JP4046309B2 - Plant monitoring device - Google Patents

Plant monitoring device Download PDF

Info

Publication number
JP4046309B2
JP4046309B2 JP6613099A JP6613099A JP4046309B2 JP 4046309 B2 JP4046309 B2 JP 4046309B2 JP 6613099 A JP6613099 A JP 6613099A JP 6613099 A JP6613099 A JP 6613099A JP 4046309 B2 JP4046309 B2 JP 4046309B2
Authority
JP
Japan
Prior art keywords
monitoring
plant
data
plant data
reference
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.)
Expired - Fee Related
Application number
JP6613099A
Other languages
Japanese (ja)
Other versions
JP2000259223A (en
Inventor
正剛 佐久間
茂 兼本
雅弘 大川
光広 榎本
優 藤波
Original Assignee
株式会社東芝
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 株式会社東芝 filed Critical 株式会社東芝
Priority to JP6613099A priority Critical patent/JP4046309B2/en
Publication of JP2000259223A publication Critical patent/JP2000259223A/en
Application granted granted Critical
Publication of JP4046309B2 publication Critical patent/JP4046309B2/en
Anticipated expiration legal-status Critical
Application status is Expired - Fee Related legal-status Critical

Links

Images

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a plant monitoring apparatus that monitors an operating state based on a process signal of a large-scale plant such as a nuclear power plant, a thermal power plant, or a chemical plant, and detects and diagnoses the occurrence of an abnormality at an early stage.
[0002]
[Prior art]
In a large-scale plant such as a nuclear power plant, many process signals are measured for the purpose of monitoring the performance of the plant and the soundness of various systems and devices constituting the plant. Since it is difficult for a plant operator to constantly monitor all of a large amount of signals, many plants are provided with a monitoring system that captures process signals using a computer and detects abnormal changes in the plant.
[0003]
These monitoring systems often use threshold judgment that is regarded as abnormal when a signal to be monitored exceeds a certain threshold. Further, when the monitoring target signal is y, when this is predicted from y itself or other process quantities x1, x2,... Including itself using a known characteristic function f (x1, x2,...) A method of detecting an abnormality by using a prediction error, that is, δy = y−f (x1, x2,...) As a monitoring index and comparing them with an abnormality detection threshold is also used.
[0004]
On the other hand, these determination results need to be displayed to inform the operator, but due to the restriction that they are displayed as a two-dimensional graph on the display device, the monitoring target signal y and its predicted value f (x1, x2,...) The prediction error δy is compared and shown as a function of time, or one of the signals used for input xi is plotted on the horizontal axis and the monitored signal y is plotted on the vertical axis.
[0005]
[Problems to be solved by the invention]
The following problems have been pointed out in the above-described conventional plant monitoring apparatus. That is, when a physical model such as a mass energy balance model is used as the characteristic function used for predicting the monitoring target signal, the prediction accuracy is insufficient and a desirable result cannot be obtained. Also, a characteristic function obtained by fitting from normal data with a linear or nonlinear regression model may be used. However, when a signal used as an input is inappropriate, the prediction accuracy is insufficient, which is not preferable. In particular, if the process quantity varies in a similar manner throughout the plant, it becomes difficult to select which signal is used for input.
[0006]
Furthermore, when the fluctuation range of the input signal of the characteristic function used for prediction is expanded to an unexpected level, the prediction performance may be degraded. If the abnormality determination is performed while ignoring the fluctuation range of the input signal, an erroneous determination cannot be avoided.
[0007]
On the other hand, since the relationship between the signal used for input and the monitoring target signal is conventionally viewed in the range of a two-dimensional plot, when the monitoring target signal changes depending on a plurality of input signals, a single input Unpredictable behavior can occur when plotting with signals.
[0008]
In addition, since disturbances are generated even in a normal state due to instrument calibration and surveillance in the plant, this is regarded as abnormal, and erroneous determination occurs.
[0009]
It is an object of the present invention to provide a plant monitoring apparatus that can perform abnormality monitoring based on a characteristic function for favorable prediction, and can appropriately display information for monitoring.
[0010]
[Means for Solving the Problems]
The invention according to claim 1 is a plant data input device that inputs a plurality of types of plant data from a plant, a data storage device that stores plant data provided from the plant data input device, and a data storage device that stores the data. A reference creation device for creating a monitoring reference based on normal plant data, and the plant data stored in the data storage device are read, and normality / abnormality determination of the determination target plant data is performed using other plant data. Input and perform using the difference between the predicted value predicted from the linear or nonlinear regression model created from normal plant data and the plant data to be judged, and a small number of principal component signals from the given plant data And extract two of the principal component signals to both axes on the plane as a three-dimensional graph. Perform a regression analysis of the determination target plant data by Shimesuru further accordance monitoring criteria given from said reference generating apparatus, the plant data of said monitored if the plant data of said monitored meet the monitoring criteria If all the other monitoring target plant data satisfies the monitoring criteria corresponding to the plant data, only the means for acquiring the monitoring target plant data is used. A monitoring processing device that determines that the plant is abnormal in other cases, and a display device that displays the plant data and the monitoring reference given to the monitoring processing device together with the determination result. Is.
[0011]
In the plant monitoring device, a monitoring reference is generated from normal plant data in the reference generation device, and the monitoring target plant data given from the data storage device and the obtained monitoring reference are compared in the monitoring processing device. As a result of the comparison, for example, if a threshold value is exceeded, it is determined as abnormal, and the determination result is displayed on the display device together with the plant data to be monitored and the monitoring reference. In this way, various abnormalities that occur in the system, equipment, etc. in the plant can be detected early.
[0012]
For this monitoring standard, for example, a linear or non-linear regression model created from normal plant data is used, and the prediction error is compared with a threshold value. By this method, it is possible to increase the accuracy in abnormality monitoring, and it is possible to perform monitoring with few erroneous determinations.
[0017]
In the invention according to claim 2 , as a function of two signals selected from a plurality of input signals used for determination by the display device as a determination target signal, a color is displayed on the three-dimensional display graph according to whether it is normal or not. They are displayed differently and are displayed with the curved surface of a linear or non-linear regression model used for determination as a reference.
[0018]
In the plant monitoring apparatus, two signals are selected from the input signals, and these are taken on both axes on the plane, and the behavior of the monitoring target signal can be appropriately displayed by a three-dimensional graph having the prediction target signal on the vertical axis. . By displaying the predicted value of the regression model used for determination in this three-dimensional graph as a curved surface graph and further displaying the observed signal to be determined as a point in the graph, it is possible to intuitively determine whether the plant behavior is normal or abnormal. Judgment can be made. In addition, by displaying the points that have become abnormal in different colors, it is possible to easily determine the presence or absence of an abnormality in the three-dimensional graph.
[0019]
Furthermore, the invention according to claim 3 displays the fluctuation range of the input signal used by the display device for creating the reference regression model on a two-dimensional graph, and the input signal used for the input at the time of determination is the fluctuation range of the reference regression model. Are displayed on the same two-dimensional graph with different colors depending on whether they are included in a certain error range.
[0020]
In the plant monitoring device, the fluctuation range of the input signal is displayed on the two-dimensional graph in order to determine that there is no error in the determination according to the reference regression model obtained based on the normal plant data. As a result, it is possible to prevent erroneous determination.
[0021]
Further, the invention according to claim 4 is the normal / abnormal determination result only when the input signal at the time of determination deviates from the fluctuation range of the input signal used by the monitoring processor for creating the reference regression model within a certain error range. It is something that will be forgotten.
[0022]
In the above plant monitoring apparatus, only when the input signal deviates from the behavior range of the input signal used for creating the reference regression model within a certain error range, the determination itself is forgotten. This makes it possible to prevent erroneous determination from occurring.
[0023]
Further, the invention according to claim 5 is the normal / abnormal judgment result only when the input signal at the judgment time deviates from the fluctuation range of the input signal used by the monitoring processor for creating the reference regression model within a certain error range. The reference regression model is learned again using the plant data.
[0024]
In the above-mentioned plant monitoring apparatus, at the same time as judging, the plant data exceeding the behavior range of the input signal used for learning of the reference regression model is learned again using the plant data. This makes it possible to expand the application range of the standard regression model step by step.
[0025]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows an overall configuration of an embodiment of a plant monitoring apparatus of the present invention. A plant data input apparatus 1 for inputting a plurality of plant data synchronized in time from a plant, and the plant data input apparatus 1 A data storage device 2 for storing the given plant data, a reference creation device 3 for creating a monitoring reference based on normal plant data, and the plant data stored in the data storage device 2 are read and the reference creation device 3 The monitoring processing device 4 that determines normality / abnormality according to the monitoring standard given from the above, and the display device 5 that displays the plant data and the monitoring standard on a display device such as a CRT.
[0026]
In the above configuration, the signal obtained from the plant is converted into time-series data by the plant data input device 1, transmitted to the data storage device 2, and stored there. The data storage device 2 continuously stores a certain period of plant data continuously. Here, a plurality of plant data transmitted from the plant data input device 1 are continuously stored cyclically at a plurality of different sampling intervals in the form as shown in FIG. It is created so that data up to the point is saved.
[0027]
In addition, the reference creation device 3 identifies linear or nonlinear regression models using data at arbitrary sampling intervals stored in the data storage device 2, and as shown in FIG. The obtained standard regression coefficient, prediction error at the time of identification, determination criterion at the time of abnormality determination, and additional plant data at the time of learning are stored. The prediction error at the time of identification is used in the abnormality determination processing device together with a determination criterion for abnormality determination that is input in advance.
[0028]
Further, the monitoring processing device 4 performs an abnormality determination process using data at an arbitrary sampling interval of the data storage device 2 and the monitoring reference given from the reference creating device 3. The display device 5 displays an abnormality determination result, a reference regression model, and the like.
[0029]
Hereinafter, embodiments of the individual apparatuses will be described in detail. The reference creation device 3 creates a prediction model of the monitoring target signal from the normal plant data. The monitoring target signal is an individual signal taken into the apparatus. Assuming that a specific monitoring target signal is y (t), this is obtained from m input signals x1 (t)... Xm (t),
y (t) = f (x1 (t),... xm (t)) (1)
Predict in the form of For linear regression models, omit t and
y = a + b1 * x1 + …… + bm * xm (2)
It becomes. In the case of a quadratic equation:
It becomes the form.
[0030]
As shown in FIG. 4, the monitoring target signal can be predicted from m inputs using a linear, nonlinear, neuro-network, or the like. The coefficient here is based on normal plant data.
[Expression 1]
That is, it can be determined to minimize the mean square error. Which model is used is determined so as to minimize the weighted average of the prediction error I (a, b1,..., Bm) and the complexity (number of coefficients) of each model.
[0031]
As another method of this prediction method, principal component analysis as shown in FIG. 5 can also be used. Here, 2 to 3 small numbers of signals (Ui, i = 1, 2,...) Are extracted from m input signals (Xi, i = 1, 2,... M) by principal component analysis. Using this extracted time series signal (U), the monitoring target signal T is changed to Ta = f (U) (5)
Predict with. Here, (Ti, i = 1, 2,...) Is a small number of principal component signals obtained by principal component analysis of the monitoring target signal (yi, i = 1, 2,... K). When there is one monitoring target signal, T = Y in this embodiment. As the prediction function f (U), any of the linear, non-linear, and neural networks described above can be applied. The abnormality determination is performed based on whether or not the deviation between the observed principal component signal T and its predicted value Ta exceeds a certain threshold value. Such extraction of principal component signals is particularly useful for monitoring large-scale plants for the following reasons.
[0032]
In large-scale plants, the monitoring signal of interest is affected by a number of different process signals and fluctuates. However, it is not independently influenced by all process signals and is usually affected by a small number of state variables. For example, in a nuclear power plant, normal operation is controlled by output and flow rate, and many observation signals behave in a complex manner, but when the principal components are extracted, they are represented by a small number of state variables corresponding to the output and flow rate. be able to. By predicting the plant state with such a small number of signals, it is possible to create a reference regression model that can be universally applied in various operation cycles.
[0033]
Furthermore, if the plant follows different output change curves in different operating cycles, the observation itself will behave differently between the cycles, but by extracting and evaluating this principal component, the state between the cycles will be Can be determined to be equivalent.
[0034]
Furthermore, this embodiment can also monitor a plurality of monitoring target signals as a group at the same time. As shown in FIG. 5, the principal component analysis is applied to k monitoring target signals to generate a small number of monitoring target signals T. The monitoring target signal T can be monitored by predicting the least squares with the prediction function f (U).
[0035]
In creating the above reference model, it is assumed that the signal used is standardized in advance to an average of zero and a fluctuation range of −1 to 1 to ensure stability in numerical calculation.
[0036]
The reference creation device 3 stores the coefficient of the above-described reference regression model, the reference data used for identification, the magnitude of the prediction error at the time of identification (standard deviation σ), and the normalization constant at the time of identification as a database for monitoring processing. Provided to apparatus 4.
[0037]
Further, the monitoring processing device 4 calculates a prediction error δy by the following equation for the determination target signal y (t) at the monitoring time given from the data storage device 2 and m input signals xm (t), Compare with threshold.
[0038]
δy = y−f (x1,... xm) (6)
As this threshold value, k times the error σ at the time of creating the reference model or a fixed value given in advance is used.
[0039]
In the present embodiment, it is also possible to make an abnormality determination using a hypothesis test method such as a sequential probability ratio test using the fact that the prediction error δy is normally zero on average and given by the variance σ.
[0040]
Since the prediction error of the regression model is used in the determination by this apparatus, it is important to avoid the erroneous determination by checking the application range of the model. In order to confirm the application range of the model, as shown in FIG. 6, the fluctuation range of the input data used for identification of the reference model is displayed with a certain tolerance, and the input data at the time of monitoring falls within that range. Whether it is included or not is determined. If it is not within the allowable range, the monitoring processor 4 defers obtaining a determination result.
[0041]
Moreover, it replaces with the said method, and when it is not contained in an allowable range, this plant data can be added to reference | standard data, and a reference | standard regression model can be learned again. This makes it possible to expand the monitoring range as the monitoring time increases.
[0042]
Next, details of the display device 5 will be described. In the conventional display device, the time series trend of the monitoring target signal is shown in comparison with the predicted value, or the time trend of the prediction error is displayed. These display methods are effective in determining at what point the abnormality occurred, but it is impossible to know what input value this predicted value was predicted from, so the operator's intuitive understanding Disturb.
[0043]
Two-dimensional correlation display is also performed with one of the input signals on the horizontal axis and the monitored signal on the vertical axis. This indicates the behavior of the monitored signal affected by multiple input signals. Often not enough to understand.
[0044]
On the other hand, in a large-scale plant, a relatively small number of principal component signals can be obtained by taking the principal component of the input signal by taking advantage of the fact that the behavior of the monitored signal is not influenced by all of the input signals that can be affected. From this, the behavior of the monitoring target signal can be predicted. By displaying two of these principal component signals on both axes on the plane and displaying the behavior of the monitoring signal as a three-dimensional graph with the vertical axis, the correlation between input and output can be grasped intuitively. .
[0045]
In the present embodiment, as shown in FIG. 7, the monitoring target data is displayed as dots along with the reference curve calculated from the reference regression model on a three-dimensional graph. Further, since it is difficult to intuitively grasp the deviation from the reference phase on the three-dimensional screen, abnormal data in the determination, that is, points that are more than a certain distance from the reference curved surface are displayed with different colors. It is also important to determine whether or not the data at the time of monitoring is included in the allowable error range of the input data at the time of creating the reference regression model. This is shown on a two-dimensional graph as shown in FIG. And whether it is included in the allowable range or not can be determined with different colors.
[0046]
Furthermore, FIG. 8 shows the plant behavior in different operating cycles on a three-dimensional graph. Different paths are taken on the three-dimensional graph, and as shown in FIG. 9, there is a difference just by looking at the time change of the monitoring signal, but as shown in FIG. 8, the same reference curved surface is taken on the three-dimensional graph. It can be determined as normal.
[0047]
On the other hand, FIG. 10 as inputs all the monitoring signals of the plant, the abnormality detection by the deviation between the observed value by predicting the individual monitoring signal y by the linear and non-linear regression model after performing principal component analysis An example is shown. In the case of a sensor failure that often occurs in a plant, the output deviation is increased by the corresponding signal, and the sensor failure can be identified and detected. Here, the same signal is used as an input when predicting the monitoring target signal y, and the number of inputs and outputs is the same, and the extracted principal components are common.
[0048]
A method different from the above method is shown in FIG. Although it is an example of the same sensor failure diagnosis, in order to improve the drawback of using the same signal as the output for the input, the same input signal corresponding to each monitoring target signal is omitted, and the principal component analysis is performed only with the remaining signals, Create a regression model and make a prediction. Thereby, when a sensor failure occurs, the signal is also used as an input signal, so that the disadvantage of the method shown in FIG.
[0049]
【The invention's effect】
In the present invention, long-term behavior from start to stop of a plant can be monitored with higher accuracy than a reference model obtained from a linear or non-linear regression model in which plant data is created by a reference creation device.
[0050]
Therefore, according to the present invention, various abnormalities that occur in systems, equipment, etc. in a large-scale plant can be detected at an early stage, and the loss of the soundness of a large-scale plant in operation can be prevented. it can.
[Brief description of the drawings]
FIG. 1 is a configuration diagram showing an embodiment of a plant monitoring apparatus according to the present invention.
FIG. 2 is a diagram showing a data storage form of the data storage device of the present invention.
FIG. 3 is a diagram showing a data storage form of the reference creation device of the present invention.
FIG. 4 is a diagram showing a multivariate regression model based on a linear, non-linear, and neuro network for prediction of a monitoring signal according to the present invention.
FIG. 5 is a diagram showing a regression prediction model of a principal component of an input signal and a principal component of a monitoring signal according to the present invention.
FIG. 6 is a diagram showing a fluctuation range of an input signal used for creating a model of the present invention and a fluctuation range of an input signal at the time of monitoring.
FIG. 7 is a diagram showing a display example of a monitoring result by a three-dimensional graph in which the input signal of the present invention is taken on two horizontal axes and the monitoring signal is taken on a vertical axis.
FIG. 8 is a diagram showing a display example in which operation histories of different operation cycles of the present invention are compared on a three-dimensional graph.
FIG. 9 is a diagram showing a display example in which operation histories of different operation cycles of the present invention are compared with time trends.
FIG. 10 is a diagram showing a configuration example of sensor diagnosis for predicting m signals of the present invention using a regression model of m input signals.
FIG. 11 is a diagram illustrating a configuration example of sensor diagnosis in which principal component analysis is performed for each prediction target signal according to the present invention.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Plant data input device 2 Data storage device 3 Standard | standard preparation apparatus 4 Monitoring processing apparatus 5 Display apparatus

Claims (5)

  1. A plant data input device that inputs multiple types of plant data from the plant, a data storage device that stores plant data given from this plant data input device, and normal plant data that is stored in this data storage device A reference creation device that creates a monitoring reference based on the data, and reads the plant data stored in the data storage device, inputs normality / abnormality of the determination target plant data, and inputs other plant data, and the normal plant The difference between the predicted value predicted from the linear or nonlinear regression model created from the data and the target plant data is extracted, and a small number of principal component signals are extracted from the given multiple plant data. Two of the signals are taken on both axes on the plane and displayed as a three-dimensional graph. Perform a regression analysis of the determination target plant data, further, according to the monitoring criteria given from said reference generating apparatus, when the plant data of said monitored meet the monitoring criteria is determined to be normal plant data of said monitored In the case where the monitoring standard is not satisfied and all other monitoring target plant data satisfies the monitoring standard corresponding to the plant data, only the means for acquiring the monitoring target plant data is determined to be abnormal, Otherwise, a plant monitoring device comprising a monitoring processor that determines that the plant is abnormal, and a display device that displays the plant data and monitoring criteria given to the monitoring processor together with the determination result .
  2. The display device displays a determination target signal as a function of two signals selected from a plurality of input signals used for determination, and displays a different color on the three-dimensional display graph depending on whether the signal is normal or not. plant monitoring apparatus according to claim 1, characterized in that so as to display, based on the curved surface of the linear or non-linear regression model used for.
  3. The display device displays the fluctuation range of the input signal used to create the reference regression model on a two-dimensional graph, and the input signal used for input at the time of determination is included in the fluctuation range of the reference regression model within a certain error range. or, by different colors depending on whether, plant monitoring apparatus according to claim 2, characterized in that so as to display the same two-dimensional graph.
  4. Only when the input signal at the time of determination deviates from the fluctuation range of the input signal used for creating the reference regression model by the monitoring processor, the normal / abnormal determination result is forgotten. The plant monitoring apparatus according to claim 1 .
  5. Only when the monitoring processing device deviates from the fluctuation range of the input signal used for creating the reference regression model and the input signal at the time of the judgment deviates within a certain error range, the monitoring processing apparatus sees to obtain the normal / abnormal judgment result, and further 5. The plant monitoring apparatus according to claim 4, wherein the reference regression model is learned again using the plant data.
JP6613099A 1999-03-12 1999-03-12 Plant monitoring device Expired - Fee Related JP4046309B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP6613099A JP4046309B2 (en) 1999-03-12 1999-03-12 Plant monitoring device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP6613099A JP4046309B2 (en) 1999-03-12 1999-03-12 Plant monitoring device

Publications (2)

Publication Number Publication Date
JP2000259223A JP2000259223A (en) 2000-09-22
JP4046309B2 true JP4046309B2 (en) 2008-02-13

Family

ID=13306989

Family Applications (1)

Application Number Title Priority Date Filing Date
JP6613099A Expired - Fee Related JP4046309B2 (en) 1999-03-12 1999-03-12 Plant monitoring device

Country Status (1)

Country Link
JP (1) JP4046309B2 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018230645A1 (en) * 2017-06-14 2018-12-20 株式会社東芝 Anomaly detection device, anomaly detection method, and program

Families Citing this family (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6735541B2 (en) * 2001-02-16 2004-05-11 Exxonmobil Research And Engineering Company Process unit monitoring program
US20020183971A1 (en) * 2001-04-10 2002-12-05 Wegerich Stephan W. Diagnostic systems and methods for predictive condition monitoring
GB0307406D0 (en) * 2003-03-31 2003-05-07 British Telecomm Data analysis system and method
ITBO20030711A1 (en) * 2003-11-24 2005-05-25 Gd Spa Method for the predictive maintenance of a component
JP4516306B2 (en) * 2003-11-28 2010-08-04 株式会社日立製作所 How to collect storage network performance information
CN101535909B (en) * 2006-09-28 2012-08-29 费舍-柔斯芒特系统股份有限公司 Abnormal situation prevention in a heat exchanger
JP5067542B2 (en) * 2007-04-27 2012-11-07 オムロン株式会社 Composite information processing apparatus, composite information processing method, program, and recording medium
JP4922265B2 (en) * 2008-08-19 2012-04-25 株式会社東芝 Plant monitoring apparatus and plant monitoring method
JP5363927B2 (en) * 2009-09-07 2013-12-11 株式会社日立製作所 Abnormality detection / diagnosis method, abnormality detection / diagnosis system, and abnormality detection / diagnosis program
JP2011145846A (en) * 2010-01-14 2011-07-28 Hitachi Ltd Anomaly detection method, anomaly detection system and anomaly detection program
JP2011174766A (en) * 2010-02-23 2011-09-08 Chugoku Electric Power Co Inc:The Apparatus deterioration evaluation support method and apparatus deterioration evaluation support device
JP5439265B2 (en) * 2010-04-20 2014-03-12 株式会社日立製作所 Abnormality detection / diagnosis method, abnormality detection / diagnosis system, and abnormality detection / diagnosis program
JP5988419B2 (en) 2012-01-11 2016-09-07 インターナショナル・ビジネス・マシーンズ・コーポレーションInternational Business Machines Corporation Prediction method, prediction system, and program
CA2815161A1 (en) * 2013-05-06 2014-11-06 Hydro-Quebec Quantitative analysis of signal related measurements for trending and pattern recognition
KR101468142B1 (en) * 2013-08-29 2014-12-05 비앤에프테크놀로지 주식회사 Prediction method of plant health status, and a computer-readable storage medium having program to perform the same
FR3019592B1 (en) * 2014-04-03 2016-04-22 Snecma Method and device for monitoring a parameter of a rotor motor
JP6378419B2 (en) * 2015-03-12 2018-08-22 株式会社日立製作所 Machine diagnostic device and machine diagnostic method
JP6150934B1 (en) * 2016-10-17 2017-06-21 三菱重工業株式会社 Information processing method, information processing apparatus, program, and information processing system

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018230645A1 (en) * 2017-06-14 2018-12-20 株式会社東芝 Anomaly detection device, anomaly detection method, and program

Also Published As

Publication number Publication date
JP2000259223A (en) 2000-09-22

Similar Documents

Publication Publication Date Title
EP0855061B1 (en) An expert system for testing industrial processes and determining sensor status
US7853339B2 (en) Statistical signatures used with multivariate analysis for steady-state detection in a process
Jones-Farmer et al. An overview of phase I analysis for process improvement and monitoring
JP2008198213A (en) Sensor validation apparatus and method
KR101316486B1 (en) Error detection method and system
AU748987B2 (en) Ultrasensitive surveillance of sensors and processes
JP5284503B2 (en) Diagnostic system and method for predictive condition monitoring
JP5364265B2 (en) Apparatus and method for abnormal event detection in continuous industrial process operation
Orchard A particle filtering-based framework for on-line fault diagnosis and failure prognosis
US7373283B2 (en) Monitoring and fault detection system and method using improved empirical model for range extrema
Russell et al. Data-driven methods for fault detection and diagnosis in chemical processes
EP0906593B1 (en) Industrial process surveillance system
Russell et al. Fault detection in industrial processes using canonical variate analysis and dynamic principal component analysis
JP2009536971A (en) Application of abnormal event detection (AED) technology to polymer process
Chen et al. Dynamic process fault monitoring based on neural network and PCA
DE102011102034A1 (en) Online adjustment of a process-analytical model with effective process operation
JP4071449B2 (en) Sensor abnormality detection method and sensor abnormality detection device
Kruger et al. Improved principal component monitoring of large-scale processes
Kruger et al. Statistical monitoring of complex multivatiate processes: with applications in industrial process control
Khoo et al. A synthetic double sampling control chart for the process mean
JP4394286B2 (en) Multidimensional method and system for statistical process management
Norvilas et al. Intelligent process monitoring by interfacing knowledge-based systems and multivariate statistical monitoring
Shams et al. Fault detection, identification and diagnosis using CUSUM based PCA
JP4626852B2 (en) Communication network failure detection system, communication network failure detection method, and failure detection program
JP4289602B2 (en) Process monitoring method

Legal Events

Date Code Title Description
A621 Written request for application examination

Free format text: JAPANESE INTERMEDIATE CODE: A621

Effective date: 20040917

RD02 Notification of acceptance of power of attorney

Free format text: JAPANESE INTERMEDIATE CODE: A7422

Effective date: 20050315

RD04 Notification of resignation of power of attorney

Free format text: JAPANESE INTERMEDIATE CODE: A7424

Effective date: 20050324

A977 Report on retrieval

Free format text: JAPANESE INTERMEDIATE CODE: A971007

Effective date: 20061206

A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20061212

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20070209

A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 20070821

A521 Written amendment

Free format text: JAPANESE INTERMEDIATE CODE: A523

Effective date: 20071010

A911 Transfer of reconsideration by examiner before appeal (zenchi)

Free format text: JAPANESE INTERMEDIATE CODE: A911

Effective date: 20071029

TRDD Decision of grant or rejection written
A01 Written decision to grant a patent or to grant a registration (utility model)

Free format text: JAPANESE INTERMEDIATE CODE: A01

Effective date: 20071116

A61 First payment of annual fees (during grant procedure)

Free format text: JAPANESE INTERMEDIATE CODE: A61

Effective date: 20071119

FPAY Renewal fee payment (prs date is renewal date of database)

Free format text: PAYMENT UNTIL: 20101130

Year of fee payment: 3

LAPS Cancellation because of no payment of annual fees