CN106127242A - Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof - Google Patents
Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof Download PDFInfo
- Publication number
- CN106127242A CN106127242A CN201610444365.2A CN201610444365A CN106127242A CN 106127242 A CN106127242 A CN 106127242A CN 201610444365 A CN201610444365 A CN 201610444365A CN 106127242 A CN106127242 A CN 106127242A
- Authority
- CN
- China
- Prior art keywords
- data
- year
- module
- model
- training
- 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
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
Abstract
The present invention discloses a kind of year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof, including data input module, model construction module, training and debugging module, integration module and data outputting module.Data input module comprises and reads in the former data of discharge site website from database file and meteorological master data is gone forward side by side line number Data preprocess;Model construction module includes that obtaining training dataset is and builds multidimensional SVM Extreme Precipitation forecast model;Training and debugging module determine training data and inspection data, are trained multidimensional SVM model and debug, determining the parameter of each model;The average relative error of the predictive value that integration module calculates each model determines weight, carries out based on D S evidence theory integrated;Extreme Precipitation predictive value in the coming year is stored in data base or file by data outputting module, it is provided that inquiry and Analysis Service.
Description
Technical field
The present invention relates to Prediction of Precipitation technology, be specifically related to a kind of year of based on integrated study Extreme Precipitation prognoses system and
Its Forecasting Methodology.
Background technology
Along with global warming, frequency and intensity that extreme weather events occurs all increased, relevant meteorological calamity
Evil is also increasing, and natural environment and social economy are all caused serious impact, therefore to Extreme Precipitation by extreme precipitation event
Prediction, can successfully manage climate change, and the defense work to natural disaster also has pivotal role.
Extreme precipitation event refers to for a specific place and time, it average of the state substantial deviation of precipitation
State, it is common that refer to that decades one meet the most a-hundred-year small probability event.The most generally use percentile definition extreme
Precipitation threshold value, exceeding this threshold value is considered as extreme value, and this event may be considered extreme precipitation event, and then threshold is passed through in calculating
The precipitation of value or number of days etc., be analyzed extreme precipitation event inquiring into.
The number of times that Extreme Precipitation occurs is relevant with many time-varying factors, such as weather, population etc., although extremely dropping certain
Water event has a randomness, but to a region, in a period of time, Extreme Precipitation has certain regularity.Cause
This, it may be considered that the various factors of Extreme Precipitation, by Extreme Precipitation seasonal effect in time series is analyzed, set up year Extreme Precipitation
Forecast model.
Aspect abroad, a lot of seminar define corresponding extreme climate index.Wherein ETCCDMI (Expert Team
On Climate Change Detection Monitoring and Indices) by (two-pronged) of double forks
Approach application has arrived in the research of extreme weather events, defined 27 about extreme temperature and the index of Extreme Precipitation,
The research of worldwide extreme climate is widely applied.Alexander etc. utilize these index analysis global all
The extreme temperature in area and the situation of change of precipitation, it is provided that the global situation of the most comprehensive extreme weather events variation characteristic.
Karl etc. utilize percentage threshold to define the Extreme precipitation index of the U.S., and by the Extreme Precipitation of a series of exponent pair U.S.
Event is made that to be researched and analysed.Aspect at home, Zhai Pan Mao etc. is defined as extreme precipitation event by the 95% of intra day ward sequence
Threshold value, found by trend analysis, in the second half in 20th century, the generation frequency of extreme precipitation event is western and the Changjiang river in China
In The Middle And Lower Reaches dramatically increases, and substantially reduces in North China.5 poles that STATDEX research project that what chapter was firm utilize proposes
Jiangxi Province's Daily rainfall amount is calculated by end Precipitation Index, utilizes method of least square that each index is made trend analysis, and utilizes
Kendall-tau method inspection linear trend significance, result shows, the Extreme precipitation index of the most of station in Jiangxi Province in
Ascendant trend, and all indexes the nineties in 20th century change significantly increase.
Although recent domestic scholar has carried out a lot of research to extreme precipitation event, observing, simulate, estimate etc. each
Aspect has been achieved for remarkable progress, but yet suffers from some shortcomings in method and content.
Research great majority to extreme precipitation event are all from meteorological angle at present, with sunykatuib analysis and qualitative analysis
The research carried out for main method.It sets up statistical relationship according to existing precipitation data, it is impossible to ensure following precipitation shape
State is equally applicable, Extreme Precipitation is simulated analysis and has certain limitation, and most research is primarily upon going through
History precipitation data is observed, and the considering of various factors for Extreme Precipitation is the most relatively short of.At present by data mining skill
The work that art is applied to the Extreme Precipitation mutation analysis in hydrology field is also little, and research method is also few, currently mainly has the time
The method such as sequence analysis, artificial neural network.And, owing to data information is limited, method is incorrect, model imperfection in research
Etc. reason, cause the problems such as the predictive ability to Extreme Precipitation is strong not, precision of prediction is the highest..
Summary of the invention
Goal of the invention: it is an object of the invention to solve the deficiencies in the prior art, it is provided that a kind of based on integrated
The year Extreme Precipitation prognoses system and Forecasting Methodology thereof practised, by the correlation technique of data mining, is carried out extreme precipitation event
Quantitative analysis and predictive study.
Technical scheme: a kind of based on integrated study year Extreme Precipitation prognoses system of the present invention, including be sequentially connected with
Data input module (100), model construction module (200), training and debugging module (300), integration module (400) and data are defeated
Go out module (500);Wherein:
Described data input module (100) reads in the former data of discharge site website and ground from data base or EXCEL file
Weather station, face master data, and the data read in are carried out process include carrying out intra day ward and year meteorological data at missing values
Reason, calculates year Extreme Precipitation intensity by the intra day ward of website, is standardized all of data processing;
Described model construction module (200) obtains training dataset i.e. meteorological data feature from data input module (100)
With year extreme precipitation, then choose different Meteorological Characteristics data as input variable, year extreme precipitation become as output
Amount, builds multidimensional SVM Extreme Precipitation forecast model;
Described training and debugging module (300) determine training data and inspection data, utilize training several to multidimensional SVM model
It is trained and debugs, determining the parameter of each model, and utilizing inspection data that integrated SVM model is predicted precision and extensive
Ability test;
Described integration module (400), after training and debugging module complete work, obtains multiple prediction data and calculating changes
The average relative error of multiple prediction data, then calculates the credibility of each value, carries out finally according to D-S theory composition rule
Synthesis, obtains the predictive value of integrated model;
Extreme Precipitation predictive value in the coming year is stored in data base or file by described data outputting module (500), for user
Check and analyze.
The invention also discloses the Forecasting Methodology of a kind of year of based on integrated study Extreme Precipitation prognoses system, include successively
Following steps:
I, described data input module (100) read in from data base or EXCEL file the former data of discharge site website and
Surface weather station's master data, and the data read in are extracted, clean and verified, and process shortage of data value etc.,
To normalized time series data;Then, check, add up and analyze the time series data after cleaning, then extreme to year
Precipitation and year-climate data characteristics data carry out pretreatment;
II, model construction module (200) obtains training dataset i.e. meteorological data feature and year extreme precipitation, then selects
Taking different Meteorological Characteristics data as input variable, year, extreme precipitation was as output variable, builds multidimensional SVM and extremely drops
Water forecast model;
III, training and debugging module (300) determine training data and inspection data, utilize training several to multidimensional SVM model
It is trained and debugs, determining the parameter of each model, and utilizing inspection data that integrated SVM model is predicted precision and extensive
Ability test;
IV, integration module (400) is after step III terminates, and calculates average relative error d of multiple predictive valuei, then count
Calculate the credibility of each value:
c(mi) it is predictive value miWeight, miFor predicting the outcome of model i, diFor average relative error, finally according to D-S
Theoretical composition rule synthesizes, and last composite value M synthesizes according to the following rules:
V, Extreme Precipitation predictive value in the coming year is stored in data base or file by data outputting module (500), looks into for user
See and analyze.
Further, in step II, model construction module step is as follows:
A, in step I year extreme precipitation and meteorological characteristic be standardized processing;
B, structure multidimensional SVM Extreme Precipitation forecast model: from training set, produce n training subset, at n by method of bootstrapping
Using year meteorological data as input vector in individual training subset, year, extreme precipitation was as output vector, constructed multidimensional SVM pole
End Precipitation Forecasting Model.
Further, the training described in step III and debugging forecast model, carry out in accordance with the following steps:
A, the parameter in multidimensional SVM model is encoded, by binary string encoding mechanism, object of study is converted into
The string structure being made up of in certain sequence special symbol;
B, according to the SVM fitness function evaluation to training of the object function i.e. SVM training result;
C, genetic manipulation, i.e. select, intersect and make a variation, the defect individual that in selected population, fitness is strong, two parents
Individual a part of structure is replaced mutually, generates new individuality, and some individual genic value is entered row stochastic change, produces new
Individuality, make individual corresponding solution Step wise approximation globally optimal solution;
D, judge whether to reach maximum evolutionary generation, if it is, draw the model parameter of optimum, if it is not, then forward to
Step C.
Further, the integration module described in step IV, carry out in accordance with the following steps:
A, integration module (400), after step III terminates, calculate average relative error d of multiple predictive valuei;
B, calculate the credibility of each predictive value:
Synthesize finally according to D-S theory composition rule.
Beneficial effect: the present invention for discharge site year Extreme Precipitation time series data prediction, by integrated study
The support vector machine composition improved with genetic algorithm, leads to Hydrological Time Series with meteorological data characteristic data set in integrated study
Cross the different types of subset of method divide into several classes, then build same amount of SVM prediction model, calculated by heredity
Method is to each prediction model parameters optimizing, and the result of last gained asks D-S fusion value to utilize integrated study at same data set
The study module that upper use is different, can help to reduce variance, thus obtain more accurate result.
Owing to Extreme Precipitation receives impact and the restriction of many factors, conventional can not reach merely with history extreme precipitation
Preferably prediction effect, compares with other Forecasting Methodology, and the present invention has higher prediction accuracy, and preferably can expand
Malleability and practical value.
Accompanying drawing explanation
Fig. 1 is the overall structure schematic diagram of prognoses system in the present invention;
Fig. 2 is the structural representation of integrated study in the present invention;
Fig. 3 is the workflow diagram of integration module in the present invention;
Fig. 4 is the overall workflow figure of the present invention;
Fig. 5 is the detailed operational flow diagrams of invention.
Detailed description of the invention
Below technical solution of the present invention is described in detail, but protection scope of the present invention is not limited to described enforcement
Example.
As it is shown in figure 1, year of based on integrated study Extreme Precipitation prognoses system of invention, including data input module 100,
Model construction module 200, training and debugging module 300, integration module 400, data outputting module 500.
Wherein, data input module 100 is extreme for obtaining Jian County, Jiangxi Province from data base or other data files
Precipitation and meteorological data include that Extreme Precipitation intensity line number Data preprocess of going forward side by side will zoom in and out process by initial data.Table
1 is Jian County 1951-2013 Extreme Precipitation data, and table 2 is Jian County 1951-2013 meteorological data, including 24 class gas
As factor.
Table 1 Jian County 1951-2013 Extreme Precipitation data
Time | 90% tantile | Extreme Precipitation number of days | Extreme Precipitation total amount | Extreme Precipitation intensity |
1951 | 10.5 | 64 | 1491.6 | 23.3062 |
1951 | 10.7 | 148 | 3846.4 | 25.9891 |
1953 | 18.3 | 148 | 5418.8 | 36.6135 |
… | … | … | … | … |
… | … | … | … | … |
2011 | 8.5 | 37 | 674.4 | 18.2270 |
2012 | 15.1 | 37 | 1203.8 | 32.5351 |
2013 | 10.8 | 37 | 1211.3 | 32.7378 |
Table 2 Jian County 1951-2013 meteorological data
Above-mentioned initial data is standardized,xi' for the data after standardization, xiFor former data,For average, s is standard deviation, and the data after standardization are shown in Table 3.
Table 3 standardized data
Time | Year Extreme Precipitation intensity | Minimum relative humidity | Extreme Precipitation total amount | Year Extreme Precipitation |
1951 | -1.2003 | -0.4920 | … | -7.07439 |
1951 | 1.3290 | -0.5078 | … | -0.1009 |
1953 | 2.4467 | -0.4604 | … | 0.8124 |
… | … | … | … | … |
… | … | … | … | … |
2011 | -0.9260 | -0.5473 | … | -0.7984 |
2012 | -0.5496 | -0.5315 | … | 0.4693 |
2013 | -0.5443 | -0.5631 | … | -0.0084 |
Described model construction module 200 includes obtaining training dataset i.e. by integrated study support vector machine, by data
Collection divide into several classes, then chooses different Meteorological Characteristics data as input variable, and year, extreme precipitation was as output variable,
Build multidimensional SVM Extreme Precipitation forecast model.
Here with precipitation, average water air pressure, maximum daily precipitation three class meteorological factor sets up SVM as input vectoriModel
As a example by, shown in input and output vector structure table 4.
Table 4 input output vector structure design
Described training and debugging module 300 determine training data and inspection data, and utilization training is several enters multidimensional SVM model
Row training and debugging, determine the parameter of each model, and utilize inspection data that integrated SVM model is predicted precision and extensive energy
Force inspecting.
Genetic algorithm is utilized to obtain the SVM optimizediModel parameter combines: c=30.27, g=2.95, p=0.0434, c are
Penalty coefficient, it is for the complexity of balance model and empiric risk, and g is the parameter of RBF kernel function, and p is insensitive loss letter
The parameter of number.Described integration module 400 draws different weights to the multiple prediction data obtained according to the quality of prediction effect,
Obtain D-S fusion value, using D-S fusion value as the predictive value of integrated model.
Extreme Precipitation predictive value in the coming year is stored in data base or file by described data outputting module 500, looks into for user
See and analyze.
As in figure 2 it is shown, the structural representation of integrated study in Fa Ming, carry out as follows:
A, given data set comprise n sample.This data set is sampled d time with putting back to, and produces the training of d sample
Collection, the sample not entering into this training set in former data sample ultimately forms inspection set (test set);
B, training set is input in integrated each grader so that training data used in grader is all
The subset of former training data;
Grader, each grader output predictive value is trained on C, respectively test set sample after these are sampled;
D, the basic classification device obtained based on different characteristic subset is integrated.
As it is shown on figure 3, the workflow of integration module is followed successively by the present invention:
From input module, obtain training dataset, i.e. year extreme precipitation and meteorological dataset, then carry out feature choosing
Selecting, by bootstrapping, method obtains data subset 1~n;By data set divide into several classes, each data set input SVM is predicted mould
In type, then choosing different Meteorological Characteristics data as input variable, year, extreme precipitation was as output variable, built multidimensional
SVM Extreme Precipitation forecast model;Utilize genetic algorithm to carry out parameter optimization on each SVM model, and model parameter is entered
Row sum-equal matrix selects the model parameter of optimum;Effect quality according to SVM model prediction provides different weights, then asks D-S to melt
Conjunction value, is then worth incoming output module by D-S fusion.
As shown in Figure 4, the present invention year of based on integrated study Extreme Precipitation prognoses system groundwork step be:
By in the intra day ward of discharge site and the incoming input module of Meteorological Characteristics key element of weather station, and data are entered
Row pretreatment;
Obtaining training dataset from data input module, i.e. year extreme precipitation and meteorological dataset, then count
According to collection, it is thus achieved that different data subsets;By data set divide into several classes, each data set is inputted in SVM forecast model, so
After choose different Meteorological Characteristics data as input variable, year, extreme precipitation was as output variable, built multidimensional SVM pole
End Precipitation Forecasting Model;Genetic algorithm is utilized to carry out parameter optimization on each SVM model;Effect according to SVM model prediction
Quality provides different weights, then seeks D-S fusion value, then D-S fusion is worth incoming output module.
As it is shown in figure 5, invention year of based on integrated study Extreme Precipitation prognoses system detailed operation flow process be:
Obtaining training dataset from data input module, i.e. year extreme precipitation and meteorological dataset, then count
Select according to collection, obtain data subset 1~n by Bootstrap sampling algorithm;By data set divide into several classes, by each
In individual data set input SVM forecast model, then choose different Meteorological Characteristics data as input variable, year extreme precipitation
As output variable, build multidimensional SVM Extreme Precipitation forecast model;Genetic algorithm is utilized to join on each SVM model
Number optimizing, encodes the parameter in multidimensional SVM model, to the SVM fitness function evaluation trained, selects, intersects, becomes
Different, select the model parameter of optimum;Effect according to SVM model prediction determines weight, finally draws pre-according to D-S composition rule
Survey result.
Claims (5)
1. one kind year of based on integrated study Extreme Precipitation prognoses system, it is characterised in that include the data input being sequentially connected with
Module (100), model construction module (200), training and debugging module (300), integration module (400) and data outputting module
(500), wherein,
Described data input module (100) reads in the former data of discharge site website and ground gas from data base or EXCEL file
As station master data, and carry out intra day ward and year meteorological data are carried out missing values process to the data read in, pass through website
Intra day ward calculate year Extreme Precipitation intensity, all of data are standardized process and obtain normalized data;
Described model construction module (200) obtains training dataset i.e. meteorological data feature and year from data input module (100)
Extreme precipitation, then chooses different Meteorological Characteristics data as input variable, and year, extreme precipitation was as output variable, structure
Build multidimensional SVM Extreme Precipitation forecast model;
Described training and debugging module (300) determine training data and inspection data, and utilization training is several to be carried out multidimensional SVM model
Training and debugging, determine the parameter of each model, and utilize inspection data that integrated SVM model is predicted precision and generalization ability
Inspection;
Described integration module (400), after training and debugging module complete work, obtains multiple prediction data and calculating changes multiple
The average relative error of prediction data, then calculates the credibility of each value, closes finally according to D-S theory composition rule
Become, obtain the predictive value of integrated model;
Extreme Precipitation predictive value in the coming year is stored in data base or file by described data outputting module (500), checks for user
And analysis.
2. a Forecasting Methodology for year of based on integrated study according to claim 1 Extreme Precipitation prognoses system, it is special
Levy and be: comprise the following steps successively:
I, data input module (100) reads in the former data of discharge site website and Ground Meteorological from data base or EXCEL file
Stand master data, and the data read in are extracted, clean and verified, and process shortage of data value etc., standardized
Time series data;Then, check, add up and analyze clean after time series data, then to year extreme precipitation with
And year-climate data characteristics data carry out pretreatment;
II, model construction module (200) obtains training dataset i.e. meteorological data feature and year extreme precipitation, then chooses not
Same Meteorological Characteristics data are as input variable, and year, extreme precipitation was as output variable, built multidimensional SVM Extreme Precipitation pre-
Survey model;
III, training and debugging module (300) determine training data and inspection data, and utilization training is several to be carried out multidimensional SVM model
Training and debugging, determine the parameter of each model, and utilize inspection data that integrated SVM model is predicted precision and generalization ability
Inspection;
IV, integration module (400) is after step III terminates, and calculates average relative error d of multiple predictive valuei, then calculate each
The credibility of value:
c(mi) it is predictive value miWeight, miFor predicting the outcome of model i, diFor average relative error, theoretical finally according to D-S
Composition rule synthesizes, and last composite value M synthesizes according to the following rules:
V, Extreme Precipitation predictive value in the coming year is stored in data base or file by data outputting module (500), check for user and
Analyze.
Year of based on integrated study the most according to claim 2 Extreme Precipitation prognoses system Forecasting Methodology, its feature exists
In, in step II, model construction module step is as follows:
A, in step I year extreme precipitation and meteorological characteristic be standardized processing;
B, structure multidimensional SVM Extreme Precipitation forecast model: from training set, produce n training subset, n instruction by method of bootstrapping
Using year meteorological data as input vector in white silk subset, year, extreme precipitation was as output vector, constructs multidimensional SVM and extremely drops
Water forecast model.
Year of based on integrated study the most according to claim 2 Extreme Precipitation prognoses system Forecasting Methodology, its feature exists
In, the training described in step III and debugging forecast model, carry out in accordance with the following steps:
A, the parameter in multidimensional SVM model is encoded, by binary string encoding mechanism, object of study is converted into by spy
Determine the string structure that symbol forms in certain sequence;
B, according to the SVM fitness function evaluation to training of the object function i.e. SVM training result;
C, genetic manipulation, i.e. select, intersect and make a variation, the defect individual that in selected population, fitness is strong, two parent individualities
A part of structure replace mutually, generate new individuality, some individual genic value entered row stochastic change, produce new
Body, makes individual corresponding solution Step wise approximation globally optimal solution;
D, judge whether to reach maximum evolutionary generation, if it is, draw the model parameter of optimum, if it is not, then forward step to
C。
Year of based on integrated study the most according to claim 1 Extreme Precipitation prognoses system Forecasting Methodology, its feature exists
In, the integration module described in step IV, carry out in accordance with the following steps:
A, integration module (400), after step III terminates, calculate average relative error d of multiple predictive valuei;
B, calculate the credibility of each predictive value:
Synthesize finally according to D-S theory composition rule.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610444365.2A CN106127242A (en) | 2016-06-21 | 2016-06-21 | Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610444365.2A CN106127242A (en) | 2016-06-21 | 2016-06-21 | Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106127242A true CN106127242A (en) | 2016-11-16 |
Family
ID=57470959
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610444365.2A Pending CN106127242A (en) | 2016-06-21 | 2016-06-21 | Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106127242A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108123834A (en) * | 2017-12-18 | 2018-06-05 | 佛山市米良仓科技有限公司 | Log analysis system based on big data platform |
CN108375808A (en) * | 2018-03-12 | 2018-08-07 | 南京恩瑞特实业有限公司 | Dense fog forecasting procedures of the NRIET based on machine learning |
CN108595814A (en) * | 2018-04-18 | 2018-09-28 | 北京大学 | A kind of precipitation generator based on day time scale |
CN109508810A (en) * | 2018-09-29 | 2019-03-22 | 天津大学 | A kind of system based on realization monthly average hydrology volume forecasting |
CN109614744A (en) * | 2018-12-28 | 2019-04-12 | 华东交通大学 | A kind of precipitation quantity measuring method and system based on big data |
CN110346844A (en) * | 2019-07-15 | 2019-10-18 | 南京恩瑞特实业有限公司 | Quantitative Precipitation estimating and measuring method of the NRIET based on cloud classification and machine learning |
CN111473777A (en) * | 2020-03-06 | 2020-07-31 | 李�杰 | Hydrological monitoring system and method |
CN112800634A (en) * | 2021-04-07 | 2021-05-14 | 水利部交通运输部国家能源局南京水利科学研究院 | Rainfall estimation method and system coupling dry-wet state identification and multi-source information fusion |
CN113361766A (en) * | 2021-06-03 | 2021-09-07 | 南京信息工程大学 | Multi-mode rainfall estimation method integrating machine learning |
CN114754973A (en) * | 2022-05-23 | 2022-07-15 | 中国航空工业集团公司哈尔滨空气动力研究所 | Wind tunnel force measurement test data intelligent diagnosis and analysis method based on machine learning |
CN111538935B (en) * | 2019-12-26 | 2023-08-25 | 北京玖天气象科技有限公司 | Fine precipitation fusion method, system, electronic equipment and storage medium based on terrain features and multi-source mode products |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102760210A (en) * | 2012-06-19 | 2012-10-31 | 南京理工大学常熟研究院有限公司 | Adenosine triphosphate binding site predicting method for protein |
CN102880755A (en) * | 2012-09-25 | 2013-01-16 | 河海大学 | Method and system for quantitatively forecasting extreme rainfall |
CN103077347A (en) * | 2012-12-21 | 2013-05-01 | 中国电力科学研究院 | Combined type intrusion detecting method on basis of data fusion of improved core vector machine |
-
2016
- 2016-06-21 CN CN201610444365.2A patent/CN106127242A/en active Pending
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102760210A (en) * | 2012-06-19 | 2012-10-31 | 南京理工大学常熟研究院有限公司 | Adenosine triphosphate binding site predicting method for protein |
CN102880755A (en) * | 2012-09-25 | 2013-01-16 | 河海大学 | Method and system for quantitatively forecasting extreme rainfall |
CN103077347A (en) * | 2012-12-21 | 2013-05-01 | 中国电力科学研究院 | Combined type intrusion detecting method on basis of data fusion of improved core vector machine |
Non-Patent Citations (4)
Title |
---|
NAYAK MA ET AL: "《Prediction of extreme rainfall event using weather pattern Recognition and support vector machine classifier》", 《THEORETICAL & APPLIED CLIMATOLOGY》 * |
唐娟等: "《一种极端降水预测方法研究》", 《信息技术》 * |
李干琼等: "《SV因子分析框架下的农产品市场短期预测》", 31 August 2013, 北京:中国经济出版社 * |
樊继慧: "《基于SVM组合分类器的局域气象预测研究》", 《中国优秀硕士学位论文全文数据库基础科学辑》 * |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108123834A (en) * | 2017-12-18 | 2018-06-05 | 佛山市米良仓科技有限公司 | Log analysis system based on big data platform |
CN108375808A (en) * | 2018-03-12 | 2018-08-07 | 南京恩瑞特实业有限公司 | Dense fog forecasting procedures of the NRIET based on machine learning |
CN108595814A (en) * | 2018-04-18 | 2018-09-28 | 北京大学 | A kind of precipitation generator based on day time scale |
CN108595814B (en) * | 2018-04-18 | 2021-06-04 | 北京大学 | Method for simulating daily rainfall by using measured perennial rainfall data |
CN109508810A (en) * | 2018-09-29 | 2019-03-22 | 天津大学 | A kind of system based on realization monthly average hydrology volume forecasting |
CN109614744B (en) * | 2018-12-28 | 2022-11-22 | 浙江理工大学 | Big data-based precipitation detection method and system |
CN109614744A (en) * | 2018-12-28 | 2019-04-12 | 华东交通大学 | A kind of precipitation quantity measuring method and system based on big data |
CN110346844A (en) * | 2019-07-15 | 2019-10-18 | 南京恩瑞特实业有限公司 | Quantitative Precipitation estimating and measuring method of the NRIET based on cloud classification and machine learning |
CN111538935B (en) * | 2019-12-26 | 2023-08-25 | 北京玖天气象科技有限公司 | Fine precipitation fusion method, system, electronic equipment and storage medium based on terrain features and multi-source mode products |
CN111473777A (en) * | 2020-03-06 | 2020-07-31 | 李�杰 | Hydrological monitoring system and method |
CN112800634B (en) * | 2021-04-07 | 2021-06-25 | 水利部交通运输部国家能源局南京水利科学研究院 | Rainfall estimation method and system coupling dry-wet state identification and multi-source information fusion |
CN112800634A (en) * | 2021-04-07 | 2021-05-14 | 水利部交通运输部国家能源局南京水利科学研究院 | Rainfall estimation method and system coupling dry-wet state identification and multi-source information fusion |
CN113361766A (en) * | 2021-06-03 | 2021-09-07 | 南京信息工程大学 | Multi-mode rainfall estimation method integrating machine learning |
CN113361766B (en) * | 2021-06-03 | 2024-04-09 | 南京信息工程大学 | Multi-mode precipitation prediction method integrated with machine learning |
CN114754973A (en) * | 2022-05-23 | 2022-07-15 | 中国航空工业集团公司哈尔滨空气动力研究所 | Wind tunnel force measurement test data intelligent diagnosis and analysis method based on machine learning |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106127242A (en) | Year of based on integrated study Extreme Precipitation prognoses system and Forecasting Methodology thereof | |
CN106650767B (en) | Flood forecasting method based on cluster analysis and real-time correction | |
CN108009674A (en) | Air PM2.5 concentration prediction methods based on CNN and LSTM fused neural networks | |
CN112506990B (en) | Hydrological data anomaly detection method based on spatiotemporal information | |
Han et al. | Intelligent decision model of road maintenance based on improved weight random forest algorithm | |
CN108733631A (en) | A kind of data assessment method, apparatus, terminal device and storage medium | |
CN104408153A (en) | Short text hash learning method based on multi-granularity topic models | |
CN104835103A (en) | Mobile network health evaluation method based on neural network and fuzzy comprehensive evaluation | |
CN106067094A (en) | A kind of dynamic assessment method and system | |
CN106960358A (en) | A kind of financial fraud behavior based on rural area electronic commerce big data deep learning quantifies detecting system | |
CN109146144A (en) | Based on the landslide disaster prediction technique for improving Bayesian network | |
CN107563567A (en) | Core extreme learning machine Flood Forecasting Method based on sparse own coding | |
CN110059852A (en) | A kind of stock yield prediction technique based on improvement random forests algorithm | |
CN104915805A (en) | Evaluation method for river ecological restoration effect | |
CN108874959A (en) | A kind of user's dynamic interest model method for building up based on big data technology | |
CN105678481A (en) | Pipeline health state assessment method based on random forest model | |
CN104881689A (en) | Method and system for multi-label active learning classification | |
Li et al. | Research and application of random forest model in mining automobile insurance fraud | |
CN107463993A (en) | Medium-and Long-Term Runoff Forecasting method based on mutual information core principle component analysis Elman networks | |
CN112735097A (en) | Regional landslide early warning method and system | |
CN109947652A (en) | A kind of improvement sequence learning method of software defect prediction | |
CN116109195B (en) | Performance evaluation method and system based on graph convolution neural network | |
CN108090515A (en) | A kind of environmental rating appraisal procedure based on data fusion | |
CN110047015A (en) | A kind of water total amount prediction technique merging KPCA and thinking Optimized BP Neural Network | |
CN104835073A (en) | Unmanned aerial vehicle control system operation performance evaluating method based on intuitionistic fuzzy entropy weight |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161116 |