CN107133398A - A kind of river ethic Forecasting Methodology based on complex network - Google Patents
A kind of river ethic Forecasting Methodology based on complex network Download PDFInfo
- Publication number
- CN107133398A CN107133398A CN201710291000.5A CN201710291000A CN107133398A CN 107133398 A CN107133398 A CN 107133398A CN 201710291000 A CN201710291000 A CN 201710291000A CN 107133398 A CN107133398 A CN 107133398A
- Authority
- CN
- China
- Prior art keywords
- msub
- mrow
- node
- corporations
- complex network
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Hardware Design (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of river ethic Forecasting Methodology, primarily directed to non-avaible hydrology website(PUB:Prediction Ungaged Basins)Runoff Forecast.Its general principle is that the topological property of hydrology Time-space serial is excavated using complex network, carries out Runoff Forecast to non-avaible website on this basis.Runoff complex network model is built according to the footpath flow data of monitoring network website, Newman fast algorithms are utilized on this basis(FN)Carry out corporations' excavation, based on the result that corporations excavate, the selection of both candidate nodes is carried out, this method considers that basin is divided and the correlation between PUB, general character node and the class node of property node two are chosen as both candidate nodes, finally website run-off to be predicted is predicted using grafting.The present invention proposes a kind of association both considered in runoff data topology structure, it is contemplated that a kind of Runoff Forecast method to non-avaible website of footpath flow data in itself.
Description
Technical field
The present invention relates to complex network application field, more particularly to a kind of discharge of river prediction side based on complex network
Method.
Background technology
River is all played an important role at numerous aspects such as the hydrology, water resources management, environment and the ecosystem, so
And, the problem of still suffering from many for the assessment and prediction of river ethic.Because river ethic is by weather shape
Condition and the complicated non-linear process of geomorphic feature interaction.For example, river ethic is not only by the time and spatially
The many factors such as influence, land use parameter, hydrology edphic factor, the geographical statistics property of rainfall distribution all can be to river
Run-off produces influence.
It is existing to be mainly the contact existed between identification river ethic on discharge of river quantifier elimination.But it is existing
Research mostly depend on it is specific the problem of and involved specific element, also existed on river ethic forecasting research
Many problems.Such as existing hydrologic forecast model is mostly relative complex, it is necessary to rely on too many parameter and data to be divided
Analysis, simultaneously because the deviation that deviation and model that data exist in itself exist in itself causes the prediction to runoff cumbersome and not
It is certain reliable;Although existing some model correct algorithms based on bias correction reduce the mistake of prediction to a certain extent
Difference, but this method is for understanding that River Basin Hydrology mechanism does not have any help;From the point of view of another angle, existing mould
Type is the hydrological model for some specific region, such as Xinanjiang model etc. mostly, applies them to wider basin
Scope still there are problems that, therefore lack a kind of pervasive hydrology frame system of unification.
Therefore, it is intended that can be analyzed from discharge of river is macroscopically set up in space and temporal contact between them
Some associations that may be present and influence, and it is desirable that imply the association existed between them by recognizing, find their topology knots
Contact on structure, obtains the description of relation that may be present between discharge of river characteristic and corresponding weather, landforms, so that diameter
Stream prediction is helpful.
The content of the invention
The technical problems to be solved by the invention are to be directed to the defect being related in background technology there is provided one kind based on complexity
The discharge of river Forecasting Methodology of network, using the topological correlation between complex network excavation discharge of river data spatially, so that
The absolute dependence of diameter flow data in itself has been broken away to a certain extent;Meanwhile, go to build river footpath using complex network thought
Flow network, is conducive to carrying out research and analysis in extensive basin.
The present invention uses following technical scheme to solve above-mentioned technical problem:
A kind of river ethic Forecasting Methodology based on complex network, is comprised the steps of:
Step A), discharge of river space-time data is modeled using complex network abstract method, runoff complex web is formed
Network model;
Step B), corporations' excavation is carried out to runoff complex network model using Newman fast algorithms, basin is completed and divides;
Step C), by basin divide based on choose general character node and property node;
Step D), general character node and property node based on selection, utilize runoff of the grafting to non-avaible hydrology website
Amount is predicted.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, the step
Rapid detailed step 1) is as follows:
Step is A.1), for the footpath flow data of each Hydrologic monitoring station, be by the geographical position of corresponding monitoring station is abstract
Node;
Step is A.2), using the correlation between the run-off sequence between two nodes as evaluate two nodes between whether
In the presence of the standard on even side, runoff complex network model is set up:If correlation is more than default relevance threshold, it is believed that corresponding node
Between exist even side, otherwise it is assumed that between node be not present even side.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, the step
Suddenly using Pearson's coefficient as the correlation between the run-off sequence between two nodes in A.2), calculation formula is as follows:
Wherein XiRepresent the flow-through period sequence of node i, wherein XjRepresent node j flow-through period sequence;For sequence
Arrange XiAnd sequence XjBetween Pearson correlation coefficients, cov (Xi,Xj) it is Xi, XjBetween covariance,For XiStandard deviation,For
XjStandard deviation.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, the step
Rapid B) detailed step it is as follows:
Step is B.1), initialization runoff complex network model is N number of corporations, and N is the nodes of runoff complex network model
Mesh, i.e., each node is exactly an independent corporations;
Initialize eijAnd ai, meet it
ai=ki/2m
Wherein, eijThe ratio shared by side between the point of corporations where the point and node j of corporations where node i, aiFor
The ratio on the side of corporations, k where being connected to node iiFor the degree of node i, m is side number total in network;
Step is B.2), the connected corporations pair in side are associated with, makes modularity Q increases at most or reduces minimum, after merging
Modularity increment Delta Q calculation formula is as follows:
Δ Q=eij+eji-2aiaj=2 (eij-aiaj)
Step is B.3), to corresponding element eijUpdate:Will be with i, the related corresponding e of row and column of j corporationsijIt is added conduct
New eij;
Step is B.4), minute book secondary module degree Q value and its corresponding corporations' partition structure;
Step is B.5), repeat step B.2) to step B.4), until whole runoff complex network model is all merged into
For a corporations;
Step is B.6), the result that the selecting module degree Q corresponding corporations' partition structure of maximum is divided as basin.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, the step
Rapid C) in choose two maximum nodes of internal connection number in each corporations and, as the general character node of this corporation, choose and treat and be pre-
Two nearest nodes of nodal distance are surveyed as the property node of node to be predicted.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, this method
Choose general character both candidate nodes and each two of characteristic both candidate nodes.
It is used as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention, the step
Rapid D) detailed step it is as follows:
Step is D.1), set general character factor-alpha;
Step is D.2), the inside Connected degree size according to general character both candidate nodes is divided the weight general character both candidate nodes
Match somebody with somebody, connection number is the number of general character both candidate nodes and the node being connected in its correspondence corporation inside general character both candidate nodes, makes general character
Both candidate nodes o inside connection number is inline_ko, then general character both candidate nodes o weight woCalculation formula it is as follows:
Wherein, koFor general character both candidate nodes o degree,Number is connected for the inside of all general character both candidate nodes
Sum;
Step is D.3), weight distribution is carried out to characteristic both candidate nodes using anti-distance weighting distribution method, i.e.,
Wherein, wqFor characteristic both candidate nodes q weight, dqFor the distance between characteristic both candidate nodes q and node to be predicted;
Step is D.4), general character part is added to the predicted value obtained to node to be predicted with characteristic part, it is specific to calculate public
Formula is as follows:
P=α ∑s wofo+(1-α)∑wqfq
Wherein, P is predicts the outcome, foAnd fqRespectively general character both candidate nodes o and characteristic both candidate nodes q average area list
Position run-off.
It is described common as a kind of further prioritization scheme of river ethic Forecasting Methodology based on complex network of the present invention
Sex factor α takes 0.2.
The present invention uses above technical scheme compared with prior art, with following technique effect:
1. universality:On the one hand, using complex network discharge of river data are predicted simply complex network to data
A variety of hydrology variables can also be predicted, such as rainfall, flood by an instantiation being predicted using such a method
Water characteristic and evaporation capacity etc., on the other hand, this method is not based on the basal conditions in any specific basin, so it can
To be applied in any one basin perimeter, meanwhile, based on this basin universality, set up one it is extensive global
Watershed Hydrologic Models are also attemptable.
2. robustness:Utilize the spy between complex network community method for digging discovery runoff spatial sequence on the topology
Levy.When carrying out Runoff Forecast, topological structure characteristic is taken into account, so that the hydrologic properties of watershed have with research
Certain directive function.At present because climate change and mankind's activity are to hydrology-hydraulic model and its to water circulation, water ring
Border and water disaster cause influence, therefore rely on the prediction of historical data if only only from data value, without considering
The implicit topological features of data, then certain error will certainly be brought to discharge of river prediction, and this method is conceived to
Topological structure between data, it is intended to find the implicit architectural feature of hydrographic data, thus go to study the water regime in basin with
And hydrology mechanism, therefore with certain robustness.
Brief description of the drawings
Fig. 1 shows the overall algorithm flow chart of the present invention;
Fig. 2 shows that the runoff complex network of the present invention builds flow chart;
Fig. 3 shows that the basin of the present invention is divided --- Newman corporations mining algorithm flow chart;
Fig. 4 shows the both candidate nodes type map of the present invention;
Fig. 5 shows the grafting prediction flow chart of the present invention.
Embodiment
Technical scheme is described in further detail below in conjunction with the accompanying drawings:
Fig. 1 is illustrated that the overall algorithm flow chart of the present invention.Runoff Forecast proposed by the present invention based on complex network
Method includes four basic steps substantially:Runoff complex network is built, Newman fast algorithms, and both candidate nodes are chosen and utilized
Grafting carries out discharge of river prediction.The input of this algorithm is several known discharge of river time serieses, is output as unknown
Website to be predicted footpath flow data.
Fig. 2 is illustrated that the runoff complex network of the present invention builds block diagram.Comprise the following steps that:
Step 1:Choose suitable hydrographic data
The hydrographic data of suitable satisfactory monitoring station is chosen from database as needed, the standard of selection is footpath
The time limit scope and the time span of Inflow Sequence of flow data.It is assumed that have chosen N number of satisfactory runoff monitoring station, Mei Gejing
The corresponding flow-through period sequence in stream monitoring station is expressed as X=(x1,x2,x3,......xT)。
Step 2:Node abstraction
The abstract of interior joint of the present invention is as node, to be then total to the geographical position of the corresponding monitoring station of flow-through period sequence
There is N number of node, correspond to X1,X2,X3......XN。
Step 3:Calculate coefficient R
The correlation criterion chosen in the present invention is Pearson's coefficient:
Wherein XiRepresent the flow-through period sequence of node i, wherein XjRepresent node j flow-through period sequence;To each node
XiIts Pearson correlation coefficients with other N-1 node, calculation formula such as above formula are calculated respectively.
Step 4:Set relevance threshold TS
Relevance threshold TS setting will consider side density p (t):
The number of network edge when wherein n (t) represents TS=t, if threshold value TS settings is excessive, causes n (t) too small, from
And many isolated nodes are caused, if threshold value TS settings is too small, cause n (t) too big so that network is almost complete network,
These situations are all useless, therefore the coefficient correlation calculated according to step 2 for the research of network and the analysis of node,
The density profile of coefficient correlation is drawn, a suitable TS span is chosen.
Relevance threshold TS selection also needs to consider the presence on distant associated side simultaneously, and so-called distant correlation is referred to i.e.
Make the node of geographical position wide apart, its correlation still very high node.It is noted that should not lose during threshold value selection
Lose these and connect side.
Step 5:Side it is abstract
For node i, j, if itsThen think there is even side, i.e. adjacency matrix A between the two nodes
Correspondence position is 1, is otherwise 0.I.e.
So far, runoff complex network model builds and terminated.
The uniformity between the division of basin is excavated according to corporations, using corporations' method for digging come to runoff complex network mould
Type carries out corporations divisions, and the result that corporations excavate is to connect relatively very close between the node of each group of inside, but each
Connection between group utilizes the target of corporations method for digging progress basin divisions comparatively than sparse.
Corporations' division is carried out using Newman fast algorithms, this is a kind of agglomerative algorithm based on greedy algorithm thought.It
Each node is considered into a corporations first, a corporations are then progressively merged into, the principle of merging is towards corporations' mould
The maximum direction of lumpiness increase or the direction for reducing minimum, then the maximum corporations of respective modules degree, which divide, then corresponds to optimal society
Group divides, i.e., the result that basin is divided.
Fig. 3 is illustrated that the Newman fast algorithm flow charts of the present invention.Specific such as following steps:
Step 1:Initialization
Initialization network is N number of corporations, i.e., each node is exactly an independent corporations.Initialize eijAnd aiMeet
ai=ki/2m
Wherein, kiFor the degree of node i, m is side number total in network.
Step 2:Merge network
The connected corporations pair in side are once associated with, and calculate the modularity increment after merging:
Δ Q=eij+eji-2aiaj=2 (eij-aiaj)
According to the principle of greedy algorithm, merge be carried out every time along making Q increase at most or reduce minimum direction.
After merging every time, to corresponding element eijUpdate:Will be with i, the related corresponding e of row and column of j corporationsijBe added as newly
eij。
This step is repeated, constantly merges corporations, until whole network all merges as a corporations.
Step 3:Select optimal corporations' division result
During corporations are merged, corporations' division result is correspond to after each merging, a correspondence is selected
The corporations for local maxima Q values divide, then it is exactly optimal network community structure that its is corresponding.
Fig. 4 is illustrated that the both candidate nodes type map of the present invention.The selection of both candidate nodes considers two factors:General character candidate
Node and characteristic both candidate nodes.So-called general character both candidate nodes refer to that the basin inside corporations extremely related in a certain basin is represented
Node, two nodes for choosing internal connection number maximum in each corporations are used as the general character node of this corporation.Because for network
For node, the degree of node represents its correlation with remaining node, if the degree of a node is bigger, demonstrates it at this
It is more important inside individual corporations;The selection of property node is then according to " close things is always similar ", therefore the standard of selection
Be then with nodal distance to be predicted, distance it is nearer, then it is assumed that it is more similar to node to be predicted, choose with nodal distance to be predicted most
Two near nodes as node to be predicted property node.
What Fig. 5 was provided is the grafting prediction flow chart of the present invention.Comprise the following steps that:
Step 1:Set the general character factor
Because have chosen general character both candidate nodes and characteristic both candidate nodes, therefore first have to set the general character factor, typically take
0.2。
Step 2:Weight distribution between general character both candidate nodes
Weight distribution between general character both candidate nodes is divided according to its significance level inside corporations, i.e.,
Inside Connected degree size according to general character both candidate nodes is allocated the weight general character both candidate nodes, and general character is waited
Select intra-node to connect the number that number is general character both candidate nodes and the node being connected in its correspondence corporation, make general character both candidate nodes o
Inside connection number be inline_ko, then general character both candidate nodes o weight woCalculation formula it is as follows:
Wherein, koFor general character both candidate nodes o degree,Number is connected for the inside of all general character both candidate nodes
Sum.
Step 3:Weight distribution between characteristic both candidate nodes
The weight distributions of characteristic both candidate nodes according to be distance it is more near then it is more important in prediction, using anti-distance weighting
Distribution method, i.e.,
Wherein, wqFor characteristic both candidate nodes q weight, dqFor the distance between characteristic both candidate nodes q and node to be predicted.
Step 4:Addition calculation
General character part is added to the predicted value obtained to node to be predicted with characteristic part.Specific formula for calculation is as follows:
P=α ∑s wofo+(1-α)∑wqfq
Wherein, P is predicts the outcome, foAnd fqRespectively general character both candidate nodes o and characteristic both candidate nodes q average area list
Position run-off.
Those skilled in the art of the present technique are it is understood that unless otherwise defined, all terms used herein (including skill
Art term and scientific terminology) with the general understanding identical meaning with the those of ordinary skill in art of the present invention.Also
It should be understood that those terms defined in such as general dictionary should be understood that with the context of prior art
The consistent meaning of meaning, and unless defined as here, will not be explained with idealization or excessively formal implication.
Above-described embodiment, has been carried out further to the purpose of the present invention, technical scheme and beneficial effect
Describe in detail, should be understood that the embodiment that the foregoing is only the present invention, be not limited to this hair
Bright, within the spirit and principles of the invention, any modification, equivalent substitution and improvements done etc. should be included in the present invention
Protection domain within.
Claims (8)
1. a kind of river ethic Forecasting Methodology based on complex network, it is characterised in that comprise the steps of:
Step A), discharge of river space-time data is modeled using complex network abstract method, runoff complex network mould is formed
Type;
Step B), corporations' excavation is carried out to runoff complex network model using Newman fast algorithms, basin is completed and divides;
Step C), by basin divide based on choose general character node and property node;
Step D), general character node and property node based on selection, using grafting the run-off of non-avaible hydrology website is entered
Row prediction.
2. the river ethic Forecasting Methodology based on complex network according to right 1, it is characterised in that the step 1)
Detailed step is as follows:
Step is A.1), it is for the footpath flow data of each Hydrologic monitoring station, the geographical position of corresponding monitoring station is abstract for node;
Step is A.2), the correlation between the run-off sequence between two nodes whether there is as between two nodes of evaluation
Even the standard on side, sets up runoff complex network model:If correlation is more than default relevance threshold, it is believed that between corresponding node
In the presence of even side, otherwise it is assumed that the company of being not present side between node.
3. the river ethic Forecasting Methodology based on complex network according to right 2, it is characterised in that the step is A.2)
Middle use Pearson coefficient is as the correlation between the run-off sequence between two nodes, and calculation formula is as follows:
<mrow>
<msub>
<mi>&rho;</mi>
<mrow>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
</mrow>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mi>cov</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>,</mo>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
<mrow>
<msub>
<mi>&sigma;</mi>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
</msub>
<msub>
<mi>&sigma;</mi>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
</msub>
</mrow>
</mfrac>
<mo>=</mo>
<mfrac>
<mrow>
<mi>&Sigma;</mi>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
</mrow>
<msqrt>
<mrow>
<mi>&Sigma;</mi>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>X</mi>
<mi>i</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
<mi>&Sigma;</mi>
<msup>
<mrow>
<mo>(</mo>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
<mo>-</mo>
<mover>
<msub>
<mi>X</mi>
<mi>j</mi>
</msub>
<mo>&OverBar;</mo>
</mover>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
</mfrac>
</mrow>
Wherein XiRepresent the flow-through period sequence of node i, wherein XjRepresent node j flow-through period sequence;For sequence XiWith
Sequence XjBetween Pearson correlation coefficients, cov (Xi,Xj) it is Xi, XjBetween covariance,For XiStandard deviation,For XjMark
It is accurate poor.
4. the river ethic Forecasting Methodology based on complex network according to right 3, it is characterised in that the step B)
Detailed step is as follows:
Step is B.1), initialization runoff complex network model is N number of corporations, and N is the interstitial content of runoff complex network model, i.e.,
Each node is exactly an independent corporations;
Initialize eijAnd ai, meet it
ai=ki/2m
Wherein, eijThe ratio shared by side between the point of corporations where the point and node j of corporations where node i, aiFor connection
The ratio on the side of corporations, k where to node iiFor the degree of node i, m is side number total in network;
Step is B.2), the connected corporations pair in side are associated with, makes modularity Q increases at most or reduces minimum, the module after merging
The calculation formula for spending increment Delta Q is as follows:
Δ Q=eij+eji-2aiaj=2 (eij-aiaj)
Step is B.3), to corresponding element eijUpdate:Will be with i, the related corresponding e of row and column of j corporationsijBe added as newly
eij;
Step is B.4), minute book secondary module degree Q value and its corresponding corporations' partition structure;
Step is B.5), repeat step B.2) to step B.4), until whole runoff complex network model all merges as one
Individual corporations;
Step is B.6), the result that the selecting module degree Q corresponding corporations' partition structure of maximum is divided as basin.
5. the river ethic Forecasting Methodology based on complex network according to right 4, it is characterised in that the step C) in
Two maximum nodes of internal connection number in each corporations are chosen, as the general character node of this corporation, to choose and nodal point separation to be predicted
From property node of the two nearest nodes as node to be predicted.
6. the river ethic Forecasting Methodology based on complex network according to right 5, it is characterised in that this method is chosen altogether
Property both candidate nodes and each two of characteristic both candidate nodes.
7. the river ethic Forecasting Methodology based on complex network according to right 6, it is characterised in that the step D)
Detailed step is as follows:
Step is D.1), set general character factor-alpha;
Step is D.2), the inside Connected degree size according to general character both candidate nodes is allocated the weight general character both candidate nodes,
Connection number is the number of general character both candidate nodes and the node being connected in its correspondence corporation inside general character both candidate nodes, makes general character candidate
Node o inside connection number is inline_ko, then general character both candidate nodes o weight woCalculation formula it is as follows:
<mrow>
<msub>
<mi>w</mi>
<mi>o</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mi>i</mi>
<mi>n</mi>
<mi>l</mi>
<mi>i</mi>
<mi>n</mi>
<mi>e</mi>
<mo>_</mo>
<msub>
<mi>k</mi>
<mi>o</mi>
</msub>
</mrow>
<mrow>
<munder>
<mo>&Sigma;</mo>
<mi>p</mi>
</munder>
<mi>i</mi>
<mi>n</mi>
<mi>l</mi>
<mi>i</mi>
<mi>n</mi>
<mi>e</mi>
<mo>_</mo>
<msub>
<mi>k</mi>
<mi>p</mi>
</msub>
</mrow>
</mfrac>
</mrow>
Wherein, koFor general character both candidate nodes o degree,Number sum is connected for the inside of all general character both candidate nodes;
Step is D.3), weight distribution is carried out to characteristic both candidate nodes using anti-distance weighting distribution method, i.e.,
<mrow>
<msub>
<mi>w</mi>
<mi>q</mi>
</msub>
<mo>=</mo>
<mfrac>
<mrow>
<mn>1</mn>
<mo>/</mo>
<msub>
<mi>d</mi>
<mi>q</mi>
</msub>
</mrow>
<mrow>
<mi>&Sigma;</mi>
<mn>1</mn>
<mo>/</mo>
<msub>
<mi>d</mi>
<mi>q</mi>
</msub>
</mrow>
</mfrac>
</mrow>
Wherein, wqFor characteristic both candidate nodes q weight, dqFor the distance between characteristic both candidate nodes q and node to be predicted;
Step is D.4), general character part is added to the predicted value obtained to node to be predicted with characteristic part, specific formula for calculation is such as
Under:
P=α ∑s wofo+(1-α)∑wqfq
Wherein, P is predicts the outcome, foAnd fqRespectively general character both candidate nodes o and characteristic both candidate nodes q average area unit footpath
Flow.
8. the river ethic Forecasting Methodology based on complex network according to right 7, it is characterised in that the general character factor
α takes 0.2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710291000.5A CN107133398B (en) | 2017-04-28 | 2017-04-28 | River runoff prediction method based on complex network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710291000.5A CN107133398B (en) | 2017-04-28 | 2017-04-28 | River runoff prediction method based on complex network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107133398A true CN107133398A (en) | 2017-09-05 |
CN107133398B CN107133398B (en) | 2020-09-01 |
Family
ID=59715790
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710291000.5A Active CN107133398B (en) | 2017-04-28 | 2017-04-28 | River runoff prediction method based on complex network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107133398B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108876021A (en) * | 2018-05-31 | 2018-11-23 | 华中科技大学 | A kind of Medium-and Long-Term Runoff Forecasting method and system |
CN109359199A (en) * | 2018-08-27 | 2019-02-19 | 平安科技(深圳)有限公司 | Fund manager's group dividing method, system, computer equipment and storage medium |
CN109753362A (en) * | 2019-01-14 | 2019-05-14 | 中国水利水电科学研究院 | A kind of confluence Method of Scheduling Parallel of hydrological distribution model |
CN110083780A (en) * | 2019-04-25 | 2019-08-02 | 上海理工大学 | Personalized recommendation method based on community division in complex network model |
CN110110391A (en) * | 2019-04-16 | 2019-08-09 | 清华大学 | A kind of flood estimation method in the scarce measuring runoff data basin based on region division |
CN110363350A (en) * | 2019-07-15 | 2019-10-22 | 西华大学 | A kind of regional air pollutant analysis method based on complex network |
CN111340288A (en) * | 2020-02-25 | 2020-06-26 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering space-time correlation |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7624081B2 (en) * | 2006-03-28 | 2009-11-24 | Microsoft Corporation | Predicting community members based on evolution of heterogeneous networks using a best community classifier and a multi-class community classifier |
CN102867106A (en) * | 2012-08-14 | 2013-01-09 | 贵州乌江水电开发有限责任公司 | Method and system for predicting short-term running water |
CN103914619A (en) * | 2014-03-31 | 2014-07-09 | 河海大学 | River network modeling method |
-
2017
- 2017-04-28 CN CN201710291000.5A patent/CN107133398B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7624081B2 (en) * | 2006-03-28 | 2009-11-24 | Microsoft Corporation | Predicting community members based on evolution of heterogeneous networks using a best community classifier and a multi-class community classifier |
CN102867106A (en) * | 2012-08-14 | 2013-01-09 | 贵州乌江水电开发有限责任公司 | Method and system for predicting short-term running water |
CN103914619A (en) * | 2014-03-31 | 2014-07-09 | 河海大学 | River network modeling method |
Non-Patent Citations (2)
Title |
---|
AHRENS B: "Distance in spatial interpolation of daily", 《HYDROLOGY AND EARTH SYSTEM SCIENCES》 * |
崔楠,吴学文: "复杂网络在水文中的应用研究综述", 《水资源与水工程学报》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108876021A (en) * | 2018-05-31 | 2018-11-23 | 华中科技大学 | A kind of Medium-and Long-Term Runoff Forecasting method and system |
CN109359199A (en) * | 2018-08-27 | 2019-02-19 | 平安科技(深圳)有限公司 | Fund manager's group dividing method, system, computer equipment and storage medium |
CN109753362A (en) * | 2019-01-14 | 2019-05-14 | 中国水利水电科学研究院 | A kind of confluence Method of Scheduling Parallel of hydrological distribution model |
CN109753362B (en) * | 2019-01-14 | 2020-03-24 | 中国水利水电科学研究院 | Convergence parallel scheduling method of distributed hydrological model |
CN110110391A (en) * | 2019-04-16 | 2019-08-09 | 清华大学 | A kind of flood estimation method in the scarce measuring runoff data basin based on region division |
CN110083780A (en) * | 2019-04-25 | 2019-08-02 | 上海理工大学 | Personalized recommendation method based on community division in complex network model |
CN110083780B (en) * | 2019-04-25 | 2023-07-21 | 上海理工大学 | Community based on complex network model partitioned personalized recommendation method |
CN110363350A (en) * | 2019-07-15 | 2019-10-22 | 西华大学 | A kind of regional air pollutant analysis method based on complex network |
CN110363350B (en) * | 2019-07-15 | 2023-10-10 | 西华大学 | Regional air pollutant analysis method based on complex network |
CN111340288A (en) * | 2020-02-25 | 2020-06-26 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering space-time correlation |
CN111340288B (en) * | 2020-02-25 | 2024-04-05 | 武汉墨锦创意科技有限公司 | Urban air quality time sequence prediction method considering time-space correlation |
Also Published As
Publication number | Publication date |
---|---|
CN107133398B (en) | 2020-09-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107133398A (en) | A kind of river ethic Forecasting Methodology based on complex network | |
Mehdizadeh et al. | Hybrid artificial intelligence-time series models for monthly streamflow modeling | |
CN109658695B (en) | Multi-factor short-term traffic flow prediction method | |
Cigizoglu | Application of generalized regression neural networks to intermittent flow forecasting and estimation | |
Noori et al. | Assessment of input variables determination on the SVM model performance using PCA, Gamma test, and forward selection techniques for monthly stream flow prediction | |
Kisi | Modeling solar radiation of Mediterranean region in Turkey by using fuzzy genetic approach | |
Shamim et al. | A comparison of artificial neural networks (ANN) and local linear regression (LLR) techniques for predicting monthly reservoir levels | |
Kumar et al. | Artificial neural network model for rainfall-runoff-A case study | |
Agarwal et al. | Simulation of runoff and sediment yield using artificial neural networks | |
CN112541302B (en) | Air quality prediction model training method, air quality prediction method and device | |
Bruen et al. | Functional networks in real-time flood forecasting—a novel application | |
Tao et al. | The feasibility of integrative radial basis M5Tree predictive model for river suspended sediment load simulation | |
Ouali et al. | Fully nonlinear statistical and machine‐learning approaches for hydrological frequency estimation at ungauged sites | |
Khu et al. | Incorporating multiple observations for distributed hydrologic model calibration: An approach using a multi-objective evolutionary algorithm and clustering | |
Üneş et al. | Estimation of groundwater level using artificial neural networks: a case study of Hatay-Turkey | |
CN107274030A (en) | Runoff Forecast method and system based on hydrology variable year border and monthly variation characteristic | |
CN110276477B (en) | Flood forecasting method based on hierarchical Bayesian network and incremental learning | |
Kwin et al. | Rainfall-runoff modeling using dynamic evolving neural fuzzy inference system with online learning | |
CN109063759A (en) | A kind of neural network structure searching method applied to the more attribute forecasts of picture | |
Yu et al. | Comparison between response surface models and artificial neural networks in hydrologic forecasting | |
Cheng et al. | A new indirect multi-step-ahead prediction model for a long-term hydrologic prediction | |
Sihag et al. | Evaluation of tree regression analysis for estimation of river basin discharge | |
Yadav et al. | Application of Artificial Neural Network and Genetic Algorithm Based Artificial Neural Network Models for River Flow Prediction. | |
Panahi et al. | Insights into enhanced machine learning techniques for surface water quantity and quality prediction based on data pre-processing algorithms | |
Atashi et al. | Distributed Muskingum model with a Whale Optimization Algorithm for river flood routing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right |
Effective date of registration: 20210106 Address after: 230088 Room 101, R & D office building 2, Kunpeng Industrial Park, 539 Wangjiang West Road, high tech Zone, Hefei City, Anhui Province Patentee after: ANHUI JINHAIDIER INFORMATION T Address before: 211100 No. 8 West Buddha Road, Jiangning District, Jiangsu, Nanjing Patentee before: HOHAI University |
|
TR01 | Transfer of patent right |