CN113408383A - Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA - Google Patents
Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA Download PDFInfo
- Publication number
- CN113408383A CN113408383A CN202110638740.8A CN202110638740A CN113408383A CN 113408383 A CN113408383 A CN 113408383A CN 202110638740 A CN202110638740 A CN 202110638740A CN 113408383 A CN113408383 A CN 113408383A
- Authority
- CN
- China
- Prior art keywords
- data
- audible noise
- bounded
- distribution
- mwkpca
- 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
- 238000009826 distribution Methods 0.000 title claims abstract description 46
- 238000000034 method Methods 0.000 title claims abstract description 43
- 230000008859 change Effects 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 4
- 230000005540 biological transmission Effects 0.000 claims description 36
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 230000008030 elimination Effects 0.000 claims description 8
- 238000003379 elimination reaction Methods 0.000 claims description 8
- 230000003203 everyday effect Effects 0.000 claims description 3
- 238000007405 data analysis Methods 0.000 abstract description 6
- 238000012549 training Methods 0.000 description 9
- 230000002354 daily effect Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000011160 research Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 230000005236 sound signal Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 150000002500 ions Chemical class 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 235000018185 Betula X alpestris Nutrition 0.000 description 1
- 235000018212 Betula X uliginosa Nutrition 0.000 description 1
- 244000274847 Betula papyrifera Species 0.000 description 1
- 235000009113 Betula papyrifera Nutrition 0.000 description 1
- 235000009109 Betula pendula Nutrition 0.000 description 1
- 235000010928 Betula populifolia Nutrition 0.000 description 1
- 235000002992 Betula pubescens Nutrition 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 239000012212 insulator Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Signal Processing (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The audible noise invalid data judgment method based on the bounded beta (g, h) distribution and MWKPCA judges the invalid data of the collected audible noise data day by day through the bounded beta (g, h) distribution, and eliminates the invalid data outside the bounded beta (g, h) distribution judgment standard range; for processed audible noise data, T of MWKPCA principal element subspace is used2Judging invalid data in the audible noise data set day by day according to the statistic and the SPE statistic of the residual error subspace; respectively solving L95, L50 and L5 percentiles of a data set which is not subjected to invalid data rejection, a data set which is bounded and distributed by beta (g, h) and a data set which is subjected to invalid data rejection by MWKPCA, and determining the change condition of the percentile before and after data processing, thereby determining whether the rejection of invalid data influences the subsequent audible noise data analysis. The judging method of the invention is more comprehensive and simultaneously movesThe data is segmented in a window form, so that the influence of the change of meteorological factors on the judgment result is reduced.
Description
Technical Field
The invention relates to the technical field of power transmission line audible noise evaluation, in particular to an audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA.
Background
The audible noise of the power transmission line is one of design standards of the power transmission line, and influences the wire selection, the corridor width, the insulator string length and the wire arrangement mode of the power transmission line. However, in the process of collecting the audible noise of the power transmission line, a large amount of background noise exists, and is very easily interfered by the background noise, and if the audible noise of the power transmission line is smaller than the background noise, the background noise has a high probability to become invalid data or abnormal values in the data set, namely invalid data. Invalid data may affect the evaluation of the transmission line.
The research types of the audible noise of the existing power transmission line are divided into the research of an alternating current power transmission line and the research of a direct current power transmission line, wherein the audible noise of the alternating current power transmission line can be divided into pure sound with the frequency being a multiple of 50Hz and broadband noise with the frequency range being concentrated in a range of 400Hz to 10kHz, positive ions in a positive polarity interval move outwards along a radial direction under the action of an alternating electric field to apply pressure to an air layer, negative ions formed by electrons attached to molecules in a negative polarity interval move outwards along the radial direction to apply pressure to the air layer, and therefore sound waves are transmitted outwards periodically.
The existing research mode of the audible noise effective data of the power transmission line comprises the steps of converting 8kHz components of sound into A sound level through an empirical formula and a machine learning algorithm so as to indirectly obtain the audible noise effective data of the power transmission line; and removing background noise in the collected sound signals through wavelet decomposition and a limited-length unit impact response filter, and directly obtaining effective data of audible noise of the power transmission line.
The above methods for removing background noise all directly process collected sound signals, and in most cases, process invalid data in a data set, and do not consider that sound signals collected at a certain sampling time are a result of the combined action of noises belonging to different frequency bands, and the sound emitted by a background noise sound source is at different frequencies, so a method capable of determining invalid data and considering noise of each frequency band is urgently needed.
Disclosure of Invention
In order to solve the technical problem, the invention provides an audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA, which utilizes the bounded beta (g, h) distribution to sequentially judge 10 audible noise component invalid data of 16 Hz-8 kHz of octave; meanwhile, an MWKPCA method is utilized to carry out overall invalid data judgment on audible noise of 10 frequency bands of 16Hz to 8kHz of octave. Compared with the traditional method for judging invalid data by using single-dimensional data, the judging method has more comprehensive judging standard, and meanwhile, the data is segmented in a moving window mode, so that the influence of the change of meteorological factors on the judging result is reduced.
The technical scheme adopted by the invention is as follows:
a method for audible noise inefficiency (audi) data decision based on bounded beta (g, h) distribution and MWKPCA, comprising the steps of:
step 1: carrying out invalid data judgment on the collected audible noise data through a bounded beta (g, h) distribution, and rejecting invalid data outside a bounded beta (g, h) distribution judgment standard range;
step 2: regarding the audible noise data processed in the step 1, the T of the MWKPCA principal element subspace is used as the whole audible noise component of each frequency band2The SPE statistic of the statistic and the residual error subspace is used for judging invalid data in the audible noise data set;
and step 3: respectively solving L95, L50 and L5 percentiles of a data set which is not subjected to invalid data rejection, a data set which is bounded and distributed by beta (g, h) and a data set which is subjected to invalid data rejection by MWKPCA, and determining the change condition of the percentile before and after data processing, thereby determining whether the rejection of invalid data influences the subsequent audible noise data analysis.
In the step 1, the collected audible noise data is decomposed into 10 audible noise components of 16Hz, 31.5Hz, 63Hz, 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, 8000Hz, which are used as a data set for abnormal data rejection, the audible noise data of each frequency band is distributed by using bounded beta (g, h), and invalid data of each frequency band is rejected day by day.
In the step 1, the audible noise data is audible noise data of the alternating current transmission line, and the decomposed audible noise components of 10 frequency bands including 16Hz to 8kHz are divided in an octave form.
In the step 2, the invalid data in the audible noise data set is determined in the form of a moving window, wherein the moving window is in the form of: and when data are collected, the audible noise data quantity of the transmission line every day is used as the moving step length of the moving window function.
In the step 3, the percentiles L95, L50 and L5 refer to numerical values arranged in the first 95%, 50% and 5% positions when data are arranged from large to small.
The invention discloses an audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA, which has the following technical effects:
1) the method has good running speed and accuracy when invalid data of each frequency band are removed in sequence through bounded beta (g, h) distribution, meanwhile, the method is used for judging the invalid data without the data conforming to normal distribution, the method has wider practicability, and the bounded beta (g, h) distribution is used for removing the invalid data before MWKPCA, so that the subsequent MWKPCA can obtain more accurate training data.
2) When the MWKPCA is used for judging invalid data in the data set, the elimination of abnormal data has a good effect no matter whether the characteristics in the data set are in a linear relation or a nonlinear relation. And the audible noise of the power transmission line is long in acquisition time span and large in meteorological factor change, and the judgment standard can be gradually updated in the day-by-day abnormal data judgment process by using a mobile window form, so that the negative influence on the judgment result caused by the change of the meteorological factor is prevented.
3) The method respectively eliminates the data which are not subjected to invalid data elimination, adopts bounded beta (g, h) distribution to eliminate abnormal data, adopts MWKPCA to eliminate the data sets of three stages after the abnormal data, and compares L5, L50 and L95 percentiles of the data sets to determine that the elimination of the invalid data does not influence the subsequent data analysis. And after the invalid data are removed, the accuracy of the prediction result is found to be improved by using a LightGBM algorithm to predict the audible noise A level of the electric transmission line by using 8000Hz components of the audible noise and meteorological factors, after the invalid data are removed from the data set by using the method disclosed by the invention, the RMSE error of the prediction result is reduced to 2.68 from 3.20, and the MAE error is reduced to 1.85 from 2.17.
4) The invention is innovated in detection means, adopts a day-by-day judgment method, and reduces the problems of repeated work and non-uniform judgment standards when the judgment is carried out manually. The bounded beta (g, h) distribution is utilized to sequentially judge 10 audible noise component invalid data of 16 Hz-8 kHz of octave, and the problem that the data are required to obey normal distribution in the traditional invalid data judging method is solved. The MWKPCA is used for carrying out overall invalid data judgment on audible noise of 10 frequency bands of 16Hz to 8kHz of octave, compared with the traditional single-dimensional data, the method has the advantages that the judgment standard is more comprehensive, the influence of the change of meteorological factors on the judgment result is reduced by segmenting the data in a moving window mode, and a new way is provided for extracting the effective data of the audible noise of the power transmission line.
Drawings
Fig. 1 is a first diagram of noise invalid data of a power transmission line according to the present invention;
fig. 2 is a second diagram of the noise invalid data of the transmission line according to the present invention.
FIG. 3 is a first diagram illustrating the effect of the bounded β (g, h) distribution culling on abnormal data according to the present invention;
FIG. 4 is a diagram illustrating the effect of the bounded β (g, h) distribution culling on abnormal data according to the present invention.
FIG. 5 is T of MWKPCA method proposed by the present invention2Statistical audible noise valid data decision graph.
Fig. 6 is a diagram of SPE statistic audible noise effective data determination in the MWKPCA method proposed by the present invention.
Detailed Description
A method for audible noise inefficiency (audi) data decision based on bounded beta (g, h) distribution and MWKPCA, comprising the steps of:
step 1: carrying out invalid data judgment on the collected audible noise data day by day through a bounded beta (g, h) distribution, and rejecting invalid data which is out of a bounded beta (g, h) distribution judgment standard range; invalid data is also referred to as "outliers".
Step 2: regarding the audible noise data processed in the step 1, the T of the MWKPCA principal element subspace is used as the whole audible noise component of each frequency band2The SPE statistics of the statistics and the residual error subspace judge invalid data in the audible noise data set in a moving window mode;
and step 3: respectively solving L95, L50 and L5 percentiles of a data set which is not subjected to invalid data rejection, a data set which is bounded and distributed by beta (g, h) and a data set which is subjected to invalid data rejection by MWKPCA, and determining the change condition of the percentile before and after data processing, thereby determining whether the rejection of invalid data influences the subsequent audible noise data analysis.
The bounded β (g, h) distribution is described in references [1], [2], and [3] of the prior art:
document [1 ]: forest white birch, beta distribution statistical method of reporters error [ J ], proceedings of the chinese institute of metrology, 2004(02):10-15, document [2 ]: ringhong birch, Panfeng, self-service method of repeated measurement data beta distribution estimation [ J ]. university of Beijing university of science and engineering, 2004(11): 947-.
Document [3 ]: seedui, Chihang, Wuxiong, etc. statistical analysis of radio interference under good weather of 1000kV ultra-high voltage alternating current single-circuit transmission line [ J ] China Motor engineering report, 2016,36(03): 861) 870.
The bounded beta (g, h) distribution is a method for expressing the data distribution rule, and the (g, h) parameter of the data set to be detected is estimated by utilizing the polymorphism of the beta (g, h), and the distribution interval of the data is further obtained by the following process
And further eliminating invalid data.
Wherein,
the mean value of u is represented by,
the variance of u is represented by the variance of u,
is the mean of the samples and σ is the standard deviation of the samples.
MWKPCA is the kernel principal component analysis of moving window, MW (moving Window) moving window, which is a way to segment data set, the moving window in the invention segments data in daily unit, then segments data in daily unitData were analyzed by chronological order in days. KPCA (Kernal Principal Component analysis) kernel Principal Component analysis maps daily data to be detected in the invention to a high-dimensional feature space through a kernel function, centralizes the data in the high-dimensional feature space, and then obtains a feature vector and a feature value of a data matrix, thereby further calculating T of a Principal element subspace in the Principal element subspace2Statistical sum T2A statistic threshold, and SPE statistics for the residual subspace and SPE statistics threshold for the residual subspace.
MWKPCA refers to the description in the prior art documents [4], [5] and [6 ]:
document [4 ]: zhuxiong, Zhan Han, Yang Chun festival MWPCA blast furnace abnormity monitoring algorithm [ J ] based on Gaussian mixture model, chemical science and newspaper, 2021,72(03): 1539-supplement 1548.
Document [5 ]: li Zhenhua, Li Chun Yan, Zhang Zhu electronic voltage transformer error state prediction based on Q-ARMA (J) Chinese science: technical science, 2018,48(12): 1401-plus 1412.
Document [6 ]: zhang Zhu, Chen Qin, Hu Chen, et al, evaluating the metric of electronic transducers on-line based on VN-MWPCA [ J ]. Measurement,2018,130:1-7.
Training data: selecting n groups of data of which each noise component is in a plurality of value intervals with the largest data quantity as initial training data, updating the training data in the process of carrying out invalid data judgment day by day, adding the data judged to be normal on the current day into the training data, and removing the data of corresponding groups from the previous training data so as to form the training data of the next day.
Invalid data: the invalid data in the invention is divided into two parts: the first part is invalid data in the single-dimensional data, i.e. a point in a certain period of time whose value is far greater or much smaller than the average value of the piece of data. The second part is invalid data in the multidimensional data, namely points which are far away from the center of the data cluster in the multidimensional space in which the multidimensional data is positioned within a certain time period. The invalid data in the single-dimensional data or the multi-dimensional data mostly result from errors in recording and calculation or accidental changes of test methods and test environments. The accuracy of subsequent research is affected.
The invention means that: and segmenting the data by taking the day as a unit, and judging the invalid data by taking the day as a unit in the invalid data judging process.
In the step 1, the collected audible noise data is decomposed into 10 audible noise components of 16Hz, 31.5Hz, 63Hz, 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, 8000Hz, which are used as a data set for abnormal data rejection, the audible noise data of each frequency band is distributed by using bounded beta (g, h), and invalid data of each frequency band is rejected day by day.
In the step 1, the audible noise data is audible noise data of the alternating current transmission line, and the decomposed audible noise components of 10 frequency bands including 16Hz to 8kHz are divided in an octave form.
In the step 2, the invalid data in the audible noise data set is determined in the form of a moving window, wherein the moving window is in the form of: and when data are collected, the audible noise data quantity of the transmission line every day is used as the moving step length of the moving window function.
In step 2, each frequency band of the audible noise is an octave, and the frequency center includes: 16Hz, 31.5Hz, 63Hz, 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, 8000 Hz.
T2Statistics, SPE statistics contains four parts: t is2Statistical threshold, SPE statistical threshold, T of each sample point2The statistical value and the SPE statistical value of each sampling point; for exceeding T2And (4) eliminating the data with the statistical threshold or the data exceeding the SPE statistical threshold. Wherein:
T2the statistic threshold is obtained from the training data of the daily data, T2The statistical value is obtained from the test data of daily data;
the SPE statistic threshold is obtained by training data of daily data, and the SPE statistic value is obtained by testing data of the daily data.
In the step 3, the percentiles L95, L50 and L5 refer to numerical values arranged in the first 95%, 50% and 5% positions when data are arranged from large to small.
Verification of the examples:
in order to further analyze the judgment of the audible noise invalid data of the power transmission line based on the bounded beta (g, h) distribution and MWKPCA, the invalid data of 1000kV alternating current power transmission line noise data collected at the night (0 point to 6 points) of 9-25-2016-2-16-2015 is eliminated.
Firstly, bounded beta (g, h) distribution is used, audible noise invalid data of the power transmission line with 10 frequency bands from 16Hz to 8kHz are removed day by day in sequence, then the audible noise of the power transmission line with 10 frequency bands from 16Hz to 8kHz is taken as a whole, MWKPCA is used for judging the integral invalid data, data sets without invalid data removal are respectively obtained, the bounded beta (g, h) distribution is used for carrying out invalid data removal, the MWKPCA is used for carrying out invalid data removal, L95, L50 and L5 percentiles of the three-stage data sets are used for judging whether the removing method of the invalid data used by the invention can influence the removing of the audible noise invalid data of the subsequent power transmission line, and the result proves that: the method provided by the invention can not cause interference on the data analysis of audible noise of the subsequent power transmission line on the basis of eliminating invalid data.
Fig. 1 is a graph of sound level data at 30 a days 10 months 10 years 2015.
Fig. 2 is a graph of sound level data for 11/1/2015 day a.
The dotted black circles in fig. 1 and 2 indicate invalid data for invalid interpretation of sound level audible noise data on 1000kV ac transmission line a at 10/30 th 2015 and at 11/1 st 2015.
Fig. 3 is a graph of the trend of sound level data at 30 days a 10 months 10 years 2015 after invalid data culling using bounded β (g, h) distributions day by day as proposed by the present invention.
Fig. 4 is a graph of the trend of sound level data at 11/1/2015 after invalid data culling using a bounded β (g, h) distribution on a daily basis as proposed by the present invention.
Compared with the traditional invalid data eliminating method which needs the data set to obey normal distribution, the bounded beta (g, h) distribution used by the method can also judge the invalid data in the data set without the data set obeying normal distribution.
FIG. 5 is a diagram of T using the MWKPCA principal element subspace proposed by the present invention2And the statistic value is the judgment result of judging the overall invalid data of the audible noise of the power transmission line with 10 frequency bands from 16Hz to 8 kHz.
Fig. 6 is a result of determining the total invalid data of the audible noise of the power transmission line in 10 frequency bands of 16Hz to 8kHz by using the SPE statistic of the MWKPCA residual subspace provided by the present invention.
Table 1 shows the L95, L50, L5 percentiles of three-stage datasets for datasets without invalid data culling, datasets with bounded β (g, h) distribution for invalid data culling, and datasets with MWKPCA for invalid data culling.
TABLE 1 original data set, bounded beta (g, h) distribution abnormal data elimination, MWKPCA abnormal data elimination, three stages of noise percentile of each frequency band
Through comparison of L95, L50 and L5 percentiles of the three stages, the percentile difference value of audible noise data of each frequency band in the three invalid data eliminating stages is not more than 2dB, so that the method for judging the audible noise invalid data of the power transmission line has no influence on subsequent data analysis.
Claims (5)
1. A method for determining audible noise-null data based on a bounded β (g, h) distribution and MWKPCA, comprising the steps of:
step 1: carrying out invalid data judgment on the collected audible noise data through a bounded beta (g, h) distribution, and rejecting invalid data outside a bounded beta (g, h) distribution judgment standard range;
step 2: regarding the audible noise data processed in the step 1, the T of the MWKPCA principal element subspace is used as the whole audible noise component of each frequency band2Statistics and residual subSpatial SPE statistics for determining invalid data in the audible noise data set;
and step 3: and respectively solving L95, L50 and L5 percentiles of the data sets which are not subjected to invalid data elimination, the data sets which are bounded by beta (g, h) and subjected to invalid data elimination, and the data sets which are subjected to invalid data elimination by MWKPCA, and determining the change condition of the percentiles before and after data processing.
2. The method of claim 1 for audible noise null data decision based on bounded β (g, h) distribution and MWKPCA, characterized by: in the step 1, the collected audible noise data is decomposed into 10 audible noise components of 16Hz, 31.5Hz, 63Hz, 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, 8000Hz, which are used as a data set for abnormal data rejection, the audible noise data of each frequency band is distributed by using bounded beta (g, h), and invalid data of each frequency band is rejected day by day.
3. The method of claim 1 for audible noise null data decision based on bounded β (g, h) distribution and MWKPCA, characterized by: in the step 1, the audible noise data is audible noise data of the alternating current transmission line, and the decomposed audible noise components of 10 frequency bands including 16Hz to 8kHz are divided in an octave form.
4. The method of claim 1 for audible noise null data decision based on bounded β (g, h) distribution and MWKPCA, characterized by: in the step 2, the invalid data in the audible noise data set is determined in the form of a moving window, wherein the moving window is in the form of: and when data are collected, the audible noise data quantity of the transmission line every day is used as the moving step length of the moving window function.
5. The method of claim 1 for audible noise null data decision based on bounded β (g, h) distribution and MWKPCA, characterized by: in the step 3, the percentiles L95, L50 and L5 refer to numerical values arranged in the first 95%, 50% and 5% positions when data are arranged from large to small.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110638740.8A CN113408383B (en) | 2021-06-08 | 2021-06-08 | Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110638740.8A CN113408383B (en) | 2021-06-08 | 2021-06-08 | Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113408383A true CN113408383A (en) | 2021-09-17 |
| CN113408383B CN113408383B (en) | 2022-10-21 |
Family
ID=77677233
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110638740.8A Active CN113408383B (en) | 2021-06-08 | 2021-06-08 | Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113408383B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114034375A (en) * | 2021-10-26 | 2022-02-11 | 三峡大学 | System and method for measuring noise of ultra-high voltage transmission line |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104634443A (en) * | 2013-11-11 | 2015-05-20 | 国家电网公司 | Method for processing corona audible noise of high-voltage direct-current power transmission line under high background noise |
| CN106776480A (en) * | 2015-11-25 | 2017-05-31 | 中国电力科学研究院 | A kind of elimination method of radio interference in-site measurement exceptional value |
| CN111505449A (en) * | 2020-03-24 | 2020-08-07 | 中国电力科学研究院有限公司 | Method and system for judging direct-current corona-starting field intensity of split conductor |
-
2021
- 2021-06-08 CN CN202110638740.8A patent/CN113408383B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104634443A (en) * | 2013-11-11 | 2015-05-20 | 国家电网公司 | Method for processing corona audible noise of high-voltage direct-current power transmission line under high background noise |
| CN106776480A (en) * | 2015-11-25 | 2017-05-31 | 中国电力科学研究院 | A kind of elimination method of radio interference in-site measurement exceptional value |
| CN111505449A (en) * | 2020-03-24 | 2020-08-07 | 中国电力科学研究院有限公司 | Method and system for judging direct-current corona-starting field intensity of split conductor |
Non-Patent Citations (2)
| Title |
|---|
| ZHANG ZHU等: "Evaluating the meteringerror of electronic transformers on-line based on VN-MWPCA", 《MEASUREMENT》 * |
| 朱雄卓等: "基于高斯混合模型的MWPCA高炉异常监测算法", 《化工学报》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114034375A (en) * | 2021-10-26 | 2022-02-11 | 三峡大学 | System and method for measuring noise of ultra-high voltage transmission line |
| CN114034375B (en) * | 2021-10-26 | 2024-06-11 | 三峡大学 | Ultra-high voltage transmission line noise measurement system and method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113408383B (en) | 2022-10-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110895526A (en) | Method for correcting data abnormity in atmosphere monitoring system | |
| CN111562108A (en) | Rolling bearing intelligent fault diagnosis method based on CNN and FCMC | |
| EP0909390A2 (en) | Statistical pattern analysis methods of partial discharge measurments in high voltage insulation | |
| CN109633368A (en) | The method of duration power quality disturbances containing distributed power distribution network based on VMD and DFA | |
| CN115982602B (en) | Photovoltaic transformer electrical fault detection method | |
| CN116865269B (en) | Wind turbine generator system high harmonic compensation method and system | |
| CN117290802B (en) | Host power supply operation monitoring method based on data processing | |
| CN110704987B (en) | Bearing abnormal state assessment method based on similar working condition of failure data mining | |
| CN112380992B (en) | Method and device for evaluating and optimizing accuracy of monitoring data in machining process | |
| CN112207631B (en) | Method for generating tool detection model, method, system, device and medium for detecting tool detection model | |
| CN111291918B (en) | Rotating machine degradation trend prediction method based on stationary subspace exogenous vector autoregression | |
| CN113918642A (en) | Data filtering, monitoring and early warning method based on power Internet of things equipment | |
| CN110930057A (en) | Quantitative evaluation method for reliability of distribution transformer test result based on LOF algorithm | |
| CN113408383B (en) | Audible noise invalid data judgment method based on bounded beta (g, h) distribution and MWKPCA | |
| CN116522268A (en) | Line loss anomaly identification method for power distribution network | |
| CN115392284A (en) | Site micro-vibration source identification method based on machine learning | |
| CN114397569A (en) | Circuit breaker fault arc detection method based on VMD parameter optimization and sample entropy | |
| CN116307938B (en) | Health state assessment method for feeding system of machining center | |
| CN109840386B (en) | Damage identification method based on factor analysis | |
| CN111339986A (en) | Frequency law mining method and system for equipment based on time domain/frequency domain analysis | |
| CN117172601A (en) | Non-invasive load monitoring method based on residual total convolution neural network | |
| CN117216528A (en) | Bridge performance early warning method based on environment multi-factor-frequency enhancement mapping model | |
| CN116773961A (en) | Transmission line corrosion detection method based on vibration signal high-frequency characteristic analysis | |
| CN113537156B (en) | Vibration data cleaning method based on interval standard deviation and spectrum analysis | |
| CN114090949A (en) | Robust vibration signal characteristic value calculation method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |