CN107178146B - A kind of city planting ductwork dredging robot and method of work based on BP neural network - Google Patents
A kind of city planting ductwork dredging robot and method of work based on BP neural network Download PDFInfo
- Publication number
- CN107178146B CN107178146B CN201710375597.1A CN201710375597A CN107178146B CN 107178146 B CN107178146 B CN 107178146B CN 201710375597 A CN201710375597 A CN 201710375597A CN 107178146 B CN107178146 B CN 107178146B
- Authority
- CN
- China
- Prior art keywords
- vehicle body
- dredging
- track
- neural network
- motor group
- 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.)
- Expired - Fee Related
Links
Classifications
-
- E—FIXED CONSTRUCTIONS
- E03—WATER SUPPLY; SEWERAGE
- E03F—SEWERS; CESSPOOLS
- E03F9/00—Arrangements or fixed installations methods or devices for cleaning or clearing sewer pipes, e.g. by flushing
- E03F9/002—Cleaning sewer pipes by mechanical means
-
- 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
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Water Supply & Treatment (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Hydrology & Water Resources (AREA)
- Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)
Abstract
The invention discloses a kind of city planting ductwork dredging robots and method of work based on BP neural network,Including vehicle body,Body bottom is provided with tire,Battery pack is provided on vehicle body,BP computing modules and motor group,Storage storehouse is provided in vehicle body,Vehicle foreside is inclined-plane,The both sides on inclined-plane are provided with track of sliding track,Vehicle foreside is provided with perching knife and the silt dredging roller that can be slided on track of sliding track,Vehicle foreside is provided with infrared detector,Infrared detector connects BP computing modules,BP computing modules connect silt dredging roller and perching knife by motor group,Present apparatus structure is smaller,It can enter in narrow municipal network pipeline and be cleared up,Not only solve the human safety issues of artificial dredging,It avoids staff and is exposed to runway for a long time,Substantially increase staff's inherently safe coefficient,And solve the problems, such as using Large-scale High-Pressure rinse vehicle can not thorough cleaning pipe network mud.
Description
Technical field
The invention belongs to pipe dredging fields, and in particular to a kind of city planting ductwork dredging robot based on BP neural network
And method of work.
Background technology
The important means that city planting ductwork drains flooded fields as the main foundation facility of urban construction and urban flood defence, with people
Life have close contact, while also have great importance to the expanding economy in city.As modern city not
The important infrastructure that can or lack, city planting ductwork maintenance work are of great significance as urban economy development, are municipal water dirts
The major measure of dye prevention, water drainage and flood control.City planting ductwork drains off floodwaters main as urban water supply, water prevention and cure of pollution and prevention waterlogging
Means if often there is the phenomenon that pipeline siltation, block, may result in sewage and cannot exclude, and make troubles to people's life
While also pollute the environment.
It often will appear the weak enterprise of some laws and regulations consciousness in city planting ductwork or personal disorderly row leave about, lead to a large amount of dirts
Water, a large amount of fragments and waste etc. enter city planting ductwork, it is easy to pipe network be caused to silt up, such as clear up, on the one hand can produce not in time
Raw mechanical stress damages pipe network structure, and on the one hand accumulation can influence pipe network blowdown flow velocity for a long time, and flood season easily causes row
It is dirty city to be caused to see extra large phenomenon not in time, there are a large amount of greasy dirts and toxic and harmful substance in another aspect sludge, not in time
Cleaning easily causes secondary pollution to environment.
City planting ductwork desilting mode mainly has winch dredging, the dredging of thorough cut machine, high-pressure sewer flushing vehicle dredging, artificial dredging at present
Deng, but it is all there are a variety of different defects, such as artificial dredging：It takes long, human cost is excessively high, occupies road easily causes traffic
It blocks；Main still to be dredged by manpower at present in the maintenance of many Municipal pipe networks in China, the tool used also more passes
System is simple, leads to that maintenance work efficiency is low, intensity is big, waste of manpower.
City planting ductwork is mostly undersized at present, and personnel are not easily accessible, and can only pass through manually or mechanically dredging, man efficiency
Lowly, and mechanical chipping is excessively coarse omits too many, and sundries is excessive in pipeline, and situation is complicated, and cleaning is inconvenient.
The learning process of BP algorithm (back-propagation algorithm), by the forward-propagating of information and two mistakes of backpropagation of error
Cheng Zucheng.Each neuron of input layer is responsible for receiving from extraneous input information, and pass to each neuron of middle layer；Middle layer
Internal information process layer, be responsible for information transformation, according to the demand of information change ability, middle layer can be designed as single hidden layer or
The more hidden layer configurations of person；The last one hidden layer is transmitted to the information of each neuron of output layer, after further treatment after, complete primary learn
The forward-propagating processing procedure of habit, by output layer outwardly output information handling result.When reality output and desired output are not inconsistent
When, into the back-propagation phase of error.Error corrects each layer weights by output layer in a manner that error gradient declines, to
The successively anti-pass of hidden layer, input layer.Information forward-propagating in cycles and error back propagation process are that each layer weights are constantly adjusted
Whole process and the process of neural network learning training, this process is performed until the error of network output, and be reduced to can be with
Until the degree of receiving or preset study number.
Invention content
The purpose of the present invention is to overcome the above shortcomings and to provide a kind of city planting ductwork dredging machines based on BP neural network
People and method of work are solved because city planting ductwork caliber is too small, be cannot be introduced into and are used artificial dredging, but artificial dredging takes
Long, human cost is excessively high, occupies the problems such as road easily causes traffic jam.
In order to achieve the above object, a kind of city planting ductwork dredging robot based on BP neural network, including vehicle body, vehicle body
Bottom is provided with tire, and battery pack, BP computing modules and motor group are provided on vehicle body, storage storehouse, vehicle body are provided in vehicle body
Forepart is inclined-plane, and the both sides on inclined-plane are provided with track of sliding track, and vehicle foreside is provided with perching knife and can be slided on track of sliding track
Silt dredging roller, vehicle foreside is provided with infrared detector, and infrared detector connection BP computing modules, BP computing modules pass through electricity
Unit connects silt dredging roller and perching knife.
The both sides of the vehicle body are both provided with track, and mechanical arm is both provided on two tracks, is set on mechanical arm useful
In the sliding rail fixed block slided in orbit, the distal end of mechanical arm is provided with for the fixed roller of hinged silt dredging roller.
The perching knife is fixed on body bottom by sliding rail, and perching knife can stretch in body bottom.
The infrared detector is two, is respectively arranged with the both sides above vehicle foreside.
The battery pack, BP computing modules and motor group are arranged at the rear portion of vehicle body.
A kind of control method of the city planting ductwork dredging robot based on BP neural network, includes the following steps：
Step 1, infrared detector acquires the information in pipeline, and sends information to BP computing modules and calculated, and counts
It calculates result and is sent to motor group；
Step 2, motor group control machinery arm is run downwards along track of sliding track, so as to which silt dredging roller be driven to extend downwardly,
After dropping to operating position, silt dredging roller starts rotation and carries out silt dredging；
Step 3, when infrared detector finds to have large-scale foreign matter in pipeline, motor group control machinery arm resets, and controls
Perching knife enters large-scale spatula for removing foreign bodies in storage storehouse；
Step 4, after large-scale foreign matter is rooted out, electric unit allocation perching knife is withdrawn, and silt dredging roller continues after dropping to operating position
Operation.
In the step 1, the method for work of BP computing modules includes the following steps：
The first step, system initialization；
Second step calculates linear combination coefficient to l layer units j,
This layer of output is calculated,
Sigmoid functions are taken,
If l is the first hidden layer, have
If l be output layer, set output as
Third walks, and carries out backwards calculation, and partial gradient is calculated to l layer units jIf l is output layer,
If taking sigmoid functions, have
If l is hidden layer, have
4th step carries out modified weight, if n is calculation times, then modified weight formula is,
N=n+1
Take new samples x1x2Return to step two, it is primary to each sample learning in sample set, you can to complete.
Compared with prior art, the device of the invention roots out large-scale foreign matter by perching knife, by silt dredging roller process mud,
The large-scale foreign matter and mud rooted out are collected in storage storehouse, present apparatus structure is smaller, can enter in narrow municipal network pipeline
It is cleared up, not only solves the human safety issues of artificial dredging, avoided staff and be exposed to runway for a long time, greatly
Improve staff's inherently safe coefficient greatly, and solve using Large-scale High-Pressure rinse vehicle can not thorough cleaning pipe network mud
The problem of.
The method of the present invention acquires signal by infrared detector, and perching knife and silt dredging is controlled to roll after being calculated by BP algorithm
Vehicle control is combined by cylinder work, this method with BP algorithm, and dredging robot is made to have study, adaptive, evolution and heredity
Engineering, it is about 0.05 to be computed learning rate, and present method solves machineries to artificial dependence, enhances the autonomous energy of machinery
Power improves work efficiency.
Description of the drawings
Fig. 1 is the structure diagram of invention；
Fig. 2 is the computation model of BP algorithm in the present invention；
Fig. 3 is the computing unit of neural network in the present invention；
Wherein, 1, battery pack；2nd, BP computing modules；3rd, motor group；4th, tire；5th, storage storehouse；6th, sliding rail fixed block；7th, it inhales
Silt roller；8th, perching knife；9th, infrared detector；10th, track；11st, mechanical arm；12nd, sliding rail slideway；13rd, fixed roller；14th, vehicle body.
Specific embodiment
The present invention will be further described below in conjunction with the accompanying drawings.
Referring to Fig. 1, a kind of city planting ductwork dredging robot based on BP neural network, including vehicle body 14,14 bottom of vehicle body
Tire 4 is provided with, battery pack 1, BP computing modules 2 and motor group 3 are provided on vehicle body 14, storage storehouse 5 is provided in vehicle body 14,
14 forepart of vehicle body is inclined-plane, and the both sides on inclined-plane are provided with track of sliding track 12, and 14 forepart of vehicle body is provided with perching knife 8 and can be in sliding rail
The silt dredging roller 7 slided on track 12,14 forepart of vehicle body are provided with infrared detector 9, and infrared detector 9 connects BP computing modules
2, BP computing modules 2 connect silt dredging roller 7 and perching knife 8 by motor group 3.
The both sides of vehicle body 14 are both provided with track 10, and mechanical arm 11 is both provided on two tracks 10, is set on mechanical arm 11
The sliding rail fixed block 6 for being slided in track 10 is equipped with, the distal end of mechanical arm 11 is provided with consolidating for hinged silt dredging roller 7
Determine roller bearing 13.
Perching knife 8 is fixed on 14 bottom of vehicle body by sliding rail, and perching knife 8 can stretch in 14 bottom of vehicle body.
Preferably, infrared detector 9 is two, is respectively arranged with the both sides of 14 front upper of vehicle body, battery pack 1, BP meters
It calculates module 2 and motor group 3 is arranged at the rear portion of vehicle body 14.
Referring to Fig. 2 and Fig. 3, a kind of control method of the city planting ductwork dredging robot based on BP neural network, including with
Lower step：
Step 1, infrared detector 9 acquires the information in pipeline, and sends information to BP computing modules and calculated,
Result of calculation is sent to motor group 3；
Step 2,3 control machinery arm 11 of motor group along track of sliding track 12 downwards run, so as to drive silt dredging roller 7 to
Lower extension, after dropping to operating position, silt dredging roller 7 starts rotation and carries out silt dredging；
Step 3, when infrared detector 9 finds to have large-scale foreign matter in pipeline, 3 control machinery arm 11 of motor group resets, and
Control perching knife 8 enters large-scale spatula for removing foreign bodies in storage storehouse 5；
Step 4, after large-scale foreign matter is rooted out, motor group 3 controls perching knife 8 to withdraw, after silt dredging roller 7 drops to operating position
It continues to run with.
The method of work of BP computing modules includes the following steps：
The first step, system initialization；
Second step calculates linear combination coefficient to l layer units j,
This layer of output is calculated,
Sigmoid functions are taken,
If l is the first hidden layer, haveWherein x is sample；
If l be output layer, set output asWherein y is sample；
Third walks, and carries out backwards calculation, and partial gradient is calculated to l layer units jIf l is output layer,
If taking sigmoid functions, have
If l is hidden layer, have
4th step carries out modified weight, if n is calculation times, then modified weight formula is,
N=n+1
Take new samples x1x2Return to step two, it is primary to each sample learning in sample set, you can to complete.
Embodiment：
First, dredging robot maximum travel distance calculating process：
Because vertical shaft stock size is 300mm, the sludge volume stored to robot is maximum, then
VMudLmax=Vmax
I.e.：
2nd, BP neural network calculates：
eij=rij/rij
Obtain reciprocal matrix E=(eij)n×nThen
W is obtained by 3 iteration(3)=[0.159 0.170 0.170 0.042 0.127 0.098 0.130 0.068]T
Iteration precision is 0.001
The completion standard of BP algorithm of neural network (back-propagation algorithm)：
To in sample set, each sample learning is primary, calculates the 2- norms of a desired output and reality output, and will
It is accumulated to accumulation norm,
After the completion of one epoch, each sample has carried out primary study, while accumulates norm,
E=0.05
Last travel distance calculates：
A3=SL；L=89m.
The present invention solves the shorter limitation of daily dredging distance, this equipment, which once can at most advance, clears up 89 meters.
Claims (6)
1. a kind of city planting ductwork dredging robot based on BP neural network, which is characterized in that including vehicle body (14), vehicle body (14)
Bottom is provided with tire (4), and battery pack (1), BP computing modules (2) and motor group (3), vehicle body (14) are provided on vehicle body (14)
Inside it is provided with storage storehouse (5), vehicle body (14) forepart is inclined-plane, and the both sides on inclined-plane are provided with track of sliding track (12), before vehicle body (14)
Portion is provided with perching knife (8) and the silt dredging roller (7) that can be slided on track of sliding track (12), and vehicle body (14) forepart is provided with infrared
Detector (9), infrared detector (9) connection BP computing modules (2), BP computing modules (2) connect silt dredging by motor group (3) and roll
Cylinder (7) and perching knife (8)；
The perching knife (8) is fixed on vehicle body (14) bottom by sliding rail, and perching knife (8) can stretch in vehicle body (14) bottom.
A kind of 2. city planting ductwork dredging robot based on BP neural network according to claim 1, which is characterized in that institute
The both sides for stating vehicle body (14) are both provided with track (10), and mechanical arm (11), mechanical arm (11) are both provided on two tracks (10)
On be provided with sliding rail fixed block (6) for sliding in the track (10), the distal end of mechanical arm (11) is provided with hingedly to inhale
The fixed roller (13) of silt roller (7).
A kind of 3. city planting ductwork dredging robot based on BP neural network according to claim 1, which is characterized in that institute
It is two to state infrared detector (9), is respectively arranged with the both sides of vehicle body (14) front upper.
A kind of 4. city planting ductwork dredging robot based on BP neural network according to claim 1, which is characterized in that institute
State the rear portion that battery pack (1), BP computing modules (2) and motor group (3) are arranged at vehicle body (14).
5. a kind of control method of city planting ductwork dredging robot based on BP neural network described in claim 1, feature
It is, includes the following steps：
Step 1, infrared detector (9) acquires the information in pipeline, and sends information to BP computing modules and calculated, and counts
It calculates result and is sent to motor group (3)；
Step 2, motor group (3) control machinery arm (11) is run downwards along track of sliding track (12), so as to drive silt dredging roller
(7) it extends downwardly, after dropping to operating position, silt dredging roller (7) starts rotation and carries out silt dredging；
Step 3, when infrared detector (9) finds to have large-scale foreign matter in pipeline, motor group (3) control machinery arm (11) resets,
And perching knife (8) is controlled to enter large-scale spatula for removing foreign bodies in storage storehouse (5)；
Step 4, after large-scale foreign matter is rooted out, motor group (3) control perching knife (8) is withdrawn, and silt dredging roller (7) drops to operating position
After continue to run with.
6. a kind of control method of city planting ductwork dredging robot based on BP neural network according to claim 5,
It is characterized in that, in the step 1, the method for work of BP computing modules includes the following steps：
The first step, system initialization；
Second step, wjiFor weights, j is current layer unit, and i is last layer unit, and k is next layer unit, and when i is 0, weights are
Threshold value calculates linear combination coefficient to l layer units j,
This layer of output is calculated,
Sigmoid functions are taken,
If l is the first hidden layer, have
If l be output layer, set output as
Third walks, and carries out backwards calculation, and partial gradient is calculated to l layer units jIf l is output layer,
If taking sigmoid functions, have
If l is hidden layer, have
4th step carries out modified weight, if n is calculation times, then modified weight formula is,
N=n+1
Take new samples x1x2Return to step two, it is primary to each sample learning in sample set, you can to complete.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710375597.1A CN107178146B (en) | 2017-05-24 | 2017-05-24 | A kind of city planting ductwork dredging robot and method of work based on BP neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710375597.1A CN107178146B (en) | 2017-05-24 | 2017-05-24 | A kind of city planting ductwork dredging robot and method of work based on BP neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107178146A CN107178146A (en) | 2017-09-19 |
| CN107178146B true CN107178146B (en) | 2018-06-08 |
Family
ID=59831433
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710375597.1A Expired - Fee Related CN107178146B (en) | 2017-05-24 | 2017-05-24 | A kind of city planting ductwork dredging robot and method of work based on BP neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107178146B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110344494B (en) * | 2019-07-15 | 2020-11-10 | 岳西县双节路桥工程有限公司 | Drainage pipe is with device of decontaminating |
| CN111160485B (en) * | 2019-12-31 | 2022-11-29 | 中国民用航空总局第二研究所 | Regression training-based abnormal behavior detection method and device and electronic equipment |
| CN111360006A (en) * | 2020-04-14 | 2020-07-03 | 夏邦传 | Bottom sludge cleaning device for sewage purification tank |
| CN113266559B (en) * | 2021-05-21 | 2022-10-28 | 华能秦煤瑞金发电有限责任公司 | Neural network-based wireless detection method for concrete delivery pump blockage |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2611539B1 (en) * | 1987-02-24 | 1989-09-08 | Corefic | MACHINE FOR INTERNAL CLEANING OF PIPES |
| CN203821582U (en) * | 2013-12-07 | 2014-09-10 | 张冬 | Intelligent sewer cleaning vehicle |
| CN104060683A (en) * | 2014-06-24 | 2014-09-24 | 北京建筑大学 | Pneumatic underground drainage pipeline dredging device |
| CN105297881A (en) * | 2015-11-16 | 2016-02-03 | 北京建筑大学 | Pneumatic type underground drainage pipeline desilting robot with desilting screw conveyor |
-
2017
- 2017-05-24 CN CN201710375597.1A patent/CN107178146B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| FR2611539B1 (en) * | 1987-02-24 | 1989-09-08 | Corefic | MACHINE FOR INTERNAL CLEANING OF PIPES |
| CN203821582U (en) * | 2013-12-07 | 2014-09-10 | 张冬 | Intelligent sewer cleaning vehicle |
| CN104060683A (en) * | 2014-06-24 | 2014-09-24 | 北京建筑大学 | Pneumatic underground drainage pipeline dredging device |
| CN105297881A (en) * | 2015-11-16 | 2016-02-03 | 北京建筑大学 | Pneumatic type underground drainage pipeline desilting robot with desilting screw conveyor |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107178146A (en) | 2017-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107178146B (en) | A kind of city planting ductwork dredging robot and method of work based on BP neural network | |
| Wang et al. | Regional prediction of ground-level ozone using a hybrid sequence-to-sequence deep learning approach | |
| CN105973207B (en) | Large-scale shallow water lake wawter bloom gathers the moisturizing algae control method in waters | |
| CN111160520A (en) | BP neural network wind speed prediction method based on genetic algorithm optimization | |
| CN107386418B (en) | A kind of urban road rainwater recycle reuse method and system | |
| CN107909206A (en) | A kind of PM2.5 Forecasting Methodologies based on deep structure Recognition with Recurrent Neural Network | |
| CN103008334A (en) | In-situ rainwater leaching repair system for contaminated soil | |
| CN106772695B (en) | A kind of Along Railway wind speed forecasting method for merging many air measuring station measured datas | |
| CN106530172A (en) | Method of dividing watershed pollution control unit | |
| CN107818410A (en) | A kind of photovoltaic module dust stratification degree long-distance intelligent diagnostic method | |
| CN111626382A (en) | Rapid intelligent identification method and system for cleanliness of vehicle on construction site | |
| US11694005B2 (en) | Method, system and equipment for vegetation restoration or rehabilitation based on machine learning | |
| CN107071053A (en) | A kind of road ponding monitoring system based on WSAN | |
| CN111400973A (en) | Method for constructing flow-water surface width relation curve based on hydrologic monitoring data | |
| CN114529058A (en) | Prediction method and system for rock block distribution in filled earth stratum | |
| CN207048027U (en) | A kind of city planting ductwork dredging robot | |
| CN107730882A (en) | Congestion in road forecasting system and method based on artificial intelligence | |
| CN109098267B (en) | Mountain building complex rainwater prevention, control and utilization system and method based on GIS | |
| M García-Barcina et al. | Modelling the faecal coliform concentrations in the Bilbao estuary | |
| Nasrulin et al. | Environmental monitoring in construction for hydro-technical structures and land reclamation facilities of Uzbekistan | |
| CN116167608A (en) | Comprehensive improvement method for urban waterlogging ponding depth risk area | |
| CN113762729A (en) | Comprehensive evaluation method for urban drainage system modification project | |
| CN114241299A (en) | Method and system for detecting garbage density of block chain water area based on genetic algorithm | |
| CN208280717U (en) | A kind of ecotecture for having organic Agricultural Water and producing breeding function and interior circular treatment function | |
| Yang et al. | Short-term passenger flow forecast of urban rail transit based on BP neural network |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180608 Termination date: 20190524 |