CN105912789B - A kind of sensor selection method for degraded data - Google Patents
A kind of sensor selection method for degraded data Download PDFInfo
- Publication number
- CN105912789B CN105912789B CN201610232071.3A CN201610232071A CN105912789B CN 105912789 B CN105912789 B CN 105912789B CN 201610232071 A CN201610232071 A CN 201610232071A CN 105912789 B CN105912789 B CN 105912789B
- Authority
- CN
- China
- Prior art keywords
- sensor
- failure
- degradation
- data
- state
- 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.)
- Active
Links
- 238000010187 selection method Methods 0.000 title claims abstract description 20
- 230000015556 catabolic process Effects 0.000 claims abstract description 61
- 238000006731 degradation reaction Methods 0.000 claims abstract description 61
- 238000000034 method Methods 0.000 claims abstract description 20
- 238000004458 analytical method Methods 0.000 claims abstract description 13
- 238000012546 transfer Methods 0.000 claims abstract description 7
- 230000007704 transition Effects 0.000 claims description 7
- 239000000463 material Substances 0.000 claims description 3
- 238000009825 accumulation Methods 0.000 claims description 2
- 238000012423 maintenance Methods 0.000 description 10
- 238000012545 processing Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 5
- 230000007547 defect Effects 0.000 description 3
- 238000005299 abrasion Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007850 degeneration Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 231100000331 toxic Toxicity 0.000 description 1
- 230000002588 toxic effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/30—Circuit design
- G06F30/32—Circuit design at the digital level
- G06F30/33—Design verification, e.g. functional simulation or model checking
- G06F30/3323—Design verification, e.g. functional simulation or model checking using formal methods, e.g. equivalence checking or property checking
Landscapes
- Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention discloses a kind of sensor selection methods based on variance for degraded data.In the method, based on to the degradation characteristics analysis under fault mode of the same race, degradation modes consistency, original state ambiguity, degenerate four whole monotonicity, failure state consistency key properties are analyzed and described, and specificity analysis basis is foundation according to this, devise health and the failure state calculating based on degradation characteristics analysis, state transfer amount calculates, failure state amount of variation calculates, consistent section rate calculating of failing, finally, using the section rate that unanimously fails as Primary Reference amount, sensor signal is neatly selected using two methods.
Description
Technical Field
The invention relates to a variance-based sensor selection method for degraded data.
Background
In order to improve reliability, safety and efficiency, equipment maintenance has received increasing attention in many areas, especially in the area of complex equipment, such as aerospace, power plants, large industrial plants, etc., where safety-related or outage and failure costs are high. Currently, a conventional logistics Maintenance system is usually a subsequent Maintenance (CM) system or a Time-based Maintenance (TBM) system when a fault Maintenance decision is made, that is, Maintenance is performed after an equipment fault or failure occurs or equipment Maintenance is performed by determining a Time period unrelated to an actual degradation condition of the equipment in advance as a period. Therefore, the actual health state of the equipment is rarely considered, and generally, a method of performing system detection by using a traditional instrument to obtain data and then performing fault judgment manually is adopted, so that the efficiency is low, the reliability is poor, and potential slight faults of system equipment are difficult to find. Therefore, a concept of predictive Maintenance (PHM) is proposed, which mainly represents that a technique of Predictive and Health Management (PHM) is proposed. The technology is characterized in that a series of sensors capable of collecting data related to the health state of equipment are mounted on the equipment to monitor the state, then the health state information of the equipment and the available residual life are predicted through fault diagnosis and fault prediction technologies, and finally decision and management are carried out through the information, so that the maintenance of the equipment is carried out from the perspective of life cycle maintenance.
While different failure modes may be associated with different sensors in processing condition monitoring data, current practice often uses all of the sensor data in processing one failure mode, which causes two major problems. Firstly, irrelevant sensor data not only can not provide effective information, but also has great possibility of causing great negative influence on expected results along with the increase of the occupied data proportion, thereby reducing the effectiveness of the information flow of the whole system and even prompting the whole system to provide error decision, and causing great safety problems and property loss. And secondly, irrelevant sensor data is used as toxic data, so that the data processing capacity of the whole system is greatly tired, the storage and calculation resources of the system are wasted, and the effectiveness of the system is reduced. Since the required state monitoring data is generally degradation data of the equipment in the fields of prediction and health management, a sensor selection method for the degradation data plays a very important role in the whole system implementation.
The only solution to the above problem is to precisely select the relevant sensor data when processing a certain failure mode, and then perform subsequent processing. Currently, there are limited studies on this problem, and three main ideas are included. Firstly, the manual observation method is selected manually according to a certain characteristic, the advantages are that the method is simple and the result is accurate, the defects are that a professional is needed, the processing speed is low, and the result reliability is poor for qualitative analysis. And secondly, an exhaustive search method is adopted, the sensors are combined and exhausted, and the evaluation indexes are designed to select a better sensor combination, so that the method has the advantages of accurate result and better reliability, and has the defects of difficult evaluation index design, huge calculation amount, long time consumption and lower cost performance. And thirdly, a quantitative analysis method, which designs performance indexes and threshold values from a certain angle through analysis and finally compares the performance indexes and the threshold values to select, has the advantages of high processing speed and good reliability, and has the defects that the difference of result accuracy is obvious along with different methods, and professional knowledge can be fully utilized.
Therefore, accurately and reliably finding sensor data associated with a current failure mode while ensuring that system storage and computational resources are used in small amounts is a difficult problem that needs to be addressed by those skilled in the art.
Disclosure of Invention
The invention mainly aims to solve the problem of lack of effective sensor signal selection aiming at sensor data with degradation property, and provides an effective sensor signal selection method under a certain fault mode.
In order to achieve the purpose, the invention adopts the following technical scheme: a variance-based sensor selection method for degraded data, characterized by the steps of: s1, aiming at the selected degradation mode and aiming at each unit sensor, calculating health and failure states; s2, calculating the state average transfer quantity and the failure state variance quantity; s3, calculating the consistent failure interval rate; and S4, selecting a sensor according to the consistent failure rate interval.
The invention has the beneficial effects that: the invention provides a sensor signal selection method aiming at degradation data in a certain fault mode. The method is based on analyzing the degradation characteristics of the equipment units, thereby fully utilizing the degradation knowledge of the equipment units and ensuring the correctness and reliability of the selection result; meanwhile, an index capable of reflecting the correlation degree of the sensor signal and the fault mode is constructed on the basis of simple parameters such as variance and the like, so that the occupation of storage and calculation resources is small, and the speed is high; and finally, the sensor signals are flexibly selected by taking the index as a reference, so that the selection method has better flexibility, practicability and expansibility. Therefore, the method provided by the invention has the characteristics of accurate result, high reliability and small calculation amount.
Drawings
FIG. 1 is a flow chart of a sensor selection method of the present invention;
FIG. 2 is a diagram of a set of sample sensor signals for an embodiment of a sensor selection method of the present invention;
FIGS. 3-1, 3-2 are exemplary sample graphs of sensor signals for embodiments of the sensor selection method of the present invention;
FIG. 4 is a graphical representation of consistent failure interval rate results for an embodiment of the sensor selection method of the present invention.
Detailed Description
The variance-based sensor selection method for degraded data of the present embodiment includes: the method comprises the steps of selecting data set description in a degradation mode, selecting degradation characteristic analysis in the degradation mode, calculating health and failure states, calculating state transfer quantity, calculating failure state variance quantity and consistent failure interval rate based on the degradation characteristic analysis, setting an interval rate threshold or a selection proportion, and selecting a sensor according to the threshold or the proportion. The degradation characteristic analysis comprises analysis of four main characteristics of degradation mode consistency, initial state uncertainty, degradation integral monotonicity and failure state consistency; the consistency of the degradation mode is the basis of other three characteristics and is also the basis of calculating the consistent failure interval rate, the uncertainty of the initial state and the consistency of the failure state are the basis of calculating the health and the failure states, the consistency of the failure state is the basis of calculating the variance of the failure states, and the monotonicity of the degradation whole is the basis of calculating the state transfer quantity; wherein the degradation characteristic is a basis for calculating respective digital quantities, which are preconditions for sensor selection.
Wherein:
the data set in this certain degeneration mode is described below,
given data set
D＝{Uniti|i＝1，2，...，n} (1)
The data set D comprises n equipment units of the same type under the same degradation mode, each Unit is provided with the same data structure, and any Unit in the data set D is
Uniti＝{Sensorj i|j＝1，2，...，m} (2)
Each equipment UnitiAll contain m Sensor signals sensorsj iFor each sensor signal
Arbitrary Sensor for ith equipment unitj iEach sensor signal comprising a length q associated with the associated celliTime series data ofThese sensor data in time series record the course of the unit from an unknown state of health to a state of failure in the same operating mode.
The degradation mode consistency specifically means that degradation signals of equipment units acquired by sensors arranged at the same acquisition point are similar in morphology; the initial state uncertainty specifically refers to that the equipment has different degrees of initial abrasion and different internal material stress structures in the initial operation stage, and although the abrasion and the like are unknown, the equipment is considered to be in a normal or healthy state; the overall monotonicity of the degradation specifically means that the global trend of the sensor signal is present and unidirectional; the failure state consistency specifically means that when the equipment unit is degraded to a failure state, the sensor signal capable of reflecting the health state of the equipment should converge to a certain constant value or a small interval, and the constant value or the interval can be regarded as a failure threshold.
The health and failure state calculations, for each sensor, are calculated as follows,
wherein S ishealthIs the average of the first few data of each sensor, SfailureThe average value is the average value of a plurality of data at the tail end of each sensor, the average value is used for reducing noise interference, and the average number of failure data is smaller than that of health data, the tail end has obvious degradation trend compared with the initial degradation trend, and the large number of data cannot reflect the real failure state of the tail end.
The state transition amount is calculated as follows,
wherein,indicates the state transition amount of the ith cell jth sensor, AjRepresenting the average state transition for the jth sensor. The average here is per unit average; the n equipment units are the same type of equipment, and the number and the types of the sensors arranged on the equipment units are the same, so that the physical quantity measured by the jth sensor of all the units is the same, and therefore, the averaging of the physical quantity is to average the measured same physical quantity.
The failure state variance measure is calculated as follows,
wherein,means, V, representing the quantity at the end of the jth sensorjIndicating the failure state variance measure for the jth sensor.
The consistent failure interval rate is calculated as follows,
wherein R isjIs the consistent failure interval rate of the jth sensor, and in order to prevent the denominator from being zero, a small positive number epsilon 1 is added to the denominator, and the value is generally 10-4。
The interval rate threshold is set, and the sensor selection is carried out according to the interval rate threshold, the specific implementation form is as follows,
SensorSet＝{Sensorj|ε0＜Rj＜ε} (10)
wherein, SensorSet is the final sensor signal set to be selected, epsilon is the set interval rate threshold, epsilon 0 is a very small positive integer, and the value is generally 10-5With the aim of converting R intojThe sensor signal of 0 is filtered out.
The specific implementation of setting the selection proportion and selecting the sensors according to the selection proportion is as follows:
1) rejecting the sensors with zero interval rate (consistent failure interval rate is also referred to as interval rate for short), and arranging the rest sensors from small to large according to the interval rate;
2) accumulating and summing the interval rates of all the sensors, and calculating the proportion of the interval rates of all the sensors;
3) accumulating from the proportion with smaller interval rate until the proportion accumulated value reaches the set value;
4) the sensor signal included in the proportional-accumulation value is the sensor signal to be selected.
The embodiments of the present invention will be described in detail below. It should be emphasized that the following description is merely exemplary in nature and is not intended to limit the scope of the invention or its application.
Referring to FIG. 1, in one embodiment, a method of variance-based sensor selection for degradation data includes: describing a data set under a degradation mode of a high-pressure compressor of an XX model engine, analyzing degradation characteristics under a certain degradation mode, calculating health and failure states, calculating state transfer quantity, calculating failure state variance quantity and consistent failure interval rate based on the degradation characteristic analysis, setting an interval rate threshold or a selection proportion, and selecting a sensor according to the threshold or the proportion. The degradation characteristic analysis comprises analysis of four main characteristics of degradation mode consistency, initial state uncertainty, degradation integral monotonicity and failure state consistency; the consistency of the degradation mode is the basis of other three characteristics and is also the basis of calculating the consistent failure interval rate, the uncertainty of the initial state and the consistency of the failure state are the basis of calculating the health and the failure states, the consistency of the failure state is the basis of calculating the variance of the failure states, and the monotonicity of the degradation whole is the basis of calculating the state transfer quantity; wherein the degradation characteristic is a basis for calculating respective digital quantities, which are preconditions for sensor selection. In fig. 1, the text in the oval box is an explanation of the flow, and does not belong to the component of the flow.
In a preferred embodiment, a given data set contains 100 equipment units in the same failure mode, each equipment unit containing 21 sensor signals, and different equipment units having different lifetimesThe length of life. Therefore, in formulae (1) and (2), n is 100, m is 21, and q in formula (3)iIf the number n is 100, the number n is the value.
Referring to fig. 2, the data of 21 sensor signals of all units in the data set is shown, and the data is not clearly shown due to the large amount of data, so that only the general trend of the data is shown, and the numbers 1-21 below the line coordinates in the figure are the sensor numbers. Referring to fig. 3-1 and 3-2, the degradation signals of 100 equipment units detected by two typical sensors in a data set are shown, the numbers of the sensors are 12 and 14 respectively (fig. 3-1 is an enlarged version of the number 12 of the sensor in fig. 2, and fig. 3-2 is an enlarged version of the number 14 of the sensor in fig. 2, which shows a difference due to the fact that points are more spatial and smaller), and the basic characteristics of the degradation data detected by the sensors can be observed more clearly through the graph. Wherein, sensor number 12 compares to conform to four degradation characteristics analyzed: namely, the sensors have different values at the starting points, have larger dispersion amplitude and represent an uncertain initial state; the values of the sensors at the tail point are consistent and are constant values or converge to a small interval, and the failure states are consistent; the whole presents a unidirectional trend, which represents the whole monotony; the degradation data of different units are similar in shape, and represent consistent degradation modes. While sensor number 14 does not meet the degradation characteristic of failure state consistency, i.e., the value of the sensor at the end point is divergent, not a constant value or does not converge into a small region.
In the preferred embodiment, consistent failure interval rates for 21 sensors can be calculated according to equations (4), (5), (6), (7), (8) and (9), with specific values as shown in table 1.
Referring to fig. 4, a size distribution of the coincidence failure interval rate is represented, the horizontal axis represents the numbers of 21 sensors, and the vertical axis represents the calculated R value of each sensor, i.e., the coincidence interval rate. The consistent failure interval rate distribution of the sensor can be visually seen, the consistent failure interval rate distribution has obvious size distribution and is easy to distinguish, and generally, the smaller the consistent failure interval rate is, the higher the association degree of the sensor and the degradation mode is.
Table 1 consistent failure interval rates for sensors
| Sensor with a sensor element | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
| Interval rate | 0.000 | 0.046 | 0.078 | 0.019 | 0.000 | 0.000 | 0.032 | 0.182 | 1.794 | 0.000 | 0.021 |
| Sensor with a sensor element | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19 | 20 | 21 | |
| Interval rate | 0.027 | 0.172 | 2.884 | 0.029 | 0.000 | 0.058 | 0.000 | 0.000 | 0.036 | 0.045 |
In the preferred embodiment, after a consistent failure interval rate is calculated, there are two methods for sensor selection. If the first threshold method is adopted, assuming that the threshold is set to be ∈ 0.25, the sensor selection results are as follows
If the second method is used, steps 1) and 2) are performed first assuming a set ratio of 5%), the results are shown in Table 2
TABLE 2 corresponding ratio of sorted sensors
| Sensor with a sensor element | 4 | 11 | 12 | 15 | 7 | 20 | 21 |
| Ratio of | 0.36％ | 0.38％ | 0.50％ | 0.53％ | 0.59％ | 0.65％ | 0.83％ |
| Sensor with a sensor element | 2 | 17 | 3 | 13 | 8 | 9 | 14 |
| Ratio of | 0.84％ | 1.06％ | 1.43％ | 3.17％ | 3.35％ | 33.08％ | 53.21％ |
The above ratios are added up from small to large, and the addition ratio is 4.69% for number 2 and 5.76% for number 17, so that the sensor selection results are as follows.
The foregoing is a more detailed description of the invention in connection with specific/preferred embodiments and is not intended to limit the practice of the invention to those descriptions. It will be apparent to those skilled in the art that various substitutions and modifications can be made to the described embodiments without departing from the spirit of the invention, and these substitutions and modifications should be considered to fall within the scope of the invention.
Claims (7)
1. A variance-based sensor selection method for degraded data, characterized by the steps of:
s0, analyzing the degradation characteristics of the equipment units;
s1, aiming at the selected degradation mode and aiming at each unit sensor, calculating health and failure states;
s2, calculating the state average transfer quantity and the failure state variance quantity;
the average transition amount of the state is calculated as follows,
wherein, ShealthIs the average of the first few data of each sensor, SfailureIs the average of several data at the end of each sensor,indicates the state transition amount of the ith cell jth sensor, AjRepresents the average state transition amount of the jth sensor;
the failure state variance measure is calculated as follows,
wherein,means, V, representing the quantity at the end of the jth sensorjRepresenting a failure state variance measure for the jth sensor;
s3, calculating the consistent failure interval rate;
the consistent failure interval rate is calculated as follows,
wherein R isjIs the consistent failure interval rate of the jth sensor, and to prevent the denominator from being zero, 10 is added to the denominator-4，AjTo representAverage state transition, V, of the jth sensorjRepresenting a failure state variance measure for the jth sensor;
and S4, selecting a sensor according to the consistent failure rate interval.
2. The variance-based sensor selection method for degradation data of claim 1, wherein: in step S4, the method of selecting a sensor is one of the following methods:
s4-1, setting an interval rate threshold value, and selecting a sensor according to the threshold value;
s4-2, selecting interval rate proportion, and selecting the sensor according to the proportion.
3. The variance-based sensor selection method for degradation data of claim 1, wherein: the degradation characteristic analysis comprises analysis of degradation mode consistency, initial state uncertainty, degradation overall monotonicity and failure state consistency.
4. The variance-based sensor selection method for degradation data according to claim 3, wherein the degradation mode consistency specifically means that device unit degradation signals acquired by sensors arranged at the same acquisition point are morphologically similar; the initial state uncertainty specifically refers to the fact that the equipment has different degrees of initial wear and different material internal stress structures in the initial operation stage, and the wear and the material internal stress structures are not known but are considered to be normal or healthy states; the overall monotonicity of the degradation specifically means that the global trend of the sensor signal is present and unidirectional; the failure state consistency specifically means that when the equipment unit is degraded to a failure state, the sensor signal capable of reflecting the health state of the equipment should converge to a certain constant value or a small interval, and the constant value or the interval can be regarded as a failure threshold.
5. The variance-based sensor selection method for degradation data of claim 1, wherein the health and failure states are calculated, the health and failure states for each sensor being calculated as follows,
wherein s ispData of a sensor p, ShealthIs the average of the first few data of each sensor, SfailureThe average value is the average value of a plurality of data at the tail end of each sensor, the average value is used for reducing noise interference, and the average number of failure data is smaller than that of health data, the tail end has obvious degradation trend compared with the initial degradation trend, and the large number of data cannot reflect the real failure state of the tail end.
6. The method of claim 2, wherein the setting of the interval rate threshold and the selection of the sensor in accordance therewith are implemented in a manner such that,
SensorSet＝{Sensorj|ε0＜Rj＜ε}
where SensorSet is the set of sensor signals to be selected finally, RjIs the consistent failure interval rate of the jth sensor, ε is the set interval rate threshold, ε 0 is a positive integer 10-5With the aim of converting R intojThe sensor signal of 0 is filtered out.
7. The variance-based sensor selection method for degradation data according to claim 2, wherein the setting of the selection ratio and the implementation of the sensor selection accordingly are,
1) eliminating sensors with zero interval rate, and arranging the rest sensors from small to large according to the interval rate;
2) accumulating and summing the interval rates of all the sensors, and calculating the proportion of the interval rates of all the sensors;
3) accumulating from the proportion with smaller interval rate until the proportion accumulated value reaches the set value;
4) the sensor signal included in the proportional-accumulation value is the sensor signal to be selected.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610232071.3A CN105912789B (en) | 2016-04-14 | 2016-04-14 | A kind of sensor selection method for degraded data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610232071.3A CN105912789B (en) | 2016-04-14 | 2016-04-14 | A kind of sensor selection method for degraded data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105912789A CN105912789A (en) | 2016-08-31 |
| CN105912789B true CN105912789B (en) | 2019-06-21 |
Family
ID=56746882
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610232071.3A Active CN105912789B (en) | 2016-04-14 | 2016-04-14 | A kind of sensor selection method for degraded data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105912789B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102369448A (en) * | 2009-03-27 | 2012-03-07 | 特英普科技有限公司 | Device and method for locating partial discharges |
| US8812943B2 (en) * | 2011-09-23 | 2014-08-19 | Fujitsu Limited | Detecting data corruption in medical binary decision diagrams using hashing techniques |
-
2016
- 2016-04-14 CN CN201610232071.3A patent/CN105912789B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102369448A (en) * | 2009-03-27 | 2012-03-07 | 特英普科技有限公司 | Device and method for locating partial discharges |
| US8812943B2 (en) * | 2011-09-23 | 2014-08-19 | Fujitsu Limited | Detecting data corruption in medical binary decision diagrams using hashing techniques |
Non-Patent Citations (2)
| Title |
|---|
| Degeneration Methods in Intelligent Building Control System Design;Arakaki J 等;《Information Technology for Balanced Manufacturing Systems》;20061231;第469-478页 |
| 基于振动传感器性能退化数据的可靠性评估水;秦丽 等;《传感器与微系统》;20090930;第28卷(第9期);第53-55页 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105912789A (en) | 2016-08-31 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106373025B (en) | Power information acquisition system based on outlier detection anti-electricity-theft monitoring method in real time | |
| CN106780121B (en) | Power consumption abnormity identification method based on power consumption load mode analysis | |
| CN105071983B (en) | Abnormal load detection method for cloud calculation on-line business | |
| Zhang et al. | Health state assessment of bearing with feature enhancement and prediction error compensation strategy | |
| US10852357B2 (en) | System and method for UPS battery monitoring and data analysis | |
| CN113126019B (en) | Remote estimation method, system, terminal and storage medium for error of intelligent ammeter | |
| CN108733966A (en) | A kind of multidimensional electric energy meter field thermodynamic state verification method based on decision woodlot | |
| CN102252843A (en) | Assessment method for rolling bearing performance variation | |
| CN109670611A (en) | A kind of power information system method for diagnosing faults and device | |
| Tang et al. | Rolling bearing remaining useful life prediction via weight tracking relevance vector machine | |
| CN103678886A (en) | Satellite Bayesian Network health determination method based on ground test data | |
| CN103902798B (en) | Data preprocessing method | |
| Atzmueller et al. | Anomaly detection and structural analysis in industrial production environments | |
| Sun et al. | Flow measurement-based self-adaptive line segment clustering model for leakage detection in water distribution networks | |
| Cui et al. | A novel robust dual unscented particle filter method for remaining useful life prediction of rolling bearings | |
| Karuppusamy | Machine learning approach to predictive maintenance in manufacturing industry-a comparative study | |
| CN113204894B (en) | Construction method and application of electric energy metering abnormality diagnosis model | |
| CN114597886A (en) | Power distribution network operation state evaluation method based on interval type two fuzzy clustering analysis | |
| CN108825482B (en) | Fault detection method and system for axial plunger pump of airplane | |
| JP2015137866A (en) | Water leakage amount estimation device, method, and system | |
| CN105912789B (en) | A kind of sensor selection method for degraded data | |
| CN117312972A (en) | Method for identifying health state of scraper conveyor speed reducer | |
| CN117113247A (en) | Drainage system abnormality monitoring method, equipment and storage medium based on two-classification and clustering algorithm | |
| CN112345972A (en) | Power failure event-based power distribution network line transformation relation abnormity diagnosis method, device and system | |
| CN116467551A (en) | Water supply pipe network leakage positioning method and system based on correlation coefficient |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |