阅读说明：本技术 基于环境气象因素的热斑监测装置与预测方法 (Hot spot monitoring device and prediction technique based on environment meteorological factor ) 是由 邵怡 徐红伟 陈芳芳 郭梦浩 吴苏阳 刘彪 于 2019-08-02 设计创作，主要内容包括：本发明公开了一种基于环境气象因素的热斑监测装置与预测方法。PC端将气象站采集的采集到的气象信息与热斑信息进行相关性分析并建立热斑预测模型；利用热斑预测模型,通过对每块光伏板进行默认编号,从而实现对热斑的初步定位,再通过监测机器人对预测到的光伏板进行监测,将拍摄到的图像处理后实现对热斑的判别,可以实现准确且快速的确定热斑的位置以及热斑的情况。本方法通过分析环境气象因素与热斑的关系,达到对热斑的预测以及初步的定位,通过进一步的监测,能够实现提高全自动热斑监测与预测的精确度与效率以及相应人工的减少。(The invention discloses a kind of hot spot monitoring device and prediction technique based on environment meteorological factor.The collected weather information that weather station acquires and hot spot information are carried out correlation analysis and establish hot spot prediction model by the end PC；Utilize hot spot prediction model, by carrying out default number to every piece of photovoltaic panel, to realize the Primary Location to hot spot, the photovoltaic panel predicted is monitored by monitoring robot again, the case where by realizing the differentiation to hot spot after the image procossing taken, position and the hot spot of accurate and quick determining hot spot may be implemented.This method reaches the prediction and preliminary positioning to hot spot by the relationship of analysis environment meteorological factor and hot spot, by further monitoring, can be realized and improves full-automatic hot spot monitoring and the accuracy of prediction and efficiency and corresponding artificial reduction.)

1. a kind of hot spot monitoring device based on environment meteorological factor, which is characterized in that including photovoltaic module (1), weather station (2), data collecting instrument (3), monitoring robot (4), docking platform (5), mobile platform (6) and the end PC (7), photovoltaic module (1) It favours ground by support frame to place, photovoltaic module (1) is mainly made of the photovoltaic panel that multiple row is arranged in parallel, each column photovoltaic panel Including multiple photovoltaic group strings, photovoltaic module (1) upper end is equipped with grooved tracks I (8)；

Docking platform (5) bottom is connected by vertical bar with mobile platform (6), docking platform (5) side and photovoltaic module (1) one Side fits, and docking platform (5) upper end is equipped with the grooved tracks II to connect with the grooved tracks I (8) of photovoltaic module (1) upper end；

Monitoring robot (4) include monitoring robot shell (401), motion module, monitoring modular, the first battery (404) and Second battery (405), the first battery (404), the second battery (405) are respectively that motion module and monitoring modular provide electricity Source；Motion module mainly by motor (402), motor driver (403), motion-control module (406), driving wheel (409) and from Driving wheel (408) composition is located at monitoring robot shell (401) bottom and is equipped with close to edge along grooved tracks I (8) and groove The driving wheel (409) and driven wheel (408) that track II slides, the first battery (404), the second battery (405), motor (402), motor driver (403), motion-control module (406) are respectively positioned in monitoring robot shell (401), motor (402) Output shaft is connected with driving wheel (409), and driving wheel (409) is connected with driven wheel (408) by synchronous belt, motion-control module (406) it is connected by motor driver (403) with motor (402), motion-control module (406) is connected with the end PC (7), and moves Control module (406) is equipped with the GPS module for monitoring photovoltaic panel GPS information；

Monitoring modular is mainly by the first lighting module (410), the second lighting module (413), the first thermal infrared imager (411), Two thermal infrared imagers (414) and image transmission module (412) composition, the first thermal infrared imager (411) are located at outside monitoring robot Shell (401) bottom centre, the first lighting module (410), the second lighting module (413) are located at the first thermal infrared imager (411) it is infrared that one end of two sides, monitoring robot shell (401) separate driving wheel (409) and driven wheel (408) is equipped with second Thermal imaging system (414), and the second thermal infrared imager (414) not monitored robot shells (401) blocks；Image transmission module (412) it is installed on monitoring robot shell (401) bottom；

It nearby and at the inclined surface of face photovoltaic module (1) is provided with weather station (2) positioned at photovoltaic module (1), in weather station (2) Between be fixed with thermometer screen (301), is placed with data collecting instrument (3) inside thermometer screen (301), data collecting instrument and the end PC (7) phase Even.

2. a kind of hot spot monitoring device based on environment meteorological factor according to claim 1, which is characterized in that monitoring Before, monitoring robot (4) is flush-mounted in the grooved tracks II of docking platform (5) by driving wheel (409) and driven wheel (408), Mobile platform (6) bottom is equipped with idler wheel, and mobile platform (6) drives docking platform (5) to be moved to photovoltaic under step motor control Component (1) side, the monitoring robot (4) being located on docking platform (5) slide into grooved tracks I (8) along grooved tracks II；Prison It surveys robot (4) to slide under the control of motion-control module (406) along the grooved tracks I (8) of photovoltaic module (1), movement control Molding block (406) is rotated in the case where the end PC is controlled by motor driver (403) driving motor (402), and motor (402) drives actively Driven wheel (408) rotation is driven while taking turns (409) rotation；Monitoring robot (4) on photovoltaic module (1) in sliding process, First thermal infrared imager (411) and the second thermal infrared imager (414) transmit the image of acquisition by image transmission module (412) To the end PC (7).

3. a kind of hot spot monitoring device based on environment meteorological factor according to claim 1, which is characterized in that data are adopted Collect instrument (3) and the data of acquisition are transmitted to by the end PC (7) by wireless acquisition module；Data collecting instrument (3) acquisition data include Environment meteorological factor data, environment meteorological factor data are mainly by irradiation level, environment temperature, humidity, air pressure, wind speed, wind direction, wind Pressure and rainfall composition.

4. a kind of hot spot monitoring device based on environment meteorological factor according to claim 1, which is characterized in that first shines Bright module (410), the second lighting module (413) are for providing illumination when the first thermal infrared imager (411) acquisition image；First is red Outer thermal imaging system (411) is used to shoot photovoltaic module (1) region of monitored robot (4) covering, the second thermal infrared imager (414) For shooting photovoltaic module (1) region of not monitored robot (4) covering；And pass through the first thermal infrared imager (411) or the Two thermal infrared imagers (414) obtain the temperature of hot spot in photovoltaic module (1).

5. the prediction technique of the hot spot monitoring device based on any environment meteorological factor of Claims 1 to 4, feature exist In, comprising the following steps:

1) there are the pictures of hot spot as samples pictures, the hot spot predicted value of samples pictures and every samples pictures correspondence for acquisition Related Environmental Factors data as sample data；

2) hot spot prediction model is established；

3) picture that monitoring robot acquires is inputted into hot spot prediction model as picture to be measured in real time, passes through hot spot prediction model There are the pictures to be measured of hot spot and corresponding photovoltaic panel to number for output；

4) to there are the pictures to be measured of hot spot to carry out image procossing, hot spot edge is extracted using Canny edge detection method.

6. the prediction technique of the hot spot monitoring device of environment meteorological factor according to claim 5, which is characterized in that the step It is rapid 1) specifically:

1.1) there are the pictures of hot spot as samples pictures for selection from the picture that monitoring robot (4) history acquires；

1.2) temperature that hot spot in every samples pictures is obtained by thermal infrared imager is obtained in samples pictures by image segmentation Hot spot region after calculate hot spot size in every samples pictures, pass through data collecting instrument (3) and obtain every samples pictures institute Locate the Related Environmental Factors data under time environment；

1.3) the hot spot predicted value of every samples pictures and the corresponding Related Environmental Factors data of every samples pictures are as sample Sample data is done data normalization processing, converts data to the value between [0,1] by notebook data；Then by sample data point For training sample and test sample；

Wherein hot spot predicted value is hot spot size and hot spot temperature.

7. the prediction technique of the hot spot monitoring device of environment meteorological factor according to claim 6, which is characterized in that the step It is rapid 1.2) in Related Environmental Factors data screen and obtain from environment meteorological factor data: by the heatmap in Python Each environment meteorological factor data are carried out correlation analysis with hot spot temperature and obtain corresponding correlation thermodynamic analysis figure by function； Environment meteorological factor data corresponding to the related coefficient for being greater than relevance threshold in correlation thermodynamic analysis figure are retained, are retained Environment meteorological factor data be irradiation level, temperature, humidity, wind speed, wind direction and rainfall, six reservation rings relevant to hot spot Border meteorologic factor data are as Related Environmental Factors data.

8. the prediction technique of the hot spot monitoring device of environment meteorological factor according to claim 5, which is characterized in that the step It is rapid 2) specifically:

2.1) it is used to judge the hot spot prediction model of hot spot predicted value based on the Sequential model construction in Keras；

2.2) sample data is divided into training sample and test sample, hot spot prediction model, instruction is repeatedly trained using training sample Use mean square poor (MSE), root-mean-square error (RMSE), mean absolute error (MAE) or cross entropy as loss letter when practicing Number, is trained using Adam optimization algorithm, to obtain completing the hot spot prediction model after training；

2.3) input data of the Related Environmental Factors data as hot spot prediction model, after model calculates, output layer output Hot spot size and hot spot temperature, that is, hot spot predicted value.

9. the prediction technique of the hot spot monitoring device of environment meteorological factor according to claim 5, which is characterized in that the step It is rapid 4) specifically:

4.1) every piece of photovoltaic module (1) is numbered with N-L, N is photovoltaic module (1) serial number, and L is light in photovoltaic module (1) Lie prostrate the columns of plate；In hot spot position fixing process, the end PC (7) control monitoring robot (4) on photovoltaic module (1) with a column photovoltaic Plate is the movement of unit monitoring objective, using the image of monitoring robot (4) shooting as picture transfer to be measured to passing through the end PC (7)；

4.2) the corresponding Related Environmental Factors data of picture to be measured are inputted hot spot prediction by data collecting instrument (3) by the end PC (7) Model, output valve are zero picture to be measured for indicating input there is no hot spot, and monitoring robot carries out next column photovoltaic group string Monitoring, output valve are not zero, and indicate that the picture to be measured of input there are hot spot, retains there are the picture to be measured of hot spot and output is deposited The GPS information of photovoltaic panel is corresponded in the picture to be measured of hot spot.

10. the prediction technique of the hot spot monitoring device of environment meteorological factor according to claim 5, which is characterized in that described Step 5) specifically:

5.1) to there are the pictures to be measured of hot spot to carry out image procossing: there are the pictures to be measured of hot spot to convert after gray processing For gray scale picture, the grey level histogram of gray scale picture is obtained according to gray value, later by histogram equalization processing enhancing ash Spend the contrast and brightness of picture, i.e. contrast effect in enhancing gray scale picture between hot spot and photovoltaic panel；

5.2) using the extraction of Canny edge detection method, there are the hot spot edges of the picture to be measured of hot spot:

5.2.1 it) is filtered: will be filtered through step 5.1) treated picture by the Gaussian function of discretization；

5.2.2) calculate image gradient: the gradient magnitude by calculating picture enhances the gray scale at hot spot edge and hot spot peripheral region Value difference is different；

5.2.3) non-maxima suppression: non-maxima suppression is carried out to each pixel in picture, if the gradient of current pixel is strong Degree is greater than the gradient intensity of two neighboring pixel on positive and negative gradient direction, then retains current pixel as marginal point, it is on the contrary then by Inhibit；

5.2.4) dual threshold screens: the marginal point using pixel gradient more than high threshold rejects as edge pixel and is less than Low threshold Marginal point, marginal point between high threshold and Low threshold is for connecting edge pixel, to complete the heat of picture to be measured Spot edge extracts.

Technical field

The invention belongs to technical field of photovoltaic modules, and in particular to a kind of hot spot monitoring device based on environment meteorological factor With prediction technique.

Background technique

With the continuous aggravation of climate change influenced in the world, in order to avoid negative effect grows in intensity, more Replace the mankind to the satisfaction of energy demand with clean energy resource come more country selections.And optical energy power is as in clean energy resource Indispensable a part, the also favor by extensive country.Meanwhile with scientific and technological level step up with it is technologic not Disconnected optimization, it is also further perfect by the technology and material of optical energy power, so that more and more country selections are sent out by photovoltaic panel Power station provides the demand of a part or most of electric energy, and expands the model of solar power generation in such a way that economy is effective It encloses, to reduce the utilization to non-renewable resources.

China is as the maximum photovoltaic market in the whole world, for photovoltaic industry, needs to obtain advanced technical support The advantage of technological precedence is kept, and actively deeply expands the advantage and technology of photovoltaic.And the core photovoltaic panel in photovoltaic industry, Different types of hot spot can be generated because of the adjoint influence of environment weather advantage, the shadow that the generation of hot spot generates electricity to photovoltaic panel Sound will be very fatal.

China's major part photo-voltaic power generation station and unrealized monitoring and prediction to hot spot, so as to cause because of hot spot reason The phenomenon that Generation Rate declines and generates is seen everywhere.Want to promote generating efficiency, be unable to do without the influence of hot spot, so actual light It lies prostrate in the environment of power station, it should pay attention to the monitoring and prediction to hot spot.

Summary of the invention

For overcome the deficiencies in the prior art, the hot spot monitoring device based on environment meteorological factor that the invention proposes a kind of With prediction technique, for monitoring the hot spot situation of multiple row photovoltaic panel, and pass through the relationship between hot spot and environment meteorological factor Prediction to hot spot.

The technical solution adopted by the invention is as follows:

One, a kind of hot spot monitoring device based on environment meteorological factor

The present invention includes photovoltaic module, weather station, data collecting instrument, monitoring robot, docking platform, mobile platform and PC End, photovoltaic module favour ground by support frame and place, and photovoltaic module is mainly made of the photovoltaic panel that multiple row is arranged in parallel, often Column photovoltaic panel includes multiple photovoltaic group strings, and photovoltaic module upper end is equipped with grooved tracks I；

Docking platform bottom is connected by vertical bar with mobile platform, and docking platform side is affixed with photovoltaic module side It closes, docking platform upper end is equipped with the grooved tracks II to connect with the grooved tracks I of photovoltaic module upper end；

Monitoring robot includes monitoring robot shell, motion module, monitoring modular, the first battery and the second electric power storage Pond, the first battery, the second battery are respectively that motion module and monitoring modular provide power supply；Motion module mainly by motor, Motor driver, motion-control module, driving wheel and driven wheel composition, are located at monitoring robot outer casing bottom and close to edge Equipped with the driving wheel and driven wheel along grooved tracks I and the sliding of grooved tracks II, the first battery, the second battery, electricity are mechanical, electrical Machine driver, motion-control module are respectively positioned in monitoring robot shell, and motor output shaft is connected with driving wheel, driving wheel with from Driving wheel is connected by synchronous belt, and motion-control module is connected to the motor by motor driver, motion-control module and the end PC It is connected；And motion-control module is equipped with the GPS module for monitoring photovoltaic panel GPS information.

Monitoring modular is mainly by the first lighting module, the second lighting module, the first thermal infrared imager, the second thermal infrared imager It is formed with image transmission module, the first thermal infrared imager is located at monitoring robot outer casing bottom center, the first lighting module, second Lighting module is located at the first thermal infrared imager two sides, and monitoring robot shell is installed far from one end of driving wheel and driven wheel There is the second thermal infrared imager, and the not monitored robot shells of the second thermal infrared imager are blocked；Image transmission module is installed on prison Survey robot shells bottom；

It is provided with weather station near photovoltaic module and at the inclined surface of face photovoltaic module, is fixed among weather station Thermometer screen is placed with data collecting instrument inside thermometer screen, and data collecting instrument is connected with the end PC.

Before monitoring, monitoring robot is flush-mounted in the grooved tracks II of docking platform by driving wheel and driven wheel, mobile Mesa base is equipped with idler wheel, and mobile platform drives docking platform to be moved to photovoltaic module side, is located under step motor control Monitoring robot on docking platform slides into grooved tracks I along grooved tracks II；Monitoring robot is in motion-control module It is slided under control along the grooved tracks of photovoltaic module I, motion-control module drives electricity by motor driver in the case where the end PC is controlled Machine rotation, motor drive driven wheel rotation while driving driving wheel rotation；Monitoring robot sliding process on photovoltaic module In, the image of acquisition is transmitted to the end PC by image transmission module by the first thermal infrared imager and the second thermal infrared imager.

The data of acquisition are transmitted to the end PC by wireless acquisition module by data collecting instrument；The data of data collecting instrument acquisition Including environment meteorological factor data, environment meteorological factor data are mainly by irradiation level, environment temperature, humidity, air pressure, wind speed, wind To, wind pressure and rainfall composition.

First lighting module, the second lighting module are for providing illumination when the first thermal infrared imager acquisition image；First is red Outer thermal imaging system is used to shoot the photovoltaic module region of monitored robot covering, and the second thermal infrared imager is not monitored for shooting The photovoltaic module region of robot covering；And heat in photovoltaic module is obtained by the first thermal infrared imager or the second thermal infrared imager The temperature of spot.

Two, the prediction technique of the hot spot monitoring device based on above-mentioned environment meteorological factor

The following steps are included:

1) there are the pictures of hot spot as samples pictures, the hot spot predicted value and every samples pictures of samples pictures for acquisition Corresponding Related Environmental Factors data are as sample data；

2) hot spot prediction model is established；

3) picture that monitoring robot acquires is inputted into hot spot prediction model as picture to be measured in real time, is predicted by hot spot There are the pictures to be measured of hot spot and corresponding photovoltaic panel to number for model output；

4) to there are the pictures to be measured of hot spot to carry out image procossing, hot spot edge is extracted using Canny edge detection method.

The step 1) specifically:

1.1) there are the pictures of hot spot as samples pictures for selection from the picture that monitoring robot history acquires；

1.2) temperature that hot spot in every samples pictures is obtained by thermal infrared imager, obtains sample graph by image segmentation The hot spot size in every samples pictures is calculated behind hot spot region in piece, and every samples pictures institute is obtained by data collecting instrument Locate the Related Environmental Factors data under time environment；

1.3) the hot spot predicted value and the corresponding Related Environmental Factors data work of every samples pictures of every samples pictures For sample data, sample data is done into data normalization processing, converts data to the value between [0,1]；Then by sample number According to being divided into training sample and test sample；

Wherein hot spot predicted value is hot spot size and hot spot temperature.

Related Environmental Factors data in the step 1.2) are screened from environment meteorological factor data to be obtained: being passed through Heatmap function in Python obtains each environment meteorological factor data with hot spot temperature progress correlation analysis corresponding Correlation thermodynamic analysis figure；Environment weather corresponding to the related coefficient of relevance threshold will be greater than in correlation thermodynamic analysis figure Factor data retains, and the environment meteorological factor data of reservation are irradiation level, temperature, humidity, wind speed, wind direction and rainfall, six guarantors The environment meteorological factor data relevant to hot spot stayed are as Related Environmental Factors data.

The step 2) specifically:

2.1) it is used to judge the hot spot prediction model of hot spot predicted value based on the Sequential model construction in Keras；

2.2) sample data is divided into training sample and test sample, repeatedly hot spot is trained to predict mould using training sample Type uses mean square poor (MSE), root-mean-square error (RMSE), mean absolute error (MAE) or cross entropy as damage when training Function is lost, is trained using Adam optimization algorithm, to obtain completing the hot spot prediction model after training；

2.3) input data of the Related Environmental Factors data as hot spot prediction model, after model calculates, output layer Export hot spot size and hot spot temperature, that is, hot spot predicted value.

The step 4) specifically:

4.1) every piece of photovoltaic module is numbered with N-L, N is photovoltaic module serial number, and L is photovoltaic panel in photovoltaic module Columns；In hot spot position fixing process, the end PC control monitoring robot monitors mesh as unit of on photovoltaic module by a column photovoltaic panel Mark movement, using the image of monitoring robot shooting as picture transfer to be measured to passing through the end PC；

4.2) the corresponding Related Environmental Factors data of picture to be measured are inputted hot spot by data collecting instrument and predict mould by the end PC Type, output valve are zero picture to be measured for indicating input there is no hot spot, and monitoring robot carries out the prison of next column photovoltaic group string It surveys, output valve is not zero, and indicating the picture to be measured inputted, there are hot spots, retains there are the picture to be measured of hot spot and exports presence The picture to be measured of hot spot corresponds to the GPS information of photovoltaic panel, to complete the accurate positionin of hot spot.

The step 5) specifically:

5.1) to there are the pictures to be measured of hot spot to carry out image procossing: there are the pictures to be measured of hot spot after gray processing It is converted into gray scale picture, the grey level histogram of gray scale picture is obtained according to gray value, is increased later by histogram equalization processing The contrast and brightness of strong lime degree picture, i.e. contrast effect in enhancing gray scale picture between hot spot and photovoltaic panel；

5.2) using the extraction of Canny edge detection method, there are the hot spot edges of the picture to be measured of hot spot:

5.2.1 it) is filtered: will be filtered through step 5.1) treated picture by the Gaussian function of discretization；

5.2.2) calculate image gradient: the gradient magnitude by calculating picture enhances hot spot edge and hot spot peripheral region Grey value difference；

5.2.3) non-maxima suppression: non-maxima suppression is carried out to each pixel in picture, if the ladder of current pixel The gradient intensity that intensity is greater than two neighboring pixel on positive and negative gradient direction is spent, then retains current pixel as marginal point, it is on the contrary Then it is suppressed；

5.2.4) dual threshold screens: the marginal point using pixel gradient more than high threshold is rejected as edge pixel less than low The marginal point of threshold value, the marginal point between high threshold and Low threshold is for connecting edge pixel, to complete picture to be measured Hot spot edge extracting.

The beneficial effects of the present invention are:

1) present invention is derived from influence of the environment meteorological factor for hot spot, to pass through analysis environment meteorological factor and hot spot Relationship, reach the prediction and preliminary positioning to hot spot；The method predicted using hot spot, by being carried out to every piece of photovoltaic panel Default number can be with to realize to the Primary Location of hot spot, then the monitoring by carrying out hot spot to the photovoltaic panel of prediction positioning Improve the accuracy of full-automatic hot spot monitoring and prediction；While the judgement of making prediction property of the case where being hot spot, and have one Fixed reference value.

2) data that the present invention is provided by multiple sensors, and can be achieved with by simple system and device to photovoltaic The monitoring and prediction of hot spot in power station, a large amount of data can be realized the promotion of detection and forecasting accuracy in real time, and It is reduced while realizing that efficiency improves corresponding artificial.

3) the advantages of present invention is realized by simple system and device, is suitable for general photo-voltaic power generation station, and with can hold The core concept of supervention exhibition is adapted.

Detailed description of the invention

Fig. 1 is the schematic diagram of apparatus of the present invention；

Fig. 2 is photovoltaic module default number schematic diagram；

Fig. 3 is the structural schematic diagram of robot and track；

Fig. 4 is the motion module structural schematic diagram of robot；

Fig. 5 is the monitoring modular structural schematic diagram of robot；

Fig. 6 is the structural schematic diagram of data collecting instrument；

Fig. 7 is photovoltaic panel schematic diagram and circled is hot spot figure；

Fig. 8 is hot spot monitoring hardware system block diagram；

Fig. 9 is hot spot forecasting system block diagram；

Figure 10 is the system block diagram that hot spot is extracted；

Figure 11 is overall workflow figure of the invention.

In figure: 1. photovoltaic modulies, 2. weather stations, 3. data collecting instruments, 4. monitoring robots, 5. docking platforms, 6. movements Platform, the end 7.PC, 8. grooved tracks I, 301. thermometer screens, 401. monitoring robot shells, 402. motors, 403. motor drivens Device, 404. first batteries, 405. second batteries, 406. motion-control modules, 407. synchronous belts, 408. driven wheels, 409. driving wheels, 410. first lighting modules, 411. first thermal infrared imagers, 412. image transmission modules, 413. second illuminations Module, 414. second thermal infrared imagers.

Specific embodiment

Present invention will be further explained below with reference to the attached drawings and examples, so that the understanding of the present invention is clearer. It needs to stress, only introduce main contents of the invention herein, some known functions and detailed description will be neglected Slightly.

Apparatus of the present invention are all suitable in general photo-voltaic power generation station, and general photo-voltaic power generation station is in the laying about photovoltaic panel When all use several pieces from top to bottom, the neighbour mode in left and right connects in a row, and carries out the laid parallel of multiple row.It is simultaneously It is in photovoltaic panel under the conditions of optimum illumination, certain fixed angle, and the placement for photovoltaic panel will be formed with ground The reflection because of photovoltaic panel to sunlight should be prevented and lead to the inaccurate problem of weather station work.

As shown in Figure 1, the present invention includes photovoltaic module 1, weather station 2, data collecting instrument 3, monitoring robot 4, docks and put down Platform 5, mobile platform 6 and the end PC 7, photovoltaic module 1 favour ground by support frame and place, and 1 upper end of photovoltaic module is equipped with groove Track I 8.

As shown in Fig. 2, photovoltaic module 1 is mainly made of the photovoltaic panel that multiple row is arranged in parallel, each column photovoltaic panel includes multiple Photovoltaic group string is numbered every piece of photovoltaic module 1 with N-L, and N is 1 serial number of photovoltaic module, and L is photovoltaic panel in photovoltaic module 1 Columns；In hot spot position fixing process, the end PC 7 controls monitoring robot 4 and monitors as unit of on photovoltaic module 1 by a column photovoltaic panel Target movement, using image that monitoring robot 4 is shot as picture transfer to be measured to passing through the end PC 7；Then robot is clapped The image taken the photograph is handled, if there is no the monitorings that hot spot can carry out next photovoltaic group string after image procossing, if it exists Hot spot is that can determine the positioning of hot spot, to accurately find out the position where hot spot.

As shown in figure 3,5 bottom of docking platform is connected by vertical bar with mobile platform 6,5 side of docking platform and photovoltaic 1 side of component fits, and 5 upper end of docking platform is equipped with the grooved tracks II to connect with the grooved tracks I 8 of 1 upper end of photovoltaic module.

As shown in figure 4, monitoring robot 4 includes monitoring robot shell 401, motion module, monitoring modular, the first electric power storage Pond 404 and the second battery 405, the first battery 404, the second battery 405 are respectively that motion module and monitoring modular provide Power supply；Motion module is mainly by motor 402, motor driver 403, motion-control module 406, driving wheel 409 and driven wheel 408 Composition is slided positioned at 401 bottom of monitoring robot shell and being equipped with close to edge along grooved tracks I 8 and grooved tracks II Driving wheel 409 and driven wheel 408, the first battery 404, the second battery 405, motor 402, motor driver 403, movement control Molding block 406 is respectively positioned in monitoring robot shell 401, and 402 output shaft of motor is connected with driving wheel 409, driving wheel 409 with from Driving wheel 408 is connected by synchronous belt, and motion-control module fills 406 containing GPS module for monitoring photovoltaic panel GPS information, Motion-control module 406 is connected by motor driver 403 with motor 402, and motion-control module 406 is connected with the end PC 7.

As shown in figure 5, monitoring modular is mainly by the first lighting module 410, the second lighting module 413, the first infrared thermal imagery Instrument 411, the second thermal infrared imager 414 and image transmission module 412 form, and the first thermal infrared imager 411 is located at monitoring robot 401 bottom centre of shell, the first lighting module 410, the second lighting module 413 are located at 411 two sides of the first thermal infrared imager, Monitoring robot shell 401 is equipped with the second thermal infrared imager 414, and second far from one end of driving wheel 409 and driven wheel 408 The not monitored robot shells 401 of thermal infrared imager 414 are blocked；Image transmission module 412 is installed on monitoring robot shell 401 Bottom；First lighting module 410, the second lighting module 413 are for providing illumination when the first thermal infrared imager 411 acquisition image； First thermal infrared imager 411 is used to shoot 1 region of photovoltaic module that monitored robot 4 covers, and the second thermal infrared imager 414 is used In 1 region of photovoltaic module that the not monitored robot 4 of shooting covers.

As shown in fig. 6, being located at photovoltaic module 1 nearby and being provided with weather station 2, gas at the inclined surface of face photovoltaic module 1 As being fixed with thermometer screen 301 among station 2, data collecting instrument 3, data collecting instrument and 7 phase of the end PC are placed with inside thermometer screen 301 Even.Data collecting instrument 3 has thermometer screen 301 to be protected, and prevents because of non-artificial destruction caused by meteorologic factor, and ensure The stability and safety of data transmission.

As shown in fig. 7, show 1 schematic diagram of photovoltaic panel taken by photovoltaic panel monitoring robot 4 shown by the end PC 7, Circled is expressed as hot spot simultaneously.

Before monitoring, monitoring robot 4 is flush-mounted in the grooved tracks II of docking platform 5 by driving wheel 409 and driven wheel 408 Interior, 6 bottom of mobile platform is equipped with idler wheel, and mobile platform 6 drives docking platform 5 to be moved to photovoltaic module under step motor control 1 side, the monitoring robot 4 on docking platform 5 slide into grooved tracks I 8 along grooved tracks II；Monitoring robot 4 exists It is slided under the control of motion-control module 406 along the grooved tracks I 8 of photovoltaic module 1, motion-control module 406 controls at the end PC It is rotated by 403 driving motor 402 of motor driver down, motor 402 drives driving wheel 409 to drive driven wheel while rotation 408 rotations；Monitoring robot 4 is on photovoltaic module 1 in sliding process, the first thermal infrared imager 411 and the second thermal infrared imager The image of acquisition is transmitted to the end PC 7 by image transmission module 412 by 414；

The data of acquisition are transmitted to the end PC 7 by wireless acquisition module by data collecting instrument 3；What data collecting instrument 3 acquired Data include environment meteorological factor data, and environment meteorological factor data are mainly by irradiation level, environment temperature, humidity, air pressure, wind Speed, wind direction, wind pressure and rainfall composition；The data that data collecting instrument 3 acquires further include the temperature of hot spot in photovoltaic module 1, photovoltaic 1 bottom of component is equipped with the paster component being connected with data collecting instrument 3 by wireless transport module, and data collecting instrument 3 passes through patch Piece component obtains the temperature of hot spot in photovoltaic module 1.

As Fig. 8, Fig. 9, Figure 10 and Figure 11 respectively indicate the hardware system of hot spot monitoring, the system block diagram that hot spot is predicted, heat The work flow diagram and overall workflow figure of the invention that spot extracts.