司 | Classification model training method for water turbine cavitation recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2019043600A1 (en) | Remaining useful life estimator | |
| CN118794690B (en) | Bearing fault diagnosis method for multi-source wind turbine | |
| JP7340265B2 (en) | Abnormality detection device, abnormality detection method, and program | |
| Ruiz-Cárcel et al. | Combination of process and vibration data for improved condition monitoring of industrial systems working under variable operating conditions | |
| CN113834657A (en) | A Bearing Fault Early Warning and Diagnosis Method Based on Improved MSET and Spectral Characteristics | |
| CN120030802B (en) | Data prediction method and system for thermal power plant equipment | |
| Yu | A nonlinear probabilistic method and contribution analysis for machine condition monitoring | |
| WO2020090767A1 (en) | Abnormality diagnostic device, abnormality diagnostic method, and program | |
| CN119780587A (en) | A Fault Diagnosis Method and System for Photovoltaic Inverter | |
| CN107851294A (en) | Condition-based preventive maintenance device and method for large-scale operation system | |
| CN115392109A (en) | LSTM multivariable time series anomaly detection method based on generative model | |
| CN107782551A (en) | Method for evaluating damage degree of mechanical part | |
| CN119150154A (en) | Construction method, diagnosis method, system and equipment of bolt fault diagnosis model | |
| Egaji et al. | A data mining based approach for electric motor anomaly detection applied on vibration data | |
| Atta et al. | Detection and diagnosis of bearing faults under fixed and time-varying speed conditions using persistence spectrum and multi-scale structural similarity index | |
| CN118656586A (en) | Vulnerability analysis and toughness assessment method and system for substation structures under strong earthquakes | |
| CN117806233A (en) | Intelligent monitoring system and method for numerical control machine tool | |
| Qian et al. | Bearing performance degradation evaluation using recurrence quantification analysis and auto-regression model | |
| Hamadache et al. | Improving signal-to-noise ratio (SNR) for inchoate fault detection based on principal component analysis (PCA) | |
| CN118968733A (en) | A monitoring and early warning system based on array optical fiber sensing technology | |
| CN118857737A (en) | A sliding bearing fault diagnosis system and method | |
| Liu et al. | Gearbox failure diagnosis based on vector autoregressive modelling of vibration data and dynamic principal component analysis | |
| Oddan | Multivariate statistical condition monitoring | |
| Astolfi et al. | Wind turbine drive-train condition monitoring through tower vibrations measurement and processing | |
| Randrianandraina et al. | Wind turbine generator bearing anomaly detection and explanation using rrcf approach |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18851790 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 18851790 Country of ref document: EP Kind code of ref document: A1 |