CN105203150A - Data abnormal point systematic error detecting method for chemical device instrument - Google Patents
Data abnormal point systematic error detecting method for chemical device instrument Download PDFInfo
- Publication number
- CN105203150A CN105203150A CN201510579123.XA CN201510579123A CN105203150A CN 105203150 A CN105203150 A CN 105203150A CN 201510579123 A CN201510579123 A CN 201510579123A CN 105203150 A CN105203150 A CN 105203150A
- Authority
- CN
- China
- Prior art keywords
- data
- wavelet
- point
- abnormal point
- quartile
- 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.)
- Granted
Links
Abstract
The invention provides a data abnormal point systematic error detecting method for a chemical device instrument. According to the detecting method, a wavelet method and a boxplot method are combined, and chemical process instrument data within a certain period of time are selected to be used as sample data. The method includes the steps that first, wavelet transformation is conducted on the sample data, and wavelet transformation coefficients of the data under various dimensions are obtained according to the local characteristic of a time-frequency domain of the sample data; then abnormal value detection is carried out on the wavelet transformation coefficients under various dimensions through the boxplot method, the positions of abnormal point systematic error points of original instrument data are determined through the one-one mapping relation, and thus data abnormal point pattern mistake error detection of the chemical process instrument is achieved. According to the detecting method, the time-frequency domain local characteristic of the data is applied, the defect of misjudgment to abnormal points in the prior art is overcome, the applicability is higher, and dependence on data normal distribution in an existing method is avoided.
Description
Technical field
The present invention relates to the data processing method of chemical plant installations instrument, particularly relate to a kind of chemical plant installations instrumented data abnormal point-type lapse error detection method.
Background technology
Measurement data is the basis of chemical process being implemented to control and Optimum Operation accurately.In order to grasp the operation characteristic of chemical process, improving production operation, often will analyze the measurement data from certain unit or even complex process device.But generally, the measurement data of chemical process unavoidably exists various error, make measurement data accurately can not meet the physics and chemistry rule of process inherence, as stoichiometric relationship, mass balance and heat balance etc.Fast, reliable error detecting algorithm, all important in inhibitings are operated for theoretical analysis and actual production.
According to the difference of measuring error producing method, stochastic error and the large class of human error two can be divided into.The former produces by the impact of enchancement factor, and obeying certain statistical law, is inevitable.Latter may, the fault such as fluctuation of service, equipment leakage malfunctioning by measurement instrument cause.In general, the human error of dynamic system is divided into two classes: system deviation type human error and abnormal point-type human error.System deviation type human error refers to, within a period of time, measured value is some fixing amplitude more high or low than actual value always, may be caused by the factor such as instrument drift, process leaks; Abnormal point-type human error refers to there is the point of other data of substantial deviation in measurement data, such as, spike etc. in measuring-signal, may be caused by factors such as instrument measurements.
Mostly suppose instrumented data Normal Distribution to the detection method of the abnormal point-type human error of instrumented data at present, adopt 3 σ methods, the data namely outside Data distribution8 average positive and negative deviation 3 σ are defined as exceptional value.But actual instrument data generally do not meet normal distribution, so the abnormal point-type human error point accuracy that method detects is poor.Another kind of box figure method (boxplot) does not rely on the hypothesis of instrumented data Normal Distribution, and carry out abnormity point judgement by the quartile of Data distribution8, versatility is stronger.But owing to not considering the dependence between time upper adjacent data, erroneous judgement is easily occurred to the detection of abnormity point.Wavelet method, by data are transformed to frequency field from time domain, namely according to the change of instrumented data frequency spectrum, judges the change of instrumented data local.
Summary of the invention
For the problem of the existing detection method poor accuracy to the abnormal point-type human error of instrumented data, wavelet method combines with box figure method by the present invention, proposes a kind of chemical plant installations instrumented data abnormal point-type lapse error detection method.
The present invention adopts following technical scheme:
A kind of chemical plant installations instrumented data abnormal point-type lapse error detection method, comprising:
Step 1: collect sample data, a station server is configured in the pulpit of chemical enterprise, server is connected with the live database server of process units, live database server gathers the data from the instrument of production scene, a real time data is recorded every a sampling period, the data in continuous collection multiple sampling period, then form sample data, and this sample data is as the input of abnormal point-type lapse error detection method;
Step 2: choose suitable wavelet transform function, comprises Haar small echo, dbN wavelet systems, symN wavelet systems, Morlet is little involves Meyer small echo;
Step 3: sampling different frequency scale resolution carries out wavelet transformation to sample data, obtains the wavelet conversion coefficient under different scale, the different local circumstances in reflected sample data time domain;
Step 4: the wavelet conversion coefficient under each yardstick is analyzed, to become soon point in the corresponding time domain of radio-frequency component;
Step 5: to the wavelet conversion coefficient application box figure method of each yardstick obtained under wavelet transformation, wavelet conversion coefficient, as input data, calculates five statistics of input data, i.e. data minimum value, maximal value, median, first quartile Q
1and the 3rd quartile Q
2, wherein first quartile Q
1the median of the data between minimum value and median, the 3rd quartile Q
2it is the median of data between maximal value and median;
Step 6: the spacing I calculating quartile
qR=Q
3-Q
1, calculate the bound of the wavelet conversion coefficient of each yardstick, wherein the upper limit is defined as HIGH=Q
3+ m*I
qR, lower limit is defined as LOW=Q
1-n*I
qR, wherein, m and n is fixed coefficient;
Step 7: as data x > HIGH or x < LOW, x is abnormity point, and record moment corresponding to current x value;
Step 8: the moment that the abnormity point detected according to box figure method is corresponding, by the mapping relations one by one of data point in wavelet transformation time-frequency domain, determines the position of the abnormal point-type human error point of original instrumented data, realizes abnormal point-type lapse error detection.
The beneficial effect that the present invention has is:
A kind of chemical plant installations instrumented data of the present invention abnormal point-type lapse error detection method, wavelet method is combined with box figure method, choose chemical industry process instrument data in a period of time, first wavelet transformation is carried out to data, according to the local characteristics of data time-frequency domain, obtain the wavelet conversion coefficient of data under each yardstick, then by box figure method, rejecting outliers is carried out to the wavelet conversion coefficient under each yardstick, by the relation mapped one by one, determine the position of the abnormal point-type human error point of original instrumented data, realize the abnormal point-type lapse error detection of chemical process instrumented data, the time-frequency domain local characteristics of application data of the present invention, overcome the erroneous judgement of prior art to abnormity point, applicability is stronger, overcome the dependence to data normal distribution in existing method.
Embodiment
Below the present invention is specifically described:
A kind of chemical plant installations instrumented data abnormal point-type lapse error detection method, comprising:
Step 1: collect sample data, a station server is configured in the pulpit of chemical enterprise, server is connected with the live database server of process units, live database server gathers the data from the instrument of production scene, a real time data is recorded every a sampling period, the data in continuous collection multiple sampling period, then form sample data, and this sample data is as the input of abnormal point-type lapse error detection method;
Step 2: choose suitable wavelet transform function, comprises Haar small echo, dbN wavelet systems, symN wavelet systems, Morlet is little involves Meyer small echo etc.;
Step 3: sampling different frequency scale resolution carries out wavelet transformation to sample data, obtains the wavelet conversion coefficient under different scale, the different local circumstances in reflected sample data time domain, general employing level Four change of scale;
Step 4: the wavelet conversion coefficient under each yardstick is analyzed, radio-frequency component to become soon point often in corresponding time domain, as steep front, rear edge and spike pulse etc., is the abnormal point-type human error needing to detect;
Step 5: to the wavelet conversion coefficient application box figure method of each yardstick obtained under wavelet transformation, wavelet conversion coefficient, as input data, calculates five statistics of input data, i.e. data minimum value, maximal value, median, first quartile Q
1and the 3rd quartile Q
2, wherein first quartile Q
1the median of the data between minimum value and median, the 3rd quartile Q
2it is the median of data between maximal value and median;
Step 6: the spacing I calculating quartile
qR=Q
3-Q
1, calculate the bound of the wavelet conversion coefficient of each yardstick, wherein the upper limit is defined as HIGH=Q
3+ m*I
qR, lower limit is defined as LOW=Q
1-n*I
qR, wherein, m and n is fixed coefficient;
Step 7: as data x > HIGH or x < LOW, x is abnormity point, and record moment corresponding to current x value;
Step 8: the moment that the abnormity point detected according to box figure method is corresponding, by the mapping relations one by one of data point in wavelet transformation time-frequency domain, determines the position of the abnormal point-type human error point of original instrumented data, realizes abnormal point-type lapse error detection.
This chemical plant installations instrumented data abnormal point-type lapse error detection method, combines wavelet method with box figure method, realizes the abnormal point-type lapse error detection of chemical process instrumented data.The time-frequency domain local characteristics of application data of the present invention, overcomes the erroneous judgement of prior art to abnormity point, and applicability is stronger, overcomes the dependence to data normal distribution in existing method.
Certainly, above-mentioned explanation is not limitation of the present invention, and the present invention is also not limited in above-mentioned citing, and the change that those skilled in the art make in essential scope of the present invention, remodeling, interpolation or replacement also should belong to protection scope of the present invention.
Claims (1)
1. a chemical plant installations instrumented data abnormal point-type lapse error detection method, is characterized in that, comprising:
Step 1: collect sample data, a station server is configured in the pulpit of chemical enterprise, server is connected with the live database server of process units, live database server gathers the data from the instrument of production scene, a real time data is recorded every a sampling period, the data in continuous collection multiple sampling period, then form sample data, and this sample data is as the input of abnormal point-type lapse error detection method;
Step 2: choose suitable wavelet transform function, comprises Haar small echo, dbN wavelet systems, symN wavelet systems, Morlet is little involves Meyer small echo;
Step 3: sampling different frequency scale resolution carries out wavelet transformation to sample data, obtains the wavelet conversion coefficient under different scale, the different local circumstances in reflected sample data time domain;
Step 4: the wavelet conversion coefficient under each yardstick is analyzed, to become soon point in the corresponding time domain of radio-frequency component;
Step 5: to the wavelet conversion coefficient application box figure method of each yardstick obtained under wavelet transformation, wavelet conversion coefficient, as input data, calculates five statistics of input data, i.e. data minimum value, maximal value, median, first quartile Q
1and the 3rd quartile Q
2, wherein first quartile Q
1the median of the data between minimum value and median, the 3rd quartile Q
2it is the median of data between maximal value and median;
Step 6: the spacing I calculating quartile
qR=Q
3-Q
1, calculate the bound of the wavelet conversion coefficient of each yardstick, wherein the upper limit is defined as HIGH=Q
3+ m*I
qR, lower limit is defined as LOW=Q
1-n*I
qR, wherein, m and n is fixed coefficient;
Step 7: as data x > HIGH or x < LOW, x is abnormity point, and record moment corresponding to current x value;
Step 8: the moment that the abnormity point detected according to box figure method is corresponding, by the mapping relations one by one of data point in wavelet transformation time-frequency domain, determines the position of the abnormal point-type human error point of original instrumented data, realizes abnormal point-type lapse error detection.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510579123.XA CN105203150B (en) | 2015-09-11 | 2015-09-11 | A kind of chemical plant installations instrumented data exception point-type lapse error detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510579123.XA CN105203150B (en) | 2015-09-11 | 2015-09-11 | A kind of chemical plant installations instrumented data exception point-type lapse error detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105203150A true CN105203150A (en) | 2015-12-30 |
| CN105203150B CN105203150B (en) | 2017-11-10 |
Family
ID=54950957
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510579123.XA Active CN105203150B (en) | 2015-09-11 | 2015-09-11 | A kind of chemical plant installations instrumented data exception point-type lapse error detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105203150B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763346A (en) * | 2018-05-15 | 2018-11-06 | 中南大学 | A kind of abnormal point processing method of sliding window box figure medium filtering |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102288286A (en) * | 2011-06-16 | 2011-12-21 | 中国科学院沈阳自动化研究所 | Method for analyzing and evaluating measure point precision of gearbox in vibration acceleration sensor |
| CN102488516A (en) * | 2011-12-13 | 2012-06-13 | 湖州康普医疗器械科技有限公司 | Nonlinear electroencephalogram signal analysis method and device |
| US20140078163A1 (en) * | 2012-09-14 | 2014-03-20 | Software Ag | Systems and/or methods for statistical online analysis of large and potentially heterogeneous data sets |
| CN104316317A (en) * | 2014-10-08 | 2015-01-28 | 西北工业大学 | Gear system multi-fault diagnosis method based on COM assemblies |
-
2015
- 2015-09-11 CN CN201510579123.XA patent/CN105203150B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102288286A (en) * | 2011-06-16 | 2011-12-21 | 中国科学院沈阳自动化研究所 | Method for analyzing and evaluating measure point precision of gearbox in vibration acceleration sensor |
| CN102488516A (en) * | 2011-12-13 | 2012-06-13 | 湖州康普医疗器械科技有限公司 | Nonlinear electroencephalogram signal analysis method and device |
| US20140078163A1 (en) * | 2012-09-14 | 2014-03-20 | Software Ag | Systems and/or methods for statistical online analysis of large and potentially heterogeneous data sets |
| CN104316317A (en) * | 2014-10-08 | 2015-01-28 | 西北工业大学 | Gear system multi-fault diagnosis method based on COM assemblies |
Non-Patent Citations (1)
| Title |
|---|
| 尹杨: ""基于小波支持向量机的发动机外部管路振动故障诊断研究"", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763346A (en) * | 2018-05-15 | 2018-11-06 | 中南大学 | A kind of abnormal point processing method of sliding window box figure medium filtering |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105203150B (en) | 2017-11-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110927521B (en) | Single-ended traveling wave fault positioning method and device | |
| JP6467186B2 (en) | System and method for detecting, correcting, and inspecting bad data in a data stream | |
| CN108593260B (en) | Optical cable line fault positioning and detecting method and terminal equipment | |
| CN104620119B (en) | The apparatus and method for monitoring power network | |
| JP2018528433A (en) | Remote sensing to derive calibrated power measurement results | |
| KR101300047B1 (en) | Partial discharging signal detector of power system | |
| CN109375060B (en) | Method for calculating fault waveform similarity of power distribution network | |
| CN109034400B (en) | Transformer substation abnormal measurement data prediction platform system | |
| CN106537443B (en) | System and method for classifying in situ sensor response data patterns indicative of grid anomaly severity | |
| CN101566663B (en) | Method for positioning voltage drop source of power distribution system | |
| US9960862B2 (en) | Method and device for detecting standing-wave ratio | |
| CN107765206B (en) | Method for evaluating running state of power quality monitoring device | |
| CN103983881A (en) | Fault diagnosis method and device for nuclear detector | |
| CN106855990B (en) | Nuclear power unit instrument channel measurement error demonstration method | |
| US20220156586A1 (en) | System for monitoring a circuit breaker | |
| Adu-Gyamfi et al. | Reliability of probe speed data for detecting congestion trends | |
| CN104569886A (en) | Calibrating method for signal detection equipment based on time-frequency parameter standard instrument | |
| CN112379213B (en) | Fault detection method and system | |
| CN105203150A (en) | Data abnormal point systematic error detecting method for chemical device instrument | |
| CN111693828B (en) | Transformer substation space partial discharge positioning system and method based on MLE | |
| CN109873822B (en) | Device and method for detecting firewall rule change based on Beidou subnanosecond high-precision time service | |
| CN114355114B (en) | GIS partial discharge detection method, device, medium and terminal equipment | |
| CN116430181A (en) | GIS external partial discharge identification method, device, equipment and medium | |
| WO2011089421A1 (en) | Apparatus and method for measuring a phasor of an electrical power system signal | |
| CN102937676A (en) | Method and system for implementing early warning of load harmonic injection pollution |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Wang Chunli Inventor after: Li Chuankun Inventor after: Gao Xinjiang Inventor before: Zhu Jianfeng Inventor before: Wang Chunli Inventor before: Li Chuankun Inventor before: Gao Xinjiang |
|
| CB03 | Change of inventor or designer information | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20170927 Address after: Yanan City, Shandong province Qingdao City three road 266071 No. 218 Applicant after: Qingdao Safety Engineering Research Institute of Sinopec Co., Ltd. Applicant after: Sinopec Corp. Address before: 100728 Chaoyangmen street, Beijing, No. 22, No. Applicant before: Sinopec Corp. Applicant before: Qingdao Safety Engineering Research Institute of Sinopec Co., Ltd. |
|
| TA01 | Transfer of patent application right | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |