CN106094969A - The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of - Google Patents
The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of Download PDFInfo
- Publication number
- CN106094969A CN106094969A CN201610327033.6A CN201610327033A CN106094969A CN 106094969 A CN106094969 A CN 106094969A CN 201610327033 A CN201610327033 A CN 201610327033A CN 106094969 A CN106094969 A CN 106094969A
- Authority
- CN
- China
- Prior art keywords
- image
- value
- module
- threshold
- radiance
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05F—SYSTEMS FOR REGULATING ELECTRIC OR MAGNETIC VARIABLES
- G05F1/00—Automatic systems in which deviations of an electric quantity from one or more predetermined values are detected at the output of the system and fed back to a device within the system to restore the detected quantity to its predetermined value or values, i.e. retroactive systems
- G05F1/66—Regulating electric power
- G05F1/67—Regulating electric power to the maximum power available from a generator, e.g. from solar cell
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
- Y02E10/00—Energy generation through renewable energy sources
- Y02E10/50—Photovoltaic [PV] energy
- Y02E10/56—Power conversion systems, e.g. maximum power point trackers
Landscapes
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Automation & Control Theory (AREA)
- Image Analysis (AREA)
Abstract
A kind of maximum power point of photovoltaic power generation system that the present invention provides follows the trail of diagnostic techniques, and described maximum power point of photovoltaic power generation system is followed the trail of the composition of diagnostic techniques system and included: radiance acquisition module, image processing module, MPPT module, signal conversion module；The process that realizes of described technology is: first having to gather radiance intensity image, acquisition system primary structure is made up of light intensity sensor and temperature sensor；Radiance graphical analysis, is divided into the visibly different part of several gray value on computers with partitioning algorithm by light elongated image pattern, and selecting Otsu method and maximum entropy method (MEM) is the partitioning algorithm of core, carries out accurate image on a display screen show by being partitioned into the strongest intense part；The present invention has good powerinjected method effect can improve work efficiency.
Description
Technical field
The present invention relates to new forms of energy photovoltaic generating system field, specifically a kind of maximum power point of photovoltaic power generation system is followed the trail of
Image aided diagnosis technique.
Background technology
Nowadays the world develops rapidly, and tellurian resource gradually decreases, and the minimizing of resource will necessarily affect environment and the mankind
Life, so the use of new forms of energy has obtained the attention in the world, and China's the most just progressively sane row on the road of new forms of energy
Moving, modal is the generation of electricity by new energy such as wind-force and photovoltaic.Electricity generation system maximum power point tracking i.e. MPPT technique, is too
An important key technology in sun energy photovoltaic generation, it refers to change in the extraneous factor condition such as temperature, intensity of illumination
Time, system can remain at maximum power output.No matter further, since photovoltaic solar panels or the conversion efficiency of battery
The highest and expensive, the input at initial stage is relatively big, farthest makes full use of energy produced by photovoltaic generation, improves energy
Amount conversion efficiency, reduction cost are that this technology exists and the key factor of development.Can if there is deviation occurs in work efficiency,
Do not reach maximal efficiency, then the loss caused will become higher.
Image aided diagnosis technique mainly uses a kind of technology being called image segmentation, and it is a kind of important image procossing
Technology, its application is increasingly wider, has penetrated into engineering, industry, health care, Aero-Space, military affairs, scientific research, safety guarantor
The various aspects such as defend, national economy and national economy play increasing effect.Therefore in new forms of energy, particularly
In photovoltaic generating system, whether this technology can also accomplish the combination with photovoltaic generating system and application, reaches more preferable effect.
At present in photovoltaic generating system, although existing much about the application technology of maximum power tracing, but still
Do not have one can be by artificial the most intuitively it can be seen that whether current power point is peak power, also one day from early to
In evening, sunlight exposure rate differs, and each stage all can exist respective maximum power point, continues just in case emergent power is followed the trail of to exist
Error and must not correct in time, be the most just more difficult to reach efficient.In actual application system, the radiant intensity of natural light and big
During the light transmittance of gas is in dynamically changing, this efficient application just brought to photovoltaic generating system brings difficulty, Er Qiexian
A lot of photovoltaic generating systems obtained by power results be not necessarily maximum point i.e. error at that time, these problems are still deposited
?.
For the problems referred to above, the present invention proposes the image auxiliary diagnosis skill that a kind of maximum power point of photovoltaic power generation system is followed the trail of
Art, the most whether it can solve manually can to reach both-end and verify with the problem of intuitive judgment mutually and finally give maximum power point also
Reduce the effect of loss, thus reach maximal efficiency.
Summary of the invention
For present stage photovoltaic generating system maximum power tracing accuracy and can careful observation problem, the present invention is open
The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of.Photovoltaic maximum power is allowed to follow the trail of
The stability of technology obtains bigger lifting and efficiency, and image processing techniques is applied to maximum power tracing technology, thus can
With from wing auxiliary diagnosis current operating state.
It is characteristic of the invention that and can observe by the naked eye the different light intensity regions that image splits, it is judged that be currently maximum
The position of power points, it is achieved than the application more preferably powerinjected method of other situations.Follow the trail of compared with a lot of pure devices and may meet pole
Value point, more can find current value point both peak powers by naked eyes in the observation of the computer of terminal demonstration image segmentation result
Point, both for auxiliary diagnosis.
The present invention is to be achieved by the following scheme:
The present invention gathers radiance intensity by CCD, through partitioning algorithm at Computer display current radiance gray value, such that it is able to
Facilitating manual observation and combine MPPT and determine every time the gray value at residing maximum power point place, the two carries out complementary examining
Break and obtain final result.The present invention can carry out the seizure of current peak power under various different weather conditions and follow the trail of, logical
Comprehensively thus whether help system diagnosis realizes maximum power tracing to cross both image procossing and MPPT, it might even be possible to play repeatedly
Whether checking current power is in maximum.The present invention assists the step that realizes of diagnosis effect:
Step 1: carry out maximum power tracing by former MPPT and obtain current power points；
Step 2: after image acquisition and algorithm, shows the segmentation image of current radiance intensity on a display screen；
Step 3: obtain maximum power point gray value accounting position (optimum gradation on segmentation image by the contrast of upper two steps
In the region of value)；
Step 4: owing to change occurs in light intensity, also there is change, the gray scale of segmentation image in the current power point that MPPT is tracked
Value also can become, but the gray value accounting position on segmentation image is substantially without there is variation, and both-end result compares；
Step 5: if finding that gray value accounting position occurs in that obvious deviation in segmentation image, then illustrate that MPPT is followed the trail of
Maximum power point necessarily occur in that error；
Step 6: if there is error, adjusts the angle of solar panels by motor, is adjusted the gray value accounting position of segmentation image
Accounting position to standard；
Step 7: achieve both-end and verify mutually and finally give maximum power point and reduce the effect of loss.
Accompanying drawing explanation
Below in conjunction with the accompanying drawings the present invention is further detailed.
Fig. 1 is the operation principle flow chart of the present invention,
Fig. 2 is the overall structure naive model figure of the present invention；
Fig. 3 is the auxiliary diagnosis principle figure of the present invention；
Fig. 4 is the partitioning algorithm flow chart of the present invention；
Detailed description of the invention:
As described in Figure 1, whole system can be divided into two parts, and with thick dashed line in scheming as boundary, right-hand component is the core of the present invention
Part, CCD industrial camera, by accepting lamp, carries out the light data acquisition of correspondence to it, and in left half, light intensity passes
The data collected are sent to signaling conversion circuit pair by sensor, temperature sensor (dual sensor), current/voltage testing circuit
It is changed, and this change-over circuit is sent to monolithic core after the data transmitted of dual sensor and testing circuit being changed
Control circuit carries out maximum power tracing hence into MPPT system；After the data conversion that CCD industrial camera is transmitted by other end
It is sent to image pick-up card, carries out image segmentation algorithm calculating, by the gray value of radiance situation corresponding for current power clearly
Display is on computers.Then the result of MPPT be correspond to light intensity gray value in the segmentation image manifested on computer display
Accounting location aided rebroadcast diagnose, judge after being verified by both-end this power points whether on current radiance point of maximum intensity, for peak power
Point；If it is not, then adjust solar panel angle to realize assist control maximum power tracing to optimal point.
In as described in Figure 2,1 is CCD industrial camera, and 2 is solar panels.Its actual operation principle can be worked along both lines: at one is
The data such as MPPT value after solar panels photovoltaic generation carries out MPPT and run are transferred to experience database at computer
Upper set, is by the segmentation figure of face display on computers after image pick-up card and partitioning algorithm operation at another by the CCD beginning
As result.Then, by experience database and the corresponding comparison of segmentation image, it is judged that maximum power point and optimum gradation value
Light intensity gray value accounting position in segmentation image is the most often in that optimum, if not both occurred some so that very big partially
Difference, controls the motor under solar panel by computer and adjusts solar panel angle to optimum.This most well embodies at image
Reason technology well assists diagnosis on maximum power tracing, and the two complements each other, and significantly can improve efficiency.
The voltage that parameter has the light intensity value L of light intensity sensor as described in Figure 3, temperature value T, MPPT of temperature sensor obtain
The radiance gray value that value and performance number and image split: experience database is with temperature value T for X-coordinate axle, with light intensity value
L is Y coordinate axle (interchangeable), the maximum power point seat numerical value obtained with selected MPPT, sets up data base.Each temperature
The corresponding MPPT value (including magnitude of voltage and performance number) of degree T light intensity L, and updated by size, will be current maximum
MPPT value store, auxiliary diagnosis effect: while setting up experience database, by record different temperature value T and
MPPT value corresponding to light intensity value L input.And MPPT value each time all by with as in corresponding temperature value T light intensity value L feelings
The radiance gray value split under condition carries out diagnosis and reaches a conclusion, and conclusion judges whether this power is in light intensity gray value exactly
Accounting position, and this auxiliary diagnosis just needs to be additionally carried out.
It is the step of the core partitioning algorithm of assistant diagnosis system in the present invention such as Fig. 4, with which kind of partitioning algorithm？First
It is envisioned that a kind of thresholding method, it is a kind of image partition method the most ancient and all the fashion, the usual base of the method
In entirety or the local message of image, if pixel being divided into Ganlei by choosing one or more threshold value, thus realize image
Single target or multiple Target Segmentation.When using threshold method that image is split, the most all to its exist certain it is assumed that
I.e. based on some specific iconic model.Conventional iconic model can be described as follows: sets the gray scale fraction of target and background in image
Cloth is single-peak response, there is the dependency of height, and target and the back of the body in target and background region between the gray value of neighbor
On gray value, larger difference is there is between the pixel of scape intersection.As image meets above-mentioned condition, then its grey level histogram can
Approximation is considered as mixed by two unimodal histogram corresponding respectively to target and background forming；Size phase such as the two distribution
Closely, and average is apart from each other and mean square deviation is less, then this rectangular histogram is double-hump characteristics, and such image can be obtained by thresholding method
Preferably segmentation effect.Thresholding method, because of simple effective, be widely used in image processing field, occurred in that various
Threshold Segmentation Algorithm, wherein affect bigger mainly having: histogram method and rectangular histogram converter technique, minimum error method, maximum entropy
Method, maximum variance between clusters, fuzzy binary images, minimum cross entropy method, dynamic thresholding method.Although there being numerous Threshold segmentations at present
Method, and these methods are the most relatively easy, but generally it is based only upon the half-tone information of image because of it and the sky that have ignored between pixel
Between dependency, therefore the method is only applicable to the highest occasion of prescription to image segmentation.For those exist noise jamming,
The image that contrast is relatively low, due to its gray value do not exist exist between notable difference or each object grey scale value large range of heavy
Folded, its grey level histogram not in obvious double-hump characteristics, so time such as use threshold method, then be difficult to obtain and split knot accurately
Really.Image in actual application is frequently not single goal image, when carrying out image segmentation with thresholding method, in order to extract
Each target in image, it must be determined that how each threshold value.Otsu method and maximum entropy method (MEM) are the widest of image processing field utilization
Two kinds of general thresholding methods, both thresholding methods can the most simply expand to multi-threshold segmentation, mainly
Illustrate and use what KMTOA solved multi thresholds problem to implement process.
Otsu method is fairly simple, and deriving based on method of least square thought develops.Its ultimate principle is: according to
The gray feature of image, is classified as different classifications, and the variance of all kinds of is the biggest, means that the difference of all kinds of is the biggest；When
Partial pixel is by wrong timesharing, and the difference of all kinds of will diminish.So that inter-class variance takes the segmentation of maximum is wrong point rate
Little segmentation.Being N for given image I, such as its sum of all pixels comprised, its gray level sum is L, then gray level i institute
Frequency Pi occurred may be defined as:
WhereinRepresent gray levelThe frequency occurred.
Such as imageBy threshold valueIt is divided intoTwo parts, whereinComprise gray level, andThen wrap
Containing gray level, thenWithThe total probability occurredCan be respectively defined as:
WithGray averageMay be defined as:
As remembered imageOverall intensity average be, then it is easy to draw:
WithInter-class variance be defined as:
Obtain the threshold value of maximumIt is required optimal threshold.
Maximum entropy method (MEM) realizes simple and segmentation effect is good, causes the extensive concern of Chinese scholars and becomes one and have
Representational entropy partitioning algorithm.It is similar with Otsu method, if the gray scale space of image I is, threshold valueBy imageIt is divided intoWithTwo parts, whereinComprise gray level, andComprise gray level, then imageEntropy can
It is defined as:
Wherein:
MakeObtain the threshold value of maximumIt is required optimal threshold.
When using Otsu method and maximum entropy method (MEM) to carry out image segmentation, need to choose so that inter-class variance in threshold space
Or image entropy obtains the threshold value of maximum, ask specifically comprising the following steps that of m threshold value based on KMTOA
Step 1: initialize population and algorithm parameter.If population scale is Popsize, then Popsize m of stochastic generation dimension to
Amount is to constitute initial population；Owing to the span of gray level is 0 ~ 255, so the most one-dimensional value of vector is between 0 ~ 255
Integer；The frequency that the gray level in each stage of statistical picture occurs, obtains corresponding rectangular histogram；
Step 2: calculate the adaptive value of each individuality based on formula, and pick out current population most has individual Xbest.In order to select
Select out each optimal threshold of m, must use following two formulas (1) or (2):
(1)
Or(2)
Wherein:
Step 3: for each individuality, calculate its acceleration that attracts accordingly, repels or fluctuate with KMTOA；
Step 4: calculate individual speed mobile individuality.After mobile individuality, it is necessary to individuality is rounded behavior, and sentences
Whether disconnected each dimension is between 0 ~ 255；As crossed the border, process of crossing the border the most accordingly；
Step 5: the optimum individual in population is performed elite retention strategy, to prevent colony from degenerating；
Step 6: whether evaluation algorithm terminates.If terminated, find out optimal threshold, and press according to this value
Carry out image according to following formula (3) to split to obtain segmentation result, without terminating then to return step 2, and segmentation result here
It it is exactly the last image for assisting diagnosis that face shows on computers.
(3)
Wherein:Represent pixel in image to be splitGray value.
Claims (6)
1. the image aided diagnosis technique of a maximum power point of photovoltaic power generation system tracking is characterized in that four module: radiance
Acquisition module, signal conversion module, image processing module, MPPT module；Radiance at that time are accurately received by radiance acquisition module
The radiance data collected are transmitted to image processing module and MPPT mould by intensity and temperature afterwards by signaling conversion circuit
Block, then show that (gray value is that diagnosis the important of radiance intensity depends on to various gray value by image processing module by partitioning algorithm
According to) and MPPT module judge currently whether be in maximum power point (radiance intensity is the strongest then closer to maximum power point).
Radiance acquisition module the most according to claim 1, it is characterised in that its four core components: light intensity sensor, temperature
Degree sensor and voltage and current detection circuit, CCD industrial machine etc., this module is the gate of whole system.
The most according to claim 2, CCD industrial machine is characterized in that it is light can to become electric charge and stored by electric charge
And transfer, it is possible to being taken out by the electric charge of storage and make voltage change, it is the whole image processing module most important beginning, adopts
Collection light intensity is formed image and is split by algorithm followed by and manifest.
The most according to claim 1, signal conversion module includes signaling conversion circuit, it is characterised in that by radiance acquisition module
The collection data sent carry out conversion and are transmitted further to image processing module and MPPT module.
Image processing module the most according to claim 1, it is characterised in that use certain partitioning algorithm run after sustainable
Manifest on computer display and split light intensity gray value accounting position (in the region of optimum gradation value) in image, and this light intensity
Gray value accounting position is current corresponding maximum power point, and by consulting other documents and data, adopted here divides
Cutting algorithm is thresholding method；Thresholding method is a kind of image partition method, and the method is typically based on the information of image, passes through
If choosing one or more threshold value pixel is divided into Ganlei, thus realizing the single target of image or multiple Target Segmentation, adopting
When image being split with threshold method, the most all it is existed certain it is assumed that i.e. based on some specific iconic model, threshold
Value split-run has effectively been widely used in image processing field because of simple, has occurred in that various Threshold Segmentation Algorithm, right
In this world, its impact is bigger, although have numerous threshold segmentation methods at present, and these methods are the most relatively easy, only
It is applicable to the highest occasion of prescription to image segmentation, the image that noise jamming, contrast are relatively low is existed for those, by
Not existing in its gray value and there is large range of overlap between notable difference or each object grey scale value, its grey level histogram is not in bright
Aobvious double-hump characteristics, so time such as use threshold method, then be difficult to obtain segmentation result accurately, the image in actual application is often
It not single goal image, when carrying out image segmentation with thresholding method, in order to extract each target in image, it is necessary to really
Fixed how each threshold value, wherein the present invention selects the most commonly used threshold that Otsu method and maximum entropy method (MEM) both image processing field are used
Value split-run, both thresholding methods can the most simply expand to multi-threshold segmentation, is using Otsu method and maximum
When entropy method carries out image segmentation, need to choose in threshold space so that inter-class variance or image entropy obtain the threshold value of maximum,
Specifically comprising the following steps that of m threshold value is asked based on KMTOA
Step 1: initialize population and algorithm parameter, if population scale is Popsize, then Popsize m of stochastic generation tie up to
Amount is to constitute initial population；Owing to the span of gray level is 0 ~ 255, so the most one-dimensional value of vector is between 0 ~ 255
Integer；The frequency that the gray level in each stage of statistical picture occurs, obtains corresponding rectangular histogram；
Step 2: calculate the adaptive value of each individuality based on formula, and pick out current population most have individual Xbest, in order to select
Select out each optimal threshold of m, must use following two formulas (1) or (2):
(1)
Or(2)
Wherein:
Step 3: for each individuality, calculate its acceleration that attracts accordingly, repels or fluctuate with KMTOA；
Step 4: calculate individual speed mobile individuality, after mobile individuality, it is necessary to individuality is rounded behavior, and sentences
Whether disconnected each dimension is between 0 ~ 255；As crossed the border, process of crossing the border the most accordingly；
Step 5: the optimum individual in population is performed elite retention strategy, to prevent colony from degenerating；
Step 6: whether evaluation algorithm terminates, if terminated, finds out optimal threshold, and according to this value according to
Following formula (3) carries out image and splits to obtain segmentation result, and without terminating then to return step 2, and segmentation result here is just
It is the last image for assisting diagnosis that face shows on computers:
(3)
Wherein:Represent pixel in image to be splitGray value.
6. it is the core of whole invention according to image processing module described in claim 1 or 5, it is characterised in that auxiliary diagnosis is current
MPPT module show whether system tracks is in maximum power point, thus decides whether to adjust solar panel angle to current optimal
Point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610327033.6A CN106094969A (en) | 2016-05-17 | 2016-05-17 | The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610327033.6A CN106094969A (en) | 2016-05-17 | 2016-05-17 | The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106094969A true CN106094969A (en) | 2016-11-09 |
Family
ID=57230005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610327033.6A Pending CN106094969A (en) | 2016-05-17 | 2016-05-17 | The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106094969A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193320A (en) * | 2017-06-06 | 2017-09-22 | 湘潭大学 | A kind of local shades Photovoltaic array MPPT controls based on Molecule Motion Theory optimized algorithm |
| CN107217977A (en) * | 2017-05-23 | 2017-09-29 | 李海英 | The curtain state transition method detected based on wind-force |
| CN109120230A (en) * | 2018-07-19 | 2019-01-01 | 苏州热工研究院有限公司 | A kind of solar battery sheet EL image detection and defect identification method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201479053U (en) * | 2009-08-18 | 2010-05-19 | 河海大学 | Maximum power point tracking device of photovoltaic power generation |
| CN103472884A (en) * | 2013-08-14 | 2013-12-25 | 浙江工业大学 | Global maximum power point tracking method for photovoltaic array |
| JP2015141438A (en) * | 2014-01-27 | 2015-08-03 | 株式会社 日立産業制御ソリューションズ | Photovoltaic power generation amount prediction apparatus and area division map display method |
-
2016
- 2016-05-17 CN CN201610327033.6A patent/CN106094969A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201479053U (en) * | 2009-08-18 | 2010-05-19 | 河海大学 | Maximum power point tracking device of photovoltaic power generation |
| CN103472884A (en) * | 2013-08-14 | 2013-12-25 | 浙江工业大学 | Global maximum power point tracking method for photovoltaic array |
| JP2015141438A (en) * | 2014-01-27 | 2015-08-03 | 株式会社 日立産業制御ソリューションズ | Photovoltaic power generation amount prediction apparatus and area division map display method |
Non-Patent Citations (2)
| Title |
|---|
| 王子明: ""光伏电站自动跟踪系统与MPPT的研究"", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 范朝冬 等: ""基于改进斜分Otsu法的回转窑火焰图像分割"", 《自动化学报》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107217977A (en) * | 2017-05-23 | 2017-09-29 | 李海英 | The curtain state transition method detected based on wind-force |
| CN107217977B (en) * | 2017-05-23 | 2018-03-20 | 杨丽 | Curtain state transition method based on wind-force detection |
| CN108119051A (en) * | 2017-05-23 | 2018-06-05 | 李海英 | Curtain state transition method based on wind-force detection |
| CN108119051B (en) * | 2017-05-23 | 2018-11-06 | 盐城市金良窗饰有限公司 | Curtain state transition method based on wind-force detection |
| CN107193320A (en) * | 2017-06-06 | 2017-09-22 | 湘潭大学 | A kind of local shades Photovoltaic array MPPT controls based on Molecule Motion Theory optimized algorithm |
| CN109120230A (en) * | 2018-07-19 | 2019-01-01 | 苏州热工研究院有限公司 | A kind of solar battery sheet EL image detection and defect identification method |
| CN109120230B (en) * | 2018-07-19 | 2019-10-11 | 苏州热工研究院有限公司 | A kind of solar battery sheet EL image detection and defect identification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112884064B (en) | Target detection and identification method based on neural network | |
| CN110210539B (en) | RGB-T image saliency target detection method based on multi-level depth feature fusion | |
| CN108121991B (en) | Deep learning ship target detection method based on edge candidate region extraction | |
| CN102324019B (en) | Method and system for automatically extracting gesture candidate region in video sequence | |
| CN109785236B (en) | Image super-resolution method based on super-pixel and convolutional neural network | |
| CN108960141A (en) | Pedestrian's recognition methods again based on enhanced depth convolutional neural networks | |
| CN107730477A (en) | Substation equipment infrared intelligent monitoring system and monitoring image feature extracting method | |
| CN114821069B (en) | Construction semantic segmentation method for remote sensing image of double-branch network fused with rich-scale features | |
| CN104166999B (en) | Cloud cluster extracting method based on strength layering of foundation cloud pictures | |
| CN106094969A (en) | The image aided diagnosis technique that a kind of maximum power point of photovoltaic power generation system is followed the trail of | |
| CN110188217A (en) | Image duplicate checking method, apparatus, equipment and computer-readable storage media | |
| CN104966285A (en) | Method for detecting saliency regions | |
| CN104966291A (en) | Cloud cluster automatic detection method based on foundation cloud atlas | |
| CN103761736B (en) | A kind of image partition method based on Bayes's harmony degree | |
| CN105426926A (en) | Method and device for performing detection and classification on active-matrix organic light emitting diode (AMOLED) | |
| CN103093241B (en) | Based on the remote sensing image nonuniformity cloud layer method of discrimination of homogeneity process | |
| Xiang et al. | Double-branch fusion network with a parallel attention selection mechanism for camouflaged object detection | |
| CN107766838B (en) | Video scene switching detection method | |
| Tang | A digitalization-based image edge detection algorithm in intelligent recognition of 5G smart grid | |
| CN104021534A (en) | Shredded paper splicing method | |
| CN114494284A (en) | Scene analysis model and method based on explicit supervision area relation | |
| CN113469286A (en) | Spacecraft multi-local component detection method based on regional convolutional neural network | |
| Li et al. | Carnet: a lightweight and efficient encoder-decoder architecture for high-quality road crack detection | |
| Yuan et al. | Salient Object Detection Based on Improved PoolNet. | |
| CN106156776A (en) | A kind of illumination recognition methods in traffic video monitoring |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20161109 |