CN208985185U - Air cooler dust stratification state perception system based on deep learning and infrared image identification - Google Patents
Air cooler dust stratification state perception system based on deep learning and infrared image identification Download PDFInfo
- Publication number
- CN208985185U CN208985185U CN201821563437.6U CN201821563437U CN208985185U CN 208985185 U CN208985185 U CN 208985185U CN 201821563437 U CN201821563437 U CN 201821563437U CN 208985185 U CN208985185 U CN 208985185U
- Authority
- CN
- China
- Prior art keywords
- air cooling
- air
- cooling tubes
- input terminal
- tubes condenser
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Radiation Pyrometers (AREA)
Abstract
The utility model relates to the air cooler dust stratification state perception systems based on deep learning and infrared image identification, including steam turbine to be connected by turbine discharge with steam discharge distribution pipe input terminal；Inlet temperature sensor is arranged in air cooling tubes condenser tube bank inside, and outlet temperature sensor is arranged in air cooling tubes condenser tube bank outside, and inlet temperature sensor and outlet temperature sensor output end are connected through temperature transfer bus with data acquisition module input terminal；Inlet velocity sensor arrangement restrains inside in air cooling tubes condenser.The beneficial effects of the utility model are: the utility model considers that thermal power station's direct air cooling system exhaust temperature high-frequency, big fluctuation of floating occurs with the variation of ambient air temperature, the high feature of high ambient temperature period exhaust steam pressure, based on heat balance principle, water temperature and hot channel exit water temperature of the joint external sensor by measurement simulation turbine discharge, the indirect analogs air cooling tubes condenser operating status such as fin inlet and outlet wind-warm syndrome.
Description
Technical field
The utility model relates to air cooler dust stratification state aware technologies, more particularly to a kind of air cooler dust stratification state aware
System.
Background technique
Air cooling tubes condenser is the important link of Steam Power Circulation generating set.Water cooling was all made of before nineteen thirty-nine, he
Working condition directly affect the safety and economy of entire unit.Either currently account for national total power generation 73.1%
Fired power generating unit, it is contemplated that 60% and 48% will be accounted for the year two thousand twenty and the year two thousand thirty, or by or will extensive development the sun
Can thermal electric generator group, needed for fire coal and solar energy resources be mainly distributed on the northeast, northwest and North China (" three Norths ") in China
Area, however " three Norths " area is the poor water shortage of water resource or the area Shao Shui, and coal-fired and solar energy resources and water is caused to provide
There are the contradictions in geographical configuration for source distribution, but air cooler causes cold end using the much smaller air of thermal capacity as cooling medium
Heat-sinking capability falls sharply, and causes the heat consumption height that generates electricity, less economical.In order to strengthen air side heat transfer process, cloth on the base tube of air cooler
Fin is set, this is China is dry, dust storm is big, " three Norths " area more than fugitive dust, and suspended particulate substance entrained by air is easily in wing
It gathers, generates black dirt (dust stratification layer) on piece pipe.The presence of black dirt reduces the heat exchange property of air cooler, improves operating cost,
Unit exhaust steam pressure under same environmental condition can be made to raise 8~12kPa, increase coal consumption for power generation about 12~18g/kWh, and increased straight
The security risk of cold group.
It is analyzed from the construction of thermal power plant, due to the water resources shortage in China, saves production, domestic water as design
Top priority, therefore more preferably solve the problems, such as this in the development air cooling technique of thermal power station energetically.But air-cooled heat exchanger dust stratification is
The problem of being difficult to avoid that, especially temperature change be big and dust storm more than area, influenced by dust stratification, air-flow rushes when load variations are big
It hits and the reasons such as the caused vibration of other factors, the heat transfer effect of air-cooled heat exchanger is frequently subjected to influence and reduce or even can send out
Raw failure.Because the dust stratification of Air-Cooling Island will cause, the coefficient of heat transfer is reduced, dirty thickness degree increases, and the dust stratification processing of Air-Cooling Island is non-
Often important, but in terms of air cooling tubes condenser operation, live operation judges the accumulation state of black dirt and right by experience substantially
The influence of Steam Turbine exhaust steam pressure and generated energy, the maintenance clean as air cooler black dirt are also entirely by virtue of experience, to cause
Not in time, black dirt accumulation is serious for cleaning, endangers the situation of unit safety operation frequent occurrence.
Utility model content
Purpose of the utility model is to overcome the above-mentioned shortcomings and provide a kind of air cooler dust stratification state perception systems.
Air cooler dust stratification state perception system based on deep learning and infrared image identification, including steam turbine pass through steamer
Machine steam discharge is connected with steam discharge distribution pipe input terminal；Inlet temperature sensor is arranged in air cooling tubes condenser tube bank inside, outlet temperature
Sensor arrangement restrains outside in air cooling tubes condenser, and inlet temperature sensor and outlet temperature sensor output end are passed through temperature
Defeated bus is connected with data acquisition module input terminal；Inlet velocity sensor arrangement restrains inside in air cooling tubes condenser, out one's intention as revealed in what one says
Fast sensor arrangement restrains outside in air cooling tubes condenser, and inlet velocity sensor and air outlet velocity sensor output are through wind speed
Transfer bus is connected with data acquisition module input terminal；Thermal infrared imager is arranged in air cooling tubes condenser tube bank outside different angle
Place, thermal infrared imager output end are connected through image data transfer bus with data acquisition module input terminal；Data acquisition module is defeated
Outlet is connected through data transmission bus with industrial personal computer input terminal；Flowmeter output end is through water-supply-pipe and condensate tank input terminal phase
Even；Power distribution cabinet output end is connected through power supply line with data acquisition module, industrial personal computer, air cooling blower fan input terminal.
As preferred: the thermal infrared imager is installed on windward side and the leeward of air cooling tubes condenser tube bank.
The beneficial effects of the utility model are:
First is that the utility model is high with the variation appearance of ambient air temperature in view of thermal power station's direct air cooling system exhaust temperature
Frequency, big fluctuation of floating, the high feature of high ambient temperature period exhaust steam pressure are based on heat balance principle: turbine discharge
Thermal discharge, air cooling tubes condenser heat exchange amount and cooling air recept the caloric equal feature, and joint external sensor passes through measurement simulation
The water temperature and hot channel exit water temperature of turbine discharge, fin import and export the indirect analogs air cooling tubes condensers such as wind-warm syndrome and run shape
State.
Second is that the utility model has the advantages of simple structure, rationally, cheap, exploitativeness is good, it is often more important that improves function
Rate precision of prediction, it is intended to establish more accurate air cooling tubes condenser dust stratification and cleaning forecasting system, facilitate power plant operator day
Often detection and cleaning accurately and timely adjust air cooling tubes condenser maintenance plan, for scientific generation optimization tune tube apparatus provide according to
According to realization gene-ration revenue maximizes.
Detailed description of the invention
Fig. 1 is the air cooler dust stratification state perception system schematic diagram based on deep learning and infrared image identification；
Fig. 2 is image and flow chart of data processing block diagram；
Fig. 3 is convolutional neural networks schematic diagram；
Description of symbols: 1 steam turbine, 2 turbine discharges, 3 steam discharge distribution pipes, the tube bank of 4 air cooling tubes condensers, 5 is air-cooled solidifying
Vapour device fin, 6 inlet temperature sensors, 7 outlet temperature sensors, 8 inlet velocity sensors, 9 air outlet velocity sensors, 10
Thermal infrared imager, 11 data acquisition modules, 12 industrial personal computers, 13 air cooling blower fans, 14 condensate tanks, 15 condensation pumps, the transmission of 16 temperature
Bus, 17 wind speed transfer bus, 18 image data transfer bus, 19 data transmission bus, 20 power distribution cabinets, 21 power supply lines, 22 streams
Meter.
Specific embodiment
The utility model is described further below with reference to embodiment.The explanation of following embodiments is only intended to help to manage
Solve the utility model.It should be pointed out that for those skilled in the art, not departing from the utility model principle
Under the premise of, several improvements and modifications can be made to this utility model, these improvement and modification also fall into the utility model
In scope of protection of the claims.
The air cooler dust stratification state perception system based on deep learning and infrared image identification of the utility model, for existing
The shortcomings that being unable to real-time perception air cooling system operating status proposes the air cooler product based on deep learning and infrared image identification
Grey state perception system, existing deep learning sensor model research is only to the formation mechanism and influence factor of particulate fouling
Desk study, there are no the associate feature analysed in depth specifically for air cooler black dirt between its influence factor and black dirt characteristic,
The pervasive mechanism model by experimental check is not set up.The utility model is based on gas deep learning and infrared image identifies
Air cooler dust stratification state perception system, on the basis of deep learning algorithm, using infrared image obtain information with it is air-cooled
Device running state parameter is known air-cooled solidifying in real time by deep learning algorithm and air cooling tubes condenser dust stratification thermal resistance value calculating method
Vapour device operating status realizes air cooling tubes condenser accurately perception, long-term forecasting function in short term.
The concept of the utility model is that the characteristic of heat transfer efficiency is influenced for air cooling tubes condenser fin dust stratification, for
Traditional status by the empirically determined cleaning frequency, 1) the utility model proposes be directed to dust stratification to air cooling tubes condenser operating parameter
Based on influence, real-time infrared image is obtained by thermal infrared imager, using deep learning and Infrared Image Processing Method from
Extract parameter needed for obtaining in image, and it is gathered biggish 6 variables of the degree of association with black dirt: dust stratification time, steam turbine are arranged
Vapour thermic load, exhaust temperature, ambient air temperature, air cooling tubes condenser outlet wind-warm syndrome and face velocity be combined as basic test object and
Research object, deduction fit air cooling tubes condenser operating status and black dirt thickness relationship.It is used as air cooling tubes condenser based on this
Operating parameter acquisition system and dust stratification state perception system are fitted dust stratification thickness for it and perception operating status do parameter support.
2) the utility model proposes using deep learning and infrared image processing detection air cooling tubes condenser operation and dust stratification state.Pass through
Infrared imagery technique understands the state of air cooling system in real time, and the transmission for finding, handling, prevent significant error in time can rise
To very crucial and effectively act on, and its dust stratification degree can be obtained in real time by the infrared image processing to air cooling tubes condenser；
Since northern certain areas arid lacks water, working environment is severe, and phenomena such as tightness decline or bursting by freezing occurs often in finned-tube bundle.
These phenomena of the failure can be remotely detected by infrared imaging mode to bring great convenience to usually maintenance work, simultaneously
Finned tube surface temperature field distribution can also be understood in real time by infrared imagery technique, monitoring air cooling tubes condenser temperature difference temperature rise changes,
Air cooling system working condition can be understood in real time.
Air cooler dust stratification state perception system based on deep learning and infrared image identification referring to Fig.1, it includes: steamer
Machine 1 is connected by turbine discharge 2 with 3 input terminal of steam discharge distribution pipe；Inlet temperature sensor 6 is arranged in air cooling tubes condenser pipe
4 inside of beam, outlet temperature sensor 7 are arranged in 4 outside of air cooling tubes condenser tube bank, and inlet temperature sensor 6 and outlet temperature pass
7 output end of sensor is connected through temperature transfer bus 16 with 11 input terminal of data acquisition module；Inlet velocity sensor 8 is arranged in
4 inside of air cooling tubes condenser tube bank, air outlet velocity sensor 9 are arranged in 4 outside of air cooling tubes condenser tube bank, inlet velocity sensor 8
It is connected through wind speed transfer bus 17 with 11 input terminal of data acquisition module with 9 output end of air outlet velocity sensor；Infrared thermal imagery
Instrument 10 is arranged in air cooling tubes condenser and restrains at 4 outside different angles, and 10 output end of thermal infrared imager is through image data transfer bus
18 are connected with 11 input terminal of data acquisition module；11 output end of data acquisition module is through data transmission bus 19 and industrial personal computer 12
Input terminal is connected；22 output end of flowmeter is connected through water-supply-pipe with 14 input terminal of condensate tank；20 output end of power distribution cabinet is through power supply
Line 21 is connected with data acquisition module 11, industrial personal computer 12,13 input terminal of air cooling blower fan.
The power distribution cabinet 20 is the power supply core of live all appts, includes mainly distribution enclosure box, is spaced apart, ac contactor
The fire prevention discharge circuit such as device and related insurance.High-power circuit can be born by being wherein always spaced apart, and be always spaced apart and be diverted to boiler power supply
Circuit, water supply pump circuit, air cooling blower fan circuit and controller circuitry.It realizes that various pieces device arbitrarily switchs, and uses and press
Key and indicator light facilitate regular job by A.C. contactor.
The data collection system main body frame is by thermal infrared imager 10, inlet temperature sensor 6, outlet temperature sensor
7, inlet velocity sensor 8, air outlet velocity sensor 9 are constituted.Using several temperature sensors respectively to air cooling tubes condenser unit
Water tank water temperature, ambient air temperature, water inlet water temperature, water outlet water temperature, entrance wind-warm syndrome, go out one's intention as revealed in what one says wind-warm syndrome measure.By temperature
Sensor is connected serially to control system, and the transmission of data can be only realized with several lines.
The strong high resolution of 5 precision of air velocity transducer is realized to the inlet port wind on air cooling tubes condenser pipeline fin
Speed measurement, A, the B signal line passed through with control system are connect, are communicated using ModbusRtu agreement.
The flowmeter uses high-accuracy electromagnetic flowmeter.
The control system is the core devices of collection site data, and using main control chip, liquid crystal display is as screen
Show monitoring data, control and temperature collection sensor, the normal operation of air velocity transducer and flowmeter and signal output.It is logical
The mode for crossing serial ports message transmissions transfers data to host computer data processing system.It, can since lab space distance is shorter
It is flated pass transmission of data with serial port power, controller is equipped with signal transmssion line simultaneously, is sent to serial ports conversion module by universal serial bus, most
The operating parameter of air cooling tubes condenser dust stratification characteristic is sent to host computer by universal serial bus by serial ports conversion module afterwards, realizes long distance
From transmission.
11 data processing system of host computer uses the realization of windows platform labview virtual instrument software, Ke Yishi
Then the corresponding clock rate of input port is arranged as input signal source in other serial ports conversion module on configuration software
It shown in host computer, store the running state parameter of air cooler, and algorithm in matlab is called to realize prediction black dirt and calculate
The coefficient of heat transfer.By the data that data collection system is transmitted can calculate the characteristic parameter of heat-transfer surface dust stratification degree.And according to heat
Equilibrium principle, simulation turbine discharge thermal discharge, air cooling tubes condenser heat exchange and the equal calculating fin heat exchange of cooling air caloric receptivity
Amount.The operating parameter of air cooling tubes condenser dust stratification characteristic can be shown, stored in host computer, and is predicted in real time based on this kind of parameter
Air cooling tubes condenser dust stratification degree and optimal cleaning frequency, facilitate power plant's operator's routine testing and cleaning, accurately and timely
Adjust air cooling tubes condenser maintenance plan.Host computer data processing system shown by labview virtual instrument software designed image,
Image procossing interface.
The image detecting method based on deep learning network referring to described in Fig. 2 and Fig. 3 characterized by comprising it includes
Several convolutional layers, pond layer and full articulamentum.By data collecting module collected to infrared image be created as a database,
And image data most in database is trained using convolutional neural networks algorithm, with the data of remaining fraction into
Row verifying.Suitable model is erected, in order to the data and image processing of next step.Specific step is as follows: input infrared image；
The infrared image is handled, multiple training samples are generated；It is carried out using the multiple training sample based on convolutional Neural
The image recognition training of network, generates training pattern；And whether contain in the image uploaded using training pattern detection
The infrared image.
The deep learning network is specially convolutional neural networks algorithm, and the network inputs data of the model are one 4 dimensions
The infrared image of input is decomposed into 2-4 characteristic layer, having a size of (1280-60000,28-256,28-256,2- by tensor
4), respectively indicate the number 1280-60000 (number of picture is unsuitable very few, also unsuitable excessive) of a collection of picture, picture it is wide
Pixel number 28-256, high pixel number 28-256 and channel number 2-4.Multiple convolutional neural networks layers are used first
The feature extraction of image is carried out, the calculating process following steps of convolutional neural networks layer:
Convolutional layer 1: convolution kernel size a × a, convolution kernel moving step length 1-2, convolution kernel number a1, pond size a2 × a2,
Pond step-length 1-3, pond type are maximum pond, activation primitive MELU, a1 images of output, a=3-7, a1=28-256, a2
=3-7.
Convolutional layer 2: convolution kernel size b × b, convolution kernel moving step length 1-2, convolution kernel number b1, pond size b2 × b2,
Pond step-length 1-3, pond type are maximum pond, activation primitive MELU, output b1 image b=3-7, b1=28-256, b2
=3-7.
Convolutional layer 3: convolution kernel size c × c, convolution kernel moving step length 1-2, convolution kernel number c1, pond size c2 × c2,
Pond step-length 1-3, pond type are maximum pond, activation primitive Meanout, c1 images of output, c=3-7, c1=28-
256, c2=3-7.
Convolutional layer 4: convolution kernel size d × d, convolution kernel moving step length 1-2, convolution kernel number d1, pond size d2 × d2,
Pond step-length 1-3, pond type are average pond, activation primitive Meanout, d1 images of output, d=3-7, d1=28-
256, d2=3-7.
Full articulamentum: hidden layer unit number 512-1024, activation primitive Meanout.It is configured in order after 4th pond layer
2-4 layers of full articulamentum, for 3 layers, neuron number is respectively 1024,512,10, successively by the 4th pond layer output image
It is converted into corresponding one-dimensional vector, vector element number is respectively 1024,512,10, and the full articulamentum of third exports 10 elements
One-dimension array corresponds to 10 groups of image classifications of this experimental design.
Classification layer: being the third layer of full articulamentum, the hidden layer unit number 10-20, activation primitive Meanout.
Wherein MELU (x)=δ ELU (x) is the mutation of index linear unit ELU, one coefficient δ is added, and parameter δ is
What the parameter that Gaussian Profile is randomly generated obtained in conjunction with the parameter of training.Formula is as follows:
Meanout is that the parameter that Maxout is introduced is excessive, and Meanout then exists different from the place of Maxout
Reduce the parameter scale of introducing on the basis of Maxout, but do not influence trained accuracy, so Meanout is introduced separately
An outer coefficient lambda, λ needs to be adjusted by trained and experience, to improve the efficiency of model.Formula is as follows:
Zij=aTW...ij+bij,W∈Rd×m×k,a∈Rd×n,b∈Rm×k (3)
N indicates the number of input sample, and d indicates the number of a upper node layer, and m indicates the number of this node layer, and k indicates every
A hiding node layer has corresponded to k " intermediate hidden layer " nodes, this k " intermediate hidden layer " nodes are all linear convergent rates, and
Each node of Meanout is just averaged from this k " intermediate hidden layer " nodes, and result only indicates take out i-th here
It arranges, the ellipsis before subscript i indicates all rows in corresponding i-th column.
Parameter initialization, all weight matrix use random_normal (0.0-0.00,0.01-0.001), all inclined
It sets vector and uses constant (0.0-0.00).Use cross entropy as objective function, is carried out using gradient descent method
Parameter updates, and learning rate is set as 0.01-0.001.
Referring to a kind of air cooler dust stratification state computation based on deep learning and infrared image identification of Fig. 2 the utility model
Method, it is characterised in that: the information and air cooler running state parameter obtained using infrared image, by deep learning algorithm and
Air cooling tubes condenser dust stratification thermal resistance value calculating method knows air cooling tubes condenser operating status in real time, realizes that air cooling tubes condenser is smart in short term
Really perception, long-term forecasting function, entire calculation method is by following equation group quantitative descriptions:
The effect for eliminating shake is evaluated with the value of PSNR, wherein PSNR definition is:
Using R, G, B information in color image using the brightness of respective component as the gray value of grayscale image, formula is as follows
It is shown:
f1(i, j)=R (i, j) (6)
f2(i, j)=G (i, j) (7)
f3(i, j)=B (i, j) (8)
Differential operator calculates the amplitude of the gradient at a pixel (x, y) are as follows:
According to heat balance principle, turbine discharge thermal discharge, ACC heat output and cooling air caloric receptivity are equal.
Qe=K × F × Δ T0=G × Cp×ΔTa (11)
The logarithmic mean temperature difference (LMTD) of air cooling tubes condenser:
ΔTa=ta2-ta1 (13)
The difference of the total heat transfer resistance of dust stratification state and the total heat transfer resistance of clean conditions is the dirtiness resistance as caused by dust stratification,
Referred to as black dirt thermal resistance:
Wherein, wherein fi、fi-1Respectively indicate the pixel value of i-th and (i+1) frame, fK(i, j) (k=1,2,3) is indicated
The gray value of each component of coordinate (i, j) point,Qe
For turbine discharge thermal discharge, kW；K is ACC overall heat-transfer coefficient, W/ (m2·K)；F is total heat exchange area, m2；ΔT0It is flat for logarithm
Mean temperature difference, DEG C； CpFor the specific heat at constant pressure of air, kJ/ (kgK)；ΔTaFor cooling air temperature rise, DEG C；G is ventilation quantity, m3/s。
tnFor the adiabatic condensation temperature of turbine discharge, DEG C；ta1And ta2Respectively cooling air in fin channels inlet and outlet temperatures, DEG C.
RfFor black dirt thermal resistance, R is the total heat transfer resistance of dust stratification state, RcFor the total heat transfer resistance of clean conditions, KcIt is always passed for clean conditions ACC
Hot coefficient.By acquiring air cooler dust stratification state infrared image, air cooler wall temperature, air cooler entry and exit cooling air temperature are monitored
Degree, cooling wind flow, in conjunction with structural parameters such as air cooler heat exchange areas, joint type (4)~(5) carry out air cooler infrared image
Debounce processing, joint type (6)~(8) carry out image gray processing processing to obtained image, joint type (9)~(10) to image into
Row edge detection process obtains dust stratification image feature information；Air cooler dust stratification thermal resistance can be obtained in joint type (11)~(14)；With sky
Cooler dust stratification image is input quantity, and air cooler dust stratification thermal resistance is aim parameter, is established based on deep learning network and infrared image
The state recognition of air cooler dust stratification and analysis model.
Claims (2)
1. a kind of air cooler dust stratification state perception system based on deep learning and infrared image identification, it is characterised in that: including
Steam turbine (1) is connected by turbine discharge (2) with steam discharge distribution pipe (3) input terminal；Inlet temperature sensor (6) is arranged in sky
On the inside of cold bank of condenser pipes (4), outlet temperature sensor (7) is arranged on the outside of air cooling tubes condenser tube bank (4), inlet temperature sensing
Device (6) and outlet temperature sensor (7) output end are through temperature transfer bus (16) and data acquisition module (11) input terminal phase
Even；Inlet velocity sensor (8) is arranged on the inside of air cooling tubes condenser tube bank (4), and air outlet velocity sensor (9) is arranged in air-cooled solidifying
Vapour device is restrained on the outside of (4), and inlet velocity sensor (8) and air outlet velocity sensor (9) output end are through wind speed transfer bus
(17) it is connected with data acquisition module (11) input terminal；Thermal infrared imager (10) is arranged on the outside of air cooling tubes condenser tube bank (4) not
At angle, thermal infrared imager (10) output end is through image data transfer bus (18) and data acquisition module (11) input terminal phase
Even；Data acquisition module (11) output end is connected through data transmission bus (19) with industrial personal computer (12) input terminal；Flowmeter (22)
Output end is connected through water-supply-pipe with condensate tank (14) input terminal；Power distribution cabinet (20) output end is acquired through power supply line (21) and data
Module (11), industrial personal computer (12) are connected with the input terminal of air cooling blower fan (13).
2. the air cooler dust stratification state perception system according to claim 1 based on deep learning and infrared image identification,
It is characterized by: the thermal infrared imager (10) is installed on windward side and the leeward of air cooling tubes condenser tube bank (4).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201821563437.6U CN208985185U (en) | 2018-09-25 | 2018-09-25 | Air cooler dust stratification state perception system based on deep learning and infrared image identification |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201821563437.6U CN208985185U (en) | 2018-09-25 | 2018-09-25 | Air cooler dust stratification state perception system based on deep learning and infrared image identification |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN208985185U true CN208985185U (en) | 2019-06-14 |
Family
ID=66785382
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201821563437.6U Active CN208985185U (en) | 2018-09-25 | 2018-09-25 | Air cooler dust stratification state perception system based on deep learning and infrared image identification |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN208985185U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112683101A (en) * | 2020-12-04 | 2021-04-20 | 浙江理工大学 | LABVIEW-based non-uniform heat exchanger automatic cleaning system and method thereof |
-
2018
- 2018-09-25 CN CN201821563437.6U patent/CN208985185U/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112683101A (en) * | 2020-12-04 | 2021-04-20 | 浙江理工大学 | LABVIEW-based non-uniform heat exchanger automatic cleaning system and method thereof |
| CN112683101B (en) * | 2020-12-04 | 2024-04-12 | 浙江理工大学 | Non-uniform heat exchanger automatic cleaning system and method based on LABVIEW |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108875719A (en) | Air cooler dust stratification state perception system and calculation method based on deep learning and infrared image identification | |
| Ge et al. | Performance monitoring of direct air-cooled power generating unit with infrared thermography | |
| Chen et al. | Experimental research of the cross walls effect on the thermal performance of wet cooling towers under crosswind conditions | |
| Hawlader et al. | Numerical study of the thermal–hydraulic performance of evaporative natural draft cooling towers | |
| CN104048842A (en) | On-line monitoring method for heat rate of steam turbine on basis of soft measurement technology | |
| CN107577859A (en) | A kind of Condenser Fouling degree on-line monitoring method | |
| Gough et al. | Evaluating single-sided natural ventilation models against full-scale idealised measurements: Impact of wind direction and turbulence | |
| Gao et al. | Experimental study regarding the evolution of temperature profiles inside wet cooling tower under crosswind conditions | |
| CN109084613A (en) | Air cooling tubes condenser dust stratification status monitoring and cleaning control system and its regulation method based on convolutional neural networks and image recognition | |
| CN105184395B (en) | The initial conditions of fired power generating unit containing afterheat utilizing system determine method | |
| Min et al. | Characteristics of primary air condensation in indirect evaporative cooler: Theoretical analysis and visualized validation | |
| CN102338568A (en) | Online monitoring system and method for performance of condenser in power plant based on cleanness coefficient index | |
| Wu et al. | Flow and heat transfer characteristics of indirect dry cooling system with horizontal heat exchanger A-frames at ambient winds | |
| Kalogirou et al. | Development of a neural network-based fault diagnostic system for solar thermal applications | |
| Dang et al. | Crosswind influence on cooling capacity in different zones for high level water collecting wet cooling towers based on field test | |
| CN208985185U (en) | Air cooler dust stratification state perception system based on deep learning and infrared image identification | |
| CN102494895B (en) | Analyzing method for energy saving and optimization of steam turbine set of power station | |
| Zhang et al. | A favorable face velocity distribution and a V-frame cell for power plant air-cooled condensers | |
| CN104360984B (en) | One kind is based on the lower cooling tower filler performance calculation method of Two point boundary value measurement | |
| Xu et al. | Study on a novel defrost control method based on the surface texture of evaporator image with gray-level cooccurrence matrix, new characterization parameter combination and machine learning | |
| CN108692608A (en) | A kind of heat exchanger operation conditions intelligent on-line monitoring device | |
| Rounis et al. | Multiple-inlet BIPV/T modeling: Wind effects and fan induced suction | |
| Wang et al. | Numerical study on the forced convection enhancement of flat-roof integrated photovoltaic by passive components | |
| CN208606637U (en) | A kind of heat exchanger operation conditions intelligent on-line monitoring device | |
| Qin et al. | Experimental research on convection–condensation heat transfer characteristics and sootblowing operation strategy of waste heat recovery heat exchanger for dryer exhaust |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant |