CN108985163A - The safe detection method of restricted clearance based on unmanned plane - Google Patents
The safe detection method of restricted clearance based on unmanned plane Download PDFInfo
- Publication number
- CN108985163A CN108985163A CN201810595823.1A CN201810595823A CN108985163A CN 108985163 A CN108985163 A CN 108985163A CN 201810595823 A CN201810595823 A CN 201810595823A CN 108985163 A CN108985163 A CN 108985163A
- Authority
- CN
- China
- Prior art keywords
- layer
- convolutional neural
- neural networks
- bar
- restricted clearance
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64C—AEROPLANES; HELICOPTERS
- B64C39/00—Aircraft not otherwise provided for
- B64C39/02—Aircraft not otherwise provided for characterised by special use
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64D—EQUIPMENT FOR FITTING IN OR TO AIRCRAFT; FLIGHT SUITS; PARACHUTES; ARRANGEMENTS OR MOUNTING OF POWER PLANTS OR PROPULSION TRANSMISSIONS IN AIRCRAFT
- B64D45/00—Aircraft indicators or protectors not otherwise provided for
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N7/00—Television systems
- H04N7/18—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
- H04N7/183—Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast for receiving images from a single remote source
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B64—AIRCRAFT; AVIATION; COSMONAUTICS
- B64U—UNMANNED AERIAL VEHICLES [UAV]; EQUIPMENT THEREFOR
- B64U2101/00—UAVs specially adapted for particular uses or applications
Abstract
This application involves the safe detection methods of restricted clearance based on unmanned plane, it uses the information collecting device of the safe sensorcraft of restricted clearance to carry out safety monitoring information collection to the region that restricted clearance domestic demand security monitoring detects, and recognition result is exported after safety monitoring information operating safety monitoring information identifying method collected to be carried out to the identification operation of safety monitoring information in safety monitoring information identification module, the safe detection method of restricted clearance provided by the present application based on unmanned plane, personnel are replaced to detect the case where restricted clearance with unmanned plane, it can be according to the defect level of the facility in the restricted clearance detected, horse back arrangement personnel are chosen whether to safeguard, still facility is continued to use, because individual defects degree not necessarily influences the work of facility, until defect to a certain extent when repair again, it is more advantageous to promotion The effective rate of utilization of resource, and it is more safe and intelligent that the safety monitoring of restricted clearance is worked.
Description
[technical field]
This application involves safe field of detecting, more particularly to the safe detection method of the restricted clearance based on unmanned plane.
[background technique]
In energy industry, such as coal mining, now all still need by professional enters in advance inside check,
Ensure safety, this be for inspection personnel it is breakneck, inspection environment uncertainty it is very big, for another example in petrochemical industry
Petrochemical industry oil storage tank (for storing gasoline, the bavin oil and gas etc.) needs of secure context, petrochemical factory are inspected periodically, mesh
The method of preceding inspection is to place the regular hour after the petroleum or natural gas inside equal oil storage tanks evacuate, then related personnel
Gas-detecting device is tied up on the surpluses such as bamboo pole and then is stretched inside oil storage tank, the gas to check the inside is then further taken out
Content's index can further allow related personnel to enter the health condition such as tank body that oil storage tank checks oil storage tank if qualified
There are free from flaw, rusty stain etc..And this detection method gas-detecting device is put in further take out come method be it is defective, because
For when gas-detecting device is taken out gas content index can than inside when change, the standard of data
Really insufficient and detection region is also extremely limited, works in addition, personnel is allowed to enter in this uncertain environment in the presence of very big
Risk.
We can be referred to as restricted clearance to above-mentioned examined environment, restricted clearance further include various people be difficult to reach or
It is the higher building of hazard index or equipment, such as the various equipment of factory or the inside of building or is difficult to outside (such as work reached
Furnace, tower reactor, tank, storehouse, pond, tank car, pipeline, flue of factory etc.) and city tunnel, sewer, ditch, hole, well, pond, culvert, valve
The closing such as Men Jian, sewage treatment facility, semi-enclosed facility and place (such as cabin, underground concealed engineering, closed container, long-term
Unused facility or unsmooth place etc. of divulging information) and rural area storage sweet potato, potato, the well of various vegetables, cellar etc..Ventilation is not
Good mine should also treat as restricted clearance.
Based on this, a kind of new paragon that can replace or assist people to be detected to restricted clearance is needed.
[summary of the invention]
In order to solve the above technical problems, the restricted clearance safety detection side of the application being designed to provide based on unmanned plane
Method replaces personnel to detect the case where restricted clearance with unmanned plane, and the personnel of greatly reducing enter the uneasiness of restricted clearance
Quan Xing, it is also very convenient to detect, and enhances the timeliness and stability of judging result.It can be according in the restricted clearance detected
Facility defect level, choose whether horse back arrangement personnel and safeguard, still continue to use facility, because of individual defects journey
Degree not necessarily influence facility work, until defect to a certain extent when repair again, be more advantageous to promoted resource it is effective
Utilization rate, and it is more safe and intelligent that the safety monitoring of restricted clearance is worked.
The application is achieved through the following technical solutions:
The safe detection method of restricted clearance based on unmanned plane, using the information collection of the safe sensorcraft of restricted clearance
The region that device detects restricted clearance domestic demand security monitoring carries out safety monitoring information collection, and by safety monitoring collected
Information operating safety monitoring information identifying method carries out the identification operation of safety monitoring information in safety monitoring information identification module
After export recognition result.
The safe detection method of restricted clearance as described above based on unmanned plane, the information collecting device are camera,
The safety monitoring information is image information, and the safety monitoring information identification module is picture recognition module, the safety prison
Measurement information recognition methods is image-recognizing method, and the camera acquired image information is in picture recognition module with figure
As recognition methods to image information identify operation after export recognition result, described image recognition methods the following steps are included:
S21 classifies the detection scene of restricted clearance, and categorized good scarce for every a kind of probe field scape matching
The identification image for the degree of falling into is fabricated to a training dataset, the detection scene be classified as feed channel burner, support storage tank,
Inside industrial chimney, large oil tank ship is internal, inside big pipeline or coal-fired power station boiler, the defect level be heavy corrosion,
Moderate corrosion or mild corrosion, the recognition result are heavy corrosion, moderate corrosion or mild corrosion;
S22, the training data inputted in step S21 concentrate the identification image that detection scene and defect level have been classified
And carry out image sharpening and edge enhancing processing;
S23 instructs convolutional neural networks using through the step S22 training dataset that treated identifies that image is formed
Practice;
S24, by the image information acquired in step S1 input in the convolutional neural networks after step S22 training, output
Recognition result.
The safe detection method of restricted clearance as described above based on unmanned plane, to convolutional neural networks described in step S23
Be trained the following steps are included:
S231 successively carries out first time process of convolution to the identification image concentrated through step S22 treated training data,
For convolutional neural networks first layer;
S232: the processing of first time pondization is carried out to the convolutional neural networks first layer, local acknowledgement's normalized is
The convolutional neural networks second layer;
S233 carries out second of process of convolution to the convolutional neural networks second layer, is convolutional neural networks third layer;
S234 carries out second of pondization processing to the convolutional neural networks third layer, is the 4th layer of convolutional neural networks;
S235 obtains convolutional neural networks layer 5 to the 4th layer of the convolutional neural networks in a manner of connecting entirely;
S236, the convolutional neural networks layer 5 is obtained output layer in a manner of connecting entirely: output layer is divided into 6
Independent part, each part are additionally connected with an independent loss function;
S237 minimizes the value of the loss function using good identification image categorized in S21, completes to convolution mind
Training through network.
The safe detection method of restricted clearance as described above based on unmanned plane, will be through step S22 treated qualification figure
The size of picture is unified for 32 × 32 × 1, and the filter size of convolutional layer is 5 × 5, and depth 6 is supplemented without using full 0, and step-length is
1, the Output Size for obtaining the convolutional neural networks first layer is 32-5+1=28, depth 6, i.e., the described convolutional neural networks
Second layer node matrix equation has 28 × 28 × 6=4704 node, and each node and 5 × 5=25 current node layers are connected, that is, roll up
A total of 4704 × (25+1)=122304 connection of product neural network first layer;
The filter size that the convolutional neural networks second layer uses is 2 × 2, step-length 2, the convolutional neural networks
Second layer output matrix size is 14 × 14 × 6;
The convolutional neural networks second layer output matrix is the input of the convolutional neural networks third layer, the mistake of use
Filter size is 5 × 5, and depth 16 is supplemented without using full 0, and the Output Size of step-length 1, convolutional neural networks third layer is
14-5+1=10, depth 16, i.e. convolutional neural networks third layer output matrix size are 10 × 10 × 16, convolutional neural networks
Third layer parameter has 5 × 5 × 6 × 16+16=2416, there is 10 × 10 × 16 × (5 × 5+1)=41600 connection;
The input matrix size that the 4th layer of the convolutional neural networks is 10 × 10 × 16, and the filter size used is 2
× 2, step-length 2, the output matrix size that the 4th layer of the convolutional neural networks is 5 × 5 × 16;
The output matrix that the 4th layer of the convolutional neural networks is the input of the convolutional neural networks layer 5, is drawn
The vector for being directly 5 × 5 × 16 for a length, i.e., the one-dimensional space is straightened in vector form in a three-dimensional matrice indicates,
It is trained into full articulamentum, the output node number of the convolutional neural networks layer 5 is 120, that is, shares 5 × 5 × 16
× 120+120=48120 parameter.
The safe detection method of restricted clearance as described above based on unmanned plane, the information collecting device further includes being used for
To the gas sensor that the gas information in the region of restricted clearance domestic demand security monitoring detection is acquired, the safety monitoring letter
Breath is gas information, and the safety monitoring information identification module is gas identification module, the safety monitoring information identifying method
For Gas Distinguishing Method, the gas sensor gas information collected uses Gas Distinguishing Method on gas identification module
Recognition result is exported after identifying operation to gas information.
The safe detection method of restricted clearance as described above based on unmanned plane, the safe sensorcraft of restricted clearance
Including drone body and the anticollision frame being located at outside the drone body;The drone body includes rack and peace
Holder above the rack, the rack are equipped with underneath type rotor, and the camera is located on the holder, and institute
It states camera side and is equipped with the irradiation light mobile with the camera;The anticollision frame includes the first hemisphere frame, the second hemisphere
The tangent plane edge of frame and connecting rod, the first hemisphere frame and the second hemisphere frame is interconnected to form by the connecting rod
The anticollision frame.
The safe detection method of restricted clearance as described above based on unmanned plane, multiple connecting rods are along described the first half
The tangent plane edge of ball frame and the second hemisphere frame, which is uniformly distributed to be formed, connects the first hemisphere frame and the second hemisphere frame
Cylindric frame.
The safe detection method of restricted clearance as described above based on unmanned plane, the rack includes pedestal, the pedestal
Two sides are symmetrically arranged with installing frame, and the installing frame includes being connected on the pedestal to stretch out the pedestal side in splayed is isometric
The first bar and the second bar, the third bar that is connect with first bar extension end end and connect with second bar extension end
The free end of 4th bar, the third bar and the 4th bar is joined the apex angle interconnecting piece to be formed and be connect with the anticollision frame,
The underneath type rotor is located at below the interconnecting piece of first bar and the third bar and second bar and the described 4th
Below the interconnecting piece of bar;
First bar, second bar, the third bar and the 4th bar head and the tail successively hingedly form the institute of diamond shape
Installing frame is stated, and there is the retention bead being located on the pedestal between first bar and the second bar interior angle and be used for
First bar and second bar is set to be adjacent to the limit convex toward relative direction drawing on first bar and second bar
The first back-moving spring risen is equipped with second back-moving spring between the third bar and the 4th bar interior angle.
The safe detection method of restricted clearance as described above based on unmanned plane, the first hemisphere frame include:
Moulding, is equipped with a plurality of, and a plurality of moulding one end is collapsed, and the other end opens to form hemisphere frame-shaped, and the moulding collapses
One end be equipped with the card convex protruded to outside the first hemisphere frame top dome, one end that the moulding opens is equipped with to described the first half
The card that ball frame section is centrally recessed into is recessed;
Ring is collapsed, is located at one end that the moulding collapses, the gathering ring includes looping pit, around the looping pit
Annular collapses wall, collapses the equally distributed annular groove of wall circular along the annular and be provided with the outer ring of the annular wall outer wall
Screw thread, the moulding inlay card is in the annular groove and the card convex collapses in the looping pit so that a plurality of moulding one
End forms rounding state;
Ring is screwed, equipped with the fixation hole mutually fixed with rack, wall is screwed around the annular of the fixation hole and opens up
The inner ring screw thread matched on wall inner wall with the outer ring screw thread is screwed in the annular, after the gathering ring collapses the moulding
It is screwed in ring described in precession;
Expanded loop, along circumference be uniformly provided with the card hole equal with the moulding quantity and with the connecting rod quantity phase
Deng lockhole, one end that the moulding opens sticks into the card hole;
Connection ring is uniformly provided with the via hole equal with the connecting rod quantity along circumference;
Attachment screw, the attachment screw pass through the lockhole and the via hole and lock the expanded loop and the connection ring
Tightly in the connecting rod.
The safe detection method of restricted clearance as described above based on unmanned plane, the safe sensorcraft of restricted clearance
Be additionally provided in real time by the location information of the safe sensorcraft of the restricted clearance be transmitted to control terminal GPS module and
For marking the label rifle of the trouble location of restricted clearance, the label rifle is the automatic spray gun that can spray colored solution.
Compared with prior art, the application has the following advantages:
1, personnel are replaced to detect the case where restricted clearance with unmanned plane, the personnel of greatly reducing enter restricted clearance
Insecurity, it is also very convenient to detect, and enhances the timeliness and stability of judging result.It can be limited according to what is detected
The defect level of facility in space chooses whether horse back arrangement personnel and safeguards, facility still continued to use, because individually
Defect level not necessarily influences the work of facility, until defect to a certain extent when repair again, be more advantageous to promotion resource
Effective rate of utilization, and make restricted clearance safety monitoring work more safety and intelligence.
2, light-weighted anticollision frame can spring back or rotate when colliding with object, can effectively prevent unmanned plane especially
It is that rotor is not damaged, even can be from high-altitude free falling without damage, being more advantageous to unmanned plane can answer in strange
It flies in miscellaneous environment, greatly reduces the incidence of " aircraft bombing ", guarantee that unmanned plane is capable of the information of normal acquisition restricted clearance
And manipulation difficulty to unmanned plane can be substantially reduced.
3, it is connected on the output shaft of close coupled type DC brushless motor above underneath type rotor, although underneath type rotor is stablized
Property declined, but the mobility of unmanned plane is stronger, is more advantageous to it and avoids obstacle in time in complicated restricted clearance, mentions
The safety of unmanned plane is risen, and camera of being more convenient for is mounted on above unmanned plane to shoot broader upper images.
4, cylindric frame is set between the first hemisphere frame and the second hemisphere frame, so that anticollision frame is falling
During falling to the ground, on the one hand still a part of impact force can be offset by rolling, on the one hand cylindric frame can be with
The rolling that line contact type is carried out with ground prevents circular anticollision frame and ground from carrying out long-term unstable point contact and rolls
Multi-faceted irregular overturning is caused to drone body and increases and destroys probability.
[Detailed description of the invention]
In order to more clearly explain the technical solutions in the embodiments of the present application, make required in being described below to embodiment
Attached drawing is briefly described, it should be apparent that, the drawings in the following description are only some examples of the present application, for
For those of ordinary skill in the art, without creative efforts, it can also be obtained according to these attached drawings other
Attached drawing.
Fig. 1 is structural representation Fig. 1 of the embodiment of the present application 1;
Fig. 2 is structural representation Fig. 2 of the embodiment of the present application 1;
Fig. 3 is the portion the A enlarged drawing of Fig. 2;
Fig. 4 is the portion the B enlarged drawing of Fig. 2;
Fig. 5 is the decomposition texture schematic diagram of the embodiment of the present application 1;
Fig. 6 is gathering ring and the decomposition texture schematic diagram for screwing ring described in the embodiment of the present application 1;
Fig. 7 is the structural schematic diagram of rack described in the embodiment of the present application 1;
Fig. 8 is the structural schematic diagram that rifle is marked described in the embodiment of the present application 1;
Fig. 9 is the outlet schematic diagram of nozzle described in the embodiment of the present application 1;
Figure 10 is the structural schematic diagram of rack described in the embodiment of the present application 2;
Figure 11 is in the embodiment of the present application 1 in the image information figure taken in restricted clearance;
Figure 12 is image information in Figure 11 by treated the image information figure of step S22 described in embodiment 1;
Figure 13 is that 1 described image of the embodiment of the present application identifies schematic diagram;
Figure 14 is the schematic illustration of process of convolution described in the embodiment of the present application 1;
Figure 15 is the schematic illustration of the processing of pondization described in the embodiment of the present application 1.
[specific embodiment]
The technical issues of in order to keep the application solved, technical solution and beneficial effect are more clearly understood, below in conjunction with
The application is further elaborated in accompanying drawings and embodiments.It should be appreciated that specific embodiment described herein is only used
To explain the application, it is not used to limit the application.
As shown in Figures 1 to 10, the embodiment of the present application proposes the safe sensorcraft of restricted clearance, including drone body
1 and the anticollision frame 2 that is located at outside the drone body 1;The drone body 1 includes rack 11 and is mounted on described
Holder 12 above rack, the rack 11 are equipped with underneath type rotor 110, and the holder 12 is equipped with camera 121, with institute
State the mobile irradiation light 122 of camera 121 and the GPS mould for the location information of unmanned plane to be transmitted to control terminal in real time
Block;The anticollision frame 2 includes the first hemisphere frame 21, the second hemisphere frame 22 and connecting rod 23, the first hemisphere frame 21 and institute
The tangent plane edge for stating the second hemisphere frame 22 is interconnected to form the anticollision frame 2 by the connecting rod 23.Above-mentioned limited sky
Between safe sensorcraft people can be replaced to carry out information collection, the work such as safety detection to restricted clearance.
Obstacle avoidance module can also be set in the drone body, such as infrared probe or ultrasonic probe, be conducive to by
Limit space safety sensorcraft avoids obstacle in complicated restricted clearance, nevertheless, it is horizontal to the manipulation of operator
It is required that still higher, above-mentioned anticollision frame can substantially reduce the manipulation difficulty to the safe sensorcraft of restricted clearance, and more
Protect to limits the safe sensorcraft of restricted clearance.Specifically, the whole maximum gauge of the anticollision frame 2 is less than
500mm, material are carbon fiber, and light-weighted anticollision frame can spring back or rotate when colliding with object, can effectively prevent
Only the safe sensorcraft of restricted clearance especially rotor is not damaged, even can from high-altitude free falling without damage,
Being more advantageous to the safe sensorcraft of restricted clearance can fly in strange complex environment, greatly reduce " aircraft bombing "
Incidence guarantees that the safe sensorcraft of restricted clearance is capable of the information of normal acquisition restricted clearance.
In restricted clearance, often complicated condition, dark are closed, it is easier to lead to the safe sensorcraft of restricted clearance and barrier
Hinder object collision to lead to not fly, front can be illuminated with the mobile irradiation light of the camera, be more advantageous to control terminal according to
The safe sensorcraft flight of the image control restricted clearance of thecamera head, while being conducive to restricted clearance and detecting nobody safely
Machine can more clearly acquire image information.
More specifically, irradiation light selects the LED light of high intense light irradiation and is mounted on the side of camera, in addition, unmanned plane sheet
Body is equipped with high capacity cell, and preferably lithium battery is as energy resource supply.Further, winged space mini3D can be selected in the holder
Pro, holder can be used as installation, the support equipment of fixed video camera even can adjust the level of video camera and the angle of pitching.By
The data for limiting space safety sensorcraft, which are transmitted, can be selected 3DR wireless data transmission module, WIFI wireless data transmission module or 4G net
Network transmission module.Control terminal includes remote controler and monitor, and remote controler can control flying for the safe sensorcraft of restricted clearance
Row state, monitor can obtain its of the image information of the safe sensorcraft of restricted clearance, location information and acquisition immediately
His environmental information.
It is connected on the output shaft of close coupled type DC brushless motor above underneath type rotor, although underneath type rotor stability
Declined, but the mobility of the safe sensorcraft of restricted clearance is stronger, is more advantageous to it in complicated restricted clearance
Obstacle is avoided in time, promotes the safety of the safe sensorcraft of restricted clearance, and camera of being more convenient for is mounted on restricted clearance
To shoot broader upper images above safe sensorcraft.
Preferably, tangent plane edge of multiple connecting rods 23 along the first hemisphere frame 21 and the second hemisphere frame 22
It is uniformly distributed the cylindric frame to be formed and connect the first hemisphere frame 21 and the second hemisphere frame 22.Specifically, the company
Extension bar 23 is equipped with 4, cylindric frame is arranged between the first hemisphere frame and the second hemisphere frame, so that anticollision frame
During dropping into ground, on the one hand still a part of impact force can be offset by rolling, on the one hand cylindric frame
The rolling that line contact type can be carried out with ground prevents circular anticollision frame and ground from carrying out long-term unstable point contact
It rolls and is caused by multi-faceted and irregular overturning and increases and destroys probability for drone body.
In the present embodiment 1, the rack 11 includes pedestal 111, and 111 two sides of pedestal are symmetrically arranged with installing frame 112,
The installing frame 112 includes being connected on the pedestal 111 in isometric the first bar for stretching out 111 side of pedestal of splayed
It 113 and second bar 114, the third bar 115 that is connect with the 113 extension end end of the first bar and is stretched out with second bar 114
It joins to be formed and the anticollision frame 2 in the free end of 4th bar 116 of end connection, the third bar 115 and the 4th bar 116
The apex angle interconnecting piece 117 of connection, the underneath type rotor 110 are located at the interconnecting piece of first bar 113 and the third bar 115
Below the interconnecting piece of lower section and second bar 114 and the 4th bar 116.Specifically, first bar 113, the second bar
114, multiple fabrication holes are all offered on third bar 115 and the 4th bar 116, can effectively mitigates rack weight, the machine of frame-shaped
Frame is symmetrically formed four for installing the installation points of rotor in two sides of base, and structure is simple, lightweight, and be conducive to rotor can
It is installed by stablizing.
In the present embodiment 1, the first hemisphere frame 21 includes: moulding 211, is equipped with a plurality of, a plurality of 211 one end of moulding
It collapses, the other end opens to form hemisphere frame-shaped, and one end that the moulding 211 collapses is equipped with to outside 21 top dome of the first hemisphere frame
The card convex 2111 of protrusion, it is recessed that one end that the moulding 211 opens is equipped with the card being centrally recessed into 21 section of the first hemisphere frame
2112;Ring 212 is collapsed, one end that the moulding 211 collapses is located at, the gathering ring 212 includes looping pit 2121, surrounds institute
The annular for stating looping pit 2121 collapses wall 2122, the equally distributed annular groove 2123 of the annular of wall 2122 is collapsed along the annular and
It is provided with the outer ring screw thread 2124 of 1142 outer wall of annular wall, 211 inlay card of moulding is in the annular groove 2123 and institute
Card convex 2111 is stated to collapse in the looping pit 2121 so that 211 one end 5 of a plurality of moulding forms rounding state;Screw ring
213, equipped with the fixation hole 2131 fixed with 11 phase of rack, wall 2132 is screwed around the annular of the fixation hole 2131 and is opened
It is located at the annular and screws the inner ring screw thread 2133 matched on 2132 inner wall of wall with the outer ring screw thread 2124, the gathering ring
212 collapse 211 back spin of moulding screws in ring 213 into described;Expanded loop 214 is uniformly provided with and the moulding along circumference
The equal card hole 2141 of 211 quantity and the lockhole 2142 equal with 23 quantity of connecting rod, the one of the opening of moulding 211
End sticks into the card hole 2141;Connection ring 215 is uniformly provided with the via hole equal with 23 quantity of connecting rod along circumference;
Attachment screw, the attachment screw pass through the lockhole 2142 and the via hole for the expanded loop 214 and the connection ring 215
It is locked in the connecting rod 23.
The structure of the second hemisphere frame 22 is symmetrical set with the first hemisphere frame 21.Spherical structure to be limited
Space safety sensorcraft is more uniformly stressed when by colliding, more to the protection of the safe sensorcraft of restricted clearance
Comprehensively, and sound construction, it is dismounted easily.
The safe sensorcraft of restricted clearance further includes the label rifle 3 for marking the trouble location of restricted clearance.
After restricted clearance safe sensorcraft discovery trouble location, label rifle can be used, sign flag is carried out to trouble location,
Such as stamp X-shaped or triangle pattern, it might even be possible to which stamping the pattern with fluorescent effect keeps label more prominent, and label can certainly
Act near trouble location or trouble location, in conjunction with the positioning of GPS module so that personnel the facility to restricted clearance into
When row repair and maintenance can it is more intuitive, accurately and quickly find trouble location.
Preferably, the label rifle 3 is the automatic spray gun that can spray colored solution.The colored solution may be selected as red
Color solution can be added fluorescent powder in certain solution and make the distincter protrusion of indicia patterns.Specifically, the automatic spray gun packet
Include: nozzle 31 is equipped with exporting so that the colored solution sprayed is distributed in X shape for X-shaped;Hydrojet channel 32, outlet with it is described
Nozzle 31 is connected to;Solution bottle 33 is used for the colored solution of holding, the interior imbibition channel being equipped with for flowing for liquid to bottleneck
34, the solution bottle 33 is equipped with the first check valve 30 entered in bottle for air;Plunger shaft 35, with the hydrojet channel 32
Be connected to the imbibition channel 34, and be equipped between the plunger shaft 35 and 32 entrance of hydrojet channel control colored solution from
The second one-way valve 36 that the plunger shaft 35 is flowed to 31 direction of nozzle, the plunger shaft 35 and the imbibition channel 34 it
Between be equipped with and control the third check valve 37 that colored solution is flowed from the solution bottle 33 to 35 direction of plunger shaft, the piston
It is equipped with the piston 38 that can be moved back and forth in the plunger shaft 35 in chamber 35, is equipped with piston thread hole in the middle part of the piston 38;It drives
Dynamic motor 39 is equipped with and has externally threaded machine shaft 391, with control in piston thread hole described in the machine shaft precession
The reciprocating motion of the piston 38.
When work, driving motor rotate forward, drive piston in plunger shaft retraction with the solution in draw solution bottle to piston
In chamber, then driving motor is inverted, and driving piston quickly moves forward in plunger shaft, by the solution sucked in plunger shaft through nozzle
It ejects.
It is additionally provided with gas sensor on the holder 12, multiple sensors, including such as one can be integrated inside gas sensor
Carbonoxide gas sensor, gas gas sensor, freon gas dependent sensor etc., can measure the type of gas, concentration and at
Point, and it is converted into electric signal.On the one hand the safety problems such as the gaseous contamination of detection restricted clearance, gas leakage, on the one hand shift to an earlier date
The gas hazard degree for detecting restricted clearance promotes degree of safety when subsequent personnel enter restricted clearance maintenance.
It should be noted that be additionally provided with master control borad in the drone body, the camera, GPS module, irradiation light,
It driving motor, holder, gas sensor and controls the electric devices such as motor of rotor rotational and is electrically connected with master control borad to realize
Control is realized and such as switchs, rotates movement.
The present embodiment also provides the safe detection method of restricted clearance based on unmanned plane, including is visited safely using restricted clearance
The information collecting device for surveying unmanned plane carries out safety monitoring information collection to the region that restricted clearance domestic demand security monitoring detects, and
Safety monitoring information operating safety monitoring information identifying method collected is pacified in safety monitoring information identification module
Recognition result is exported after full monitoring information identification operation.Personnel are replaced to detect the case where restricted clearance with unmanned plane, greatly
Reduce the insecurity that personnel enter restricted clearance greatly, it is also very convenient to detect, and enhance judging result timeliness and
Stability.It can choose whether horse back arrangement personnel according to the defect level of the facility in the restricted clearance detected and safeguard,
Still continue to use facility because individual defects degree not necessarily influences the work of facility, until defect to a certain extent when again
It repairs, is more advantageous to the effective rate of utilization for promoting resource, and it is safer that the safety monitoring of restricted clearance is worked
And intelligence.
Correspondingly, the information collecting device is camera, and the safety monitoring information is image information, the safety prison
Measurement information identification module is picture recognition module, and the safety monitoring information identifying method is image-recognizing method, the camera shooting
Head acquired image information exports knowledge after identifying operation to image information with image-recognizing method in picture recognition module
Not as a result, of course, it is possible to by the safe sensorcraft of restricted clearance by setting route automatic cruising, can also manually in real time it is distant
Camera shooting activity is carried out in the control safe sensorcraft to restricted clearance of restricted clearance, can be as shown in figure 11 shooting in restricted clearance
The image information 600 arrived.Figure 13 is image recognition schematic diagram, the corrosion sites in figure in rectangle frame to recognize.Described image
Recognition methods the following steps are included:
S21 classifies the detection scene of restricted clearance, and categorized good scarce for every a kind of probe field scape matching
The identification image for the degree of falling into is fabricated to a training dataset, it should be noted that the defects of described defect level may include
Situations such as damage, slight crack, corrosion, the specifically such as damage of pipe leakage, oil storage tank, Coal Mine Subsidence, and the safety monitoring information
The detectable drawbacks described above of identification module and the recognition result for exporting degree of correspondence, such as different degrees of leakage, different degrees of damage
Evil etc., in the present embodiment, the detection scene be classified as feed channel burner, support storage tank, inside industrial chimney, Large Oil
Tank ship is internal, inside big pipeline or coal-fired power station boiler, and the defect level is heavy corrosion, moderate corrosion or mild corrosion,
The recognition result is heavy corrosion, moderate corrosion or mild corrosion;
S22, the training data inputted in step S21 concentrate the identification image that detection scene and defect level have been classified
And carry out image sharpening and edge enhancing processing, wherein Figure 12 is that the image information 600 taken in restricted clearance passes through step
S22 treated image information 700;
S23 instructs convolutional neural networks using through the step S22 training dataset that treated identifies that image is formed
Practice;
S24, by the image information acquired in step S1 input in the convolutional neural networks after step S22 training, output
Recognition result, in the present embodiment, the output valve of the recognition result are heavy corrosion, moderate corrosion or mild corrosion.
Wherein, convolutional neural networks are trained described in step S23 the following steps are included:
S231 successively carries out first time process of convolution to the identification image concentrated through step S22 treated training data,
For convolutional neural networks first layer, it is preferable that by through step S22 treated identification image size be unified for 32 × 32 × 1,
The filter size of convolutional layer is 5 × 5, and depth 6 is supplemented, step-length 1 without using full 0, obtains the convolutional neural networks the
One layer of Output Size is 32-5+1=28, depth 6, i.e., the described convolutional neural networks second layer node matrix equation has 28 × 28 ×
6=4704 node, each node and 5 × 5=25 current node layers are connected, i.e. convolutional neural networks first layer is a total of
4704 × (25+1)=122304 connection;
S232: the processing of first time pondization is carried out to the convolutional neural networks first layer, local acknowledgement's normalized is
The convolutional neural networks second layer, it is preferable that the filter size that the convolutional neural networks second layer uses is for 2 × 2, step-length
2, the convolutional neural networks second layer output matrix size is 14 × 14 × 6;
S233 carries out second of process of convolution to the convolutional neural networks second layer, is convolutional neural networks third layer,
Preferably, the convolutional neural networks second layer output matrix is the input of the convolutional neural networks third layer, the mistake of use
Filter size is 5 × 5, and depth 16 is supplemented without using full 0, and the Output Size of step-length 1, convolutional neural networks third layer is
14-5+1=10, depth 16, i.e. convolutional neural networks third layer output matrix size are 10 × 10 × 16, convolutional neural networks
Third layer parameter has 5 × 5 × 6 × 16+16=2416, there is 10 × 10 × 16 × (5 × 5+1)=41600 connection;
S234 carries out second of pondization processing to the convolutional neural networks third layer, is the 4th layer of convolutional neural networks,
Preference, the input matrix size that the 4th layer of the convolutional neural networks are 10 × 10 × 16, the filter size used for 2 ×
2, step-length 2, the output matrix size that the 4th layer of the convolutional neural networks is 5 × 5 × 16;
S235 obtains convolutional neural networks layer 5 to the 4th layer of the convolutional neural networks in a manner of connecting entirely, excellent
Selection of land, the output matrix that the 4th layer of the convolutional neural networks are the input of the convolutional neural networks layer 5, are straightened
The vector for being 5 × 5 × 16 for a length, i.e., the one-dimensional space is straightened in vector form in a three-dimensional matrice indicates, into
Entering full articulamentum to be trained, the output node number of the convolutional neural networks layer 5 is 120, that is, share 5 × 5 × 16 ×
120+120=48120 parameter;
S236, the convolutional neural networks layer 5 is obtained output layer in a manner of connecting entirely: output layer is divided into 6
Independent part, each part are additionally connected with an independent loss function;
S237 minimizes the value of the loss function using good identification image categorized in S21, completes to convolution mind
Training through network, this step are further to be corrected by categorized good identification picture and optimization convolutional neural networks.
Wherein, the Edge contrast method can be high-pass filtering and the airspace differential method, and enhancing processing in edge can then use
The brightness value of image adjacent picture elements is differed biggish edge and is highlighted by airspace method and frequency domain method, with efficiently
The feature for extracting image information, reinforces the accuracy of image recognition.The principle of the process of convolution is to bring input data 100 into
Convolution accounting method is converted into weight matrix 300 by convolution kernel 200, it should be noted that the convolution kernel, that is, above-mentioned
The filter of convolutional layer, the matrix is in conjunction with image, to generate the output of a convolution.The pondization processing is for subtracting
The quantity of few convolutional neural networks training parameter and the space size of image, pond are completed alone in each depth dimension,
Therefore with allowing convolutional layer to reduce operand under the premise of the depth of image remains unchanged, principle is will be to pond image data
It chooses maximum value and forms new image data 500 in the region not being overlapped in 400.
The safe detection method of the above-mentioned restricted clearance based on unmanned plane replaces the case where personnel are to restricted clearance with unmanned plane
It is detected, the personnel of greatly reducing enter the insecurity of restricted clearance, detect also very convenient intelligence, and enhance judgement
As a result timeliness and stability.Moreover, in order to improve the effective rate of utilization of resource, it can be according in the restricted clearance detected
Facility damage or defect level, choose whether horse back arrangement personnel and safeguard, still continue to use facility, wait restricted clearances
The defects of repaired when reaching the degree to need repairing again.
Convolutional neural networks can recognize that image, can be by by training simultaneously by the accuracy rate of training Statistical error result
Artificial judgement is gradually got rid of, is realized intelligent.So that judgement is more stable, quick.
In addition, the information collecting device can also include the region for detecting to restricted clearance domestic demand security monitoring
The gas sensor that gas information is acquired, the safety monitoring information are gas information, the safety monitoring information identification
Module is gas identification module, and the safety monitoring information identifying method is Gas Distinguishing Method, and the gas sensor is adopted
The gas information of collection exports recognition result after identifying operation to gas information with Gas Distinguishing Method on gas identification module.
Gas sensor can be used to measure the type of gas, concentration and ingredient, the special component in gas can be detected, such as an oxygen
Change the detection of carbon gas, the detection of methane gas, the detection of coal gas, detection of freon (R11, R12) etc., using on unmanned plane
The gas sensor of setting is acquired the gas information of restricted clearance, on the one hand detects gaseous contamination, the gas of restricted clearance
The safety problems such as body leakage, on the one hand detect the gas hazard degree of restricted clearance in advance, promote subsequent personnel and enter limited sky
Between overhaul when degree of safety.
In example 2, the present embodiment is substantially the same manner as Example 1, states for simplicity, only illustrate its primary structure with
1 difference of embodiment.Unaccounted part is same as Example 1 in the present embodiment, and difference is: first bar 113,
Second bar 114, the third bar 115 and 116 head and the tail of the 4th bar successively hingedly form the installing frame of diamond shape
112, and there is the retention bead 1131 being located on the pedestal 111 between first bar 113 and 114 interior angle of the second bar
And for first bar 113 and second bar 114 to be made first bar 113 and described second toward relative direction drawing
Bar 114 is adjacent to the first back-moving spring 1132 of the retention bead 1131, the third bar 115 and 116 interior angle of the 4th bar
Between be equipped with second back-moving spring 1133.Specifically, the first back-moving spring and the second back-moving spring are tension spring,
When the safe sensorcraft of restricted clearance is hit, especially when apex angle interconnecting piece is by impact force, two of diamond shape
The installing frame is squeezed and deformed toward opposite direction, the first back-moving spring and the second back-moving spring then with the limit
Installing frame is pulled to normal shape by male cooperation, by the buffer function of spring, restricted clearance can be effectively reduced and detect safely
The damage that drone body especially rotor is subject to when unmanned plane is hit, and there is certain shock-absorbing effect.
In conclusion the application is had the advantages that by the improvement in structure
1, personnel are replaced to detect the case where restricted clearance with unmanned plane, the personnel of greatly reducing enter restricted clearance
Insecurity, it is also very convenient to detect, and enhances the timeliness and stability of judging result.It can be limited according to what is detected
The defect level of facility in space chooses whether horse back arrangement personnel and safeguards, facility still continued to use, because individually
Defect level not necessarily influences the work of facility, until defect to a certain extent when repair again, be more advantageous to promotion resource
Effective rate of utilization, and make restricted clearance safety monitoring work more safety and intelligence.
2, light-weighted anticollision frame can spring back or rotate when colliding with object, can effectively prevent unmanned plane especially
It is that rotor is not damaged, even can be from high-altitude free falling without damage, being more advantageous to unmanned plane can answer in strange
It flies in miscellaneous environment, greatly reduces the incidence of " aircraft bombing ", guarantee that unmanned plane is capable of the information of normal acquisition restricted clearance
And manipulation difficulty to unmanned plane can be substantially reduced.
3, it is connected on the output shaft of close coupled type DC brushless motor above underneath type rotor, although underneath type rotor is stablized
Property declined, but the mobility of unmanned plane is stronger, is more advantageous to it and avoids obstacle in time in complicated restricted clearance, mentions
The safety of unmanned plane is risen, and camera of being more convenient for is mounted on above unmanned plane to shoot broader upper images.
4, cylindric frame is set between the first hemisphere frame and the second hemisphere frame, so that anticollision frame is falling
During falling to the ground, on the one hand still a part of impact force can be offset by rolling, on the one hand cylindric frame can be with
The rolling that line contact type is carried out with ground prevents circular anticollision frame and ground from carrying out long-term unstable point contact and rolls
Multi-faceted and irregular overturning is caused to drone body and increases and destroys probability.
5, after unmanned plane finds trouble location, label rifle can be used, sign flag is carried out to trouble location, such as stamp X
Shape or triangle pattern, it might even be possible to which stamping the pattern with fluorescent effect keeps label more prominent, and label can act on certainly
Near trouble location or trouble location, enable personnel to restricted clearance facility carry out repair and maintenance when it is more intuitive,
Accurately and quickly find trouble location, in conjunction with GPS module locating effect more preferably.
It is the one or more embodiments provided in conjunction with particular content as described above, does not assert the specific reality of the application
It applies and is only limited to these instructions.It is all approximate with the present processes, structure etc., identical, or under the application concept thereof
It makes several technologies to deduce, or replacement all should be considered as the protection scope of the application.
Claims (10)
1. the safe detection method of restricted clearance based on unmanned plane, which is characterized in that use the safe sensorcraft of restricted clearance
Information collecting device region that restricted clearance domestic demand security monitoring is detected carry out safety monitoring information collection, and will be acquired
Safety monitoring information operating safety monitoring information identifying method safety monitoring letter is carried out in safety monitoring information identification module
Recognition result is exported after breath identification operation.
2. the safe detection method of the restricted clearance according to claim 1 based on unmanned plane, which is characterized in that the information
Acquisition device is camera, and the safety monitoring information is image information, and the safety monitoring information identification module is image knowledge
Other module, the safety monitoring information identifying method are image-recognizing method, and the camera acquired image information is being schemed
As exporting recognition result, described image recognition methods after using image-recognizing method to identify operation to image information on identification module
The following steps are included:
S21 classifies the detection scene of restricted clearance, and matches categorized good defect journey for every a kind of probe field scape
The identification image of degree is fabricated to a training dataset;
S22, the training data concentration inputted in step S21 have gone forward side by side the identification image that detection scene and defect level have been classified
Row image sharpening and edge enhancing processing;
S23 is trained convolutional neural networks using through the step S22 training dataset that treated identifies that image is formed;
S24, by the image information acquired in step S1 input in the convolutional neural networks after step S22 training, output identification
As a result.
3. the safe detection method of the restricted clearance according to claim 2 based on unmanned plane, which is characterized in that step S23
It is described convolutional neural networks are trained the following steps are included:
S231 successively carries out first time process of convolution to the identification image concentrated through step S22 treated training data, for volume
Product neural network first layer;
S232: the processing of first time pondization is carried out to the convolutional neural networks first layer, local acknowledgement's normalized is convolution
The neural network second layer;
S233 carries out second of process of convolution to the convolutional neural networks second layer, is convolutional neural networks third layer;
S234 carries out second of pondization processing to the convolutional neural networks third layer, is the 4th layer of convolutional neural networks;
S235 obtains convolutional neural networks layer 5 to the 4th layer of the convolutional neural networks in a manner of connecting entirely;
S236, the convolutional neural networks layer 5 is obtained output layer in a manner of connecting entirely: output layer is divided into 6 independences
Part, each part is additionally connected with an independent loss function;
S237 minimizes the value of the loss function using good identification image categorized in S21, completes to convolutional Neural net
The training of network.
4. the safe detection method of the restricted clearance according to claim 2 based on unmanned plane, which is characterized in that the detection
Scene classification be feed channel burner, support storage tank, inside industrial chimney, large oil tank ship is internal, big pipeline or coal fired power plant
Boiler internal, the defect level are divided into heavy corrosion, moderate corrosion or mild corrosion, the recognition result be heavy corrosion,
Moderate corrosion or mild corrosion.
5. the safe detection method of the restricted clearance according to claim 3 based on unmanned plane, which is characterized in that will be through step
The size of S22 treated identification image is unified for 32 × 32 × 1, and the filter size of convolutional layer is 5 × 5, depth 6, no
Supplemented using full 0, step-length 1, obtain the convolutional neural networks first layer Output Size be 32-5+1=28, depth 6,
The i.e. described convolutional neural networks second layer node matrix equation has 28 × 28 × 6=4704 node, and each node and 5 × 5=25 are a
Current node layer is connected, i.e., convolutional neural networks first layer a total of 4704 × (25+1)=122304 connection;
The filter size that the convolutional neural networks second layer uses is 2 × 2, step-length 2, the convolutional neural networks second
Layer output matrix size is 14 × 14 × 6;
The convolutional neural networks second layer output matrix is the input of the convolutional neural networks third layer, the filter of use
Size is 5 × 5, and depth 16 is supplemented, step-length 1 without using full 0, and the Output Size of convolutional neural networks third layer is 14-5+
1=10, depth 16, i.e. convolutional neural networks third layer output matrix size are 10 × 10 × 16, convolutional neural networks third
Layer parameter has 5 × 5 × 6 × 16+16=2416, there is 10 × 10 × 16 × (5 × 5+1)=41600 connection;
The input matrix size that the 4th layer of the convolutional neural networks is 10 × 10 × 16, the filter size used for 2 × 2,
Step-length is 2, and the output matrix size that the 4th layer of the convolutional neural networks is 5 × 5 × 16;
The output matrix that the 4th layer of the convolutional neural networks be the convolutional neural networks layer 5 input, be straightened for
The vector that one length is 5 × 5 × 16, i.e., the one-dimensional space is straightened in vector form in a three-dimensional matrice indicates, enters
Full articulamentum is trained, and the output node number of the convolutional neural networks layer 5 is 120, that is, share 5 × 5 × 16 ×
120+120=48120 parameter.
6. the safe detection method of the restricted clearance according to claim 1 based on unmanned plane, which is characterized in that the information
Acquisition device further includes the air-sensitive biography that the gas information in the region for detecting to restricted clearance domestic demand security monitoring is acquired
Sensor, the safety monitoring information are gas information, and the safety monitoring information identification module is gas identification module, the peace
Full monitoring information recognition methods is Gas Distinguishing Method, and the gas sensor gas information collected is in gas identification module
Upper utilization Gas Distinguishing Method exports recognition result after identifying operation to gas information.
7. the safe detection method of the restricted clearance according to claim 2 based on unmanned plane, which is characterized in that described limited
Space safety sensorcraft includes drone body (1) and the anticollision frame for being located at the drone body (1) outside
(2);The drone body (1) includes the holder (12) of rack (11) and the side of being mounted on the rack, the rack (11)
It is equipped with underneath type rotor (110), the camera (121) is located on the holder (12), and the camera (121) side
Equipped with the irradiation light (122) mobile with the camera (121);The anticollision frame (2) includes the first hemisphere frame (21), second
The tangent plane edge of hemisphere frame (22) and connecting rod (23), the first hemisphere frame (21) and the second hemisphere frame (22) passes through institute
It states connecting rod (23) and is interconnected to form the anticollision frame (2), multiple connecting rods (23) are along the first hemisphere frame
(21) and the tangent plane edge of the second hemisphere frame (22) is uniformly distributed to form connection the first hemisphere frame (21) and described the
The cylindric frame of two hemisphere frames (22).
8. the safe detection method of the restricted clearance according to claim 7 based on unmanned plane, which is characterized in that the rack
It (11) include pedestal (111), pedestal (111) two sides are symmetrically arranged with installing frame (112), and the installing frame (112) includes connecting
It connects on the pedestal (111) in isometric the first bar (113) and the second bar for stretching out pedestal (111) side of splayed
(114), the third bar (115) that is connect with the first bar (113) extension end end and connect with the second bar (114) extension end
It joins to be formed and the anticollision frame in the free end of the 4th bar (116) connect, the third bar (115) and the 4th bar (116)
The apex angle interconnecting piece (117) of frame (2) connection, the underneath type rotor (110) are located at first bar (113) and the third bar
(115) below interconnecting piece and below second bar (114) and the interconnecting piece of the 4th bar (116);
First bar (113), second bar (114), the third bar (115) and the 4th bar (116) head and the tail are successively
The hinged installing frame (112) for forming diamond shape, and have between first bar (113) and second bar (114) interior angle
The retention bead (1131) that is located on the pedestal (111) and for by first bar (113) and second bar (114)
First bar (113) and second bar (114) is set to be adjacent to the first of the retention bead (1131) toward relative direction drawing
Reset spring (1132) is equipped with second back-moving spring between the third bar (115) and the 4th bar (116) interior angle
(1133)。
9. the safe detection method of the restricted clearance according to claim 7 based on unmanned plane, which is characterized in that described first
Hemisphere frame (21) includes:
Moulding (211), is equipped with a plurality of, and a plurality of moulding (211) one end is collapsed, and the other end opens to form hemisphere frame-shaped, the frame
One end that item (211) collapses is equipped with the card convex (2111) protruded to outside the first hemisphere frame (21) top dome, the moulding (211)
It is recessed (2112) that one end of opening is equipped with the card being centrally recessed into the first hemisphere frame (21) section;
It collapses ring (212), is located at one end of the moulding (211) gathering, the gathering ring (212) includes looping pit
(2121), the annular around the looping pit (2121) collapses wall (2122), collapses wall (2122) annular uniformly along the annular
The annular groove (2123) of distribution and the outer ring screw thread (2124) for being provided with the annular wall (1142) outer wall, the moulding
(211) inlay card is in the annular groove (2123) and the card convex (2111) is collapsed in the looping pit (2121) so that a plurality of
Described moulding (211) one end forms rounding state;
Ring (213) are screwed, the ring for being equipped with the fixation hole (2131) mutually fixed with rack (11), surrounding the fixation hole (2131)
Shape screws wall (2132) and is provided with the annular and screws and matches on wall (2132) inner wall with the outer ring screw thread (2124)
Inner ring screw thread (2133), the gathering ring (212) collapses moulding (211) back spin and screws in ring (213) into described;
Expanded loop (214), along circumference be uniformly provided with the card hole equal with the moulding (211) quantity (2141) and with it is described
The equal lockhole (2142) of connecting rod (23) quantity, one end that the moulding (211) is opened stick into the card hole (2141);
Connection ring (215), the via hole equal with the connecting rod (23) quantity is uniformly provided with along circumference;
Attachment screw, the attachment screw pass through the lockhole (2142) and the via hole for the expanded loop (214) and described
Connection ring (215) is locked on the connecting rod (23).
10. according to the safe detection method of the described in any item restricted clearances based on unmanned plane of claim 7-9, feature exists
In the safe sensorcraft of restricted clearance is additionally provided in real time by the position of the safe sensorcraft of the restricted clearance
Information is transmitted to the GPS module of control terminal and the label rifle (3) for marking the trouble location of restricted clearance, the label rifle
It (3) is the automatic spray gun that can spray colored solution.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810595823.1A CN108985163A (en) | 2018-06-11 | 2018-06-11 | The safe detection method of restricted clearance based on unmanned plane |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810595823.1A CN108985163A (en) | 2018-06-11 | 2018-06-11 | The safe detection method of restricted clearance based on unmanned plane |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108985163A true CN108985163A (en) | 2018-12-11 |
Family
ID=64540175
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810595823.1A Pending CN108985163A (en) | 2018-06-11 | 2018-06-11 | The safe detection method of restricted clearance based on unmanned plane |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108985163A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109671071A (en) * | 2018-12-19 | 2019-04-23 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect location and grade determination method based on deep learning |
CN109956046A (en) * | 2019-04-02 | 2019-07-02 | 北京中科利丰科技有限公司 | Unmanned plane protective device |
CN109974585A (en) * | 2019-04-18 | 2019-07-05 | 甘肃博程技术咨询服务有限公司 | A kind of elevator shaft structure recognition and data measurement unit |
CN111046908A (en) * | 2019-11-05 | 2020-04-21 | 杭州电子科技大学 | Emulsion explosive package fault real-time monitoring model based on convolutional neural network |
Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130292512A1 (en) * | 2012-05-07 | 2013-11-07 | Honeywell International Inc. | Environmental sampling with an unmanned aerial vehicle |
CN103837045A (en) * | 2014-02-24 | 2014-06-04 | 中国航天时代电子公司 | Foldable diamond-shaped missile wing spread system |
US8956635B2 (en) * | 2011-06-07 | 2015-02-17 | Fly Armor, Llc. | Insect repellent devices |
CN105751023A (en) * | 2014-12-16 | 2016-07-13 | 镇江润京机电科技有限公司 | Aluminum alloy silicon wafer surface treatment device |
US20160214715A1 (en) * | 2014-11-21 | 2016-07-28 | Greg Meffert | Systems, Methods and Devices for Collecting Data at Remote Oil and Natural Gas Sites |
CN105897042A (en) * | 2016-06-06 | 2016-08-24 | 长春工业大学 | Asymmetrical diamond-shaped hinge quadratured driving type piezoelectric stick-slip linear motor and recombination excitation method thereof |
KR101718287B1 (en) * | 2016-07-06 | 2017-04-05 | 양권석 | Method for providing safety assessment using drone |
GB201710692D0 (en) * | 2017-07-04 | 2017-08-16 | Daimler Ag | Inspection system and method for automatic visual inspection of a motor vehicle |
US20170259920A1 (en) * | 2016-03-10 | 2017-09-14 | International Business Machines Corporation | Automatic painting system with drone, user interface and computer vision |
KR20170104762A (en) * | 2016-03-08 | 2017-09-18 | 한국해양과학기술원 | Apparatus and method for detecting surface flaw of wind turbine using drone |
EP3239048A1 (en) * | 2016-04-30 | 2017-11-01 | Flyability SA | Unmanned aerial vehicle and protective outer cage therefor |
CN107316035A (en) * | 2017-08-07 | 2017-11-03 | 北京中星微电子有限公司 | Object identifying method and device based on deep learning neutral net |
CN107539660A (en) * | 2017-10-09 | 2018-01-05 | 周洪春 | A kind of container of convenient regulation accommodation space |
CN107709158A (en) * | 2015-06-15 | 2018-02-16 | 多尼克公司 | System and method for checking surface automatically |
CN207045377U (en) * | 2017-05-05 | 2018-02-27 | 嵊州市诺和玻璃制品有限公司 | A kind of glassware transport vehicle of good damping effect |
CN107816368A (en) * | 2017-11-09 | 2018-03-20 | 河南理工大学 | Mine fire the condition of a disaster UAV Intelligent detection system and detection method |
CN107886133A (en) * | 2017-11-29 | 2018-04-06 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect inspection method based on deep learning |
US9964658B2 (en) * | 2015-10-06 | 2018-05-08 | The United States Of America As Represented By The Secretary Of The Army | Method of autonomous power line detection, avoidance, navigation, and inspection using aerial crafts |
US20180144644A1 (en) * | 2016-11-23 | 2018-05-24 | Sharper Shape Oy | Method and system for managing flight plan for unmanned aerial vehicle |
CN108120566A (en) * | 2017-12-11 | 2018-06-05 | 西北工业大学 | Space station Leakage inspection localization method based on miniature ultrasonic sensor array unmanned plane |
-
2018
- 2018-06-11 CN CN201810595823.1A patent/CN108985163A/en active Pending
Patent Citations (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8956635B2 (en) * | 2011-06-07 | 2015-02-17 | Fly Armor, Llc. | Insect repellent devices |
US20130292512A1 (en) * | 2012-05-07 | 2013-11-07 | Honeywell International Inc. | Environmental sampling with an unmanned aerial vehicle |
CN103837045A (en) * | 2014-02-24 | 2014-06-04 | 中国航天时代电子公司 | Foldable diamond-shaped missile wing spread system |
US20160214715A1 (en) * | 2014-11-21 | 2016-07-28 | Greg Meffert | Systems, Methods and Devices for Collecting Data at Remote Oil and Natural Gas Sites |
CN105751023A (en) * | 2014-12-16 | 2016-07-13 | 镇江润京机电科技有限公司 | Aluminum alloy silicon wafer surface treatment device |
CN107709158A (en) * | 2015-06-15 | 2018-02-16 | 多尼克公司 | System and method for checking surface automatically |
US9964658B2 (en) * | 2015-10-06 | 2018-05-08 | The United States Of America As Represented By The Secretary Of The Army | Method of autonomous power line detection, avoidance, navigation, and inspection using aerial crafts |
KR20170104762A (en) * | 2016-03-08 | 2017-09-18 | 한국해양과학기술원 | Apparatus and method for detecting surface flaw of wind turbine using drone |
US20170259920A1 (en) * | 2016-03-10 | 2017-09-14 | International Business Machines Corporation | Automatic painting system with drone, user interface and computer vision |
EP3239048A1 (en) * | 2016-04-30 | 2017-11-01 | Flyability SA | Unmanned aerial vehicle and protective outer cage therefor |
CN105897042A (en) * | 2016-06-06 | 2016-08-24 | 长春工业大学 | Asymmetrical diamond-shaped hinge quadratured driving type piezoelectric stick-slip linear motor and recombination excitation method thereof |
KR101718287B1 (en) * | 2016-07-06 | 2017-04-05 | 양권석 | Method for providing safety assessment using drone |
US20180144644A1 (en) * | 2016-11-23 | 2018-05-24 | Sharper Shape Oy | Method and system for managing flight plan for unmanned aerial vehicle |
CN207045377U (en) * | 2017-05-05 | 2018-02-27 | 嵊州市诺和玻璃制品有限公司 | A kind of glassware transport vehicle of good damping effect |
GB201710692D0 (en) * | 2017-07-04 | 2017-08-16 | Daimler Ag | Inspection system and method for automatic visual inspection of a motor vehicle |
CN107316035A (en) * | 2017-08-07 | 2017-11-03 | 北京中星微电子有限公司 | Object identifying method and device based on deep learning neutral net |
CN107539660A (en) * | 2017-10-09 | 2018-01-05 | 周洪春 | A kind of container of convenient regulation accommodation space |
CN107816368A (en) * | 2017-11-09 | 2018-03-20 | 河南理工大学 | Mine fire the condition of a disaster UAV Intelligent detection system and detection method |
CN107886133A (en) * | 2017-11-29 | 2018-04-06 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect inspection method based on deep learning |
CN108120566A (en) * | 2017-12-11 | 2018-06-05 | 西北工业大学 | Space station Leakage inspection localization method based on miniature ultrasonic sensor array unmanned plane |
Non-Patent Citations (3)
Title |
---|
FLYABILITY: "Flyability presentation - Elios, the collision-tolerant drone for industrial inspection", 《HTTPS://WWW.YOUTUBE.COM/WATCH?V=S96Q2GXGOEE》 * |
FLYABILITY: "Flyability’s Elios: Drone inspection of the", 《HTTPS://WWW.YOUTUBE.COM/WATCH?V=DNULU9ELJQ8》 * |
XHELI RC DRONE HELICOPTER: "UDI U828 The Unbreakable Caged Quadcopter", 《HTTPS://WWW.YOUTUBE.COM/WATCH?V=IPM9PX8SZZS》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109671071A (en) * | 2018-12-19 | 2019-04-23 | 南京市测绘勘察研究院股份有限公司 | A kind of underground piping defect location and grade determination method based on deep learning |
CN109671071B (en) * | 2018-12-19 | 2023-10-31 | 南京市测绘勘察研究院股份有限公司 | Underground pipeline defect positioning and grade judging method based on deep learning |
CN109956046A (en) * | 2019-04-02 | 2019-07-02 | 北京中科利丰科技有限公司 | Unmanned plane protective device |
CN109956046B (en) * | 2019-04-02 | 2020-11-17 | 北京中科利丰科技有限公司 | Unmanned aerial vehicle protection device |
CN109974585A (en) * | 2019-04-18 | 2019-07-05 | 甘肃博程技术咨询服务有限公司 | A kind of elevator shaft structure recognition and data measurement unit |
CN111046908A (en) * | 2019-11-05 | 2020-04-21 | 杭州电子科技大学 | Emulsion explosive package fault real-time monitoring model based on convolutional neural network |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108688805A (en) | The unmanned plane detected safely applied to restricted clearance | |
CN108985163A (en) | The safe detection method of restricted clearance based on unmanned plane | |
CN109936080B (en) | Method for inspecting power transmission line by unmanned aerial vehicle | |
CN110200598B (en) | Poultry detection system and detection method for abnormal physical signs in large farm | |
CN110047241A (en) | A kind of forest fire unmanned plane cruise monitoring system | |
CN207600967U (en) | A kind of device for wind power plant blade surface defects detection | |
CN109780452A (en) | Gas based on laser telemetry technology leaks unmanned plane inspection retrieving concentration method | |
CN208044404U (en) | A kind of natural gas crusing robot | |
CN205584374U (en) | Unmanned aerial vehicle oil gas pipeline system of patrolling and examining | |
CN109711348A (en) | Intelligent monitoring method and system based on the long-term real-time architecture against regulations in hollow panel | |
CN208544425U (en) | The unmanned plane detected safely applied to restricted clearance | |
CN112506214B (en) | Operation flow of unmanned aerial vehicle autonomous fan inspection system | |
CN110274160B (en) | Pipeline inspection system based on infrared and visible light fusion image | |
CN105866626A (en) | Unmanned plane-mounted UV detection system for power transmission and transformation equipment | |
CN107069549A (en) | Rotor wing unmanned aerial vehicle power-line patrolling system | |
CN108802040A (en) | A kind of unmanned plane device and detection method for crane surface defects detection | |
CN110110112A (en) | It is a kind of based on liftable trolley around machine check method and system | |
CN107776888A (en) | A kind of integral environmentally friendly unmanned plane of empty water | |
CN205843666U (en) | A kind of depopulated helicopter three dimensional data collection and cruising inspection system | |
CN110763804A (en) | Atmospheric pollution source tracing system and method based on unmanned aerial vehicle | |
CN208544411U (en) | The unmanned plane with underneath type rotor detected safely applied to restricted clearance | |
CN208544431U (en) | The anticollision unmanned plane detected safely applied to restricted clearance | |
CN105809219A (en) | Prefabricated pipe segment quality detection and statistical system and method of tunnel prefabricated pipe segment production line | |
CN206804544U (en) | A kind of unmanned plane device for crane surface defects detection | |
CN113744230A (en) | Intelligent detection method for aircraft skin damage based on unmanned aerial vehicle 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 | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20221206 |
|
AD01 | Patent right deemed abandoned |