CN106096558B - A kind of binding reinforcing bars method neural network based - Google Patents
A kind of binding reinforcing bars method neural network based Download PDFInfo
- Publication number
- CN106096558B CN106096558B CN201610428597.9A CN201610428597A CN106096558B CN 106096558 B CN106096558 B CN 106096558B CN 201610428597 A CN201610428597 A CN 201610428597A CN 106096558 B CN106096558 B CN 106096558B
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- reinforcing
- diameter
- bar
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- 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
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/061—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using biological neurons, e.g. biological neurons connected to an integrated circuit
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Neurology (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Microelectronics & Electronic Packaging (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Basic Packing Technique (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The invention discloses a kind of binding reinforcing bars methods neural network based, this method installs radar-probing system on reinforcing-bar binding machine, to obtain the diameter data to wiretie, then successively direct wave is carried out to diameter data to handle, Dewow processing, trim data matrix, matrix gray processing, the processing of matrix gray processing, data to be identified are obtained after matrix binary conversion treatment, three layers of WAVELET PACKET DECOMPOSITION are carried out to data to be identified, third layer coefficients are reconstructed, construction feature vector, feature vector is trained using BP neural network, bar diameter data after finally obtaining training, again with this data point reuse motor output torque.The present invention overcomes tied silk existing in the prior art and is easy the problems such as twisting into two parts, is not firm on the basis of guaranteeing binding reinforcing bars efficiency, provides a kind of strong auxiliary tool for construction.
Description
Technical field
The present invention relates to a kind of construction methods, and in particular to a kind of binding reinforcing bars method neural network based.
Background technique
In building trade be laid with reinforcing bar after to the purpose that staggered reinforcing bar is attached be in order to guarantee that reinforcement location is constant,
And then ensure that reinforcing steel bar bear position does not change.If longitudinal reinforcement can be worked as when small numbers of using welding
Longitudinal reinforcement number mostly uses greatly the mode tied up when more.Transverse steel generally also uses and ties up mode.Currently,
Mainly using mode is tied up by hand when the wiretie of construction site, tie up has the physical strength and qualification of bar bender by hand
Very high requirement, and tie up by hand both laborious and time consuming.Occur a kind of to carry out binding reinforcing bars automatically currently on the market
Handheld machine reinforcing-bar binding machine, it is a kind of novel reinforcement construction intelligent electric tool.When using this machine, efficiency
Than manually can be improved 2~3 times, a large amount of manpower can be saved, substantially reducing the labor intensity of worker.
But the usage amount of existing reinforcing-bar binding machine tied silk when tying up is more much larger than manually tying up dosage, is unfavorable for pricking
The effective use of silk;In addition, the circle number around reinforcing bar can not automatically adjust when using reinforcing-bar binding machine, when banding tied silk without
The output torque of method real-time control motor and be easy to appear tied silk and twist into two parts or twist illusive situation.
Summary of the invention
For above-mentioned problems of the prior art, the object of the present invention is to provide a kind of based on neural network
Binding reinforcing bars method, this method are applied on existing reinforcing-bar binding machine, when solving the banding tied silk occurred in the prior art
It twists into two parts tied silk or twists the problems such as not firm, improve bundling quality.
In order to realize above-mentioned task, the invention adopts the following technical scheme:
A kind of binding reinforcing bars method neural network based, comprising the following steps:
Step 1 is installed radar-probing system on the bayonet of reinforcing-bar binding machine front end, and is set inside reinforcing-bar binding machine
Controller is set, radar-probing system is connect with controller, and the microprocessor of controller and reinforcing-bar binding machine is connected;
Step 2, it is in cross-shaped reinforcing bar that when construction site is constructed, reinforcing bar, which is laid with scene and chooses two, utilizes thunder
The diameter data of a wherein reinforcing bar is obtained up to detection system;
Step 3 is removed direct wave processing, Dewow processing to collected diameter data respectively, and rejects and do not exist
Then sampled point in range carries out gray processing processing, obtains grayscale image, grayscale image is normalized, will finally return
One changes that treated that grayscale image is converted into bianry image, and the data in bianry image are data to be identified;
Step 4 carries out three layers of WAVELET PACKET DECOMPOSITION to data to be identified, each coefficient of third layer is reconstructed, and extracts each
The signal of frequency range finds out the energy of each frequency range signal, is configured to feature vector using energy, and to feature vector into
Row normalized;
Step 5, by the eigendecomposition after normalization at multiple matrixs in block form, each matrix in block form is as one
Sample;The first half matrix in block form for multiple matrixs in block form that segmentation is obtained is made as training sample database, later half matrix in block form
For test sample library;
Step 6 establishes BP neural network, and multiple matrixs in block form in step 5 are input to conduct in BP neural network
The input layer number of input data, BP neural network is identical as the number of the feature vector, output layer neuron number
It is 1, hidden layer neuron number is determined using trial and error procedure;Set frequency of training, every training it is complete it is primary after record BP neural network
Training error and test error, select frequency of training when test error minimum as network training number;
After determining network training number, tested, obtain training result and test result, by correct probability 90% with
On to be tied up diameter of the data as reinforcing bar;
Then step 7 is pressed using the diameter data of another reinforcing bar described in radar-probing system obtaining step two
It is handled according to step 3 to the identical method of step 6, obtains the diameter to be tied up for being directed to the reinforcing bar;Controller is by two
The diameter to be tied up of reinforcing bar feeds back to microprocessor, output torque of the microprocessor according to diameter adjustment motor to be tied up, output
Reinforcing bar is tied up after after torque adjusting.
Further, the formula that trial and error procedure described in step 6 uses are as follows:
In above formula, m is hidden neuron number, and n is input layer number, and l is output layer neuron number, and a is less than 10
Positive integer.
Further, the frequency of training set in step 6 is 10~1000 time.
The present invention has following technical characterstic compared with prior art:
1. the present invention on the basis of guaranteeing binding reinforcing bars efficiency, overcomes tied silk existing in the prior art and is easy to twist
The problems such as disconnected, not firm, provides a kind of strong auxiliary tool for construction.
2. binding method neural network based proposed by the present invention can pass through motor in reinforcing-bar binding machine with real-time control
Electric current so that control the output torque of motor, so that automatically adjust reinforcing bar ties up elasticity, so that binding reinforcing bars mistake
Journey can remain preferable quality.
3. the present invention is small to existing reinforcing-bar binding machine structural modification, can be mounted on various types of reinforcing-bar binding machines,
With good portability.
Detailed description of the invention
Fig. 1 is the overall flow figure of the method for the present invention;
Fig. 2 is the data prediction flow chart of step 3;
Fig. 3 is the tree construction of WAVELET PACKET DECOMPOSITION;
Fig. 4 is BP neural network structure diagram in the present invention;
Fig. 5 is the schematic diagram after radar-probing system is mounted on reinforcing-bar binding machine;
Fig. 6 is the effect picture that binding reinforcing bars are carried out using the method for the present invention;
Figure label represents: 1-radar-probing system, 2-bayonets, 3-reinforcing-bar binding machines, 4-rechargeable batteries, and 5-is logical
Electric switch.
Specific embodiment
In compliance with the above technical solution, as shown, the present invention provides a kind of binding reinforcing bars method neural network based,
The following steps are included:
Step 1 is installed radar-probing system on the bayonet of reinforcing-bar binding machine front end, and is set inside reinforcing-bar binding machine
Controller is set, the data for the acquisition of dissection process radar-probing system;Radar-probing system is connect with controller, and will control
The microprocessor of device and reinforcing-bar binding machine processed connects;
As shown, be shaped like in pistol, tail portion is equipped with wire box for the structure of existing reinforcing-bar binding machine, in wire box
For tied silk coil, tied silk is supplied for binding reinforcing bars process;Its internal structure mainly include wire feed and shearing mechanism, winding mechanism,
Mechanism, microcontroller, transmission rotating device, distribution mechanism, motor etc. are wrung, pikestaff lower end can install rechargeable battery, in rifle
Trigger is exactly power-on switch at handle, for the on-off of control circuit and the working condition of reinforcing-bar binding machine.Reinforcing-bar binding machine front end
There is bayonet, the movement of bayonet controls motor by motor control, therefore by microcontroller, and bayonet can be adjusted when binding silk
Tightness.
Step 2, Selection Model, the actual conditions that reinforcing bar is laid with when being constructed according to construction site, when construction site is constructed
It is in cross-shaped reinforcing bar that reinforcing bar, which is laid with scene and chooses two, and the diameter of a wherein reinforcing bar is obtained using radar-probing system
Data;Here influence of the diameter data read by factors such as distance, signal strength or weakness is a general data, is had certain
Error, it is therefore desirable to diameter data accurately be obtained by subsequent network training, in order to the process of binding of tied silk.
Data prediction: step 3 is removed direct wave processing to collected diameter data, method is to utilize respectively
Each collected data subtract the arithmetic average for collecting total data, do so and air direct wave can be effectively reduced
With the influence of antenna coupled wave;Then Dewow processing is carried out, mainly by applying a zero phase high-pass filtering to signal come real
Existing, the chatter that signal deviates mean value can be reduced by doing so;Next obtained data matrix is trimmed, is rejected not in model
Interior sampled point is enclosed, to reduce operand;Then gray processing processing is carried out, grayscale image is obtained, place is normalized to grayscale image
The value range of data is set in [0,1] by reason;Bianry image finally is converted by the grayscale image after normalized, is made
The value of entire data matrix only has 0 and 1;The data in bianry image are data to be identified at this time;
Step 4, extract characteristic parameter: to data to be identified carry out three layers of WAVELET PACKET DECOMPOSITION, by each coefficient of third layer into
Row reconstruct, extracts the signal of each frequency range, finds out the energy of each frequency range signal, be configured to feature vector using energy,
And feature vector is normalized;
In the present embodiment, the specific 8 frequency content signals of third layer from low to high that extract are as energy parameter.It is small
For wave packet hierarchical structure as shown in figure 3, what wherein (0,0) represented is initial data, what (3,0) represented is the 0th node of third layer
Coefficient X30, that (3,1) represent is the coefficient X of the 1st node of third layer31, other and so on.
Each coefficient for the third layer that WAVELET PACKET DECOMPOSITION obtains is reconstructed, each frequency band (low-frequency band and high frequency band) model is extracted
The signal enclosed, with S30Indicate X30Reconstruction signal, with S31Indicate X31Reconstruction signal, other and so on.
The energy for seeking each frequency band (low-frequency band and high frequency band) range signal, in order that seeking feature vector P, so as to find out returning
Vector T after one change.If S3j(j=0,1,2,3,4,5,6,7) corresponding energy is E3j(j=0,1,2,3,4,5,6,7), then
Have:
In above formula, xjk(j=0,1,2 ..., 7;K=1,2 ..., n) indicate reconstruction signal S3jDiscrete point amplitude, n
Indicate reconstruction signal S3jThe number of discrete point.
Construction feature vector: reinforcing bar can have a significant impact to the energy of each inband signal of radar return, therefore with energy
Secondary element can construct a feature vector.Feature vector P construction is as follows:
P=[E30,E31,E32,E33,E34,E35,E36,E37]
Wherein, E3jIt (j=0,1,2,3,4,5,6,7) is the energy of each frequency range signal.
In addition, due to the characteristic of neural network transmission function needing that P is normalized.It enables
Vector T is exactly the feature vector after normalization.
Step 5, the eigendecomposition after the normalization that step 4 is obtained is at multiple matrixs in block form, each piecemeal
Matrix is as a sample;The first half matrix in block form for multiple matrixs in block form that segmentation is obtained is latter as training sample database
Half matrix in block form is as test sample library.
In the present embodiment, by the eigendecomposition after normalization at multiple 8x8 rank square matrixes;The first half that segmentation is obtained
Matrix in block form where columns is as training sample database, and matrix in block form is as test sample library where later half columns.
Step 6 establishes BP neural network, the structure of BP neural network are as follows: input layer number-hidden layer neuron
The structure diagram of number-output layer neuron number, BP network of the invention is as shown in Figure 4.
To simplify data handling procedure, operability is provided, multiple matrixs in block form in step 5 are input to BP nerve
Input data is used as in network.
The input layer number of BP neural network is identical as the number of the feature vector, in the present embodiment, feature
Element number is 8 in vector P, therefore input layer number is 8;Output layer neuron number is 1, and Y is to pass through net in the figure
The bar diameter size finally measured after network training;Hidden layer neuron number is determined using trial and error procedure, under can be used to when examination is gathered
Formula is estimated, it may be assumed that
In above formula, m is hidden neuron number, and n is input layer number, and l is output layer neuron number, and a is less than 10
Positive integer;
After determining hidden layer neuron number, to obtain relatively good extensive effect, optimal training time should be selected
Number s, then network training is insufficient for deconditioning before this, and deconditioning is then overtrained after this.
In the present embodiment, frequency of training s=10~1000 are set, with 500 incrementeds, the complete primary rear record of every training
The training error and test error of lower BP neural network select frequency of training (the extensive effect at this time when test error minimum
It is best) it is used as network training number;
It after determining network training number, is tested, obtains training result and test result, training error and test error
Most of no more than 0.2mm, the correct probability of diameter to be tied up is determined with this;Correct probability is made in 90% or more data
For the diameter to be tied up of reinforcing bar;The diameter to be tied up for the reinforcing bar chosen in step 2;
Step 7 utilizes another in the reinforcing bar of cross-shaped arrangement described in radar-probing system obtaining step two
Then the diameter data of reinforcing bar is handled according to step 3 to the identical method of step 6, obtain being directed to the reinforcing bar to
Tie up diameter;The diameter to be tied up of two reinforcing bars is fed back to microprocessor by controller, is made with the diameter to be tied up of two reinforcing bars
For the accurate diameter of reinforcing bar;Bar diameter pass corresponding with motor output torque is stored in the microprocessor of reinforcing-bar binding machine
It is table, this mapping table is motor torque and bar diameter when the dynamics of tying up reaches best in different bar diameters
Corresponding relationship.Microprocessor adjusts the output torque of motor, to control the tightness of tying up of tied silk, adjusts output torque
Afterwards, bayonet is stuck on cross-shaped reinforcing bar, presses the power-on switch of reinforcing-bar binding machine, reinforcing bar is tied using tied silk
It pricks.
Above-mentioned step two to step 7 is completed to one in the staggered binding for laying reinforcing bar of cross, is needed under binding
One cross interlock reinforcing bar when, it is only necessary to reinforcing-bar binding machine is moved to next cross and is interlocked the top of reinforcing bar, step is repeated
Two to step 7.
Claims (3)
1. a kind of binding reinforcing bars method neural network based, which comprises the following steps:
Step 1 installs radar-probing system on the bayonet of reinforcing-bar binding machine front end, and control is arranged inside reinforcing-bar binding machine
Radar-probing system is connect by device processed with controller, and the microprocessor of controller and reinforcing-bar binding machine is connected;
Step 2, it is in cross-shaped reinforcing bar that when construction site is constructed, reinforcing bar, which is laid with scene and chooses two, is visited using radar
Examining system obtains the diameter data of a wherein reinforcing bar;
Step 3 is removed direct wave processing, Dewow processing to collected diameter data respectively, and rejects not in range
Then interior sampled point carries out gray processing processing, obtains grayscale image, grayscale image is normalized, finally will normalization
Treated, and grayscale image is converted into bianry image, and the data in bianry image are data to be identified;
Step 4 carries out three layers of WAVELET PACKET DECOMPOSITION to data to be identified, each coefficient of third layer is reconstructed, each frequency band is extracted
The signal of range, finds out the energy of each frequency range signal, is configured to feature vector using energy, and return to feature vector
One change processing;
Step 5, by the eigendecomposition after normalization at multiple matrixs in block form, each matrix in block form is as a sample;
The first half matrix in block form for multiple matrixs in block form that segmentation is obtained is as training sample database, and later half matrix in block form is as test
Sample database;
Step 6 establishes BP neural network, and multiple matrixs in block form in step 5 are input in BP neural network as input
The input layer number of data, BP neural network is identical as the number of the feature vector, and output layer neuron number is 1,
Hidden layer neuron number is determined using trial and error procedure;Set frequency of training, every training it is complete it is primary after record the instruction of BP neural network
Practice error and test error, selects frequency of training when test error minimum as network training number;
It after determining network training number, is tested, obtains training result and test result, by correct probability 90% or more
To be tied up diameter of the data as reinforcing bar;
Step 7, using the diameter data of another reinforcing bar in radar-probing system obtaining step two, then extremely according to step 3
The identical method of step 6 is handled, and the diameter to be tied up for being directed to another reinforcing bar is obtained;Controller is by two steel
The diameter to be tied up of muscle feeds back to microprocessor, output torque of the microprocessor according to diameter adjustment motor to be tied up, power output
Square ties up reinforcing bar after adjusting.
2. binding reinforcing bars method neural network based as described in claim 1, which is characterized in that examination described in step 6
The formula that method of gathering uses are as follows:
In above formula, m is hidden layer neuron number, and n is input layer number, and l is output layer neuron number, and a is less than 10
Positive integer.
3. binding reinforcing bars method neural network based as described in claim 1, which is characterized in that the instruction set in step 6
Practicing number is 10~1000 times.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610428597.9A CN106096558B (en) | 2016-06-16 | 2016-06-16 | A kind of binding reinforcing bars method neural network based |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610428597.9A CN106096558B (en) | 2016-06-16 | 2016-06-16 | A kind of binding reinforcing bars method neural network based |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106096558A CN106096558A (en) | 2016-11-09 |
CN106096558B true CN106096558B (en) | 2019-08-30 |
Family
ID=57235893
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610428597.9A Active CN106096558B (en) | 2016-06-16 | 2016-06-16 | A kind of binding reinforcing bars method neural network based |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106096558B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107767256A (en) * | 2017-09-15 | 2018-03-06 | 重庆市个人信用管理有限责任公司 | Assessing credit risks method based on image expression credit data and depth belief network |
CN110956230A (en) * | 2018-09-27 | 2020-04-03 | 物流及供应链多元技术研发中心有限公司 | Reinforcing steel bar contact identification using artificial intelligent vision |
CN110097087B (en) * | 2019-04-04 | 2021-06-11 | 浙江科技学院 | Automatic reinforcing steel bar binding position identification method |
CN110158967A (en) * | 2019-06-13 | 2019-08-23 | 浙江科技学院 | A kind of binding reinforcing bars method and system |
CN110321824B (en) * | 2019-06-24 | 2021-10-19 | 北京理工华汇智能科技有限公司 | Binding determination method and device based on neural network |
CN110334751B (en) * | 2019-06-24 | 2021-08-13 | 北京理工华汇智能科技有限公司 | Image processing method and device for binding nodes and terminal |
CN110328662B (en) * | 2019-06-24 | 2021-06-22 | 北京理工华汇智能科技有限公司 | Path planning method and device based on image recognition |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN2654565Y (en) * | 2003-09-09 | 2004-11-10 | 上海亚华印刷机械有限公司 | Thermoprinting machine waste aluminium foil rewinding device |
CN1561271A (en) * | 2001-09-28 | 2005-01-05 | 美克司株式会社 | Reinforcement binding machine, reel, and method of detecting rotation of reel |
CN1963824A (en) * | 2006-11-17 | 2007-05-16 | 东华大学 | Method for forecasting diameter of fibre of unwoven cloth based on extrusion method of polymer of NN |
CN101623876A (en) * | 2009-05-11 | 2010-01-13 | 东莞市飞新达精密机械科技有限公司 | Tension control method |
CN102880908A (en) * | 2012-09-11 | 2013-01-16 | 天津大学 | Method for calculating remanufactured part environmental loss based on back propagation (BP) neural network |
CN103967283A (en) * | 2014-05-27 | 2014-08-06 | 重庆大学 | Steel bar strapping machine |
CN104715295A (en) * | 2015-04-03 | 2015-06-17 | 江苏物联网研究发展中心 | Chinese chemical fertilizer price index prediction method based on BP neural network |
-
2016
- 2016-06-16 CN CN201610428597.9A patent/CN106096558B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1561271A (en) * | 2001-09-28 | 2005-01-05 | 美克司株式会社 | Reinforcement binding machine, reel, and method of detecting rotation of reel |
CN2654565Y (en) * | 2003-09-09 | 2004-11-10 | 上海亚华印刷机械有限公司 | Thermoprinting machine waste aluminium foil rewinding device |
CN1963824A (en) * | 2006-11-17 | 2007-05-16 | 东华大学 | Method for forecasting diameter of fibre of unwoven cloth based on extrusion method of polymer of NN |
CN101623876A (en) * | 2009-05-11 | 2010-01-13 | 东莞市飞新达精密机械科技有限公司 | Tension control method |
CN102880908A (en) * | 2012-09-11 | 2013-01-16 | 天津大学 | Method for calculating remanufactured part environmental loss based on back propagation (BP) neural network |
CN103967283A (en) * | 2014-05-27 | 2014-08-06 | 重庆大学 | Steel bar strapping machine |
CN104715295A (en) * | 2015-04-03 | 2015-06-17 | 江苏物联网研究发展中心 | Chinese chemical fertilizer price index prediction method based on BP neural network |
Non-Patent Citations (3)
Title |
---|
"U-TIER"全自动钢筋捆扎机;王进 等;《建筑机械》;20041231;第2卷;第1节第3段,第2节,第3节,图2、图3 |
混凝土中钢筋直径雷达检测的神经网络方法;叶爱文 等;《建筑科学与工程学报》;20081231;第25卷(第4期);第1节第2段,第2节,第3节,第5.1节 |
钢筋混凝土雷达信号的BP人工神经网络识别;谢慧才 等;《华中科技大学学报(城市科学版)》;20070331;第24卷(第1期);1-4、12 |
Also Published As
Publication number | Publication date |
---|---|
CN106096558A (en) | 2016-11-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106096558B (en) | A kind of binding reinforcing bars method neural network based | |
US20150114519A1 (en) | Harvesting and grafting of trees | |
CN109815846A (en) | Image processing method, device, storage medium and electronic device | |
CN203984911U (en) | The residual film of full film double-furrow steel rider belt-type reclaims binder | |
CN101706493A (en) | On-line water quality monitoring method of aquaculture ponds and system | |
CN107389130A (en) | Intelligent irrigation inspection car and irrigation method | |
CN115546195A (en) | Litchi late autumn tip detection method based on improved YOLOV5 algorithm and processing terminal | |
CN104715342A (en) | Electrical equipment fault handling and aid decision making method based on case-based reasoning | |
CN110158967A (en) | A kind of binding reinforcing bars method and system | |
CN108428061A (en) | Intelligent Tenders Decision system based on DEA-GA-BP and tender Evaluation Method | |
CN105744225A (en) | Method and system for analyzing crop growth remotely | |
CN204994289U (en) | Huge fungus grass integration formula binding apparatus | |
CN116110042A (en) | Tomato detection method based on CBAM attention mechanism of YOLOv7 | |
CN202184816U (en) | Kitten brain stereotaxic instrument ear rod fixing device | |
CN110060294A (en) | A kind of yield assessment method of fruit tree crop | |
CN107679615A (en) | Transformer Method of Sample Selection based on improved adaptive GA-IAGA | |
CN213180133U (en) | Novel field crop monitoring management unmanned vehicle | |
CN114722915B (en) | Fault diagnosis method and system based on ADASYN algorithm and random forest algorithm | |
CN203881568U (en) | Simple efficient soil collection device | |
DE102018251708A1 (en) | Process for the destruction of at least one plant | |
CN204929643U (en) | A kind of paddy rice harvesting binder | |
CN204624849U (en) | A kind of dress table connects electricity reason line block | |
DE102014103493A1 (en) | Toggle frequency mode for magnetic resonant multiple power transmission using alternating frequencies | |
CN106651184A (en) | Vineyard chemical fertilizer utilization efficiency measurement and calculation method and device | |
CN207488510U (en) | A kind of surface geophone placement tool |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |