CN109583566A - Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network - Google Patents
Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network Download PDFInfo
- Publication number
- CN109583566A CN109583566A CN201811380025.3A CN201811380025A CN109583566A CN 109583566 A CN109583566 A CN 109583566A CN 201811380025 A CN201811380025 A CN 201811380025A CN 109583566 A CN109583566 A CN 109583566A
- Authority
- CN
- China
- Prior art keywords
- neural network
- ammonia concentration
- layer
- pig house
- elman neural
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- QGZKDVFQNNGYKY-UHFFFAOYSA-N Ammonia Chemical compound N QGZKDVFQNNGYKY-UHFFFAOYSA-N 0.000 title claims abstract description 117
- 229910021529 ammonia Inorganic materials 0.000 title claims abstract description 58
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 30
- 238000000354 decomposition reaction Methods 0.000 title claims abstract description 30
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000007613 environmental effect Effects 0.000 claims abstract description 16
- 238000005286 illumination Methods 0.000 claims abstract description 5
- 230000000694 effects Effects 0.000 claims description 8
- 238000009423 ventilation Methods 0.000 claims description 8
- 230000005540 biological transmission Effects 0.000 claims description 6
- 210000002569 neuron Anatomy 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000008859 change Effects 0.000 claims description 3
- 210000004205 output neuron Anatomy 0.000 claims description 2
- 229920006395 saturated elastomer Polymers 0.000 claims description 2
- 230000008054 signal transmission Effects 0.000 claims 1
- 238000012544 monitoring process Methods 0.000 abstract description 7
- 238000005516 engineering process Methods 0.000 abstract description 4
- 238000012549 training Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 4
- 239000007789 gas Substances 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 101001095088 Homo sapiens Melanoma antigen preferentially expressed in tumors Proteins 0.000 description 2
- 102100037020 Melanoma antigen preferentially expressed in tumors Human genes 0.000 description 2
- 229910002092 carbon dioxide Inorganic materials 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 244000144977 poultry Species 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- RWSOTUBLDIXVET-UHFFFAOYSA-N Dihydrogen sulfide Chemical compound S RWSOTUBLDIXVET-UHFFFAOYSA-N 0.000 description 1
- 206010044565 Tremor Diseases 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000008645 cold stress Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000005034 decoration Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 229910000037 hydrogen sulfide Inorganic materials 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000009022 nonlinear effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000505 pernicious effect Effects 0.000 description 1
- 230000035935 pregnancy Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
- G01N33/0036—General constructional details of gas analysers, e.g. portable test equipment concerning the detector specially adapted to detect a particular component
- G01N33/0054—Ammonia
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Chemical & Material Sciences (AREA)
- Strategic Management (AREA)
- Artificial Intelligence (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Marketing (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Primary Health Care (AREA)
- Combustion & Propulsion (AREA)
- Marine Sciences & Fisheries (AREA)
- Mining & Mineral Resources (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Medicinal Chemistry (AREA)
- Food Science & Technology (AREA)
- Immunology (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Pathology (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
Abstract
The invention discloses ammonia concentration prediction techniques in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network, belong to ambient intelligence control technology field.Include: step 1: choosing the influence factor of the historical temperature, humidity, gas concentration lwevel and the intensity of illumination that acquire in pig house as ammonia concentration, empirical mode decomposition is carried out to the time series of ammonia concentration and 4 kinds of influence factors, respectively obtains intrinsic mode function and trend term；Step 2: Elman neural network prediction model is established respectively to the wave component under same time scale after decomposition；Step 3: being reconstructed to obtain ammonia concentration predicted value to each component prediction result, establishes the ammonia concentration prediction technique based on empirical mode decomposition Yu Elman neural network.The accuracy of pig house ammonia concentration prediction can be improved in prediction result of the invention compared with using the independent prediction technique of Elman neural network, and the performance continuously predicted can provide actual parameter for environmental monitoring in cold ground pig house and the regulation of ammonia concentration.
Description
Technical field:
The invention belongs to ambient intelligence control technology fields, more particularly to one kind is based on empirical mode decomposition and Elman mind
Ammonia concentration prediction technique in cold ground pig house through network.
Background technique:
Currently, pig raising industry gradually develops towards scale, industrialization and intensive direction.Ground pig house winter tremble with fear to protect
Temperature is typically in air-tight state, frequently results in that harmful gas concentration in pig house is excessively high, endangers the health of live pig and poultry feeders,
In, content highest, the harmfulness of ammonia are maximum.Pig house is a real-time change, complicated, non-linear and interaction and shape
At system of microclimates, give up in each environmental parameter ammonia concentration can all be impacted.Currently, the research of China's majority is ammonia
Influence of the height of concentration to pig growth state, and the research of aspect is influenced very to ammonia concentration based on environmental factor in pig house
Few, prediction model mainly uses mechanism model and empirical model, both prediction techniques show prediction result it is very unstable and
Precision of prediction is not high.
" pig house ammonia concentration forecasting system [D] of the Luo Wenbo based on Android platform, Heilungkiang Aug. 1st are land-reclaimable big for document
It learns, master thesis, predicts that ammonia concentration, the training time is too long using Elman neural network in " in 2018,
It can be only achieved target error by 7047 steps, and cannot achieve the effect that continuously to predict.Therefore present invention use experience mould first
The method that state is decomposed carries out tranquilization processing to data, reuses Elman neural network and carries out continuous precisely prediction.Meanwhile it is deep
Enter research multivariate predictive model can be determined for pig breeding industry ambient intelligence control strategy provide a reasonable approach with
Realize tranquilization ventilation.
Summary of the invention:
Goal of the invention: it for the accuracy and reliability for improving ammonia concentration prediction result, divulges information so that real-time and precise controls
System reduces influence of the ammonia to the live pig general level of the health and production capacity in giving up, the present invention is based on experience state decompose (EMD) with
Elman neural net model establishing proposes a kind of ammonia concentration combination forecasting method.
The purpose of the present invention can be achieved through the following technical solutions:
The present invention proposes a kind of based on empirical mode decomposition and ammonia concentration prediction in the cold ground pig house of Elman neural network
Method, characterized by the following steps:
Step S1: the historical temperature, humidity, gas concentration lwevel and the intensity of illumination that acquire in pig house are chosen as ammonia
The influence factor of concentration carries out empirical mode decomposition to the time series of ammonia concentration and 4 kinds of influence factors, respectively obtains intrinsic
Mode function and trend term；
Step S2: Elman neural network prediction mould is established respectively to the wave component under same time scale each after decomposition
Type；
Step S3: each component prediction result is reconstructed to obtain ammonia concentration predicted value, is established based on empirical modal point
The ammonia concentration prediction model of solution and Elman neural network.
The pig house that the step S1 is selected is closed force ventilation mode.
The step S1 specifically:
If fmIt (n) is the time series data of environmental parameter and ammonia concentration, wherein n ∈ N*, m value range are 1~5,
5 variables are expressed as, following decomposition is carried out to each environmental parameter:
(1) it initializes: another r0=fm(n), k=1,
(2) k intrinsic mode functions, d are calculatedk,
A) it initializes: h0=rk-1, j=1
B) all local extremum h are definedj-1
C) and all Local Extremums are defined, calculate envelope mean value:
Wherein Emax,j-1(t) coenvelope line, E are determined for Local modulus maximamin,j-1(t) it is determined for local minizing point
Envelope, mean value Emean,j-1It (t) is average envelope line,
D) it calculates
hj[n]=hj-1[n]-Emean,j-1(n) (2)
If e) meeting standard obtains dk=hj, then stop, otherwise j=j+1；
(3) it calculates:
rk(n)=rk-1(n)-dk(n), (3)
(4) if rkBe not it is dull, return to step (2), otherwise decompose complete,
After carrying out resolution process to each variable, available all intrinsic mode functions and trend term.
The step S2 specifically:
Elman neural network includes four layers: input layer, hidden layer accept layer and output layer, and the unit of input layer plays signal
Linear or non-linear letter can be used in transmitting effect, the linear weighting effect of output layer unit, the transmission function of implicit layer unit
Number accepts layer and is used to remember the output valve of implicit layer unit preceding layer previous moment and return to the input of network,
The non-linear state space expression of Elman network are as follows:
Y (k)=g (w3x(k)) (4)
X (k)=f (w1xc(k)+w2(u(k-1))) (5)
xc(k)=x (k-1) (6)
In formula, y is that m ties up output node vector；X is that n ties up middle layer node unit vector；U is r dimensional input vector；xcFor
N ties up feedback state vector；w3For middle layer to output layer connection weight；w2For input layer to middle layer connection weight；w1To accept
Layer arrives the connection weight of middle layer；G (*) is the transmission function of output neuron, is the linear combination of middle layer output；f(*)
For the transmission function of middle layer neuron, frequently with S function.
The detailed process of the step S3 are as follows:
To important and trend term predicted value sum, obtain final forecast sample result:
It is reconstructed using formula (7) by Elman neural network, to avoid neuron from being saturated, in input layer to input
Data are normalized, and each numerical value is scaled in [0,1] section, carry out in output layer to obtained prediction data anti-
Normalization.
The present invention is based on empirical mode decomposition in the cold ground pig house of Elman neural network ammonia concentration prediction technique it is excellent
Point is:
1. each component is decomposed and reconstructed by using EMD decomposition method, original time can be clearly given expression to
Fluctuation situation of the sequence on different time scales solves the forecasting problem of Multiple Time Scales sequence, in conjunction with Elman neural network
It is not high that the double-deck feedback arrangement can effectively solve ammonia concentration precision of prediction in the case where cold district is because of external environment interference in pig house
Problem simultaneously reduces the training time.
2. can reach the effect continuously predicted.To realize under conditions of dynamic temp compensation, to real in standardization pig house
Existing tranquilization ventilation and environmental parameter intelligentized control method etc. provide effective prediction model.
Detailed description of the invention:
Fig. 1 is first group of trained image parameter ammonia concentration, the temperature, humidity, CO chosen2Concentration, intensity of illumination when
Between sequence chart.
Fig. 2 is that pig house in 8 to 31 January in 2017 gives up outer hygrogram.
Fig. 3 is EMD_Elman combined prediction flow chart.
Fig. 4 is each modal components wave pattern after ammonia concentration and environmental parameter EMD decomposition.
Fig. 5 is (a) EMD_Elman predicted value figure compared with actual value；
(b) Elman predicted value figure compared with actual value.
Specific embodiment:
Below in conjunction with attached drawing, case study on implementation of the invention is described in detail:
The nonpregnant pregnant sow house of monitoring is closed force ventilation mode, and pig house amounts to having a size of 75m × 25m × 3m
304 nonpregnant pregnant sows.House inside wall is uniformly equipped with 6 negative-pressure air fans, and outer dimension is 800mm × 800mm × 300mm,
Blower ventilation amount is 1100m3/ h, power 0.4KW.Built-in having heaters is given up, outer cold air is given up and leads to after heater preheats
The air supply duct crossed at the top of house enters in pig house.By the on-line monitoring and intelligence control system of the exploitations such as Zhang Y, to gestation
Sow is monitored, and system includes: poultry house environmental information multi-source awareness apparatus, for monitor give up in temperature and humidity, carbon dioxide,
The information such as hydrogen sulfide, ammonia, and monitoring information is transmitted to by main controlled node by wireless sensor network；Main controlled node timing will monitoring
Data are sent to remote server.
Environment monitoring node is mounted on to the center of pig house, the height about 2.0m apart from ground in the present invention.Pig house
Interior ammonia concentration and influence factor are acquired primary by multi-source awareness apparatus on-line monitoring system per half an hour in giving up.Acquisition 23 days
(2017/1/08-2017/01/31) totally 1104 groups of data, Fig. 1 are the time series chart for the first group of trained object chosen.
As can be seen from Figure 1 6:00 or so, CO in the morning2In trend is decreased obviously, variation is compared for concentration, ammonia concentration
Greatly.Because administrative staff can carry out cleaning up excrement to pig house at this time, and all ventilation equipment openings are aerated, can divulge information always
To noon 14:00 or so.Each parameter concentration is gradually increasing after closing blower, is especially both greater than 25mg/ in night ammonia concentration
m3。
The temperature in 8 to 31 January in 2017 is as shown in Figure 2.Reachable -29 DEG C of minimum temperature, maximum temperature also at -9 DEG C,
The unlatching blower of duration in such one day, will certainly cause certain cold stress to pig.Therefore, it is necessary to ammonia concentration into
Row prediction in real time, in order to effectively open the pernicious gases such as blower removal ammonia.
Fig. 3 is the combined prediction flow chart of EMD_Elman neural network.In this specific embodiment, first to ammonia
Gas concentration and 4 kinds of environmental parameter time serieses are decomposed, and 5 intrinsic mode function IMF and 1 trend term R are respectively obtainede,
Then Elman neural network model is established to the time series under same time scale after decomposition respectively, and to a prediction result
It is reconstructed to obtain ammonia concentration predicted value, establishes the ammonia concentration based on empirical mode decomposition and Elman neural network and predict
Model.
According to formula (1)~(3), EMD decomposition is carried out to first 3 days 144 groups of data of ammonia concentration and its influence factor first.
As shown in figure 4, having respectively obtained 5 intrinsic mode functions and 1 trend term Re.According to formula (4)~(6) again respectively to each group IMFs
Carry out Elman neural network prediction.Wherein in each IMF, preceding 134 samples of 4 influence factors are as fitting input letter
Matrix number, data matrix of rear 10 data as test；Preceding 134 sample datas of ammonia concentration are as fitting output function
Matrix, objective matrix of rear 10 data as test.Finally, each component is reconstructed according to formula (7), obtain final
Ammonia concentration predicted value.
In the network training parameter of setting, study precision is 10-4, iterative steps are set as 2000, when iterating to 1990
It is 0.020 that best training error is obtained when step.
For the performance of verification algorithm, using 134 data sets as training sequence, using algorithm to remaining 970 data
Collection carries out slip heavy loads, and is compared with the data of actual acquisition.Elman model and EMD_Elman Combined model forecast knot
Fruit and initial data comparison are as shown in Figure 5.As can be seen that method proposed by the present invention has preferable estimated performance and lesser
Predict error.
The present invention selects following error assessment index: mean absolute error (MAE), average absolute percentage error (MAPE),
Standard error (RMSE), it is for statistical analysis to the prediction accuracy of two methods.Formula is as follows:
X (t) is the initial data of ammonia concentration,To predict ammonia concentration data, N is the sample size of predicted value.
Table 1 compares the prediction error and coefficient R of two kinds of prediction techniques2, as shown in Table 1, rear mold is decomposed by EMD
Type prediction result is higher than without the model prediction result accuracy of decomposition, related coefficient 0.9856, reacts actual prediction
The MAE of error size is 0.7088ppm, RMSE 0.9096ppm, shows that the model dispersion degree is low, and MAPE is 0.41 table
Its bright prediction is more accurate.To sum up, show that the model has preferable fitting and predictive ability, can satisfy ammonia concentration prediction
Required precision.Meanwhile the combined prediction result coefficient of determination R of EMD_Elman neural network2It is predicted higher than Elman, shows group
Output valve and the target value deviation for closing prediction model are smaller, are highly effective prediction techniques.
Since data collected are under the conditions of extreme cold ground, the outer temperature of house is minimum to reach -29 DEG C, at noon time-division wind
Machine is opened, and is given up interior environmental parameter influential on ammonia concentration and is had greatly changed, CO2It is even dropped to ammonia concentration
0ppm will cause certain interference to network model prediction result.It is analyzed by the foundation of model and data, empirical mode decomposition
Method can have according to time series temporal scale feature adaptively by Time Series to different time scales
Effect solves the influence under the conditions of the environmental parameter and external environmental interference of multivariable to ammonia concentration.
It is pre- based on empirical mode decomposition and ammonia concentration in the cold ground pig house of Elman neural network that the invention proposes a kind of
Survey method chooses environmental parameter CO in pig house2, temperature, humidity, intensity of illumination as prediction ammonia concentration main influence because
Element.Each component is decomposed and reconstructed by using EMD decomposition method, can clearly give expression to original time series not
With the fluctuation situation in time scale, the forecasting problem of Multiple Time Scales sequence is solved, is fed back in conjunction with Elman neural network bilayer
Structure can be improved the precision of prediction of time series and reduce the training time.Method proposed by the present invention can reach continuous prediction
Effect, be to realize under conditions of dynamic temp compensation, to realizing tranquilization ventilation and environmental parameter intelligence in standardization pig house
Energyization control etc. provides effective prediction model.
The above described is only a preferred embodiment of the present invention, be not intended to limit the present invention in any form, though
So the present invention has been disclosed as a preferred embodiment, and however, it is not intended to limit the invention, any technology people for being familiar with this profession
Member, in the range of not departing from technical solution of the present invention, when the technology contents using the disclosure above are modified or are repaired
Decorations are the equivalent embodiment of equivalent variations, but without departing from the technical solutions of the present invention, according to the technical essence of the invention
To any simple modification, equivalent change and modification made by above example, all of which are still within the scope of the technical scheme of the invention.
Claims (5)
1. ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network, feature
It is: includes the following steps:
Step S1: the historical temperature, humidity, gas concentration lwevel and the intensity of illumination that acquire in pig house are chosen as ammonia concentration
Influence factor, empirical mode decomposition is carried out to the time series of ammonia concentration and 4 kinds of influence factors, respectively obtains intrinsic mode
Function and trend term；
Step S2: Elman neural network prediction model is established respectively to the wave component under same time scale each after decomposition；
Step S3: being reconstructed to obtain ammonia concentration predicted value to each component prediction result, establish based on empirical mode decomposition with
The ammonia concentration prediction model of Elman neural network.
2. ammonia in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network according to claim 1
Concentration prediction method, it is characterised in that: the pig house that the step S1 is selected is closed force ventilation mode.
3. ammonia in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network according to claim 1
Concentration prediction method, it is characterised in that: the step S1 specifically:
If fmIt (n) is the time series data of environmental parameter and ammonia concentration, wherein n ∈ N*, m value range are 1~5, are expressed as
5 variables, following decomposition is carried out to each environmental parameter:
(1) it initializes: another r0=fm(n), k=1,
(2) k intrinsic mode functions, d are calculatedk,
A) it initializes: h0=rk-1, j=1
B) all local extremum h are definedj-1
C) and all Local Extremums are defined, calculate envelope mean value:
Wherein Emax,j-1(t) coenvelope line, E are determined for Local modulus maximamin,j-1(t) coenvelope is determined for local minizing point
Line, mean value Emean,j-1It (t) is average envelope line,
D) it calculates
hj[n]=hj-1[n]-Emean,j-1(n) (2)
If e) meeting standard obtains dk=hj, then stop, otherwise j=j+1；
(3) it calculates: rk(n)=rk-1(n)-dk(n), (3)
(4) if rkBe not it is dull, return to step (2), otherwise decompose complete,
After carrying out resolution process to each variable, available all intrinsic mode functions and trend term.
4. ammonia in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network according to claim 1
Concentration prediction method, it is characterised in that: the step S2 specifically:
Elman neural network includes four layers: input layer, hidden layer accept layer and output layer, and the unit of input layer plays signal transmission
Effect, the linear weighting effect of output layer unit, the transmission function of implicit layer unit can be used linear or nonlinear function, hold
It connects layer and is used to remember the output valve of implicit layer unit preceding layer previous moment and return to the input of network,
The non-linear state space expression of Elman network are as follows:
Y (k)=g (w3x(k)) (4)
X (k)=f (w1xc(k)+w2(u(k-1))) (5)
xc(k)=x (k-1) (6)
In formula, y is that m ties up output node vector；X is that n ties up middle layer node unit vector；U is r dimensional input vector；xcIt is anti-for n dimension
Present state vector；w3For middle layer to output layer connection weight；w2For input layer to middle layer connection weight；w1It is arrived to accept layer
The connection weight of middle layer；G (*) is the transmission function of output neuron, is the linear combination of middle layer output；F (*) is centre
The transmission function of layer neuron, frequently with S function.
5. ammonia in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network according to claim 1
Concentration prediction method, it is characterised in that: the detailed process of the step S3 are as follows:
To important and trend term predicted value sum, obtain final forecast sample result:
It is reconstructed using formula (7) by Elman neural network, to avoid neuron from being saturated, in input layer to input data
It is normalized, each numerical value is scaled in [0,1] section, anti-normalizing is carried out to obtained prediction data in output layer
Change.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811380025.3A CN109583566A (en) | 2018-11-20 | 2018-11-20 | Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811380025.3A CN109583566A (en) | 2018-11-20 | 2018-11-20 | Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109583566A true CN109583566A (en) | 2019-04-05 |
Family
ID=65923301
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811380025.3A Pending CN109583566A (en) | 2018-11-20 | 2018-11-20 | Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109583566A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555570A (en) * | 2019-09-16 | 2019-12-10 | 武汉理工大学 | Intelligent prediction method and device for gas concentration of mine limited space disaster |
| CN110763830A (en) * | 2019-12-04 | 2020-02-07 | 济南大学 | Method for predicting content of free calcium oxide in cement clinker |
| CN111144286A (en) * | 2019-12-25 | 2020-05-12 | 北京工业大学 | Urban PM2.5 concentration prediction method fusing EMD and LSTM |
| CN112149870A (en) * | 2020-08-21 | 2020-12-29 | 江苏大学 | Pig house ammonia concentration combined prediction method based on ISODATA clustering and Elman neural network |
| CN114002303A (en) * | 2021-12-31 | 2022-02-01 | 中国农业科学院农业资源与农业区划研究所 | Calibration method for gas sensing in cold-chain logistics and multi-source sensing device |
| CN115656437A (en) * | 2022-10-26 | 2023-01-31 | 东北大学 | Soft measurement method for concentration of gas easy to prepare poison based on signal multi-feature data |
| CN116029435A (en) * | 2023-01-10 | 2023-04-28 | 盐城工学院 | Environmental comfort early warning system is bred to live pig facility |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030033094A1 (en) * | 2001-02-14 | 2003-02-13 | Huang Norden E. | Empirical mode decomposition for analyzing acoustical signals |
| WO2008132066A1 (en) * | 2007-04-27 | 2008-11-06 | Siemens Aktiengesellschaft | A method for computer-assisted learning of one or more neural networks |
| CN104537444A (en) * | 2015-01-13 | 2015-04-22 | 安徽理工大学 | Gas outburst predicting method based on EMD and ELM |
| CN104899656A (en) * | 2015-06-05 | 2015-09-09 | 三峡大学 | Wind power combined predication method based on ensemble average empirical mode decomposition and improved Elman neural network |
| CN105159216A (en) * | 2015-08-31 | 2015-12-16 | 淮阴工学院 | Hen house environment ammonia gas concentration intelligent monitoring system |
| CN107016453A (en) * | 2016-12-08 | 2017-08-04 | 中国农业大学 | A kind of aquaculture dissolved oxygen prediction method and device |
-
2018
- 2018-11-20 CN CN201811380025.3A patent/CN109583566A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030033094A1 (en) * | 2001-02-14 | 2003-02-13 | Huang Norden E. | Empirical mode decomposition for analyzing acoustical signals |
| WO2008132066A1 (en) * | 2007-04-27 | 2008-11-06 | Siemens Aktiengesellschaft | A method for computer-assisted learning of one or more neural networks |
| CN104537444A (en) * | 2015-01-13 | 2015-04-22 | 安徽理工大学 | Gas outburst predicting method based on EMD and ELM |
| CN104899656A (en) * | 2015-06-05 | 2015-09-09 | 三峡大学 | Wind power combined predication method based on ensemble average empirical mode decomposition and improved Elman neural network |
| CN105159216A (en) * | 2015-08-31 | 2015-12-16 | 淮阴工学院 | Hen house environment ammonia gas concentration intelligent monitoring system |
| CN107016453A (en) * | 2016-12-08 | 2017-08-04 | 中国农业大学 | A kind of aquaculture dissolved oxygen prediction method and device |
Non-Patent Citations (4)
| Title |
|---|
| WANG XIANG 等: "Establishment of NH3-N Prediction Model in Aquaculture Water Based on ELMAN Neural Network", 《METEOROLOGICAL AND ENVIRONMENTAL RESEARCH 》 * |
| 李润求 等: "采煤工作面瓦斯涌出预测的EMD-Elman方法及应用", 《中国安全科学学报》 * |
| 罗文博: "基于Android平台的猪舍氨气浓度预测系统", 《中国优秀硕士学位论文全文数据库农业科技辑》 * |
| 谢秋菊 等: "基于神经网络猪舍氨气浓度预测方法研究", 《东北农业大学学报》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110555570A (en) * | 2019-09-16 | 2019-12-10 | 武汉理工大学 | Intelligent prediction method and device for gas concentration of mine limited space disaster |
| CN110763830A (en) * | 2019-12-04 | 2020-02-07 | 济南大学 | Method for predicting content of free calcium oxide in cement clinker |
| CN110763830B (en) * | 2019-12-04 | 2022-04-05 | 济南大学 | Method for predicting content of free calcium oxide in cement clinker |
| CN111144286A (en) * | 2019-12-25 | 2020-05-12 | 北京工业大学 | Urban PM2.5 concentration prediction method fusing EMD and LSTM |
| CN112149870A (en) * | 2020-08-21 | 2020-12-29 | 江苏大学 | Pig house ammonia concentration combined prediction method based on ISODATA clustering and Elman neural network |
| CN112149870B (en) * | 2020-08-21 | 2024-03-22 | 江苏大学 | Pig house ammonia concentration combination prediction method based on ISODATA clustering and Elman neural network |
| CN114002303A (en) * | 2021-12-31 | 2022-02-01 | 中国农业科学院农业资源与农业区划研究所 | Calibration method for gas sensing in cold-chain logistics and multi-source sensing device |
| CN114002303B (en) * | 2021-12-31 | 2022-04-05 | 中国农业科学院农业资源与农业区划研究所 | Calibration method for gas sensing in cold-chain logistics and multi-source sensing device |
| CN115656437A (en) * | 2022-10-26 | 2023-01-31 | 东北大学 | Soft measurement method for concentration of gas easy to prepare poison based on signal multi-feature data |
| CN115656437B (en) * | 2022-10-26 | 2024-06-11 | 东北大学 | Soft measurement method for concentration of easily-poisoned gas based on signal multi-feature data |
| CN116029435A (en) * | 2023-01-10 | 2023-04-28 | 盐城工学院 | Environmental comfort early warning system is bred to live pig facility |
| CN116029435B (en) * | 2023-01-10 | 2023-07-07 | 盐城工学院 | Environmental comfort early warning system is bred to live pig facility |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109583566A (en) | Ammonia concentration prediction technique in a kind of cold ground pig house based on empirical mode decomposition and Elman neural network | |
| US20240000046A1 (en) | Predictive control system and regulatory method for temperature of livestock house | |
| CN110414788B (en) | Electric energy quality prediction method based on similar days and improved LSTM | |
| CN106292802B (en) | A kind of intelligent Prediction Control System and method for fish and vegetable symbiotic system | |
| CN110070224A (en) | A kind of Air Quality Forecast method based on multi-step recursive prediction | |
| CN112527037B (en) | Greenhouse environment regulation and control method and system with environment factor prediction function | |
| CN108710947A (en) | A kind of smart home machine learning system design method based on LSTM | |
| CN109934422A (en) | Neural network wind speed prediction method based on time series data analysis | |
| CN114297907A (en) | Greenhouse environment spatial distribution prediction method and device | |
| CN116029435B (en) | Environmental comfort early warning system is bred to live pig facility | |
| CN113359486A (en) | Intelligent window system based on neural network algorithm regulation and control and regulation method | |
| CN116649160B (en) | Edible fungus strain production monitoring system and monitoring method | |
| CN109002928A (en) | A kind of electric load peak value prediction technique and device based on Bayesian network model | |
| CN110045771A (en) | A kind of fishpond water quality intelligent monitor system | |
| CN114119273A (en) | Park comprehensive energy system non-invasive load decomposition method and system | |
| CN117391482B (en) | Greenhouse temperature intelligent early warning method and system based on big data monitoring | |
| Sen | Time series prediction based on improved deep learning | |
| CN101285816A (en) | Copper matte air refining procedure parameter soft sensing instrument and its soft sensing method | |
| Ramos-Fernández et al. | Fuzzy modeling vapor pressure deficit to monitoring microclimate in greenhouses | |
| CN113962819A (en) | Method for predicting dissolved oxygen in industrial aquaculture based on extreme learning machine | |
| Sadaghat et al. | Residential building energy consumption estimation: a novel ensemble and hybrid machine learning approach | |
| CN109858576B (en) | Progressive self-feedback concentration entropy change prediction method and system for gas and storage medium | |
| CN116430930A (en) | Forest farm environment balance adjustment equipment | |
| Xie et al. | Multi-sensor data fusion based on fuzzy neural network and its application in piggery environmental control strategies | |
| Zhang et al. | An intelligent Attentional-GRU-based model for dynamic blood glucose prediction |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190405 |