CN110298498A - A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines - Google Patents
A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines Download PDFInfo
- Publication number
- CN110298498A CN110298498A CN201910514384.1A CN201910514384A CN110298498A CN 110298498 A CN110298498 A CN 110298498A CN 201910514384 A CN201910514384 A CN 201910514384A CN 110298498 A CN110298498 A CN 110298498A
- Authority
- CN
- China
- Prior art keywords
- data
- flood
- support vector
- time series
- model
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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"
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A10/00—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE at coastal zones; at river basins
- Y02A10/40—Controlling or monitoring, e.g. of flood or hurricane; Forecasting, e.g. risk assessment or mapping
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Evolutionary Biology (AREA)
- Strategic Management (AREA)
- Operations Research (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Economics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computational Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Probability & Statistics with Applications (AREA)
- Entrepreneurship & Innovation (AREA)
- Algebra (AREA)
- Game Theory and Decision Science (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Development Economics (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of Flood of small drainage area forecasting procedures that data-driven is established with support vector machines, comprising steps of (1) carries out correlation analysis to meteorological and data on flows time series, input data of the suitable time series as data-driven model is chosen；(2) algorithm of support vector machine regression modeling is used, its kernel function type and hyper parameter are adjusted, validation-cross is rolled over by k, determines the model structure and hyper parameter for establishing data-driven with support vector machines；(3) with all historical data training data driving models；(4) it combines real-time traffic data and history to postpone data on flows, generates prediction input data, input data driving model, to forecast real-time traffic, storage capacity and water level.The present invention establishes the data-driven model of the sudden flood forecasting in basin by introducing Support vector regression algorithm, solves the problems, such as that traditional conceptual model and physical model calculating need parameter excessive and influence in calculating process vulnerable to algorithm stability.
Description
Technical field
The present invention relates to a kind of Flood Forecasting Methods, and in particular to a kind of small stream that data-driven is established with support vector machines
Domain Flood Forecasting Method.
Background technique
Mountain torrents Yi Faqu small watershed is common based on small-sized massif, and the Mountainous Areas slope Shan Gao is steep, and streams is intensive, flood
Concentration time is short, and water level rises suddenly to be fallen suddenly, is broken with tremendous force, and often the short time causes disaster, at the same may cause landslide, collapse slope, avalanche and
The secondary disasters such as mud-rock flow.Since medium and small reservoirs are distributed in these regions mostly, the construction age is more early, and construction criteria is low, fortune
Segment length when row, burst mountain torrents and secondary disaster easily cause the safety problem of reservoir dam structure and reservoir area bank slope, or even cause
The risk of big dam breaking, dam, which once bursts, will more bring the casualties for being difficult to estimate and economic asset to lose.Therefore, make
A part of the quick early warning technology research of calamity is caused for mountain torrents Yi Faqu reservoir, Flood of small drainage area monitoring and prediction technical research has weight
Want meaning.
Due to the rainfall concentration of induction Flood of small drainage area, region is obvious, it is rapid to cause disaster, spatial and temporal distributions characteristic is complicated, is based on
The weather forecast uncertainty of existing meteorologic model is larger.These features all increase the difficulty of Flood of small drainage area monitoring and prediction.
Traditional conceptual model and the parameter that physical model calculating needs to obtain are more, and vulnerable to algorithm stability in calculating process
Influence.
Support vector machines is that a kind of foundation proposed by statistician Vapnik ties up theoretical and Structural risk minization original in VC
Supervised machine learning method on the basis of reason.In recent years, on the basis of support vector machines, development is modified to continuous variable
The support vector regression of prediction is widely applied in hydrologic(al) prognosis field.Huang Mutao, Tian Yong et al. are in 2011 based on certainly
It adapts to organising map and support vector regression predicts that realizing RMSE is to Hubei Province, river, Qingjian River intraday effect
55.63m3The precision of prediction of/s.Liu Zhiyong, all equality people proposed a kind of support vector regression of integrated wavelet transformation in 2014
Model, and be used for day peak flow prediction of the U.S. Baihe Dong Cha, achieve NSE be 0.928, RMSE 58.443m3/ s's
Prediction effect.Algorithm generalization ability with higher limits the upper limit of prediction error and always converges on globally optimal solution.
Cross verification can also be rolled over by k adjusts the methods of control parameter, historical data re -training continuous improvement arithmetic accuracy.But
Since middle and small river is located at the Mountain Area of data shortage mostly, flood has sudden spy strong, the concentration time is fast, leading time is short
Point is rarely used in Flood of small drainage area forecast with the Flood Forecasting Model that support vector machines establishes data-driven.
Summary of the invention
For the above-mentioned problems in the prior art, it is an object of that present invention to provide one kind to establish number with support vector machines
According to the Flood of small drainage area forecasting procedure of driving.Algorithm of support vector machine generalization ability with higher limits prediction error
The upper limit and always converge on globally optimal solution.In addition, solving tradition based on the data-driven model that machine learning algorithm proposes
Conceptual model and physical model calculating the problem of needing parameter excessive and being influenced in calculating process vulnerable to algorithm stability.
In order to solve the above technical problems, the present invention uses following technological means:
The present invention proposes a kind of Flood of small drainage area forecasting procedure that data-driven is established with support vector machines, comprising:
Step (1) collects and analyzes existing weather history and data on flows；
Step (2) carries out correlation analysis to meteorological and data on flows time series, chooses suitable time series and makees
For the input data of data-driven model；
Step (3) carries out regression modeling using algorithm of support vector machine, and adjusts to its kernel function type and hyper parameter
It is whole, validation-cross is rolled over by k, determines the model structure and hyper parameter for establishing data-driven with support vector machines；
Step (4), with all historical data training data driving models, postpone flow in conjunction with real-time traffic data and history
Data generate prediction input data, input data driving model, to forecast real-time traffic, storage capacity and water level.
Further, Flood of small drainage area forecasting procedure proposed by the invention, in step (2), correlation analysis is main
Including two parts, a part is the sample autocorrelation function for calculating time series, and another part is to calculate different delay time sequences
Related coefficient and mutual information score value between column.The specific method is as follows:
To time series { yt, t=1 ..., K delay sample autocorrelation function value under k and are defined as follows:
In formula, ckFor the sample variance of time series, K represents time series { ytQuantity.
To two time series { xt},{yt, t=1 ... K, related coefficient are defined as follows:
To two time series { xt},{yt, t=1 ... K, mutual information score value are defined as follows:
In formula, p (x), p (y) are { x respectivelytAnd { ytMarginal probability distribution function, p (x, y) is two time serieses
Joint probability distribution function is estimated by sample.
Further, Flood of small drainage area forecasting procedure proposed by the invention, in step (3), using support vector machines
The specific method is as follows for algorithm regression modeling:
Give one group of rainfall runoff training data { X, y }={ (xi,yi) | i=1 ..., m }, first by kernel function come real
Show Φ: X → F of mapping transformation for input data xiIt is mapped to more high-dimensional feature space Φ (x), the kernel function used is radial direction
Base kernel function indicates are as follows:
In formula, < Φ (xi,xj) > indicate Φ (xi) and Φ (xj) inner product, K (xi,xj) indicate xiAnd xjKernel function product, γ
Represent regulation coefficient；
Then, linear regression, regression equation are carried out in feature space Φ (x) are as follows:
Wherein, coefficient w and b can be determined by solving following optimization problem:
Constraint condition:
Wherein, wTFor the transposition of coefficient w,The complexity of constricted regression model, avoids overfitting problem.Constant C is
Weight determines balance of the objective function between fitting degree and function complexity.Determine the fitting of data
Degree is calculated by ε-loose penalty function, is indicated are as follows:
Further, Flood of small drainage area forecasting procedure proposed by the invention, in step (3), can by training data into
The hyper parameter of row adjustment refers to the ε in coefficient gamma, constraint constant C and penalty function.
Further, Flood of small drainage area forecasting procedure proposed by the invention, in step (3), k folding cross verification refers to
Rainfall and runoff training data are divided into k subset, takes and is wherein used as validation group for one group, remaining k-1 group is obtained as test set
Extensive error.It repeats k times, takes the average value of k extensive error.
The invention adopts the above technical scheme, and comparison prior art has following technical effect that
1, the present invention establishes the data-driven model of Flood of small drainage area forecast using algorithm of support vector machine, which has
Higher generalization ability limits the upper limit of prediction error and always converges on globally optimal solution.Interaction can also be rolled over by k to test
Demonstration adjusts the methods of control parameter, historical data re -training and arithmetic accuracy is continuously improved.
2, the present invention solves traditional conceptual model and needs to join with physical model calculating by establishing data-driven model
The problem of counting excessively and influencing in calculating process vulnerable to algorithm stability, can learn directly from data from input (example
Such as precipitation) to the mapping relations of output (such as flow), realize that flood forecasting calculates.
Detailed description of the invention
Fig. 1 is the operating procedure flow diagram of the method for the present invention.
Fig. 2 is the data interaction verification process of one embodiment of detection method of the present invention.
Fig. 3 is the data-driven model training and verification result schematic diagram of detection method one embodiment of the present invention.
Specific embodiment
It is described in further detail with reference to the accompanying drawings and examples:
Those skilled in the art can understand that unless otherwise defined, all terms used herein (including skill
Art term and scientific term) there is meaning identical with the general understanding of those of ordinary skill in fields of the present invention.Also
It should be understood that those terms such as defined in the general dictionary should be understood that have in the context of the prior art
The consistent meaning of meaning will not be explained in an idealized or overly formal meaning and unless defined as here.
The characteristics of for small watershed burst flood forecasting, it is pre- that the present invention uses algorithm of support vector machine to establish Flood of small drainage area
The data-driven model of report.
One embodiment of the present of invention is to carry out Flood Forecasting Model modeling to certain reservoir.The above domination set of reservoir dam site
Rain area 63.4km2, the more than dam site main long 14.4km in river, river mean inclination is than drop 17.7 ‰.Crest elevation 1972.10m, top
Wide 5.0m, long 192.20m, aggregate storage capacity 1127.72m3, check flood level 1971.81m, design flood level 1971.31m are normal to store
Water level 1968.5m.Basin locating for reservoir belongs to north subtropical monsoon climate, and dry and wet season is clearly demarcated.Mean annual precipitation
780.9mm is unevenly distributed in year, and 5~October of flood season, rainfall accounted for the 91% of annual rainfall, and 7, account within August two months 49.2%, it is withered
Season (April 11~next year) accounts for 9%.Be averaged water surface evaporation 2316.1mm for many years.The method of this flood forecasting, by using support
Vector machine algorithm establishes data-driven model, operating process such as Fig. 1, comprising steps of
Step 1 collects and analyzes existing weather history and data on flows, for the present embodiment, precipitation and data on flows
It needs to carry out inverse scale processing, chooses daily maximum stream flow, add up precipitation daily.
Step 2 carries out correlation analysis to meteorological and data on flows time series, chooses suitable time series conduct
The input data of data-driven model.
For the present embodiment, the selected time series as data-driven model input variable includes:
(1) away from it is previous be more than 3mm precipitation interval number of days；
(2) previous hour reservoir inflow；
(3) preceding 24 hours accumulative precipitation；
(4) add up precipitation in first 120 hours；
(5) add up precipitation in first 336 hours；
(6) preceding 24 hours maximum stream flows；
Maximum stream flow before (7) 48 hours；
Maximum stream flow before (8) 72 hours.
Wherein, (1) item takes its inverse.The relatively closely spaced period (24 to 48 hours) can be differentiated with the larger period in this way
Come.Other 7 take natural logrithm.Last linear transformation is to (0,1) section again, to guarantee the stability of algorithm.Become as target
The day maximum stream flow of amount (i.e. the input variable of data-driven model) similarly takes natural logrithm, makees linear transformation.Correspondingly, mould
The predicted value of type need to pass through inverse transformation, exponentiation.
Step 3 carries out regression modeling using algorithm of support vector machine, and adjusts to its kernel function type and hyper parameter
It is whole, validation-cross is rolled over by k, determines the model structure and hyper parameter for establishing data-driven with support vector machines.
For the present embodiment, one group of rainfall runoff training data { X, y }={ (x is giveni,yi) | i=1 ..., m }, first
Realize Φ: X → F of mapping transformation by input data x by kernel functioniIt is mapped to more high-dimensional feature space Φ (x), is used
Kernel function be Radial basis kernel function, indicate are as follows:
In formula, < Φ (xi,xj) > indicate Φ (xi) and Φ (xj) inner product, K (xi,xj) indicate xiAnd xjKernel function product.
Then, linear regression, regression equation are carried out in feature space Φ (x) are as follows:
Wherein, coefficient w and b can be determined by solving following optimization problem:
Constraint condition:
Wherein, wTFor the transposition of coefficient w,The complexity of constricted regression model, avoids overfitting problem.Constant C is
Weight determines balance of the objective function between fitting degree and function complexity.Determine the fitting of data
Degree is calculated by ε-loose penalty function, is indicated are as follows:
Using 5 folding validation-cross, i.e., precipitation and data on flows are divided into 5 roughly the same subsets of size.Such as Fig. 2 institute
Show, validation-cross the results show that (1) and (2) item input data is more important；
Step 4, with all historical data training data driving models, postpone flow number in conjunction with real-time traffic data and history
According to, generation prediction input data, input data driving model, to forecast real-time traffic, storage capacity and water level.
Historical data feature based on the present embodiment, model selection are hourly with first 80 days (first arrival mid or late July in May)
Data are trained.As verifying, with 134 days thereafter maximum stream flows of the model prediction after training, model accuracy result is as follows
Shown in table:
Fig. 3 shows the data-driven model training and verification result schematic diagram of the application detection method one embodiment.
By modeling result as it can be seen that being Qualify Phase in training, the performance of this supporting vector machine model is fine.NS coefficient is greater than 0.7, averagely
Error is close to 0, and mean absolute error is within the 25% of average flow rate.The fitting that mean square deviation is more laid particular stress on to peak value is slightly higher.
The above is only some embodiments of the invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (5)
1. a kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines characterized by comprising
Step (1) collects and analyzes existing history meteorology and data on flows；
Step (2) carries out correlation analysis to meteorological and data on flows time series, chooses suitable time series as number
According to the input data of driving model；
Step (3) carries out regression modeling using algorithm of support vector machine, and is adjusted to its kernel function type and hyper parameter,
Validation-cross is rolled over by k, determines the model structure and hyper parameter for establishing data-driven with support vector machines；
Step (4), with all historical data training data driving models, postpone flow number in conjunction with real-time traffic data and history
According to, generation prediction input data, input data driving model, to forecast real-time traffic, storage capacity and water level.
2. Flood of small drainage area forecasting procedure according to claim 1, which is characterized in that in step (2), correlation analysis
Including two parts, a part is the sample autocorrelation function for calculating time series, and another part is to calculate different delay time sequences
Related coefficient and mutual information score value between column；The specific method is as follows:
To time series { yt, t=1 ..., K delay sample autocorrelation function value r under kkIt is defined as follows:
In formula, ckFor the sample variance of time series, K represents time series { ytQuantity；
To two time series { xt},{yt, t=1 ... K, related coefficient are defined as follows:
To two time series { xt},{yt, t=1 ... K, mutual information score value are defined as follows:
In formula, p (x), p (y) are { x respectivelytAnd { ytMarginal probability distribution function, p (x, y) is the joint of two time serieses
Probability-distribution function is estimated by sample.
3. Flood of small drainage area forecasting procedure according to claim 1, which is characterized in that in step (3), using support to
The specific method is as follows for amount machine algorithm regression modeling:
Give one group of rainfall runoff training data { X, Y }={ (xi,yi) | i=1 ..., m }, it is reflected first by kernel function to realize
Φ: X → F of transformation is penetrated by input data xiIt is mapped to more high-dimensional feature space Φ (x), the kernel function used is radial base core
Function indicates are as follows:
In formula, < Φ (xi,xj) > indicate Φ (xi) and Φ (xj) inner product, K (xi,xj) indicate xiAnd xjKernel function product, γ represent
Regulation coefficient；
Then, linear regression, regression equation are carried out in feature space Φ (x) are as follows:
Wherein, coefficient w and b is determined by solving following optimization problem:
Constraint condition:
Wherein, wTFor the transposition of coefficient w,For the complexity of constricted regression model, overfitting problem is avoided；Constant C is
Weight determines balance of the objective function between fitting degree and function complexity；ξiFor slack variable,Certainly
The fitting degree for having determined data is calculated by ε-loose penalty function, is indicated are as follows:
4. Flood of small drainage area forecasting procedure according to claim 3, which is characterized in that in step (3), training can be passed through
The hyper parameter that data are adjusted refers to: the ε in coefficient gamma, constraint constant C and penalty function.
5. Flood of small drainage area forecasting procedure according to claim 1, which is characterized in that in step (3), k folding interaction is tested
Demonstration, which refers to, is divided into k subset for rainfall and runoff training data, takes and is wherein used as validation group for one group, remaining k-1 group is as survey
Examination collection, obtains extensive error；After repeating k times, the average value of k extensive error is taken.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910514384.1A CN110298498A (en) | 2019-06-14 | 2019-06-14 | A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910514384.1A CN110298498A (en) | 2019-06-14 | 2019-06-14 | A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110298498A true CN110298498A (en) | 2019-10-01 |
Family
ID=68028101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910514384.1A Pending CN110298498A (en) | 2019-06-14 | 2019-06-14 | A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110298498A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111027764A (en) * | 2019-12-06 | 2020-04-17 | 中国水利水电科学研究院 | Flood forecasting method suitable for runoff data lack basin based on machine learning |
| CN112434470A (en) * | 2020-12-03 | 2021-03-02 | 中国电建集团华东勘测设计研究院有限公司 | Riverway diversion port door water level data extension method, device, storage medium and equipment |
| CN112684519A (en) * | 2020-12-30 | 2021-04-20 | 北京墨迹风云科技股份有限公司 | Weather forecasting method and device, computer equipment and storage medium |
| CN113780585A (en) * | 2021-11-12 | 2021-12-10 | 江苏铨铨信息科技有限公司 | Convection cloud machine learning identification method based on satellite cloud picture |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740969A (en) * | 2016-01-21 | 2016-07-06 | 水利部交通运输部国家能源局南京水利科学研究院 | Data-driven small watershed real-time flood forecast method |
| CN108921345A (en) * | 2018-06-28 | 2018-11-30 | 杭州市水文水资源监测总站 | The river flood-peak stage Real-time Forecasting Method that precipitation forecast is coupled with support vector machines |
-
2019
- 2019-06-14 CN CN201910514384.1A patent/CN110298498A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105740969A (en) * | 2016-01-21 | 2016-07-06 | 水利部交通运输部国家能源局南京水利科学研究院 | Data-driven small watershed real-time flood forecast method |
| CN108921345A (en) * | 2018-06-28 | 2018-11-30 | 杭州市水文水资源监测总站 | The river flood-peak stage Real-time Forecasting Method that precipitation forecast is coupled with support vector machines |
Non-Patent Citations (2)
| Title |
|---|
| 吴佳文: "水文时间序列数据挖掘算法研究与应用", 《中国博士学位论文全文数据库 基础科学辑》 * |
| 尼玛旦增 等: "基于ν-SVR的雅鲁藏布江羊村站洪水预报模型", 《人民长江》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111027764A (en) * | 2019-12-06 | 2020-04-17 | 中国水利水电科学研究院 | Flood forecasting method suitable for runoff data lack basin based on machine learning |
| CN112434470A (en) * | 2020-12-03 | 2021-03-02 | 中国电建集团华东勘测设计研究院有限公司 | Riverway diversion port door water level data extension method, device, storage medium and equipment |
| CN112434470B (en) * | 2020-12-03 | 2022-05-31 | 中国电建集团华东勘测设计研究院有限公司 | River channel diversion port door water level data extension method and device, storage medium and equipment |
| CN112684519A (en) * | 2020-12-30 | 2021-04-20 | 北京墨迹风云科技股份有限公司 | Weather forecasting method and device, computer equipment and storage medium |
| CN112684519B (en) * | 2020-12-30 | 2022-07-05 | 北京墨迹风云科技股份有限公司 | Weather forecasting method and device, computer equipment and storage medium |
| CN113780585A (en) * | 2021-11-12 | 2021-12-10 | 江苏铨铨信息科技有限公司 | Convection cloud machine learning identification method based on satellite cloud picture |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110298498A (en) | A kind of Flood of small drainage area forecasting procedure for establishing data-driven with support vector machines | |
| Campolo et al. | Artificial neural network approach to flood forecasting in the River Arno | |
| Choubin et al. | Drought forecasting in a semi-arid watershed using climate signals: a neuro-fuzzy modeling approach | |
| Liong et al. | River stage forecasting in Bangladesh: neural network approach | |
| Kuok et al. | Particle swarm optimization feedforward neural network for modeling runoff | |
| CN103729550B (en) | Multiple-model integration Flood Forecasting Method based on propagation time cluster analysis | |
| Zhang et al. | Predicting hydrological signatures in ungauged catchments using spatial interpolation, index model, and rainfall–runoff modelling | |
| CN115375198B (en) | Method and system for communication joint scheduling and water quality safety guarantee of regional river and lake water systems | |
| CN106598918A (en) | Non-uniform designed flood calculation method based on quantile regression | |
| CN110334851A (en) | A kind of mixed connection step reservoir joint Flood Optimal Scheduling method that consideration divides flood storage people Wan to use | |
| Wei et al. | Multireservoir flood-control optimization with neural-based linear channel level routing under tidal effects | |
| Liu et al. | Long Short-Term Memory (LSTM) Based Model for Flood Forecasting in Xiangjiang River | |
| Tshimanga | Two decades of hydrologic modeling and predictions in the Congo River Basin: Progress and prospect for future investigations | |
| Soomlek et al. | Using backpropagation neural networks for flood forecasting in PhraNakhon Si Ayutthaya, Thailand | |
| Mandal et al. | Prediction of Wind Speed using Machine Learning | |
| Al-Fawa’Reh et al. | Intelligent methods for flood forecasting in Wadi al Wala, Jordan | |
| Li et al. | Long-lead daily streamflow forecasting using Long Short-Term Memory model with different predictors | |
| Chang et al. | Investigating the impact of the Chi‐Chi earthquake on the occurrence of debris flows using artificial neural networks | |
| Tretiakov et al. | Assessment of influences of anthropogenic and climatic changes in the drainage basin on hydrological processes in the Gulf of Ob | |
| Ishfaque et al. | Trend analysis of hydro-climatological parameters and assessment of climate impact on dam seepage using statistical and machine learning models | |
| Ceribasi et al. | Estimation of energy to be produced in hydroelectric power plants by using artificial neural networks and innovative sen method | |
| KESHTA | A MULTI-COMPONENT MODEL FOR LONG-TERM RIVER FLOW FORECASTING | |
| Panin et al. | Current variations in the wind speed vector and the rate of evaporation from the Caspian Sea surface | |
| Altayyar | ANFIS Model for Forecasting Storm Surge in New York City Using Hydro-Climatological Data | |
| Razavi | Streamflow estimation in ungauged basins using regionalization methods |
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: 20191001 |
|
| RJ01 | Rejection of invention patent application after publication |