CN112685216A - Equipment abnormity monitoring system and method based on trend analysis - Google Patents
Equipment abnormity monitoring system and method based on trend analysis Download PDFInfo
- Publication number
- CN112685216A CN112685216A CN202110151207.9A CN202110151207A CN112685216A CN 112685216 A CN112685216 A CN 112685216A CN 202110151207 A CN202110151207 A CN 202110151207A CN 112685216 A CN112685216 A CN 112685216A
- Authority
- CN
- China
- Prior art keywords
- preset
- trend
- measuring point
- measuring points
- preset measuring
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 52
- 238000012544 monitoring process Methods 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000000605 extraction Methods 0.000 claims abstract description 41
- 230000002159 abnormal effect Effects 0.000 claims abstract description 31
- 230000005856 abnormality Effects 0.000 claims abstract description 23
- 238000001514 detection method Methods 0.000 claims abstract description 20
- 238000004364 calculation method Methods 0.000 claims description 14
- 230000001052 transient effect Effects 0.000 claims description 10
- 238000010835 comparative analysis Methods 0.000 claims description 5
- 230000035772 mutation Effects 0.000 claims description 5
- 238000012795 verification Methods 0.000 claims description 3
- 238000005259 measurement Methods 0.000 description 25
- 238000007405 data analysis Methods 0.000 description 8
- 238000010219 correlation analysis Methods 0.000 description 7
- 230000007246 mechanism Effects 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000036541 health Effects 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- 230000001629 suppression Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000005192 partition Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000001174 ascending effect Effects 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 239000000498 cooling water Substances 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000002427 irreversible effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012806 monitoring device Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
The embodiment of the invention provides a system and a method for monitoring equipment abnormity based on trend analysis, wherein the system comprises: the system comprises a data acquisition unit, a feature extraction unit, a trend identification unit, a comparison analysis unit and an abnormality monitoring unit, wherein the data acquisition unit is used for acquiring real-time operation data of preset measuring points; the characteristic extraction unit is used for carrying out characteristic matching on the real-time operation data according to the type of the preset measuring point to obtain corresponding signal characteristics; the trend identification unit is used for carrying out trend identification on the signal characteristics according to a preset trend identification algorithm to obtain primitive types; the comparison and analysis unit is used for carrying out self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points; and the abnormity detection unit is used for judging whether the preset measuring point is abnormal or not according to the comparison result. By adopting the method, the trend of a large number of measuring points can be contrasted and analyzed, and customers can conveniently and visually and rapidly position abnormal equipment and investigate abnormal reasons.
Description
Technical Field
The invention relates to the technical field of equipment abnormity monitoring, in particular to equipment abnormity monitoring system and method based on trend analysis.
Background
On various industrial sites, there are a large number of critical pieces of equipment whose proper and reliable operation is critical to the operation and economic efficiency of the entire plant. If the equipment fails, huge economic losses are brought to the whole industrial operation, so that whether the equipment is abnormal or not needs to be monitored in real time. Such as: and monitoring the abnormity of each key device in steady operation of the nuclear power plant.
In order to monitor the state of each key device in the industrial field, a large number of sensors and a matched data acquisition system are usually laid on the site to monitor the acquired data in real time. The conventional monitoring system has the defects that monitoring is only carried out through a measuring point threshold, namely, if the measuring point exceeds the threshold, a corresponding alarm is generated, an engineer responds, positions a system, equipment and a part to which the alarm measuring point belongs, looks up the historical trend of the measuring point before and after the alarm time, and carries out some necessary analysis work on the basis. Firstly, the threshold setting requirement on the relevant threshold is high, the threshold is too narrow and frequently gives alarms against the actual operation rule of the equipment to cause fatigue response of field workers, and when the threshold is too wide and alarms, irreversible damage may occur to the equipment and irreparable influence is caused to the production of a factory. The optimal mode is to set a wider fixed threshold and pay close attention to the trend change condition of each measuring point on the basis of the wider fixed threshold. Therefore, engineers are required to frequently call out historical trends of key measuring points to analyze the trends of the key measuring points and analyze and judge possible reasons according to an online inspection mode, but a plurality of core devices are arranged in a common factory, the number of related measuring points related to each device is often dozens, and the workload of related personnel is obviously increased by a method that a large number of measuring points are only manually inspected online.
In view of the above problems, there is a need for an equipment abnormality monitoring system that solves the above problems.
Disclosure of Invention
Aiming at the problems in the prior art, the embodiment of the invention provides a system and a method for monitoring equipment abnormity based on trend analysis.
The embodiment of the invention provides a device abnormity monitoring system based on trend analysis, which comprises: a data acquisition unit, a feature extraction unit, a trend identification unit, a comparison and analysis unit and an abnormality monitoring unit, wherein,
the data acquisition unit is used for acquiring real-time operation data of preset measuring points;
the characteristic extraction unit is used for carrying out characteristic matching on corresponding real-time operation data according to the type of a preset measuring point to obtain signal characteristics corresponding to the preset measuring point;
the trend recognition unit is used for performing trend recognition on the signal characteristics according to a preset trend recognition algorithm to obtain a primitive type corresponding to the preset measuring point;
the comparison analysis unit is used for carrying out self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points;
and the abnormity detection unit is used for judging whether the preset measuring point is abnormal or not according to the self-comparison result and the mutual comparison result.
The embodiment of the invention provides a method for monitoring equipment abnormity based on trend analysis, which comprises the following steps:
the data acquisition unit acquires real-time operation data of preset measuring points and sends the real-time operation data to the feature extraction unit;
the characteristic extraction unit is used for carrying out characteristic matching on corresponding real-time running data according to the type of a preset measuring point to obtain signal characteristics corresponding to the preset measuring point and sending the signal characteristics to the trend identification unit;
the trend recognition unit carries out trend recognition on the signal characteristics according to a preset trend recognition algorithm to obtain a primitive type corresponding to the preset measuring point, and sends the primitive type to the comparison analysis unit;
the comparison analysis unit carries out self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points, and sends the self-comparison results and the mutual comparison results to the abnormality detection unit;
and the abnormity detection unit judges whether the preset measuring point is abnormal or not according to the self-comparison result and the mutual comparison result.
In one embodiment, the method further comprises:
the feature extraction unit is used for matching a corresponding feature extraction type according to the type of a preset measuring point and extracting the features of the real-time operation data according to the feature extraction type to obtain the signal features corresponding to the preset measuring point.
In one embodiment, the feature extraction type includes:
discrete feature extraction, smooth feature extraction and fluctuation feature extraction.
In one embodiment, the method further comprises:
performing trend hypothesis on the preset measuring points to obtain confidence threshold values;
and performing statistical calculation according to the signal characteristics, comparing the statistical calculation result with the confidence coefficient threshold value, and determining the primitive type corresponding to the preset measuring point according to the comparison result.
In one embodiment, the method further comprises:
and the trend identification unit carries out mutation verification on the signal characteristics and identifies the transient type of the preset measuring point, wherein the primitive type corresponding to the preset measuring point comprises the transient type.
In one embodiment, the method further comprises:
and acquiring associated measuring points related to the attributes of the preset measuring points in the detection equipment, acquiring primitive types of the associated measuring points, comparing the primitive types of the preset measuring points with the primitive types of the associated measuring points, and obtaining a self-comparison result of the preset measuring points according to a comparison result.
In one embodiment, the method further comprises:
acquiring a preset physical association dimension, and acquiring an association measuring point physically associated with the preset measuring point through the physical association dimension;
or acquiring normal data of a preset measuring point and a measuring point to be determined, calculating the correlation between the normal data of the preset measuring point and the normal data of the measuring point to be determined, and determining a related measuring point in the measuring point to be determined according to the correlation calculation result.
In one embodiment, the method further comprises:
acquiring associated equipment with the same type as the detection equipment, acquiring a second associated measuring point at the same position as the preset measuring point in the associated equipment, and acquiring the primitive type of the second associated measuring point;
and comparing the primitive types of the preset measuring points with the primitive types of the second associated measuring points, and obtaining the mutual comparison result of the preset measuring points according to the comparison result.
According to the equipment abnormity monitoring system and method based on trend analysis, provided by the embodiment of the invention, through deep learning by means of artificial intelligence, the trend comparison analysis of a large number of associated measuring points can be carried out, and customers can conveniently and visually and rapidly position abnormal equipment and investigate abnormal reasons.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
FIG. 1 is a block diagram of a trend analysis based equipment anomaly monitoring system in an embodiment of the present invention;
fig. 2 is a flowchart of an apparatus anomaly monitoring method based on trend analysis according to an embodiment of the present invention.
Detailed Description
Fig. 1 is a schematic structural diagram of an equipment anomaly monitoring system based on trend analysis according to an embodiment of the present invention, and as shown in fig. 1, an equipment anomaly monitoring system based on trend analysis according to an embodiment of the present invention includes: the system comprises a data acquisition unit 101, a feature extraction unit 102, a trend identification unit 103, a comparison analysis unit 104 and an abnormality monitoring unit 105.
The data acquisition unit 101 is connected to the feature extraction unit 102 and configured to acquire real-time operation data of a preset measurement point, where the real-time operation data includes, but is not limited to, data types such as temperature, pressure, flow, current, power, voltage, rotation speed, and vibration.
The feature extraction unit 102 is connected to the data acquisition unit 101 and the trend identification unit 103, and is configured to perform feature matching on the corresponding real-time operation data according to the type of the preset measurement point to obtain a signal feature corresponding to the preset measurement point.
The trend identification unit 103 is connected to the feature extraction unit 102 and the comparative analysis unit 104, and is configured to perform trend identification on the signal features according to a preset trend identification algorithm to obtain primitive types corresponding to preset measurement points, where the identified primitive types may include rising, falling, steady, up-transient, down-transient, and the like.
The comparison and analysis unit 104 is connected with the trend identification unit 103 and the anomaly monitoring unit 105, and is configured to perform self-comparison and mutual comparison according to primitive types of preset measurement points to obtain self-comparison results and mutual comparison results of the preset measurement points, where the mutual comparison is mainly performed between the same measurement points of the same type of equipment, and the concept of the same type of equipment is as follows: the equipment has the same structure and the same or similar running working conditions, the object trends of the monitoring measuring points of the equipment are always the same, and the measuring points with abnormal trends can be intuitively found and the equipment which is possibly abnormal can be positioned by comparing the same measuring points of the equipment with the same type.
The self-comparison is performed for a single device, a plurality of associated measurement point sets can exist in the single device, and repeated measurement points also exist among the associated measurement point sets. Under the unit, collected measuring points under the same equipment are split, associated and recombined by introducing two association modes of mechanism analysis and data analysis. The mechanism correlation analysis is mainly carried out on the basis of measuring point correlation by combining the physical correlation, the thermal correlation and the time correlation of equipment and the understanding of an engineer on a monitored object, and has certain objective basis and subjective experience. The data correlation analysis is mainly started from the cleaned data under the healthy operation condition of the equipment, a data analysis method is introduced to calculate positive correlation, negative correlation and nonlinear correlation among the measuring points, and the results are used as the basis of the correlation of the measuring points.
The abnormality detection unit 105 is connected to the comparison and analysis unit 104, and is configured to determine whether there is an abnormality in the preset measurement point according to the self-comparison result and the mutual-comparison result. Specifically, in the self-comparison, the unit can directly perform the overall monitoring of the running state of the equipment or the component according to the characteristic trend change of each associated measuring point, and can also match the trend rule generated by the associated measuring point in real time with the health rule base or the abnormal rule base so as to perform the equipment state monitoring. In the mutual comparison, trend recognition is carried out on data collected by the same measuring points of the same type of equipment, a redundant parameter difference recognition method is introduced, and when the trend of the measuring points of single equipment or a few equipment is different from the trend of the measuring points of other equipment of the same type, an engineer is reminded that the relevant equipment possibly has abnormal conditions, and the engineer needs to pay attention to and further analyze the abnormal conditions.
By adopting the equipment abnormity monitoring system based on trend analysis provided by the embodiment, through deep learning by means of artificial intelligence, trend comparison analysis of a large number of associated measuring points can be performed, and customers can conveniently and visually and rapidly position abnormal equipment and investigate abnormal reasons.
Based on the equipment anomaly monitoring system based on trend analysis shown in fig. 1, the present embodiment also provides an equipment anomaly monitoring method based on trend analysis, as shown in fig. 2, the method mainly includes the following steps:
step S201, the data acquisition unit acquires real-time operation data of a preset measuring point and sends the real-time operation data to the feature extraction unit.
Specifically, the data acquisition unit acquires real-time operation data of preset measuring points, wherein the real-time operation data include but are not limited to data types such as temperature, pressure, flow, current, power, voltage, rotating speed and vibration, and the real-time operation data are sent to the feature extraction unit.
Step S202, the feature extraction unit performs feature matching on corresponding real-time running data according to the type of the preset measuring point to obtain signal features corresponding to the preset measuring point, and sends the signal features to the trend identification unit.
Specifically, the feature extraction unit performs feature matching on the corresponding real-time operation data according to the type of the preset measuring point to obtain a signal feature corresponding to the preset measuring point, and sends the signal feature to the trend identification unit.
In step S203, the trend recognition unit performs trend recognition on the signal features according to a preset trend recognition algorithm to obtain primitive types corresponding to the preset measurement points, and sends the primitive types to the comparative analysis unit.
Specifically, the trend recognition unit performs trend recognition on the signal features according to a preset trend recognition algorithm to obtain primitive types corresponding to preset measuring points, the recognized primitive types can include types of ascending, descending, stable, up-transient, down-transient and the like, and the primitive types are sent to the comparative analysis unit.
S204, the comparison and analysis unit performs self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points, and sends the self-comparison results and the mutual comparison results to the abnormality detection unit.
Specifically, the comparison analysis unit performs self-comparison and mutual comparison according to primitive types of preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points, and sends the self-comparison results and the mutual comparison results to the abnormality detection unit, wherein the mutual comparison is mainly performed between the same measuring points of the same type of equipment, and the concept of the same type of equipment is as follows: the equipment has the same structure and the same or similar running working conditions, the object trends of the monitoring measuring points of the equipment are always the same, and the measuring points with abnormal trends can be intuitively found and the equipment which is possibly abnormal can be positioned by comparing the same measuring points of the equipment with the same type.
The self-comparison is performed for a single device, a plurality of associated measurement point sets can exist in the single device, and repeated measurement points also exist among the associated measurement point sets. Under the unit, collected measuring points under the same equipment are split, associated and recombined by introducing two association modes of mechanism analysis and data analysis. The mechanism correlation analysis is mainly carried out on the basis of measuring point correlation by combining the physical correlation, the thermal correlation and the time correlation of equipment and the understanding of an engineer on a monitored object, and has certain objective basis and subjective experience. The data correlation analysis is mainly started from the cleaned data under the healthy operation condition of the equipment, a data analysis method is introduced to calculate positive correlation, negative correlation and nonlinear correlation among the measuring points, and the results are used as the basis of the correlation of the measuring points.
In step S205, the anomaly detection unit determines whether there is an anomaly in the preset measurement point according to the self-comparison result and the mutual-comparison result.
Specifically, the abnormality detection unit judges whether the preset measurement point is abnormal or not according to the self-comparison result and the mutual-comparison result. Specifically, in the self-comparison, the unit can directly perform the overall monitoring of the running state of the equipment or the component according to the characteristic trend change of each associated measuring point, and can also match the trend rule generated by the associated measuring point in real time with the health rule base or the abnormal rule base so as to perform the equipment state monitoring. In the mutual comparison, trend recognition is carried out on data collected by the same measuring points of the same type of equipment, a redundant parameter difference recognition method is introduced, and when the trend of the measuring points of single equipment or a few equipment is different from the trend of the measuring points of other equipment of the same type, an engineer is reminded that the relevant equipment possibly has abnormal conditions, and the engineer needs to pay attention to and further analyze the abnormal conditions.
In addition, the comparison and analysis unit can be generated in an off-line training or on-line training mode, related experts can establish a trend rule base between each measuring point according to the correlation analysis result in the establishing process, the rule base can be a health rule base or an abnormal rule base, for example, when the inlet flow trend of pump equipment is increased, the outlet flow trend of the pump equipment is necessarily increased, the inlet temperature of bearing cooling water is also necessarily reduced, the temperature of the bearing is also necessarily reduced, the trend rule is a health rule, otherwise, the trend rule is an abnormal rule.
According to the equipment abnormity monitoring method based on trend analysis, provided by the embodiment of the invention, through deep learning by means of artificial intelligence, the trend comparison analysis of a large number of associated measuring points can be carried out, and customers can conveniently, visually and quickly position abnormal equipment and investigate abnormal reasons.
On the basis of the above embodiment, the method for monitoring equipment abnormality based on trend analysis further includes:
the feature extraction unit is used for matching a corresponding feature extraction type according to the type of a preset measuring point and extracting the features of the real-time operation data according to the feature extraction type to obtain the signal features corresponding to the preset measuring point.
In the embodiment of the invention, after the original data (real-time operation data) of the measuring points are obtained, the characteristic matching is firstly carried out, and different corresponding characteristics are matched for different measuring points according to the characteristic requirements. For example, when an abnormality occurs at a point related to electric power of the equipment, a transient change may occur, and at this time, it is necessary to match raw data as an analysis feature to extract related information, and in most cases, a transient change does not occur at a point related to flow rate, and thus, other features may be extracted as needed. In the system, the features extracted from each measuring point can be freely configured as required, and the extracted features include but are not limited to discrete features, smooth features, fluctuation features and the like. The feature extraction method comprises the following steps:
obtaining discrete features: inputting historical data of the measuring points and the feature extraction precision v, dividing the data into v intervals by adopting an equal width method, giving corresponding discrete values according to the intervals in which the continuous measuring points fall, and finally obtaining the required discrete features. The method has the advantages that the noise suppression can be carried out on the original data, and the fitness with the adopted trend identification algorithm is high. The system can freely adjust the feature extraction precision v value to flexibly adjust the noise suppression level, the smaller the v value is, the larger the noise suppression is, the more the original data is lost, and vice versa.
Extracting the fluctuation characteristics: extracting fluctuation characteristics of the selected measuring points, establishing a window with the width of w, sliding the window by the step length of l each time, and passing through a formula after sliding each time
The fluctuation characteristics within the window are calculated. Wherein xi is the ith data in the sliding window and is the mean value of the data in the sliding window.
And (3) extracting a smooth feature: and extracting smooth features from the selected measuring points, establishing a window with the width of w, wherein the sliding step length is l each time, and calculating the average value of data in the window after each sliding is the obtained smooth feature index.
According to the embodiment of the invention, different corresponding characteristics are matched through different measuring points according to the characteristic requirements, and more accurate characteristic extraction data can be obtained by corresponding to different characteristic extraction methods.
On the basis of the above embodiment, the method for monitoring equipment abnormality based on trend analysis further includes:
performing trend hypothesis on the preset measuring points to obtain confidence threshold values;
and performing statistical calculation according to the signal characteristics, comparing the statistical calculation result with the confidence coefficient threshold value, and determining the primitive type corresponding to the preset measuring point according to the comparison result.
In the embodiment of the present invention, for trend prediction of preset measurement points, trend hypothesis may be performed on the preset measurement points to obtain a confidence threshold, then statistical calculation is performed according to signal characteristics, a statistical calculation result is compared with the confidence threshold, a primitive type corresponding to the preset measurement points is determined according to the comparison result, and a transient type of the preset measurement points may also be identified by performing mutation verification on the signal characteristics, wherein the prediction of the primitive type may include primitive identification through long-term trend and primitive identification through instantaneous trend, and the specific steps may be:
primitive identification of long-term trends:
a. order event H0Event H, for the original hypothesis to indicate no monotonic trendaTo make an alternative assumption, there is a monotonic trend. Initial assumption H0If true, H0 is rejected from accepting Ha when the confidence exceeds some threshold.
b. Calculating statistical index S, and giving the discrete characteristics according to the acquisition time1，y2…ynTo indicate. Calculate y for all featuresi-yjValue j ≦ n for all i, and always let i>j。
c. Calculating a test statistic Z based on the statistical index S, wherein when the statistic formula is greater than, equal to or less than zero, the statistic formula respectively comprises the following steps:
d. in the bilateral trend test, for a given level of significance, if
If the original hypothesis H0 is unacceptable, the representation shows a monotonous trend, if the absolute value of Z is greater than or equal to 1.28, 1.64 and 2.32, the significance tests of confidence degrees of 90%, 95% and 99% are passed respectively, the original hypothesis H0 is rejected, and the alternative hypothesis Ha is accepted, so that the trend exists. At this time, the value Z is judged>0 is a trend upward, Z<0 is a trend decreasing.
Primitive identification of transient trends:
and the trend identification unit also carries out mutation check on the original data of the associated measuring points, and the base class for identifying the trend is an up transient and a down transient. U and d respectively represent the upper and lower limits, h representsThe mutation tolerance scale is defaulted to 3 and is usually given based on the distribution characteristics of the objects at each monitoring point. u = Q3+h*(Q3-Q1)，d=Q1-h*(Q3-Q1). Wherein Q is3Represents the upper quartile, Q1Representing the lower quartile. An up-transient is considered to occur when there is raw data above an upper value u and a down-transient is considered to occur when there is raw data below a lower value d.
According to the embodiment of the invention, the comprehensive trend information of the preset measuring points is obtained by performing the primitive identification of the long-term trend and the primitive identification of the instantaneous trend on the prediction correlation measurement, so that whether the preset measuring points are abnormal or not can be conveniently judged according to the trend information in the follow-up process.
On the basis of the above embodiment, the method for monitoring device abnormality based on trend analysis further includes performing self-comparison and mutual comparison according to primitive types of preset measuring points, including:
self-comparison step:
and acquiring associated measuring points related to the attributes of the preset measuring points in the detection equipment, acquiring primitive types of the associated measuring points, comparing the primitive types of the preset measuring points with the primitive types of the associated measuring points, and obtaining a self-comparison result of the preset measuring points according to a comparison result.
Acquiring the associated measuring points related to the preset measuring point attributes in the detection equipment, wherein the acquiring comprises the following steps:
acquiring a preset physical association dimension, and acquiring an association measuring point physically associated with the preset measuring point through the physical association dimension;
or acquiring normal data of a preset measuring point and a measuring point to be determined, calculating the correlation between the normal data of the preset measuring point and the normal data of the measuring point to be determined, and determining a related measuring point in the measuring point to be determined according to the correlation calculation result.
And a mutual comparison step:
acquiring associated equipment with the same type as the detection equipment, acquiring a second associated measuring point at the same position as the preset measuring point in the associated equipment, and acquiring the primitive type of the second associated measuring point;
and comparing the primitive types of the preset measuring points with the primitive types of the second associated measuring points, and obtaining the mutual comparison result of the preset measuring points according to the comparison result.
In the embodiment of the invention, mutual comparison is mainly carried out among the same measuring points of the same type of equipment, wherein the concept of the same type of equipment is as follows: the equipment has the same structure and the same or similar running working conditions, the object trends of the monitoring measuring points of the equipment are always the same, and the measuring points with abnormal trends can be intuitively found and the equipment which is possibly abnormal can be positioned by comparing the same measuring points of the equipment with the same type.
The self-comparison is performed for a single device, a plurality of associated measurement point sets can exist in the single device, and repeated measurement points also exist among the associated measurement point sets. Under the unit, collected measuring points under the same equipment are split, associated and recombined by introducing two association modes of mechanism analysis and data analysis. The mechanism correlation analysis is mainly carried out on the basis of measuring point correlation by combining the physical correlation, the thermal correlation and the time correlation of equipment and the understanding of an engineer on a monitored object, and has certain objective basis and subjective experience. The data correlation analysis is mainly started from the cleaned data under the healthy operation condition of the equipment, a data analysis method is introduced to calculate positive correlation, negative correlation and nonlinear correlation among the measuring points, and the results are used as the basis of the correlation of the measuring points. The detailed calculation mode of the relevance of each sensor based on the data analysis method is as follows:
x = { xi, i =1,2,3 … n }, Y = { yi, i =1,2,3 … n }, n is the number of samples, X and Y are the two sets of points to be analyzed.
The positive correlation and the negative correlation in the linear correlation are calculated as follows:
wherein r is a linear correlation coefficient, and is an average value of the measured points X and Y.
The nonlinear correlation is calculated as follows:
wherein p (X, Y) is the joint probability density of X and Y, and p (X), p (Y) are edge probabilities.
Assuming that D = { (xi, yi), i =1,2, …, n } is a set of ordered pairs, defining a partition G, dividing the measuring points X and Y to be analyzed into grids represented by G according to a value range, dividing the internal calculation mutual information MI (X, Y) according to each obtained grid, and extracting the MI (X, Y) maximum value in different partition modes. Wherein D | G indicates that the data D is divided in the G mode.
Finally, the nonlinear correlation calculation result is as follows:
where b (n) is the upper limit value of the grid division, b (n) = n0.6 is taken.
And finally, obtaining the associated measuring points obtained based on a data analysis mode through the linear correlation degree r and the nonlinear correlation degree MIC, wherein the threshold value of the correlation degree in the system is adjustable (default is 0.3), and if the absolute value or the MIC value of r is greater than the given associated threshold value, the measuring points to be analyzed are considered to have the correlation and need to be listed as the associated measuring points.
According to the embodiment of the invention, whether the preset measuring points are abnormal or not is judged by searching the associated measuring points and the associated equipment for self-comparison and mutual comparison, so that an accurate judgment result can be obtained.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (9)
1. An equipment anomaly monitoring system based on trend analysis, comprising: a data acquisition unit, a feature extraction unit, a trend identification unit, a comparison and analysis unit and an abnormality monitoring unit, wherein,
the data acquisition unit is used for acquiring real-time operation data of preset measuring points;
the characteristic extraction unit is used for carrying out characteristic matching on corresponding real-time operation data according to the type of a preset measuring point to obtain signal characteristics corresponding to the preset measuring point;
the trend recognition unit is used for performing trend recognition on the signal characteristics according to a preset trend recognition algorithm to obtain a primitive type corresponding to the preset measuring point;
the comparison analysis unit is used for carrying out self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points;
and the abnormity detection unit is used for judging whether the preset measuring point is abnormal or not according to the self-comparison result and the mutual comparison result.
2. A method for monitoring equipment abnormality based on trend analysis according to claim 1, the method comprising:
the data acquisition unit acquires real-time operation data of preset measuring points and sends the real-time operation data to the feature extraction unit;
the characteristic extraction unit is used for carrying out characteristic matching on corresponding real-time running data according to the type of a preset measuring point to obtain signal characteristics corresponding to the preset measuring point and sending the signal characteristics to the trend identification unit;
the trend recognition unit carries out trend recognition on the signal characteristics according to a preset trend recognition algorithm to obtain a primitive type corresponding to the preset measuring point, and sends the primitive type to the comparison analysis unit;
the comparison analysis unit carries out self-comparison and mutual comparison according to the primitive types of the preset measuring points to obtain self-comparison results and mutual comparison results of the preset measuring points, and sends the self-comparison results and the mutual comparison results to the abnormality detection unit;
and the abnormity detection unit judges whether the preset measuring point is abnormal or not according to the self-comparison result and the mutual comparison result.
3. The method for monitoring equipment abnormality based on trend analysis as claimed in claim 2, wherein the step of performing feature matching on the corresponding real-time operation data by the feature extraction unit according to the type of the preset measuring point to obtain the signal feature corresponding to the preset measuring point comprises:
the feature extraction unit is used for matching a corresponding feature extraction type according to the type of a preset measuring point and extracting the features of the real-time operation data according to the feature extraction type to obtain the signal features corresponding to the preset measuring point.
4. The method for monitoring equipment abnormality based on trend analysis according to claim 3, wherein the feature extraction type includes:
discrete feature extraction, smooth feature extraction and fluctuation feature extraction.
5. The method for monitoring equipment abnormality based on trend analysis as claimed in claim 2, wherein the trend recognition unit performs trend recognition on the signal features according to a preset trend recognition algorithm to obtain primitive types corresponding to the preset measuring points, and the method comprises the following steps:
performing trend hypothesis on the preset measuring points to obtain confidence threshold values;
and performing statistical calculation according to the signal characteristics, comparing the statistical calculation result with the confidence coefficient threshold value, and determining the primitive type corresponding to the preset measuring point according to the comparison result.
6. The method for trend analysis based equipment anomaly monitoring according to claim 5, further comprising:
and the trend identification unit carries out mutation verification on the signal characteristics and identifies the transient type of the preset measuring point, wherein the primitive type corresponding to the preset measuring point comprises the transient type.
7. The equipment abnormality monitoring method based on trend analysis according to claim 2, wherein the comparative analysis unit is used for performing self-comparison according to primitive types of the preset measuring points to obtain self-comparison results of the preset measuring points, and comprises the following steps:
and acquiring associated measuring points related to the attributes of the preset measuring points in the detection equipment, acquiring primitive types of the associated measuring points, comparing the primitive types of the preset measuring points with the primitive types of the associated measuring points, and obtaining a self-comparison result of the preset measuring points according to a comparison result.
8. The method for monitoring equipment abnormality based on trend analysis according to claim 7, wherein the acquiring of the associated measuring point in the detection equipment related to the preset measuring point attribute comprises:
acquiring a preset physical association dimension, and acquiring an association measuring point physically associated with the preset measuring point through the physical association dimension;
or acquiring normal data of a preset measuring point and a measuring point to be determined, calculating the correlation between the normal data of the preset measuring point and the normal data of the measuring point to be determined, and determining a related measuring point in the measuring point to be determined according to the correlation calculation result.
9. The method for monitoring equipment abnormality based on trend analysis as claimed in claim 2, wherein the comparative analysis unit is used for comparing the primitive types of the preset measuring points with each other to obtain the comparison result of the preset measuring points, and comprises:
acquiring associated equipment with the same type as the detection equipment, acquiring a second associated measuring point at the same position as the preset measuring point in the associated equipment, and acquiring the primitive type of the second associated measuring point;
and comparing the primitive types of the preset measuring points with the primitive types of the second associated measuring points, and obtaining the mutual comparison result of the preset measuring points according to the comparison result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110151207.9A CN112685216A (en) | 2021-02-04 | 2021-02-04 | Equipment abnormity monitoring system and method based on trend analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110151207.9A CN112685216A (en) | 2021-02-04 | 2021-02-04 | Equipment abnormity monitoring system and method based on trend analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112685216A true CN112685216A (en) | 2021-04-20 |
Family
ID=75459657
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110151207.9A Pending CN112685216A (en) | 2021-02-04 | 2021-02-04 | Equipment abnormity monitoring system and method based on trend analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112685216A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112988978A (en) * | 2021-04-27 | 2021-06-18 | 河南金明源信息技术有限公司 | Case trend analysis system in key field of public welfare litigation |
| CN114387123A (en) * | 2021-12-09 | 2022-04-22 | 中核武汉核电运行技术股份有限公司 | Data acquisition management method |
| WO2022252505A1 (en) * | 2021-06-02 | 2022-12-08 | 杭州安脉盛智能技术有限公司 | Device state monitoring method based on multi-index cluster analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011024382A1 (en) * | 2009-08-28 | 2011-03-03 | 株式会社日立製作所 | Method and device for monitoring the state of a facility |
| CN102402218A (en) * | 2011-11-10 | 2012-04-04 | 浙江工业大学 | Intelligent industrial network monitoring system |
| KR20180076172A (en) * | 2016-12-27 | 2018-07-05 | 주식회사 씨에스리 | Apparatus and method for detecting anomaly of database system |
| CN112069909A (en) * | 2020-08-11 | 2020-12-11 | 李睿飞 | Real-time sewage discharge monitoring method and device and readable storage medium |
| CN112231181A (en) * | 2020-12-08 | 2021-01-15 | 平安科技（深圳）有限公司 | Data abnormal update detection method and device, computer equipment and storage medium |
-
2021
- 2021-02-04 CN CN202110151207.9A patent/CN112685216A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011024382A1 (en) * | 2009-08-28 | 2011-03-03 | 株式会社日立製作所 | Method and device for monitoring the state of a facility |
| CN102402218A (en) * | 2011-11-10 | 2012-04-04 | 浙江工业大学 | Intelligent industrial network monitoring system |
| KR20180076172A (en) * | 2016-12-27 | 2018-07-05 | 주식회사 씨에스리 | Apparatus and method for detecting anomaly of database system |
| CN112069909A (en) * | 2020-08-11 | 2020-12-11 | 李睿飞 | Real-time sewage discharge monitoring method and device and readable storage medium |
| CN112231181A (en) * | 2020-12-08 | 2021-01-15 | 平安科技（深圳）有限公司 | Data abnormal update detection method and device, computer equipment and storage medium |
Non-Patent Citations (4)
| Title |
|---|
| 张明等: "基于动态自学习阈值和趋势滤波的机械故障智能预警方法", 《振动与冲击》 * |
| 张贝克等: "定性趋势分析在过程故障诊断中的应用研究", 《系统仿真学报》 * |
| 王建等: "基于多传感器信息融合的大坝监测数据分析", 《武汉大学学报(工学版)》 * |
| 陶伟忠等: "选煤厂煤炭输送设备异常状态智能预警系统的设计与应用", 《选煤技术》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112988978A (en) * | 2021-04-27 | 2021-06-18 | 河南金明源信息技术有限公司 | Case trend analysis system in key field of public welfare litigation |
| CN112988978B (en) * | 2021-04-27 | 2024-03-26 | 河南金明源信息技术有限公司 | Case trend analysis system in important field of public service litigation |
| WO2022252505A1 (en) * | 2021-06-02 | 2022-12-08 | 杭州安脉盛智能技术有限公司 | Device state monitoring method based on multi-index cluster analysis |
| CN114387123A (en) * | 2021-12-09 | 2022-04-22 | 中核武汉核电运行技术股份有限公司 | Data acquisition management method |
| CN114387123B (en) * | 2021-12-09 | 2024-04-09 | 中核武汉核电运行技术股份有限公司 | Data acquisition management method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112685216A (en) | Equipment abnormity monitoring system and method based on trend analysis | |
| CN106407589B (en) | Fan state evaluation and prediction method and system | |
| RU2626780C1 (en) | Method and system of remote monitoring energy installations | |
| CN112629585A (en) | Equipment on-line monitoring method and device based on multi-dimensional parameter estimation | |
| CN107851294B (en) | State-based preventive maintenance device and method for large-scale operation system | |
| CN116756595B (en) | Conductive slip ring fault data acquisition and monitoring method | |
| CN117407679B (en) | Data acquisition method and system of intelligent end screen sensor | |
| WO2018106150A1 (en) | System and method of remote object monitoring | |
| CN117060409B (en) | Automatic detection and analysis method and system for power line running state | |
| CN116660672B (en) | Power grid equipment fault diagnosis method and system based on big data | |
| CN113757093B (en) | Flash steam compressor unit fault diagnosis method | |
| CN117270514B (en) | Production process whole-flow fault detection method based on industrial Internet of things | |
| CN114239708B (en) | Combustion engine abnormity detection method based on quality control chart theory | |
| CN116086537A (en) | Equipment state monitoring method, device, equipment and storage medium | |
| CN117725541B (en) | Intelligent monitoring and fault diagnosis system for running state of annealing furnace | |
| CN117560300B (en) | Intelligent internet of things flow prediction and optimization system | |
| CN113868948A (en) | User-oriented dynamic threshold model training system and method | |
| CN113934536A (en) | Data acquisition method facing edge calculation | |
| CN113673600A (en) | Industrial signal abnormity early warning method, system, storage medium and computing equipment | |
| CN112326246A (en) | Bearing safety state online monitoring method based on periodic data and nuclear density estimation | |
| CN116910677A (en) | Industrial instrument fault diagnosis method and system | |
| CN116910128A (en) | Wind turbine generator hard and soft fault data analysis system and analysis method thereof | |
| CN115034094B (en) | Prediction method and system for operation state of metal processing machine tool | |
| CN114971308A (en) | Transformer health condition assessment method based on multi-dimensional data analysis | |
| JP2021076597A (en) | Detecting rotor anomalies by determining vibration trends during transient speed operation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210420 |
|
| RJ01 | Rejection of invention patent application after publication |