CN103617356B - A kind of adaptive online monitoring data trend abstraction method - Google Patents
A kind of adaptive online monitoring data trend abstraction method Download PDFInfo
- Publication number
- CN103617356B CN103617356B CN201310612784.9A CN201310612784A CN103617356B CN 103617356 B CN103617356 B CN 103617356B CN 201310612784 A CN201310612784 A CN 201310612784A CN 103617356 B CN103617356 B CN 103617356B
- Authority
- CN
- China
- Prior art keywords
- trend
- sequence
- trend term
- term
- data
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000012544 monitoring process Methods 0.000 title claims abstract description 16
- 230000003044 adaptive effect Effects 0.000 title claims abstract description 13
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 13
- 230000000877 morphologic effect Effects 0.000 claims abstract description 10
- 238000011161 development Methods 0.000 claims description 11
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 5
- 238000010276 construction Methods 0.000 abstract description 3
- 238000012216 screening Methods 0.000 description 9
- 238000001914 filtration Methods 0.000 description 5
- 238000012360 testing method Methods 0.000 description 4
- 230000007423 decrease Effects 0.000 description 3
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 2
- XLOMVQKBTHCTTD-UHFFFAOYSA-N Zinc monoxide Chemical compound [Zn]=O XLOMVQKBTHCTTD-UHFFFAOYSA-N 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- QHGUCRYDKWKLMG-UHFFFAOYSA-N octopamine Chemical compound NCC(O)C1=CC=C(O)C=C1 QHGUCRYDKWKLMG-UHFFFAOYSA-N 0.000 description 1
- 229960001576 octopamine Drugs 0.000 description 1
- 238000004382 potting Methods 0.000 description 1
- 238000009666 routine test Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000011787 zinc oxide Substances 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The present invention relates to a kind of adaptive online monitoring data trend abstraction method, the method combines combination form wave filter and two kinds of methods of empirical mode decomposition, first dynamic construction adaptive morphological filter structural element, described morphological filter structural element is used to be filtered, and then carry out empirical mode decomposition, then construct trend term according to the feature of handled data, carry out trending early warning finally according to described trend term.The method operational performance is high, and adaptivity is strong, and using effect is good.
Description
Technical field
The present invention relates to the extract real-time technical field of online monitoring data variation tendency, particularly a kind of power transmission and transformation
The most online trend abstraction method of equipment on-line monitoring data adaptive.
Background technology
In recent years, repair based on condition of component is widely popularized in State Grid Corporation of China, as the important technology of repair based on condition of component
The on-line monitoring of means is also developed rapidly.Online monitoring data is real-time, abundant, substantial amounts, at equipment
In the case of routine test cycle stretch-out, it it is the important channel grasping equipment state.But on-Line Monitor Device is easy
Disturbed by factors such as ambient temperature and humidity, load, grid switching operations, cause data frequent fluctuation, use single to adopt
Collection data carry out the method for state alarm and often report by mistake, have impact on the application effect of on-line monitoring.
Online monitoring data containing substantial amounts of work information, if analysing in depth, engineer applied having been had very much
It is worth.To this end, occur in that mathematical morphology, empirical mode decomposition (Empirical Mode Decomposition,
Be called for short EMD), the method such as wavelet analysis, extract in reflection equipment, the trend of long term state, and use
Gesture carries out status early warning.
Empirical mode decomposition is the one proposed in 1998 by Norden E Huang of U.S. NASA et al.
New signal analysis method.It is that the signal decomposition complicated becomes several intrinsic mode functions that EMD decomposes
(Intrinsic Mode Function is called for short IMF) and a remainder linear and.IMF reflects time sequence
The internal feature of column signal, remainder reflects trend.But, the screening process operand that EMD decomposes is very big,
When data volume increases substantially, the time cost of computing becomes geometric growth, operational performance drastically to decline, nothing
Method meets the needs of the On-line testing trend to mass data.It addition, EMD decomposes there is " crossing screening "
Phenomenon, screening IMF component out is often too much, and some components there is no clear and definite physical meaning.
King waits (2009) quietly and directly uses mathematical morphology filter to be filtered online data processing, after filtering
Data as trend.Structural element and window width thereof to choose the effect to this method most important, supervise online
The waveform morphology surveyed is varied, it is difficult to know that the structural element instantly selecting which kind of shape is more suitable for.Window width
Determining and be also difficult to solve, window width typically can obtain smoother than maximum impulse noise width great talent in data
Filter effect, but this cannot a priori know.Therefore, the method for mathematical morphology filter cannot expire adaptively
The needs of foot online monitoring data trend abstraction.
Summary of the invention
It is an object of the invention to provide a kind of adaptive online monitoring data trend abstraction method, the method is transported
Calculation performance is high, and adaptivity is strong, and using effect is good.
For achieving the above object, the technical scheme is that a kind of adaptive online monitoring data trend carries
Access method, the method first dynamic construction adaptive morphological filter structural element, use described shape filtering
Device structural element is filtered, and then carries out empirical mode decomposition, then constructs according to the feature of handled data
Trend term, carries out trending early warning finally according to described trend term.
Further, the method comprises the following steps:
(1) standard deviation of dynamic calculation original series, constructs semicircular structure using described standard deviation as radius
Element, as morphological filter structural element；
(2) use described semicircular structure element to be filtered original series processing, remove high frequency, pulse
Deng noise, obtain preliminary trend sequence data；
(3) preliminary trend sequence data empirically mode decomposition method is decomposed, obtain one group of IMF
Component and a surplus；
(4) according to the feature of handled data, select the superposition of surplus or several IMF components as becoming
Gesture item r (t)；
(5) for trend term r (t), trend term r (t) trend development in corresponding sequence time length is calculated
Speed, if trend development speed exceedes the threshold values of setting, then sends trending early warning, trend development speed V
It is defined as follows formula:
Wherein, t0It is the initial time of sequence, tnIt is the deadline of sequence, r (t0) it is t0Moment trend term
Value, r (tn) it is tnThe value of moment trend term, T is reference period.
Further, in step (4), the principle that trend term selects is as follows: be 1 year for time span
And above sequence, choose the cycle component more than 90 days in all IMF components and stack up as trend term
r(t)；For the sequence that time span is more than 3 months less than 1 year, choose the cycle in all IMF components big
Component in 30 days stacks up as trend term r (t)；It it is more than 1 month less than 3 months for time span
Sequence, choose the component more than 1 day of cycle in all IMF components and stack up as trend term r (t).
Compared to prior art, the invention has the beneficial effects as follows: operational performance is high, there is not " crossing screening "
Phenomenon, the physical meaning of screening IMF component out is clearer and more definite.The method uses combination form wave filter
Original series is filtered, eliminates noise contribution, this increase follow-up EMD screening process, simultaneously
Decrease the IMF component without clear and definite physical significance, thus substantially increase operational performance.Additionally, the present invention
The trend term of method can objectively respond Sequence Trend essence, and has the strongest adaptivity.This method filters
The structural element used is according to original series dynamic calculation, it is not necessary to manually a priori select such that it is able to dynamic
State adapts to the needs of all kinds of sequential filterings.
Accompanying drawing explanation
Fig. 1 is the workflow diagram of the inventive method.
Fig. 2 be in the embodiment of the present invention certain transformer station 110kVI section bus arrester resistive current data of A phase and
Its tendency chart.
Detailed description of the invention
Below in conjunction with the accompanying drawings and specific embodiment the present invention is described in further detail.
The present invention provides a kind of adaptive online monitoring data trend abstraction method, and the method combines combination shape
State wave filter and two kinds of methods of empirical mode decomposition, first dynamic construction adaptive morphological filter structural elements
Element, uses described morphological filter structural element to be filtered, and then carries out empirical mode decomposition, then basis
The feature structure trend term of handled data, carries out trending early warning finally according to described trend term.The method is concrete
Comprise the following steps:
(1) standard deviation of dynamic calculation original series, constructs semicircular structure using described standard deviation as radius
Element, as morphological filter structural element.
(2) use described semicircular structure element to be filtered original series processing, remove high frequency, pulse
Deng noise, obtain preliminary trend sequence data.
(3) preliminary trend sequence data empirically mode decomposition method is decomposed, obtain one group of IMF
Component and a surplus.
(4) according to the feature of handled data, select the superposition of surplus or several IMF components as becoming
Gesture item r (t).The principle that trend term selects is as follows: be 1. 1 year and above sequence for time span, it is considered to
Being affected to online monitoring data by Seasonal is cycle the longest influence factor, thus its trend part
Cycle should be greater than 90 days, chooses the cycle component more than 90 days in all IMF components and stacks up as becoming
Gesture item r (t)；2. for the sequence that time span is more than 3 months less than 1 year, all IMF components are chosen
The component more than 30 days of the middle cycle stacks up as trend term r (t)；3. for time span be 1 month with
The sequence of upper less than 3 months, it is contemplated that affected by ambient temperature and humidity and load, chooses all IMF components
The component more than 1 day of the middle cycle stacks up as trend term r (t).
(5) for trend term r (t), trend term r (t) trend development in corresponding sequence time length is calculated
Speed, if trend development speed exceedes the threshold values of setting, then sends trending early warning, trend development speed V
It is defined as follows formula:
Wherein, t0It is the initial time of sequence, tnIt is the deadline of sequence, r (t0) it is t0Moment trend term
Value, r (tn) it is tnThe value of moment trend term, T is reference period, such as 30,90,360 days, corresponding is the moon,
Season, the development speed in year.
The inventive method operational performance is high, there is not the phenomenon of " crossing screening ", screening IMF component out
Physical meaning clearer and more definite.This method uses combination form wave filter to be filtered original series, eliminates
Noise contribution, this increases follow-up EMD screening process, decreases the IMF without clear and definite physical significance simultaneously
Component, thus substantially increase operational performance.
Compare employing EMD method and the operational performance of this method below.Experiment is randomly drawed 266 zinc oxide and is kept away
Thunder device equipment, respectively the total current data of 1 month and 2 months to online acquisition (144 data every day)
Process, program use JAVA language write, CPU2.93GHZ, internal memory 2G PC on transport
OK, statistics such as table 1.
The operational performance of table 1EMD method and this method compares
| Extracting method | Time length (moon) | Average calculating operation time (ms) | Average weight number (individual) |
| FMD | 1 | 5803 | 7 |
| This method | 1 | 243 | 5 |
| EMD | 2 | 44946 | 8 |
| This method | 2 | 418 | 5.5 |
Experiment shows, the trend abstraction to 1 month data, and the average calculating operation time of EMD method is this method
23.8 times, 107 times during 2 months data, are reached.Visible, this method is greatly improved operational performance, especially
It is as the growth of data volume, the most drastically reduces average calculating operation time, performance can meet online to big data
The trend abstraction requirement of amount.
The trend term of the inventive method can objectively respond Sequence Trend essence, and has the strongest adaptivity.
The structural element that this method filtering uses is according to original series dynamic calculation, it is not necessary to manually a priori select,
It is thus possible to the needs of all kinds of sequential filtering of dynamically adapting.The determination of trend term determines the quality of trend abstraction,
Method proposes the angle from the IMF component cycle and choose which component principle as trend term, thus solve
Previous methods of having determined relies on experience to determine which component problem as trend term.This can be from following concrete reality
Example illustrates: in March, 2009, the trend of the current in resistance property of certain transformer station 110kVI section bus arrester A phase is sent out
Exhibition speed V reaches the 27.5 μ A/ month (Fig. 2), has exceeded trending early warning scope.Follow the tracks of continuously, V value April
Being the 32.9 μ A/ months, May was the 39 μ A/ months, and June was the 47.3 μ A/ months, had obvious accelerated development trend.
Using live testing to survey its current in resistance property, the trend that variation tendency is extracted with MM-EMD substantially matches.After
This arrester is carried out high-potting, by the 7uA of commissioning test in 2003 under DC test 75%U1mA
Rise to 38uA；Through strip inspection, this lightning arrester insulation bucket makes moist.
It is above presently preferred embodiments of the present invention, all changes made according to technical solution of the present invention, produced merit
When can act on the scope without departing from technical solution of the present invention, belong to protection scope of the present invention.
Claims (2)
1. an adaptive online monitoring data trend abstraction method, it is characterised in that the method is the most dynamic
State constructs adaptive morphological filter structural element, uses described morphological filter structural element to be filtered,
And then carry out empirical mode decomposition, then construct trend term, finally according to described according to the feature of handled data
Trend term carries out trending early warning；Specifically include following steps:
(1) standard deviation of dynamic calculation original series, constructs semicircular structure using described standard deviation as radius
Element, as morphological filter structural element；
(2) use described semicircular structure element to be filtered original series processing, remove high frequency, pulse
Deng noise, obtain preliminary trend sequence data；
(3) preliminary trend sequence data empirically mode decomposition method is decomposed, obtain one group of IMF
Component and a surplus；
(4) according to the feature of handled data, select the superposition of surplus or several IMF components as becoming
Gesture item r (t)；
(5) for trend term r (t), trend term r (t) trend development in corresponding sequence time length is calculated
Speed, if trend development speed exceedes the threshold values of setting, then sends trending early warning, trend development speed V
It is defined as follows formula:
Wherein, t0It is the initial time of sequence, tnIt is the deadline of sequence, r (t0) it is t0Moment trend term
Value, r (tn) it is tnThe value of moment trend term, T is reference period.
One the most according to claim 1 adaptive online monitoring data trend abstraction method, it is special
Levy and be, in step (4), trend term select principle as follows: for time span be 1 year and more than
Sequence, choose the component more than 90 days of cycle in all IMF components and stack up as trend term r (t)；
It is the sequence more than or equal to 3 months and less than 1 year for time span, chooses the cycle in all IMF components
Component more than 30 days stacks up as trend term r (t)；It is more than or equal to 1 month and little for time span
In the sequence of 3 months, choose the cycle component more than 1 day in all IMF components and stack up as trend
Item r (t).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310612784.9A CN103617356B (en) | 2013-11-27 | 2013-11-27 | A kind of adaptive online monitoring data trend abstraction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310612784.9A CN103617356B (en) | 2013-11-27 | 2013-11-27 | A kind of adaptive online monitoring data trend abstraction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103617356A CN103617356A (en) | 2014-03-05 |
| CN103617356B true CN103617356B (en) | 2016-09-07 |
Family
ID=50168059
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310612784.9A Active CN103617356B (en) | 2013-11-27 | 2013-11-27 | A kind of adaptive online monitoring data trend abstraction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103617356B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105203346A (en) * | 2015-10-23 | 2015-12-30 | 珠海格力电器股份有限公司 | Fault diagnosis method and system for range hood based on EMD (Empirical Mode Decomposition) noise reduction |
| CN107730494A (en) * | 2017-10-24 | 2018-02-23 | 河海大学 | A kind of anchor pole detection method based on variation mode decomposition |
| CN110824304B (en) * | 2019-10-16 | 2022-03-18 | 福建和盛高科技产业有限公司 | Method for analyzing insulation degradation trend of zinc oxide arrester |
| CN111121955B (en) * | 2019-12-31 | 2022-02-18 | 华侨大学 | Vibration signal wavelet analysis trend term removing method adaptive to threshold frequency |
| CN113324646A (en) * | 2021-05-26 | 2021-08-31 | 兰州交通大学 | Big wind area electrified railway contact net positive feeder line galloping positioning algorithm |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102540021A (en) * | 2010-12-20 | 2012-07-04 | 上海市电力公司 | Method for monitoring GIS (gas insulated switchgear) by aid of high-frequency partial discharge online system |
-
2013
- 2013-11-27 CN CN201310612784.9A patent/CN103617356B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102540021A (en) * | 2010-12-20 | 2012-07-04 | 上海市电力公司 | Method for monitoring GIS (gas insulated switchgear) by aid of high-frequency partial discharge online system |
Non-Patent Citations (4)
| Title |
|---|
| Fault diagnosis using pattern classification based on one-dimensional adaptive rank-order morphological filter;Han Li，et al.;《Journal of Process Control》;20121231;436-449 * |
| 基于形态奇异值分解和经验模态分解的滚动轴承故障特征提取方法;汤宝平，等.;《机械工程学报》;20100331;第46卷(第5期);37-42,48 * |
| 电容型设备绝缘在线监测与故障诊断的研究;王楠.;《中国博士学位论文全文数据库 工程科技Ⅱ辑》;20050315(第01期);C042-38 * |
| 经验模分解在信号趋势项提取中的应用;陈隽，等.;《振动、测试与诊断》;20050630;第25卷(第2期);101-104,153 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103617356A (en) | 2014-03-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103617356B (en) | A kind of adaptive online monitoring data trend abstraction method | |
| Anderson et al. | Optimal buffer storage capacity in production | |
| CN101807925B (en) | Historical data compression method based on numerical ordering and linear fitting | |
| CN104091056B (en) | A kind of laboratory equipment life forecast system and method | |
| CN111371188A (en) | Real-time early warning system for abnormal power consumption data | |
| CN103346797A (en) | Real-time compression method for gear fault signals | |
| CN110824304A (en) | Method for analyzing insulation degradation trend of zinc oxide arrester | |
| CN106405280A (en) | Intelligent transformer station online monitoring parameter trend early warning method | |
| CN104408525A (en) | Quantitative evaluation and control method of job shop scheduling risks | |
| CN114184870A (en) | Non-invasive load identification method and equipment | |
| CN104467193A (en) | Four-division line loss quasi real-time on-line monitoring scheme based on daily frequentness monitoring | |
| CN102254100A (en) | Proportional hazard rate model method for estimating operation reliability of tool | |
| Visconti et al. | Comprehensive analysis of conservation voltage reduction: A real case study | |
| EP3987624B1 (en) | Method for determining a characteristic curve of a photovoltaic (pv) string, dc/dc converter, and photovoltaic system suitable for carrying out the method | |
| CN113866532A (en) | Method and device for quantizing maximum amplitude of disturbance voltage of in-situ intelligent measuring equipment | |
| Netsanet et al. | Short term load forecasting using wavelet augmented non-linear autoregressive neural networks: A single customer level perspective | |
| CN106249022A (en) | The analysis system and method that DC control and protection system is affected by exchange transient voltage | |
| Carmona et al. | Glasso model for electric load and wind power and monte carlo scenario generation | |
| Viet et al. | A method for monitoring voltage disturbances based on discrete wavelet transform and adaptive linear neural network | |
| Valsan et al. | Fault detection and classification logic for transmission lines using multi-resolution wavelet analysis | |
| Kumar et al. | Identification of power quality events using inverse properties of S transform | |
| CN105652108A (en) | Detection method of flexible direct current power transmission system and apparatus thereof | |
| CN111181172A (en) | Power grid frequency disturbance source positioning method for scheduling master station | |
| Naik et al. | Wavelet packet transform based parameter for the analysis of short duration power quality disturbances | |
| CN104199935A (en) | Large-volume data collecting method and device and analysis and search system for swimming pool |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |