CN108038571A - A kind of Combination Forecasting method and system of cultivation water - Google Patents
A kind of Combination Forecasting method and system of cultivation water Download PDFInfo
- Publication number
- CN108038571A CN108038571A CN201711297240.2A CN201711297240A CN108038571A CN 108038571 A CN108038571 A CN 108038571A CN 201711297240 A CN201711297240 A CN 201711297240A CN 108038571 A CN108038571 A CN 108038571A
- Authority
- CN
- China
- Prior art keywords
- predicted
- prediction
- processing
- individual event
- weight coefficient
- 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
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 88
- 238000013277 forecasting method Methods 0.000 title claims abstract description 11
- 238000000034 method Methods 0.000 claims abstract description 120
- 238000012545 processing Methods 0.000 claims abstract description 100
- 238000005457 optimization Methods 0.000 claims abstract description 61
- 238000001914 filtration Methods 0.000 claims abstract description 39
- 238000000926 separation method Methods 0.000 claims abstract description 15
- 238000011156 evaluation Methods 0.000 claims description 34
- 238000001514 detection method Methods 0.000 claims description 16
- 241001123248 Arma Species 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 12
- 238000000605 extraction Methods 0.000 claims description 12
- 238000000354 decomposition reaction Methods 0.000 claims description 11
- 238000012544 monitoring process Methods 0.000 claims description 8
- 239000002131 composite material Substances 0.000 claims description 3
- 238000012360 testing method Methods 0.000 claims 2
- 239000000470 constituent Substances 0.000 description 11
- 238000010586 diagram Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 2
- 210000005036 nerve Anatomy 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- 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
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Marketing (AREA)
- General Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Animal Husbandry (AREA)
- Life Sciences & Earth Sciences (AREA)
- Agronomy & Crop Science (AREA)
- Mining & Mineral Resources (AREA)
- General Health & Medical Sciences (AREA)
- Marine Sciences & Fisheries (AREA)
- Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention proposes a kind of Combination Forecasting method and system of cultivation water, and wherein this method includes：Based on the cultivation water signal acquisition water quality characteristic component monitored；Different filtering process is carried out to the different frequency section in the water quality characteristic component, with the data to be predicted in the corresponding different frequency section of generation；For the data to be predicted of different frequency separations, choose different Forecasting Methodologies and carry out individual event prediction, to obtain multiple predicted characteristics；When the precision of individual event prediction reaches preset requirement, optimization processing is combined to each predicted characteristics, to generate prediction result；Reach default requirement when the Combinatorial Optimization is handled, the Combination Forecasting model based on prediction result generation cultivation water.The precision predicted cultivation water is improved by the scheme of the embodiment of the present invention with this.
Description
Technical field
The present invention relates to water quality prediction field, the Combination Forecasting method of more particularly to a kind of cultivation water and it is
System.
Background technology
Existing water quality cultivation has increasing need for accurate water quality data；But the characteristic first included by water quality is very
It is more, and interactional relation is complicated；In addition the precision of existing water quality prediction is inadequate, can not meet needs.
The content of the invention
For in the prior art the defects of, the present invention propose a kind of Combination Forecasting method of cultivation water and be
System, for improving the precision of water quality monitoring.
Specifically, the present invention proposes embodiment in detail below：
The embodiment of the present invention proposes a kind of Combination Forecasting method of cultivation water, including：
Signal decomposition is carried out to the cultivation water signal monitored based on multiple default feature extraction algorithms, to obtain water quality
Characteristic component；
Different filtering process is carried out according to different frequency section to the water quality characteristic component, with the corresponding different frequencies of generation
The data to be predicted in rate section；
For the data to be predicted of different frequency separations, choose different Forecasting Methodologies and carry out individual event prediction, to obtain
Multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Each predicted characteristics are corresponding with
Respective weight coefficient；
When the precision of each individual event prediction reaches preset requirement, to each predicted characteristics and the weight coefficient
The Combinatorial Optimization processing of nonlinear way is carried out, to generate prediction result；
When Combinatorial Optimization processing reaches default requirement, the comprehensive prediction result and weight coefficient generation
The Combination Forecasting model of cultivation water.
In a specific embodiment, the frequency separation includes：Low frequency section, intermediate frequency section, high frequency section；
The filtering process includes being filtered the low frequency filtering processing of processing, in described for the low frequency section
Frequency section is filtered the intermediate frequency filtering processing of processing, the High frequency filter processing of processing is filtered for the high frequency section.
In a specific embodiment, the method for the individual event prediction includes：ARMA Forecasting Methodologies, more layer multi-layers nerve
Neural network forecast method, SVM prediction method；
Wherein, the ARMA Forecasting Methodologies are to carry out individual event prediction for the data to be predicted obtained after low frequency filtering processing
Forecasting Methodology；The multilayer multilayer neural network Forecasting Methodology be for obtained data to be predicted after intermediate frequency filtering processing into
Forecasting Methodology, the SVM prediction method of row individual event prediction are for the number to be predicted obtained after High frequency filter processing
According to the Forecasting Methodology for carrying out individual event prediction.
In a specific embodiment, it is to pass through error evaluation that whether the precision of the individual event prediction, which reaches preset requirement,
Method is assessed；Wherein, the error evaluation method includes：Mean square error or evaluation absolute error or evaluation are absolute
Percentage error, log error quadratic sum are assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and with reference to special
Levy the difference between corresponding weight coefficient whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that
The precision of the individual event prediction reaches preset requirement；If judging result is more than or equal to, it is determined that the unidirectional default essence
Degree is not up to preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；This method further includes：
When the precision of individual event prediction is not reaching to preset requirement, by the error evaluation side to each prediction
After the weight of the parameter of method and each data to be predicted is adjusted, the data to be predicted after adjustment are predicted again.
In a specific embodiment, each predicted characteristics are corresponding with respective weight coefficient；The Combinatorial Optimization
Processing is realized by swarm intelligence multi-objective optimization algorithm；
Judge whether the Combinatorial Optimization processing reaches default and require to be by the cultivation water signal with monitoring
Site Detection result is compared to judge；If result and Site Detection that the result compared obtains after being handled for Combinatorial Optimization
As a result difference is less than preset value, it is determined that reach default requirement, if what the result compared obtained after being handled for Combinatorial Optimization
As a result the difference with Site Detection result is greater than or equal to preset value, it is determined that not up to default requirement；Wherein different water
Different preset values is answered in confrontation, and default requirement is corresponding with the situation of water quality；
This method further includes：
When Combinatorial Optimization processing is not reaching to default require, added by forecasting effective measure Criterion Method or broad sense
Power is that evaluation assessment or IOWGA are corresponding with each predicted characteristics after respective weight coefficient is adjusted, again to adjustment
Predicted characteristics afterwards carry out the Combinatorial Optimization processing.
The embodiment of the present invention also proposed a kind of Combination Forecasting system of cultivation water, including：
Acquisition module, for carrying out signal point to the cultivation water signal monitored based on multiple default feature extraction algorithms
Solution, to obtain water quality characteristic component；
Filter module, for carrying out different filtering process according to different frequency section to the water quality characteristic component, with
The data to be predicted in the corresponding different frequency section of generation；
Single directional prediction module, for the data to be predicted for different frequency separations, choose different Forecasting Methodologies into
Row individual event is predicted, to obtain multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Respectively
The predicted characteristics are corresponding with respective weight coefficient；
Composite module, for when the precision of each individual event prediction reaches preset requirement, to each predicted characteristics with
And the weight coefficient carries out the Combinatorial Optimization processing of nonlinear way, to generate prediction result；
Generation module, for reaching default requirement, the comprehensive prediction result and institute when Combinatorial Optimization processing
State the Combination Forecasting model of weight coefficient generation cultivation water.
In a specific embodiment, the frequency separation includes：Low frequency section, intermediate frequency section, high frequency section；
The filtering process includes being filtered the low frequency filtering processing of processing, in described for the low frequency section
Frequency section is filtered the intermediate frequency filtering processing of processing, the High frequency filter processing of processing is filtered for the high frequency section.
In a specific embodiment, the method for the individual event prediction includes：ARMA Forecasting Methodologies, more layer multi-layers nerve
Neural network forecast method, SVM prediction method；
Wherein, the ARMA Forecasting Methodologies are to carry out individual event prediction for the data to be predicted obtained after low frequency filtering processing
Forecasting Methodology；The multilayer multilayer neural network Forecasting Methodology be for obtained data to be predicted after intermediate frequency filtering processing into
Forecasting Methodology, the SVM prediction method of row individual event prediction are for the number to be predicted obtained after High frequency filter processing
According to the Forecasting Methodology for carrying out individual event prediction.
In a specific embodiment,
Whether the precision of the individual event prediction reaches preset requirement by error evaluation method to be assessed；Its
In, the error evaluation method includes：Mean square error or evaluation absolute error or evaluation absolute percent error, log error
Quadratic sum is assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and with reference to special
Levy the difference between corresponding weight coefficient whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that
The precision of the individual event prediction reaches preset requirement；If judging result is more than or equal to, it is determined that the unidirectional default essence
Degree is not up to preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；The system further includes：
First adjustment module, when the precision for being predicted when the individual event is not reaching to preset requirement, passes through the error
After assessment side is adjusted the parameter of each Forecasting Methodology and the weight of each data to be predicted, again to being treated after adjustment
Prediction data is predicted.
In a specific embodiment, each predicted characteristics are corresponding with respective weight coefficient；The Combinatorial Optimization
Processing is realized by swarm intelligence multi-objective optimization algorithm；
Judge whether the Combinatorial Optimization processing reaches default and require to be by the cultivation water signal with monitoring
Site Detection result is compared to judge；If result and Site Detection that the result compared obtains after being handled for Combinatorial Optimization
As a result difference is less than preset value, it is determined that reach default requirement, if what the result compared obtained after being handled for Combinatorial Optimization
As a result the difference with Site Detection result is greater than or equal to preset value, it is determined that not up to default requirement；Wherein different water
Different preset values is answered in confrontation, and default requirement is corresponding with the situation of water quality；
The system further includes：
Second adjustment module is effective by predicting for when Combinatorial Optimization processing is not reaching to default require
Degree Criterion Method or generalized weighted are that evaluation assessment or IOWGA are corresponding with each predicted characteristics respective weight coefficient progress
After adjustment, the Combinatorial Optimization processing is carried out to the predicted characteristics after adjustment again.
With this, the present invention proposes a kind of Combination Forecasting method and system of cultivation water, wherein this method bag
Include：Signal decomposition is carried out to the cultivation water signal monitored based on multiple default feature extraction algorithms, to obtain water quality characteristic
Component；Different filtering process is carried out according to different frequency section to the water quality characteristic component, to generate corresponding different frequency
The data to be predicted in section；For the data to be predicted of different frequency separations, it is pre- to choose different Forecasting Methodology progress individual events
Survey, to obtain multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Each prediction
Feature is corresponding with respective weight coefficient；It is special to each prediction when the precision of each individual event prediction reaches preset requirement
Sign and the weight coefficient carry out the Combinatorial Optimization processing of nonlinear way, to generate prediction result；When the Combinatorial Optimization
Processing reaches default requirement, and the nonlinear combination that the comprehensive prediction result and the weight coefficient generate cultivation water is pre-
Survey model.The precision predicted cultivation water is improved by the scheme of the embodiment of the present invention with this.
Brief description of the drawings
In order to illustrate the technical solution of the embodiments of the present invention more clearly, below will be to needed in the embodiment attached
Figure is briefly described, it will be appreciated that the following drawings illustrate only certain embodiments of the present invention, therefore be not construed as pair
The restriction of scope, for those of ordinary skill in the art, without creative efforts, can also be according to this
A little attached drawings obtain other relevant attached drawings.
Fig. 1 is a kind of flow diagram of the Combination Forecasting method for cultivation water that the embodiment of the present invention proposes；
Fig. 2 is a kind of cultivation to be predicted in a manner of nonlinear combination under the concrete application scene that the embodiment of the present invention proposes
The schematic diagram of water quality；
Fig. 3 is a kind of structure diagram of the Combination Forecasting system for cultivation water that the embodiment of the present invention proposes.
Embodiment
Hereinafter, the various embodiments of the disclosure will be described more fully.The disclosure can have various embodiments, and
It can adjust and change wherein.It should be understood, however, that：There is no the various embodiments of the disclosure are limited to spy disclosed herein
Determine the intention of embodiment, but the disclosure should be interpreted as covering in the spirit and scope for the various embodiments for falling into the disclosure
All adjustment, equivalent and/or alternatives.
Hereinafter, disclosed in the term " comprising " that can be used in the various embodiments of the disclosure or " may include " instruction
Function, operation or the presence of element, and do not limit the increase of one or more functions, operation or element.In addition, such as exist
Used in the various embodiments of the disclosure, term " comprising ", " having " and its cognate are meant only to represent special characteristic, number
Word, step, operation, the combination of element, component or foregoing item, and be understood not to exclude first one or more other
Feature, numeral, step, operation, element, component or foregoing item combination presence or one or more features of increase, numeral,
Step, operation, element, component or foregoing item combination possibility.
In the various embodiments of the disclosure, stating "or" or " at least one in A or/and B " includes what is listed file names with
Any combinations of word or all combinations.For example, " A or B " or " at least one in A or/and B " may include A, may include for statement
B may include A and B both.
The statement (" first ", " second " etc.) used in the various embodiments of the disclosure can be modified in various implementations
Various constituent element in example, but respective sets can not be limited into element.For example, presented above be not intended to limit the suitable of the element
Sequence and/or importance.The purpose presented above for being only used for differentiating an element and other elements.For example, the first user fills
Put and indicate different user device with second user device, although the two is all user apparatus.For example, each of the disclosure is not being departed from
In the case of the scope of kind embodiment, the first element is referred to alternatively as the second element, and similarly, the second element is also referred to as first
Element.
It should be noted that：, can be by the first composition member if a constituent element ' attach ' to another constituent element by description
Part is directly connected to the second constituent element, and " connection " the 3rd can be formed between the first constituent element and the second constituent element
Element.On the contrary, when a constituent element " being directly connected to " is arrived another constituent element, it will be appreciated that be in the first constituent element
And second be not present the 3rd constituent element between constituent element.
The term " user " used in the various embodiments of the disclosure, which may indicate that, to be used the people of electronic device or uses electricity
The device (for example, artificial intelligence electronic device) of sub-device.
The term used in the various embodiments of the disclosure is only used for the purpose of description specific embodiment and not anticipates
In the various embodiments of the limitation disclosure.As used herein, singulative is intended to also include plural form, unless context is clear
Chu it is indicated otherwise.Unless otherwise defined, otherwise all terms (including technical term and scientific terminology) used herein have
There is the implication identical with the various normally understood implications of embodiment one skilled in the art of the disclosure.The term
(term such as limited in the dictionary generally used) is to be interpreted as having and situational meaning in the related technical field
Identical implication and the implication of Utopian implication or overly formal will be interpreted as having, unless in the various of the disclosure
It is clearly defined in embodiment.
Embodiment 1
The embodiment of the present invention 1 discloses a kind of Combination Forecasting method of cultivation water, as shown in Figure 1, including：
Step 101, based on multiple default feature extraction algorithms carry out signal decomposition to the cultivation water signal that monitors, with
Obtain water quality characteristic component；
Specifically, the cultivation water signal monitored can be the water quality signal that field monitoring arrives, also can scene
Verify subsequent prediction result, it is believed that the cultivation water signal is the Combination Forecasting model of training cultivation water
Sample data.
Different feature extraction algorithm is chosen specifically, carrying out feature extraction and can be directed to the difference of water quality, such as can be with
The water quality signal is decomposed by EMD algorithms and/or EEMD algorithms, to extract water quality component；EMD(Empirical
Mode Decomposition, based on empirical mode decomposition), and since EMD algorithms using cubic spline interpolation obtain signal
Moment average value, there are special boundary effect, the timeliness of water quality signal decomposition can decline, and noise reduction algorithm (such as it is small
Wave conversion algorithm) specifically there is the ability of " concentration ", the system by reducing or cancelling noise produces, then can reach by reconstruct
The optimal estimation of original signal, therefore the preposition noise processor using noise reduction algorithm as EMD methods, can subsequently improve feature
The effect that extraction and dynamic are repeatedly all analyzed.
And specifically, EEMD is the abbreviation of Ensemble Empirical Mode Decomposition, Chinese is set
Empirical mode decomposition, is the deficiency for EMD methods, it is proposed that a kind of noise auxiliary data analysis method.
Specifically, can also be other feature extraction modes, however it is not limited to which listed above is several, can be according to specific
Situation and reality needs flexibly chosen.
Step 102, carry out different filtering process according to different frequency section to the water quality characteristic component, with generation pair
Answer the data to be predicted in different frequency section；
Specifically in one embodiment, as shown in Fig. 2, the frequency separation includes：Low frequency section, intermediate frequency section, height
Frequency section；The section of specific senior middle school low frequency can rule of thumb carry out drawing for three-stage according to those skilled in the art
Point；And after demarcation interval, it is filtered processing for different sections；
The filtering process includes being filtered the low frequency filtering processing of processing, in described for the low frequency section
Frequency section is filtered the intermediate frequency filtering processing of processing, the High frequency filter processing of processing is filtered for the high frequency section.
Step 103, the data to be predicted for different frequency separations, it is pre- to choose different Forecasting Methodology progress individual events
Survey, to obtain multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Each prediction
Feature is corresponding with respective weight coefficient；
The Forecasting Methodology includes：ARMA Forecasting Methodologies, multilayer multilayer neural network Forecasting Methodology, SVM prediction
Method；
Wherein, the ARMA (Auto-Regressive and Moving Average Mode, search time sequence
Important method, is formed by autoregression model (abbreviation AR models) with " being mixed " based on moving average model (abbreviation MA models))
Forecasting Methodology is that the Forecasting Methodology of individual event prediction is carried out for the data to be predicted obtained after low frequency filtering processing；The multilayer god
Obtained data to be predicted carry out the Forecasting Methodology, described of individual event prediction after neural network forecast method is for intermediate frequency filtering processing
SVM prediction method is that the Forecasting Methodology of individual event prediction is carried out for the data to be predicted obtained after High frequency filter processing.
Specifically, in addition to above-mentioned Forecasting Methodology, there can also be other Forecasting Methodologies, specifically can be according to need
Want and actual application scenarios choose existing Forecasting Methodology to carry out, no longer carry out superfluous chat herein.
Step 104, when the precision of each individual event prediction reaches preset requirement, to each predicted characteristics and described
Weight coefficient carries out the Combinatorial Optimization processing of nonlinear way, to generate prediction result；
Specifically, in one embodiment,
Whether the precision of the individual event prediction reaches preset requirement by error evaluation method to be assessed；Its
In, the error evaluation method includes：Mean square error or evaluation absolute error or evaluation absolute percent error, log error
Quadratic sum is assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and with reference to special
Levy the difference between corresponding weight coefficient whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that
The precision of the individual event prediction reaches preset requirement；If judging result is more than or equal to, it is determined that the unidirectional default essence
Degree is not up to preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；This method further includes：
When the precision of individual event prediction is not reaching to preset requirement, by the error evaluation side to each prediction
After the weight of the parameter of method and each data to be predicted is adjusted, the data to be predicted after adjustment are predicted again.
In addition, each predicted characteristics are corresponding with respective weight coefficient；Nonlinear Combinatorial Optimization processing can be logical
Swarm intelligence multi-objective optimization algorithm is crossed to realize；
Each predicted characteristics are corresponding with respective weight coefficient；The Combinatorial Optimization processing is by the more mesh of swarm intelligence
Mark optimizes algorithm to realize；
Judge whether the Combinatorial Optimization processing reaches default and require to be by the cultivation water signal with monitoring
Site Detection result is compared to judge；If result and Site Detection that the result compared obtains after being handled for Combinatorial Optimization
As a result difference is less than preset value, it is determined that reach default requirement, if what the result compared obtained after being handled for Combinatorial Optimization
As a result the difference with Site Detection result is greater than or equal to preset value, it is determined that not up to default requirement；Wherein different water
Different preset values is answered in confrontation, and default requirement is corresponding with the situation of water quality；
This method further includes：
When Combinatorial Optimization processing is not reaching to default require, added by forecasting effective measure Criterion Method or broad sense
Power is that evaluation assessment or IOWGA are corresponding with each predicted characteristics after respective weight coefficient is adjusted, again to adjustment
Predicted characteristics afterwards carry out the Combinatorial Optimization processing.
Only when Combinatorial Optimization processing, which reaches default, to be required, following step can be just carried out, namely perform following step
105.
Step 105, reach default requirement when Combinatorial Optimization processing, the comprehensive prediction result and the weight
Coefficient generates the Combination Forecasting model of cultivation water.
Specifically, the Combination Forecasting model of generation is used to carry out Combination Forecasting to water quality.This programme is
The thought that combining weights based on induced ordered weighted averaging operator together innovatory algorithm determine is the weight coefficient of combined prediction
It is unrelated with individual event Forecasting Methodology species, and precision of prediction size is closely related on each time point with each individual event Forecasting Methodology,
Precision of prediction becomes the inducement of individual event Forecasting Methodology t moment weight coefficient in i-th, namely the high preferential imparting of precision of prediction
Larger weight coefficient, actual needs are more in line with this, so as to effectively increase the precision of prediction.
Embodiment 2
The embodiment of the present invention 2 discloses a kind of Combination Forecasting system of cultivation water, as shown in figure 3, including：
Acquisition module 201, for carrying out letter to the cultivation water signal monitored based on multiple default feature extraction algorithms
Number decompose, to obtain water quality characteristic component；
Filter module 202, for carrying out different filtering process according to different frequency section to the water quality characteristic component,
With the data to be predicted in the corresponding different frequency section of generation；
Single directional prediction module 203, for the data to be predicted for different frequency separations, chooses different Forecasting Methodologies
Individual event prediction is carried out, to obtain multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；
Each predicted characteristics are corresponding with respective weight coefficient；
Composite module 204, when the precision for being predicted when each individual event reaches preset requirement, to each predicted characteristics
And the weight coefficient carries out the Combinatorial Optimization processing of nonlinear way, to generate prediction result；
Generation module 205, for reaching default requirement when Combinatorial Optimization processing, the comprehensive prediction result and
The Combination Forecasting model of the weight coefficient generation cultivation water.
In a specific embodiment,
The frequency separation includes：Low frequency section, intermediate frequency section, high frequency section；
The filtering process includes being filtered the low frequency filtering processing of processing, in described for the low frequency section
Frequency section is filtered the intermediate frequency filtering processing of processing, the High frequency filter processing of processing is filtered for the high frequency section.
In a specific embodiment,
The method of the individual event prediction includes：ARMA Forecasting Methodologies, multilayer neural network Forecasting Methodology, support vector machines are pre-
Survey method；
Wherein, the ARMA Forecasting Methodologies are to carry out individual event prediction for the data to be predicted obtained after low frequency filtering processing
Forecasting Methodology；The multilayer neural network Forecasting Methodology is to carry out list for the data to be predicted obtained after intermediate frequency filtering processing
Forecasting Methodology, the SVM prediction method of prediction be for obtained data to be predicted after High frequency filter processing into
The Forecasting Methodology of row individual event prediction.
In a specific embodiment,
Whether the precision of the individual event prediction reaches preset requirement by error evaluation method to be assessed；Its
In, the error evaluation method includes：Mean square error or evaluation absolute error or evaluation absolute percent error, log error
Quadratic sum is assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and with reference to special
Levy the difference between corresponding weight coefficient whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that
The precision of the individual event prediction reaches preset requirement；If judging result is more than or equal to, it is determined that the unidirectional default essence
Degree is not up to preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；The system further includes：
First adjustment module, when the precision for being predicted when the individual event is not reaching to preset requirement, passes through the error
After assessment side is adjusted the parameter of each Forecasting Methodology and the weight of each data to be predicted, again to being treated after adjustment
Prediction data is predicted.
In a specific embodiment,
Each predicted characteristics are corresponding with respective weight coefficient；The Combinatorial Optimization processing is by the more mesh of swarm intelligence
Mark optimizes algorithm to realize；
Judge whether the Combinatorial Optimization processing reaches default and require to be by the cultivation water signal with monitoring
Site Detection result is compared to judge；If result and Site Detection that the result compared obtains after being handled for Combinatorial Optimization
As a result difference is less than preset value, it is determined that reach default requirement, if what the result compared obtained after being handled for Combinatorial Optimization
As a result the difference with Site Detection result is greater than or equal to preset value, it is determined that not up to default requirement；Wherein different water
Different preset values is answered in confrontation, and default requirement is corresponding with the situation of water quality；
The system further includes：
Second adjustment module is effective by predicting for when Combinatorial Optimization processing is not reaching to default require
Degree Criterion Method or generalized weighted are that evaluation assessment or IOWGA are corresponding with each predicted characteristics respective weight coefficient progress
After adjustment, the Combinatorial Optimization processing is carried out to the predicted characteristics after adjustment again.
The present invention proposes a kind of Combination Forecasting method and system of cultivation water, and wherein this method includes：Base
Signal decomposition is carried out to the cultivation water signal monitored in multiple default feature extraction algorithms, to obtain water quality characteristic component；
Different filtering process is carried out according to different frequency section to the water quality characteristic component, with the corresponding different frequency section of generation
Data to be predicted；For the data to be predicted of different frequency separations, choose different Forecasting Methodologies and carry out individual event prediction, to obtain
Take multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Each predicted characteristics correspond to
There is respective weight coefficient；When the precision of each individual event prediction reaches preset requirement, to each predicted characteristics and institute
The Combinatorial Optimization processing that weight coefficient carries out nonlinear way is stated, to generate prediction result；When Combinatorial Optimization processing reaches
The Combination Forecasting model of default requirement, the comprehensive prediction result and weight coefficient generation cultivation water.
The precision predicted cultivation water is improved by the scheme of the embodiment of the present invention with this.
It will be appreciated by those skilled in the art that attached drawing is a schematic diagram for being preferable to carry out scene, module in attached drawing or
Flow is not necessarily implemented necessary to the present invention.
It will be appreciated by those skilled in the art that the module in device in implement scene can be described according to implement scene into
Row is distributed in the device of implement scene, can also carry out one or more dresses that respective change is disposed other than this implement scene
In putting.The module of above-mentioned implement scene can be merged into a module, can also be further split into multiple submodule.
The invention described above sequence number is for illustration only, does not represent the quality of implement scene.
Disclosed above is only several specific implementation scenes of the present invention, and still, the present invention is not limited to this, Ren Heben
What the technical staff in field can think change should all fall into protection scope of the present invention.
Claims (10)
1. a kind of Combination Forecasting method of cultivation water, it is characterised in that including：
Signal decomposition is carried out to the cultivation water signal monitored based on multiple default feature extraction algorithms, to obtain water quality characteristic
Component；
Different filtering process is carried out according to different frequency section to the water quality characteristic component, with the corresponding different frequency area of generation
Between data to be predicted；
For the data to be predicted of different frequency separations, choose different Forecasting Methodologies and carry out individual event prediction, it is multiple to obtain
Predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；Each predicted characteristics are corresponding with each
Weight coefficient；
When the precision of each individual event prediction reaches preset requirement, each predicted characteristics and the weight coefficient are carried out
The Combinatorial Optimization processing of nonlinear way, to generate prediction result；
When Combinatorial Optimization processing reaches default requirement, the comprehensive prediction result and weight coefficient generation cultivation
The Combination Forecasting model of water quality.
2. the method as described in claim 1, it is characterised in that the frequency separation includes：Low frequency section, intermediate frequency section, height
Frequency section；
The filtering process includes being filtered the low frequency filtering processing of processing, for the intermediate frequency zone for the low frequency section
Between be filtered the intermediate frequency filtering processing of processing, be filtered for the high frequency section High frequency filter processing of processing.
3. method as claimed in claim 1 or 2, it is characterised in that the method for the individual event prediction includes：ARMA prediction sides
Method, multilayer multilayer neural network Forecasting Methodology, SVM prediction method；
Wherein, the ARMA Forecasting Methodologies are to carry out the pre- of individual event prediction for the data to be predicted obtained after low frequency filtering processing
Survey method；The multilayer multilayer neural network Forecasting Methodology is to carry out list for the data to be predicted obtained after intermediate frequency filtering processing
Forecasting Methodology, the SVM prediction method of prediction be for obtained data to be predicted after High frequency filter processing into
The Forecasting Methodology of row individual event prediction.
4. the method as described in claim 1, it is characterised in that it is logical that whether the precision of the individual event prediction, which reaches preset requirement,
Error evaluation method is crossed to be assessed；Wherein, the error evaluation method includes：Mean square error or evaluation absolute error,
Or evaluation absolute percent error, log error quadratic sum are assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and fixed reference feature pair
Difference between the weight coefficient answered whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that it is described
The precision of individual event prediction reaches preset requirement；As judging result be more than or equal to, it is determined that the unidirectional default precision is not
Reach preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；This method further includes：
When the precision of individual event prediction is not reaching to preset requirement, by the error evaluation side to each Forecasting Methodology
Parameter and each data to be predicted weight be adjusted after, the data to be predicted after adjustment are predicted again.
5. the method as described in claim 1, it is characterised in that each predicted characteristics are corresponding with respective weight coefficient；Institute
Combinatorial Optimization processing is stated to realize by swarm intelligence multi-objective optimization algorithm；
It is the scene by the cultivation water signal with monitoring to judge whether the Combinatorial Optimization processing reaches default requirement
Testing result is compared to judge；If the result that the result compared obtains after being handled for Combinatorial Optimization and Site Detection result
Difference be less than preset value, it is determined that reach default requirement, if the result compared is obtained result after Combinatorial Optimization processing
It is greater than or equal to preset value with the difference of Site Detection result, it is determined that not up to default requirement；Wherein different water quality pair
Different preset values is answered, default requirement is corresponding with the situation of water quality；
This method further includes：
When Combinatorial Optimization processing is not reaching to default require, calculated by forecasting effective measure Criterion Method or generalized weighted
That evaluation assessment or IOWGA are corresponding with each predicted characteristics after respective weight coefficient is adjusted, again to adjustment after
Predicted characteristics carry out the Combinatorial Optimization processing.
A kind of 6. Combination Forecasting system of cultivation water, it is characterised in that including：
Acquisition module, for carrying out signal decomposition to the cultivation water signal monitored based on multiple default feature extraction algorithms,
To obtain water quality characteristic component；
Filter module, for carrying out different filtering process according to different frequency section to the water quality characteristic component, with generation
The data to be predicted in corresponding different frequency section；
Single directional prediction module, for the data to be predicted for different frequency separations, chooses different Forecasting Methodologies and carries out list
Item prediction, to obtain multiple predicted characteristics；Wherein, the mode of each individual event prediction corresponds to different frequency fields；It is each described
Predicted characteristics are corresponding with respective weight coefficient；
Composite module, when the precision for being predicted when each individual event reaches preset requirement, to each predicted characteristics and institute
The Combinatorial Optimization processing that weight coefficient carries out nonlinear way is stated, to generate prediction result；
Generation module, for reaching default requirement, the comprehensive prediction result and the power when Combinatorial Optimization processing
The Combination Forecasting model of weight coefficient generation cultivation water.
7. system as claimed in claim 6, it is characterised in that the frequency separation includes：Low frequency section, intermediate frequency section, height
Frequency section；
The filtering process includes being filtered the low frequency filtering processing of processing, for the intermediate frequency zone for the low frequency section
Between be filtered the intermediate frequency filtering processing of processing, be filtered for the high frequency section High frequency filter processing of processing.
8. system as claimed in claims 6 or 7, it is characterised in that the method for the individual event prediction includes：ARMA prediction sides
Method, multilayer multilayer neural network Forecasting Methodology, SVM prediction method；
Wherein, the ARMA Forecasting Methodologies are to carry out the pre- of individual event prediction for the data to be predicted obtained after low frequency filtering processing
Survey method；The multilayer multilayer neural network Forecasting Methodology is to carry out list for the data to be predicted obtained after intermediate frequency filtering processing
Forecasting Methodology, the SVM prediction method of prediction be for obtained data to be predicted after High frequency filter processing into
The Forecasting Methodology of row individual event prediction.
9. system as claimed in claim 6, it is characterised in that
Whether the precision of the individual event prediction reaches preset requirement by error evaluation method to be assessed；Wherein, institute
Stating error evaluation method includes：Mean square error or evaluation absolute error or evaluation absolute percent error, log error square
With assessed；
Or based on according to acquired predicted characteristics and corresponding weight coefficient and according to fixed reference feature and fixed reference feature pair
Difference between the weight coefficient answered whether less than preset value come the judgement that carries out；If judging result be less than, it is determined that it is described
The precision of individual event prediction reaches preset requirement；As judging result be more than or equal to, it is determined that the unidirectional default precision is not
Reach preset requirement；
Each data to be predicted are corresponding with respective weight coefficient；The system further includes：
First adjustment module, when the precision for being predicted when the individual event is not reaching to preset requirement, passes through the error evaluation
After side is adjusted the parameter of each Forecasting Methodology and the weight of each data to be predicted, again to be predicted after adjustment
Data are predicted.
10. system as claimed in claim 6, it is characterised in that each predicted characteristics are corresponding with respective weight coefficient；Institute
Combinatorial Optimization processing is stated to realize by swarm intelligence multi-objective optimization algorithm；
It is the scene by the cultivation water signal with monitoring to judge whether the Combinatorial Optimization processing reaches default requirement
Testing result is compared to judge；If the result that the result compared obtains after being handled for Combinatorial Optimization and Site Detection result
Difference be less than preset value, it is determined that reach default requirement, if the result compared is obtained result after Combinatorial Optimization processing
It is greater than or equal to preset value with the difference of Site Detection result, it is determined that not up to default requirement；Wherein different water quality pair
Different preset values is answered, default requirement is corresponding with the situation of water quality；
The system further includes：
Second adjustment module, for when Combinatorial Optimization processing is not reaching to default require, passing through forecasting effective measure standard
Then method or generalized weighted are that evaluation assessment or IOWGA are corresponding with respective weight coefficient to each predicted characteristics and are adjusted
Afterwards, the Combinatorial Optimization processing is carried out to the predicted characteristics after adjustment again.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711297240.2A CN108038571A (en) | 2017-12-08 | 2017-12-08 | A kind of Combination Forecasting method and system of cultivation water |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711297240.2A CN108038571A (en) | 2017-12-08 | 2017-12-08 | A kind of Combination Forecasting method and system of cultivation water |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108038571A true CN108038571A (en) | 2018-05-15 |
Family
ID=62101717
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711297240.2A Pending CN108038571A (en) | 2017-12-08 | 2017-12-08 | A kind of Combination Forecasting method and system of cultivation water |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108038571A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114340384A (en) * | 2019-08-20 | 2022-04-12 | 卡塞株式会社 | Water quality management device and method for culture pond |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103577694A (en) * | 2013-11-07 | 2014-02-12 | 广东海洋大学 | Aquaculture water quality short-time combination forecast method on basis of multi-scale analysis |
| CN103778482A (en) * | 2014-02-12 | 2014-05-07 | 中国农业大学 | Aquaculture dissolved oxygen short-term prediction method based on multi-scale analysis |
| CN107016453A (en) * | 2016-12-08 | 2017-08-04 | 中国农业大学 | A kind of aquaculture dissolved oxygen prediction method and device |
| CN108038565A (en) * | 2017-11-30 | 2018-05-15 | 常州大学 | A kind of cultivation water dissolved oxygen prediction method of Runs-test method reconstruct EEMD |
-
2017
- 2017-12-08 CN CN201711297240.2A patent/CN108038571A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103577694A (en) * | 2013-11-07 | 2014-02-12 | 广东海洋大学 | Aquaculture water quality short-time combination forecast method on basis of multi-scale analysis |
| CN103778482A (en) * | 2014-02-12 | 2014-05-07 | 中国农业大学 | Aquaculture dissolved oxygen short-term prediction method based on multi-scale analysis |
| CN107016453A (en) * | 2016-12-08 | 2017-08-04 | 中国农业大学 | A kind of aquaculture dissolved oxygen prediction method and device |
| CN108038565A (en) * | 2017-11-30 | 2018-05-15 | 常州大学 | A kind of cultivation water dissolved oxygen prediction method of Runs-test method reconstruct EEMD |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114340384A (en) * | 2019-08-20 | 2022-04-12 | 卡塞株式会社 | Water quality management device and method for culture pond |
| CN114340384B (en) * | 2019-08-20 | 2023-09-26 | 卡塞株式会社 | Water quality management device and method for culture pond |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Du et al. | H/sub/spl infin//filtering of 2-D discrete systems | |
| Kumar et al. | Fuzzy classifier for fault diagnosis in analog electronic circuits | |
| CN104915534B (en) | Electric power tower deformation analysis based on Sequence Learning and decision-making technique | |
| Andalib et al. | A fuzzy expert system for earthquake prediction, case study: the Zagros range | |
| CN108038571A (en) | A kind of Combination Forecasting method and system of cultivation water | |
| KR101992672B1 (en) | Visual impact analyzing method for overhead transmitting line | |
| CN112380126B (en) | Web system health prediction device and method | |
| Kivits et al. | On representations of linear dynamic networks | |
| CN106596005B (en) | Actively recessed mechanical condition formulates module in vibration control system | |
| CN111901134A (en) | Method and device for predicting network quality based on recurrent neural network model (RNN) | |
| Vaishali et al. | Weather prediction model using Savitzky-Golay and Kalman Filters | |
| Feng et al. | V-CNN: Data visualizing based convolutional neural network | |
| CN110345986B (en) | Multi-stress testing method based on stochastic resonance and task migration | |
| Huang | Human reliability analysis in aviation maintenance by a Bayesian network approach | |
| CN106846262A (en) | Remove the method and system of mosquito noise | |
| CN111310647A (en) | Generation method and device for automatic identification falling model | |
| Cuihua et al. | An information system security evaluation model based on AHP and GRAP | |
| Burmeister et al. | Assessing safety effects of digitization with the European Maritime Simulator Network EMSN: the sea traffic management case | |
| CN112561121A (en) | Rework trend prediction method and system based on mobile phone signaling data | |
| Arai et al. | Recursive least square: RLS method-based time series data prediction for many missing data | |
| Xie et al. | Prioritizing processes for better implementation of statistical process control techniques | |
| Echtenkamp et al. | Implementation issues for symbolic sensitivity analysis | |
| Tirnakli et al. | Damage Spreading In The Bak–Sneppen Model: Sensitivity To The Initial Conditions And Equilibration Dynamics | |
| Ranganathan et al. | Fast-converging complex adaptive algorithm for diversity wireless receivers in linearly fading channels | |
| Matta et al. | Analyzing Social Media in Crisis Management Using Expertise Feedback Modelling. |
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 |
Application publication date: 20180515 |
|
| RJ01 | Rejection of invention patent application after publication |