CN110530876B - Insulator pollution degree development prediction method based on long-term and short-term memory neural network - Google Patents
Insulator pollution degree development prediction method based on long-term and short-term memory neural network Download PDFInfo
- Publication number
- CN110530876B CN110530876B CN201910833968.5A CN201910833968A CN110530876B CN 110530876 B CN110530876 B CN 110530876B CN 201910833968 A CN201910833968 A CN 201910833968A CN 110530876 B CN110530876 B CN 110530876B
- Authority
- CN
- China
- Prior art keywords
- insulator
- hyperspectral
- pollution
- pollution degree
- degree
- 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
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/84—Systems specially adapted for particular applications
- G01N21/88—Investigating the presence of flaws or contamination
- G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
- G01N2021/8887—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges based on image processing techniques
Abstract
The invention discloses a method for predicting the development of pollution degree of an insulator based on a long-term and short-term memory neural network, which comprises the steps of obtaining a hyperspectral image shot in a local area of the insulator and extracting a hyperspectral spectral line of the hyperspectral image; inputting the hyperspectral spectral line of the insulator into a constructed insulator pollution degree hyperspectral regression model of the same pollution type to obtain the pollution degree of the insulator; and inputting the pollution degree and the prediction time sequence of the insulator into a pollution degree hyperspectral development prediction model of the same pollution type, which is constructed by adopting a long-term and short-term memory neural network algorithm, so as to obtain the predicted pollution degree of the insulator required by the engineering. According to the scheme, the pollution development of the insulator can be accurately predicted through the combination of the pollution degree hyperspectral development prediction model and the hyperspectral image, so that the problems that the operation is complicated, the manual interference is large, the future development of the pollution degree is difficult to predict and the like in the prior art are solved.
Description
Technical Field
The invention belongs to the technical field of maintenance of operation states of power transmission and transformation equipment, and particularly relates to a method for predicting development of pollution degree of an insulator based on a long-term and short-term memory neural network.
Background
The insulator of the power transmission line is exposed in the atmospheric environment for a long time, and the surface of the insulator is gradually polluted. With the rapid development of industry and agriculture, the insulator of the power transmission line faces more serious pollution accumulation problem due to the improvement of the voltage grade. The electrical performance of insulator dirt is reduced after the insulator dirt is affected with damp in humid weather conditions such as fog, dew and rain, and flashover accidents are easy to happen. Once pollution flashover happens, a large-area power failure accident is caused, the overhaul recovery time is long, inconvenience is brought to people's life and social production, and huge economic loss and potential safety hazards are brought, so that the cleaning plan is made in advance by accurately predicting the pollution degree of the insulator, and the important significance is achieved for reducing the pollution flashover accident.
At present, inspection of the pollution degree of a line insulator mainly depends on manpower, an electric worker is required to climb a pole to disassemble the insulator, salt density is measured after cleaning, and ash density is measured after filtering, drying and weighing, so that the current pollution degree is obtained, the process is complicated, random errors are easily caused depending on the manpower, and the pollution development state at the future time cannot be predicted. China is wide in regions and large in regional difference, a line insulator cleaning strategy is difficult to make according to the experience of workers and the traditional standard, and the purpose of preventing and treating pollution flashover cannot be achieved under the most economic condition.
In summary, the existing insulator pollution degree detection method has certain condition limitations, and in view of this, it is necessary to provide a method capable of accurately predicting the development of insulator pollution degree based on time series.
Disclosure of Invention
Aiming at the defects in the prior art, the insulator pollution degree development prediction method based on the long-term and short-term memory neural network solves the problem that the difficulty in detecting the insulator pollution degree is high in the prior art.
In order to achieve the purpose of the invention, the invention adopts the technical scheme that:
acquiring a hyperspectral image shot in a local area of the insulator, and extracting a hyperspectral spectral line of the hyperspectral image;
inputting the hyperspectral spectral line of the insulator into a constructed insulator pollution degree hyperspectral regression model of the same pollution type to obtain the pollution degree of the insulator;
and inputting the pollution degree and the prediction time sequence of the insulator into a pollution degree hyperspectral development prediction model of the same pollution type, which is constructed by adopting a long-term and short-term memory neural network algorithm, so as to obtain the predicted pollution degree of the insulator required by the engineering.
Further, the construction method of the pollution degree hyperspectral development prediction model comprises the following steps:
a1, in a pollution area of the same pollution type, regularly shooting hyperspectral images in local areas of m insulators at fixed points for n times, and extracting hyperspectral spectral lines of each hyperspectral image;
a2, after the insulator finishes the nth hyperspectral image shooting, taking down and cleaning the dirt in the local area of the insulator and measuring the dirt degree in the local area;
a3, constructing an insulator pollution degree hyperspectral regression model by adopting a regression algorithm according to the nth hyperspectral spectral line of each insulator and the pollution degree measured by the insulator;
a4, calculating according to the insulator pollution degree hyperspectral regression model and the hyperspectral spectral lines of 1 st to (n-1) th times of each insulator to obtain the pollution degree of 1 st to (n-1) th times of each insulator;
and A5, constructing a high-spectrum development prediction model of the pollution degree by adopting a long-short term memory neural network algorithm according to the time sequence and the pollution degrees of all insulators for 1 to n times.
Further, step a3 further includes:
a31, taking a hyperspectral spectral line obtained when each insulator is shot for the nth time as a data source, taking the measured pollution degree as a label value, and dividing the data source into a first training set and a first testing set according to a set proportion;
a32, constructing and forming an insulator pollution degree hyperspectral regression model by adopting the label values, the first training set and the regression algorithm;
and A33, inputting the first test set into a pollution degree hyperspectral regression model to optimize the model, and obtaining the pollution degree hyperspectral regression model.
Further, the regression algorithm is a partial least squares regression method or an extreme learning machine algorithm.
Further, step a5 further includes:
a51, adopting the pollution degrees and time sequences of 1 to (n-1) times of all insulators as a second training set, and adopting the pollution degrees measured by all insulators at the nth time as a second test set;
a52, constructing a filthy degree hyperspectral development prediction model according to a second training set and a long-short term memory neural network algorithm;
and A53, inputting the second test set into a pollution degree hyperspectral development prediction model for optimization, and obtaining a final pollution degree hyperspectral development prediction model.
Further, in the long-short term memory neural network algorithm, x (x) is given to a given input time series1，…，xt) The output of the long-short term memory module is:
ht＝ottanh(ct)
ot＝s(Wxoxt+Whoht-1+Wcoct+bo)
ct＝ftct-1+ittanh(Wxcxt+Whcht-1+bc)
wherein h istAnd ht-1Predicting the pollution degrees output at the t moment and the t-1 moment for the model respectively; otIs the excitation output of the output gate at time t; c. CtAnd ct-1The outputs of the t-th and t-1-th neurons, respectively; s (-) is the excitation function of hidden layer neurons; wcoAnd WhcAre all weight coefficients; b0And bcAre all deviation vectors; f. oftAnd itThe excitation outputs of the forgetting gate and the input gate are respectively.
Further, before extracting the hyperspectral spectral lines in the hyperspectral image, performing black and white correction on the hyperspectral image:
wherein R isciThe hyperspectral image is a hyperspectral image after black and white correction; sampleciIs original spectral image data; dark redciCalibrating image data for all black; whiteciThe image data is fully white scaled.
Further, the pollution types are soot, saline alkali, dust or chemical industry, and the salt and ash components and the proportion in the same pollution type area are assumed to be similar.
Further, the hyperspectral image is obtained by adopting an unmanned aerial vehicle to carry a hyperspectral image acquisition device.
The invention has the beneficial effects that: according to the scheme, the acquired hyperspectral image of the insulator and the pollution degree hyperspectral development prediction model which is constructed by adopting a long-term and short-term memory neural network algorithm and has the same pollution type as the insulator are combined with each other, and the pollution degree development can be predicted based on different time sequences, so that the insulator can be cleaned in time before reaching serious pollution threshold values in different areas, pollution flashover accidents are avoided, and the reliability and the safety of a power transmission line are improved.
The scheme adopts the mode to predict the pollution degree and can also solve the problems of complex operation, large manual interference, difficulty in predicting future development of the pollution degree and the like in the prior art. In addition, when the hyperspectral image is collected, the collection device carried by the unmanned aerial vehicle is used for collecting the hyperspectral image of the insulator, so that manpower and material resources can be greatly reduced, manual operation is reduced, and intelligent detection of insulator contamination is realized.
Drawings
Fig. 1 is a flow chart of a method for predicting insulator pollution degree development based on a long-term and short-term memory neural network.
FIG. 2 is a flow chart of a construction method of a pollution degree hyperspectral development prediction model.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Referring to fig. 1, fig. 1 shows a flow chart of a method for predicting insulator pollution development based on a long-term and short-term memory neural network; as shown in fig. 1, the method includes steps 101 to 103.
In step 101, acquiring a hyperspectral image shot in a local area of an insulator, and extracting a hyperspectral spectral line of the hyperspectral image; the optimal hyperspectral image is obtained by adopting an unmanned aerial vehicle to carry a hyperspectral image acquisition device.
In step 102, inputting hyperspectral spectral lines of the insulators into the constructed hyperspectral regression model of the contamination degrees of the insulators with the same contamination type to obtain the contamination degree of the insulators;
the types of the pollutants are soot, saline alkali, dust or chemical industry, and the main difference is that the contents of the salts and the ashes in the areas with the same pollution type are different on the assumption that the components and the proportions of the salts and the ashes are similar.
In step 103, the pollution degree and the prediction time sequence of the insulator are input into a pollution degree hyperspectral development prediction model of the same pollution type, which is constructed by adopting a long-term and short-term memory neural network algorithm, so that the predicted pollution degree of the insulator required by the engineering is obtained.
In an embodiment of the present invention, the method for constructing the pollution degree hyperspectral development prediction model in step 103 includes steps a1 to a5, and refer to fig. 2 specifically.
In step A1, in a pollution area of the same pollution type, regularly shooting hyperspectral images at local areas of m insulators at fixed points for n times, and extracting hyperspectral spectral lines of each hyperspectral image; the larger the values of n and m are, the higher the prediction accuracy of the constructed pollution degree hyperspectral development prediction model is.
Before extracting the hyperspectral lines of the hyperspectral image in the step 101 and the step A1, performing black and white correction on the hyperspectral image:
wherein R isciThe hyperspectral image is a hyperspectral image after black and white correction; sampleciIs original spectral image data; dark redciCalibrating image data for all black; whiteciThe image data is fully white scaled.
In the step A2, after the insulator finishes the nth hyperspectral image shooting, taking down and cleaning the dirt in the local area of the insulator and measuring the dirt degree in the local area;
in the step A3, according to the nth hyperspectral spectral line of each insulator and the contamination degree measured by the insulator, constructing an insulator contamination degree hyperspectral regression model by adopting a regression algorithm; the insulator pollution degree hyperspectral regression model in the step 102 is constructed by adopting the steps A1 to A3.
In practice, the step a3 of this embodiment preferably further comprises:
a31, taking a hyperspectral spectral line obtained when each insulator is shot for the nth time as a data source, taking the measured pollution degree as a label value, and dividing the data source into a first training set and a first testing set according to a set proportion;
a32, constructing and forming an insulator pollution degree hyperspectral regression model by adopting the label values, the first training set and the regression algorithm;
a33, inputting the first test set into a pollution degree hyperspectral regression model to optimize the model, and obtaining a pollution degree hyperspectral regression model; the main purpose of the optimization in the step is to remove data with large errors.
The regression algorithm used in the construction of the insulator pollution degree hyperspectral regression model in the step a32 may be a partial least squares regression method, or may also be an extreme learning machine algorithm, but is not limited to these two.
When a partial least square regression method is adopted to construct an insulator pollution degree hyperspectral regression model, the key formula is as follows:
s＝aj1x1+...+ajnxn
in the formula, s is a detection value of the pollution equivalent salt deposit density of the insulator to be detected; a isj1、aj2、...、ajnSolving parameters for an insulator pollution equivalent salt deposit density detection model obtained according to a known pollution salt deposit density value; x is the number of1、x2、...、xnIs the normalized reflectance at n bands.
In the step A4, calculating the pollution degree of each insulator from 1 st to (n-1) th according to the insulator pollution degree hyperspectral regression model and the hyperspectral spectral lines of each insulator from 1 st to (n-1) th;
in the step A5, according to the time sequence and the pollution degrees of all insulators from 1 st to nth time, a high-spectrum pollution degree development prediction model is constructed and obtained by adopting a long-short term memory neural network algorithm.
In one embodiment of the present invention, step a5 further includes:
a51, adopting the pollution degrees and time sequences of 1 to (n-1) times of all insulators as a second training set, and adopting the pollution degrees measured by all insulators at the nth time as a second test set;
a52, constructing a filthy degree hyperspectral development prediction model according to a second training set and a long-short term memory neural network algorithm;
in long-short term memory neural network algorithms, x (x) is given to a given input time series1，…，xt) The output of the long-short term memory module is:
ht＝ottanh(ct)
ot＝s(Wxoxt+Whoht-1+Wcoct+bo)
ct＝ftct-1+ittanh(Wxcxt+Whcht-1+bc)
wherein h istAnd ht-1Predicting the pollution degrees output at the t moment and the t-1 moment for the model respectively; otIs the excitation output of the output gate at time t; c. CtAnd ct-1The outputs of the t-th and t-1-th neurons, respectively; s (-) is the excitation function of hidden layer neurons; wcoAnd WhcAre all weight coefficients; b0And bcAre all deviation vectors; f. oftAnd itThe excitation outputs of the forgetting gate and the input gate are respectively.
A53, inputting the second test set into a pollution degree hyperspectral development prediction model for optimization to obtain a final pollution degree hyperspectral development prediction model; in this step, the main purpose of the optimization is to calculate more optimal model parameters.
In conclusion, the method can accurately predict the pollution development of the insulator based on the pollution degree hyperspectral development prediction model constructed by the long-term and short-term memory neural network algorithm so as to solve the problems of complex operation, large manual interference, difficulty in predicting the future development of the pollution degree and the like in the prior art; the method is used for predicting the pollution degree development of the insulator of the power transmission line, and can predict the serious area of the pollution degree in advance so as to realize the advanced and accurate cleaning of the polluted insulator.
Claims (8)
1. The insulator pollution degree development prediction method based on the long-term and short-term memory neural network is characterized by comprising the following steps of:
acquiring a hyperspectral image shot in a local area of the insulator, and extracting a hyperspectral spectral line of the hyperspectral image;
inputting the hyperspectral spectral line of the insulator into a constructed insulator pollution degree hyperspectral regression model of the same pollution type to obtain the pollution degree of the insulator;
inputting the pollution degree and the prediction time sequence of the insulator into a pollution degree hyperspectral development prediction model of the same pollution type, which is constructed by adopting a long-term and short-term memory neural network algorithm, so as to obtain the predicted pollution degree of the insulator required by engineering;
the construction method of the pollution degree hyperspectral development prediction model comprises the following steps:
a1, in a pollution area of the same pollution type, regularly shooting hyperspectral images in local areas of m insulators at fixed points for n times, and extracting hyperspectral spectral lines of each hyperspectral image;
a2, after the insulator finishes the nth hyperspectral image shooting, taking down and cleaning the dirt in the local area of the insulator and measuring the dirt degree in the local area;
a3, constructing an insulator pollution degree hyperspectral regression model by adopting a regression algorithm according to the nth hyperspectral spectral line of each insulator and the pollution degree measured by the insulator;
a4, calculating according to the insulator pollution degree hyperspectral regression model and the hyperspectral spectral lines of 1 st to (n-1) th times of each insulator to obtain the pollution degree of 1 st to (n-1) th times of each insulator;
and A5, constructing a high-spectrum development prediction model of the pollution degree by adopting a long-short term memory neural network algorithm according to the time sequence and the pollution degrees of all insulators for 1 to n times.
2. The insulator contamination degree development prediction method according to claim 1, wherein the step a3 further comprises:
a31, taking a hyperspectral spectral line obtained when each insulator is shot for the nth time as a data source, taking the measured pollution degree as a label value, and dividing the data source into a first training set and a first testing set according to a set proportion;
a32, constructing and forming an insulator pollution degree hyperspectral regression model by adopting the label values, the first training set and the regression algorithm;
and A33, inputting the first test set into a pollution degree hyperspectral regression model to optimize the model, and obtaining the pollution degree hyperspectral regression model.
3. The insulator contamination degree development prediction method according to claim 2, wherein the regression algorithm is a partial least squares regression method or an extreme learning machine algorithm.
4. The insulator contamination degree development prediction method according to claim 1, wherein the step a5 further comprises:
a51, adopting the pollution degrees and time sequences of 1 to (n-1) times of all insulators as a second training set, and adopting the pollution degrees measured by all insulators at the nth time as a second test set;
a52, constructing a filthy degree hyperspectral development prediction model according to a second training set and a long-short term memory neural network algorithm;
and A53, inputting the second test set into a pollution degree hyperspectral development prediction model for optimization, and obtaining a final pollution degree hyperspectral development prediction model.
5. The insulator filthy degree development prediction method according to claim 1 or 4, characterized in that in the long-short term memory neural network algorithm, for a given input time sequence x ═ (x ═ x)1，…，xt) The output of the long-short term memory module is:
ht＝ottanh(ct)
ot＝s(Wxoxt+Whoht-1+Wcoct+bo)
ct＝ftct-1+ittanh(Wxcxt+Whcht-1+bc)
wherein h istAnd ht-1Predicting the pollution degrees output at the t moment and the t-1 moment for the model respectively; otIs the excitation output of the output gate at time t; c. CtAnd ct-1The outputs of the t-th and t-1-th neurons, respectively; s (-) is the excitation function of hidden layer neurons; wcoAnd WhcAre all weight coefficients; boAnd bcAre all deviation vectors; f. oftAnd itThe excitation outputs of the forgetting gate and the input gate are respectively.
6. The insulator contamination degree development prediction method according to claim 1, further comprising performing black and white correction on the hyperspectral image before extracting the hyperspectral spectral lines in the hyperspectral image:
wherein R isciThe hyperspectral image is a hyperspectral image after black and white correction; sampleciIs original spectral image data; dark redciCalibrating image data for all black;Whitecithe image data is fully white scaled.
7. The insulator pollution degree development prediction method according to claim 1, wherein the pollution type is soot, saline alkali, dust or chemical engineering, and the salt and ash components and the proportion are assumed to be similar in the same pollution type area.
8. The insulator contamination degree development prediction method according to claim 1, wherein the hyperspectral image is obtained by using an unmanned aerial vehicle to carry a hyperspectral image acquisition device.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910833968.5A CN110530876B (en) | 2019-09-04 | 2019-09-04 | Insulator pollution degree development prediction method based on long-term and short-term memory neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910833968.5A CN110530876B (en) | 2019-09-04 | 2019-09-04 | Insulator pollution degree development prediction method based on long-term and short-term memory neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110530876A CN110530876A (en) | 2019-12-03 |
| CN110530876B true CN110530876B (en) | 2020-08-18 |
Family
ID=68667029
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910833968.5A Active CN110530876B (en) | 2019-09-04 | 2019-09-04 | Insulator pollution degree development prediction method based on long-term and short-term memory neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110530876B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111289854B (en) * | 2020-02-26 | 2021-05-11 | 华北电力大学 | Insulator insulation state evaluation method of 3D-CNN and LSTM based on ultraviolet video |
| CN111257242A (en) * | 2020-02-27 | 2020-06-09 | 西安交通大学 | High-spectrum identification method for pollutant components of insulator |
| CN114187783B (en) * | 2021-12-06 | 2023-10-31 | 中国民航大学 | Method for analyzing and predicting potential conflict in airport flight area |
| CN115856550B (en) * | 2022-12-19 | 2024-01-16 | 华南理工大学 | Salt fog flashover prediction method and device for composite insulator and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108072667A (en) * | 2017-09-28 | 2018-05-25 | 江苏省电力试验研究院有限公司 | Insulator contamination level detection method and system based on EO-1 hyperion |
| CN108256626A (en) * | 2016-12-28 | 2018-07-06 | 中国科学院深圳先进技术研究院 | The Forecasting Methodology and device of time series |
| CN109612947A (en) * | 2019-01-17 | 2019-04-12 | 西南交通大学 | Insulator contamination equivalent salt density detection method based on partial least-squares regression method |
| CN110096737A (en) * | 2019-03-21 | 2019-08-06 | 国网内蒙古东部电力有限公司电力科学研究院 | Insulator life-span prediction method, device, computer installation and storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102520286B (en) * | 2011-12-15 | 2014-07-09 | 国网电力科学研究院 | Hyperspectrum-based composite insulator operation state classification method |
| US10417788B2 (en) * | 2016-09-21 | 2019-09-17 | Realize, Inc. | Anomaly detection in volumetric medical images using sequential convolutional and recurrent neural networks |
| CN109799245A (en) * | 2019-03-29 | 2019-05-24 | 云南电网有限责任公司电力科学研究院 | A kind of insulator contamination degree non-contact detection method and device |
-
2019
- 2019-09-04 CN CN201910833968.5A patent/CN110530876B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108256626A (en) * | 2016-12-28 | 2018-07-06 | 中国科学院深圳先进技术研究院 | The Forecasting Methodology and device of time series |
| CN108072667A (en) * | 2017-09-28 | 2018-05-25 | 江苏省电力试验研究院有限公司 | Insulator contamination level detection method and system based on EO-1 hyperion |
| CN109612947A (en) * | 2019-01-17 | 2019-04-12 | 西南交通大学 | Insulator contamination equivalent salt density detection method based on partial least-squares regression method |
| CN110096737A (en) * | 2019-03-21 | 2019-08-06 | 国网内蒙古东部电力有限公司电力科学研究院 | Insulator life-span prediction method, device, computer installation and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting;Xingjian Shi等;《Proceedings of the 28th international conference on Neural Information Processing System》;20151231;第1卷;第802-810页 * |
| 基于ARIMA-LSTM的架空线状态数据挖掘;钟令枢;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20170215;第45页第1段，第51页最后1段，第52页第2段，第53页第2段 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110530876A (en) | 2019-12-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110530876B (en) | Insulator pollution degree development prediction method based on long-term and short-term memory neural network | |
| WO2022077605A1 (en) | Wind turbine blade image-based damage detection and localization method | |
| US20190285517A1 (en) | Method for evaluating health status of mechanical equipment | |
| CN106290388A (en) | A kind of insulator breakdown automatic testing method | |
| CN110716512A (en) | Environmental protection equipment performance prediction method based on coal-fired power plant operation data | |
| CN111539355A (en) | Photovoltaic panel foreign matter detection system and detection method based on deep neural network | |
| Thanh et al. | Insulator leakage current prediction using surface spark discharge data and particle swarm optimization based neural network | |
| CN113108918B (en) | Method for inverting air temperature by using thermal infrared remote sensing data of polar-orbit meteorological satellite | |
| Zhang et al. | Prediction of urban pm 2.5 concentration based on wavelet neural network | |
| CN109799245A (en) | A kind of insulator contamination degree non-contact detection method and device | |
| CN116612098B (en) | Insulator RTV spraying quality evaluation method and device based on image processing | |
| CN111008641A (en) | Power transmission line tower external force damage detection method based on convolutional neural network | |
| CN115170988A (en) | Power grid line fault identification method and system based on deep learning | |
| CN114863361A (en) | Distribution line bird nest detection method based on USRNet and YOLOv3 | |
| CN112070783A (en) | Wind deflection detection method and device for insulator string of power transmission line | |
| CN113989209A (en) | Power line foreign matter detection method based on fast R-CNN | |
| CN114638417A (en) | Fine-grained air quality prediction method and system | |
| CN110070520B (en) | Pavement crack detection model construction and detection method based on deep neural network | |
| CN113284103A (en) | Substation equipment defect online detection system based on space transformation fast R-CNN model | |
| Hong et al. | Drainage network flow anomaly classification based on XGBoost | |
| CN116485802B (en) | Insulator flashover defect detection method, device, equipment and storage medium | |
| CN117522872B (en) | Real-time abnormality detection method and device for photovoltaic power station | |
| CN112381560B (en) | Shared equipment product market prediction system and method | |
| CN117252840B (en) | Photovoltaic array defect elimination evaluation method and device and computer equipment | |
| Wang et al. | Crack Detection of Old Residential Buildings Based on UAV Intelligent Vision |
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 |