CN109960882A - Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency - Google Patents
Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency Download PDFInfo
- Publication number
- CN109960882A CN109960882A CN201910240549.0A CN201910240549A CN109960882A CN 109960882 A CN109960882 A CN 109960882A CN 201910240549 A CN201910240549 A CN 201910240549A CN 109960882 A CN109960882 A CN 109960882A
- Authority
- CN
- China
- Prior art keywords
- variable
- tropical cyclone
- forecast
- frequency
- characteristic value
- 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.)
- Withdrawn
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
- 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
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- General Health & Medical Sciences (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses method, apparatus, equipment and storage mediums that a kind of forecast tropical cyclone generates frequency, wherein method includes: that the generation frequency that generates of the characteristic value and tropical cyclone for variable there may be influence being generated on tropical cyclone carries out stepwise regression analysis, and it examines, if upchecked, retain variable;The variable of reservation is used, as predictand, to be forecast as characteristic value, the generation frequency of tropical cyclone generation using the method for neural network.Forecast tropical cyclone provided by the embodiment of the present invention generates frequency techniques, it is contemplated that the influence of atmosphere and ocean to tropical cyclone.It can not have to consider the mapping relations between characteristic value and predictand, generate the non-linear relation between weather ambient field without the concern for tropical cyclone.It realizes optimizing forecast model, improve accuracy of the forecast.
Description
Technical field
The present invention relates to field of computer technology, the in particular to method of forecast tropical cyclone generation frequency;It further relates to pre-
Tropical cyclone is reported to generate device, equipment and the computer readable storage medium of frequency.
Background technique
There is many tropical cyclones (Tropical cyclone in the annual whole world;TC it) generates, northwest Pacific is tropical gas
One main generation source of rotation, stronger tropical cyclone is referred to as tropical cyclone in northwest Pacific.Tropical cyclone is
Occur in a kind of deepwater severe storm in the torrid zone, is one of natural calamity most strong on the earth, while China is also full generation
Boundary a few by tropical cyclones influence one of the countries with the most serious ....The typhoon of northwest Pacific is stepped in China southeastern coastal areas
Land, typhoon bring high wind, heavy rain and tide cause very big life threat and economic damage to the people of China's littoral area
It loses.
Due to the shortage of complexity and oceanographic observation data of the tropical cyclone during occurrence and development, at present about heat
Generting machanism with cyclone obtains a specific result not yet.Heat is based primarily upon about the generting machanism of tropical cyclone at present
The trigger mechanism that the Climatic field generated with cyclone and tropical cyclone generate.And the potential generation index of tropical cyclone is exactly
The one kind proposed on this basis can forecast that tropical cyclone generates the model of frequency.Therefore at present using the potential of tropical cyclone
Index (GPI) is generated to indicate that a possibility that tropical cyclone generates size has been obtained and be widely applied.Tropical cyclone is one
A nonlinear process.Present some GPI that can show tropical cyclone generation possibility size can not be shown completely
This nonlinear variation.
Based on closely related Climatic field is generated with tropical cyclone, Gray (1979) uses the environment of six kinds of large scales
Variable constitutes an index, the i.e. empirical index number of tropical cyclone generation, to indicate seasonal variations and the space point of TC generation
Cloth feature, but the index of Gray proposition is not particularly suited for the research of climate change.In the trigger mechanism WISHE of tropical cyclone
In, tropical cyclone is regarded as the heat engine of an acting, potential strength (the potential intensity of tropical cyclone;PI)
It is exactly the mechanical efficiency of heat engine.The power for comprehensively considering tropical cyclone is gained knowledge and the basic theories of Carnot cycle, Emanuel
With the potential strength of Nolan (2004) based on tropical cyclone, the wind speed of troposphere bottom, relative humidity and vertical wind shear are proposed
Potential generation index (the Genesis Potential Index of one tropical cyclone;GPI), extensive at present
Using, but can not output data frequency is generated to tropical cyclone the change in future of preferable use pattern predict.
Recently, it has been proposed that many to be answered using machine learning techniques are relevant to climatic model and Forecast Model For Weather
With including commonly using simple mould from high-resolution simulation middle school from the relationship between climate model study orbit parameter and climatic field
Type improves prediction, and help makes a policy under extreme weather conditions, the extreme weather that detection climatic data is concentrated, and predicts the weather
The uncertainty of forecast.The technology of all these propositions is all the valuable technology of climate science and meteorology.But their all purports
Certain information are being extracted from model, or the information in model is added in other models.
Tropical cyclone is a kind of nonlinear process, and existing GPI is a kind of specific exponential function form, only wherein
Several variables that relevant ocean and atmosphere are generated to tropical cyclone: sea surface temperature, vertical wind shear, relative humidity and opposite
Vorticity may ignore other elements for being possible to affecting China tropical cyclone physical process.
Currently, the potential generation index of tropical cyclone is all based on certain physical process, pass through log in statistical significance
A kind of index that the method for regression analysis obtains, be it is a kind of all known results are counted after as a result, and used
Exponential form all almost fix, it is not high for the accuracy rate of predictions for future.
Summary of the invention
The object of the present invention is to provide a kind of methods that forecast tropical cyclone generates frequency, realize optimizing forecast model, mention
High accuracy of the forecast;It is a further object of the present invention to provide device, equipment and calculating that forecast tropical cyclone generates frequency
Machine readable storage medium storing program for executing all has above-mentioned technical effect.
In order to solve the above technical problems, the present invention provides a kind of methods that forecast tropical cyclone generates frequency, comprising:
The generation frequency that the characteristic value that variable there may be influence is generated on tropical cyclone and tropical cyclone are generated into
Row stepwise regression analysis, and examine, if upchecked, retain variable;
The variable of reservation is used, as predictand, to use nerve net as characteristic value, the generation frequency of tropical cyclone generation
The method of network is forecast.
Preferably, the generation characteristic value that variable there may be influence is generated on tropical cyclone and tropical cyclone generated
Frequency carries out stepwise regression analysis, and examines, if upchecked, retains variable, includes:
It selects and maximum variable is contributed to predictand, be added in regression equation;
Continue to find out the maximum variable of contribution in remaining variable, be added in regression equation;
It tests to variable, if passed through, it is determined that the variable is added;If not over being added just now
Variable is rejected;
It repeats the above steps to remaining variable, until there is no factor rejecting, is also introduced without the factor.
Preferably, the variable of reservation is used, as predictand, to use as characteristic value, the generation frequency of tropical cyclone generation
The method of BP neural network is forecast;The characteristic value of the variable is in the input layer of BP neural network;The life of tropical cyclone
At frequency in output layer;Mapping relations between the characteristic value of variable and the generation frequency of tropical cyclone are in hidden layer.
It preferably, is that F is examined to the method that variable is tested.
In order to solve the above technical problems, including the present invention also provides the device that a kind of forecast tropical cyclone generates frequency:
Regression analysis module, for the characteristic value and tropical cyclone of the variable there may be influence will to be generated on tropical cyclone
The generation frequency of generation carries out stepwise regression analysis, and examines, if upchecked, retains variable;
Neural network module, the generation frequency conduct for using the variable of reservation to generate as characteristic value, tropical cyclone
Predictand is forecast using the method for BP neural network.
Preferably, the regression analysis module, for the feature of the variable there may be influence will to be generated on tropical cyclone
Value carries out stepwise regression analysis with the generation frequency that tropical cyclone generates, and examines, if upchecked, retains variable, wraps
Contain: selecting and maximum variable is contributed to predictand, be added in regression equation;Continue to find out contribution most in remaining variable
Big variable is added in regression equation;It tests to variable, if passed through, it is determined that the variable is added;If do not led to
It crosses, then the variable being added just now is rejected;It repeats the above steps to remaining variable, until there is no factor rejecting, also without the factor
It introduces.
Preferably, the neural network module, the life for using the variable of reservation to generate as characteristic value, tropical cyclone
At frequency as predictand, forecast using the method for BP neural network;The characteristic value of the variable is in BP neural network
In input layer;The generation frequency of tropical cyclone is in output layer;Between the characteristic value of variable and the generation frequency of tropical cyclone
Mapping relations are in hidden layer.
Preferably, regression analysis module is that F is examined to the method that variable is tested.
In order to solve the above technical problems, the present invention also provides a kind of equipment that forecast tropical cyclone generates frequency, comprising:
Memory, for storing computer program;
Processor realizes that forecast tropical cyclone as described in any one of the above embodiments generates when for executing the computer program
The step of method of frequency.
In order to solve the above technical problems, the computer can the present invention also provides a kind of computer readable storage medium
It reads storage medium and is stored with computer program, realized when the computer program is executed by processor such as any of the above-described forecast heat
The step of generating the method for frequency with cyclone.
The method that forecast tropical cyclone generates frequency provided by the embodiment of the present invention, including will generate to tropical cyclone can
The generation frequency that the characteristic value and tropical cyclone of the variable that can be had an impact generate carries out stepwise regression analysis, and examines, if
It upchecks, then retains variable;The method of the embodiment of the present invention on it is all may on the influential variable of tropical cyclone generation into
Row screening, filters out the variable of some needs, to eliminate the variable for causing multicollinearity, and reduces neural network meter
The workload of calculation.
The variable of reservation is used, as predictand, to use BP nerve as characteristic value, the generation frequency of tropical cyclone generation
The method of network is forecast.The method of the embodiment of the present invention can not have to consider that the mapping between characteristic value and predictand is closed
System generates the non-linear relation between weather ambient field without the concern for tropical cyclone.
As it can be seen that forecast tropical cyclone provided by the present invention generates frequency techniques, it is able to achieve optimizing forecast model, improves in advance
The accuracy of report.
Forecast tropical cyclone provided by the present invention generates device, equipment and the computer readable storage medium of frequency,
All have above-mentioned technical effect.
Detailed description of the invention
It to describe the technical solutions in the embodiments of the present invention more clearly, below will be to institute in the prior art and embodiment
Attached drawing to be used is needed to be briefly described, it should be apparent that, the accompanying drawings in the following description is only some implementations of the invention
Example, for those of ordinary skill in the art, without creative efforts, can also obtain according to these attached drawings
Obtain other attached drawings.
Fig. 1 is the flow diagram of the method for forecast TC Frequency provided by the embodiment of the present invention.
Fig. 2 is that the embodiment of the present invention forecasts that the method for TC Frequency carries out the flow diagram of regression analysis.
Fig. 3 is what the method that the embodiment of the present invention forecasts TC Frequency was forecast using the method for BP neural network
Flow diagram.
Fig. 4 is the schematic diagram of the device of forecast TC Frequency provided by the embodiment of the present invention.
Fig. 5 is the schematic diagram of the equipment of forecast TC Frequency provided by the embodiment of the present invention.
Specific embodiment
Core of the invention is to provide the method for forecast TC Frequency, realizes optimizing forecast model, improves forecast
Accuracy;Another core of the invention is to provide device, equipment and the computer-readable storage medium of forecast TC Frequency
Matter all has above-mentioned technical effect.
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
Referring to FIG. 1, Fig. 1 is the process signal of the method for forecast TC Frequency provided by the embodiment of the present invention
Figure;Method with reference to Fig. 1, the forecast TC Frequency includes:
The generation frequency that the characteristic value that variable there may be influence is generated on tropical cyclone and tropical cyclone are generated into
Row stepwise regression analysis, and examine, if upchecked, retain variable;
The variable of reservation is used, as predictand, to use BP nerve as characteristic value, the generation frequency of tropical cyclone generation
The method of network is forecast.
Specifically, generating tropical cyclone there may be the variable of influence, tropical cyclone is generated comprising atmosphere and ocean
There may be the variables of influence, make a variation comprising low frequency.In the present embodiment, generated for tropical cyclone of northwestern Pacific Ocean, i.e. typhoon
Generation, include six primary conditions, be segmented into three classes.It is thermal condition first, the sea mainly 1. relatively warmed up, it is desirable that
Water temperature below sea and sea at 60 meters is greater than 26 DEG C;2. potential instability;The two conditions provide for convection current
Heating power basis.3. followed by dynamic condition mainly should have a tropical disturbance in low latitude or in boundary layer, be right
The generation of stream provides most basic condition;4. position should be greater than 5 ° of latitude, that is, there is certain coriolis force, can make
Convection current rotates.It is finally environmental condition, mainly 5. tropospheric middle part needs to guarantee certain humidity, that is,
Relative humidity between 500hPa~700hPa is greater than 50%, generally all between 70%~80%;6. wind speed is vertically cut
Become smaller, if vertical shear is big, Warm-cored structure be not easily formed inside typhoon, it is however generally that, it is desirable that 200hPa and
The vertical shear of wind speed is less than 10m/s between 850hPa.
The present embodiment, on tropical cyclone of northwestern Pacific Ocean, there may be the variable of influence, such as sea-surface temperature, the whirlpools of low layer
Degree, coriolis force, the vertical shear of wind speed, atmosphere middle layer humidity, atmosphere unstability, the background as mode such as marine superstructure variable
?.Wherein sea-surface temperature is SST;Low layer vorticity and coriolis force selection using 850hPa absolute vorticity or 1000hPa it is exhausted
Vorticity is indicated;The vector difference of the vertical shear of wind speed wind speed between 200hPa and 850hPa;Atmosphere middle layer humidity uses
The unstability of the relative humidity of 700hPa, atmosphere is indicated with CAPE.CAPE(the convective available
Potential energy) it is convective available potential energy, it is one about temperature, pressure and than a wet function.Meanwhile base
In the potential strength PI that the temperature computation of CAPE and sea surface temperature and troposphere upper layer outflow layer comes out.It also include marine superstructure
A variety of variables, such as opposite SST, the heat flux of sea surface, the depth of mixed layer, the thermal content and different temperatures of marine superstructure
Thermoisopleth depth etc..
Specifically, generating on tropical cyclone, there may be the generation frequencies that the characteristic value of the variable of influence and tropical cyclone generate
Number carries out stepwise regression analysis, and examines, if upchecked, retains variable;With stepwise regression analysis method, to above-mentioned
Each variable characteristic value and typhoon generate frequency between carry out stepwise regression analysis, to select suitable variable.For
Retained by the variable of inspection, not verified variable rejected, screened using successive Regression and reject cause it is multiple total
Linear variable.The absolute whirlpool for being 1. 850hPa for the variable that tropical cyclone of northwestern Pacific Ocean is finally selected in the present embodiment
Degree;2. the vertical velocity of 500hPa;3. latent heat flux;4. the mean temperature of mixed layer and 5. 26 DEG C of isothermal depth.
Specifically, the variable retained is used, as predictand, to use as characteristic value, the generation frequency of tropical cyclone generation
The method of neural network is forecast.The present embodiment is the absolute vorticity for using variable 850hPa, and the vertical velocity of 500hPa is dived
Heat flux, the mean temperature of mixed layer, characteristic value of 26 DEG C of isothermal depth as mode, the frequency that typhoon generates is as pre-
Report amount is obtained forecasting model using the method for BP neural network or is forecast using the forecasting model of acquisition.
Specifically, the implementation steps of BP neural network are mainly realized by Python code, deep learning frame is
Tensorflow.Tensorflow is most widely used in deep learning field, provides the network model of a variety of maturations, calculates effect
Rate is relatively preferable.
Forecasting model is obtained using the method for neural network or is forecast using the forecasting model of acquisition.Specifically, making
The method for using BP neural network obtains forecasting model and refers to the generation for using the variable retained to generate as characteristic value, tropical cyclone
Frequency forms the model for being forecast as predictand, using the method and historical data of BP neural network, which makees
It is directly forecast using the data newly observed for product for user.The forecast mould obtained is utilized using the method for neural network
Type is forecast.Specifically, carrying out forecast use and BP neural network method using the forecasting model of the acquisition.
In conclusion forecast tropical cyclone provided by the embodiment of the present invention generates frequency techniques, it is contemplated that atmosphere and sea
Influence of the ocean to tropical cyclone.It can not have to consider the mapping relations between characteristic value and predictand, without the concern for tropical gas
Rotation generates the non-linear relation between weather ambient field.It realizes optimizing forecast model, improve accuracy of the forecast.
On the basis of the above embodiments, Fig. 2 and Fig. 3 is please referred to, Fig. 2 is the method that embodiment forecasts TC Frequency
Carry out the flow diagram of regression analysis;The present embodiment forecasts the method that tropical cyclone generates frequency, will generate to tropical cyclone
The generation frequency generated there may be the characteristic value of the variable of influence and tropical cyclone carries out stepwise regression analysis, and examines, such as
Fruit is upchecked, then retains variable.
Specifically include:
It selects and maximum variable is contributed to predictand, be added in regression equation;
Continue to find out the maximum variable of contribution in remaining variable, be added in regression equation;
It tests to variable, if passed through, it is determined that the variable is added;If not over being added just now
Variable is rejected;
It repeats the above steps to remaining variable, until there is no factor rejecting, is also introduced without the factor.
With reference to the specific step of Fig. 2, include:
(1) all variables: X1, X2 are got out ..., Xn;
(2) factor Xk of maximum contribution is selected;Y=aXk;
(3) the maximum variable X m of contribution factor in remaining variable is selected;
(4) it judges whether there is and is used as GPI variable if not if it is progress (5) from Xm different before;
(5) if there is Xm, Y=aXk+bXm;
(6) judge whether to examine by F;If by carrying out (7), if not by carrying out (8);
(7) Xm, Y=aXk+bXm is added;It carries out (3);
(8) Xm, Y=aXk are removed;It carries out (3);It repeats the above steps, until there is no factor rejecting, is also introduced without the factor
Until.
Referring to FIG. 3, Fig. 3 is the method that the method that the embodiment of the present invention forecasts TC Frequency uses BP neural network
The flow diagram forecast.The present embodiment forecasts the method that tropical cyclone generates frequency, uses the variable of reservation as spy
The generation frequency that value indicative, tropical cyclone generate is forecast as predictand using the method for BP neural network;The variable
Characteristic value is in the input layer of BP neural network;The generation frequency of tropical cyclone is in output layer;The characteristic value of variable and the torrid zone
Mapping relations between the generation frequency of cyclone are in hidden layer.
It is forecast using the method for BP neural network, specifically includes:
(1) the first initialization for carrying out network.The node of input layer is determined according to the characteristic value and predictand that have determined
Number is 5, and the number of nodes of output layer is 1, and the number of nodes of hidden layer is according to formula s=sqrt (0.43mn+0.12nn+2.54m+
0.77n+0.35)+0.51 (m is the number of input layer, and n is the number of output layer) is determined as 4, gives input layer at random and hides
The ω 2 and the parameters such as threshold value beta and learning rate between weight ω 1 and threshold value b and hidden layer and output layer between layer, is learned
Coefficient selection is practised between 0.1~0.8.
(2) output of hidden layer calculates, and according to the characteristic value x of input layer, weight ω 1 and threshold value b calculate the defeated of hidden layer
It is worth out.
(3) output of output layer calculates, and according to the output valve of hidden layer, exports weight ω 2 and threshold value beta, calculates output knot
Fruit y.
(4) loss function is calculated according to the observation y ' of the y of network output and desired output.
(5) Widrow-Hoff learning rules are used, the weight and number of threshold values of neural network are adjusted, until loss function
Reach minimum.
Trained data and verification data are splitted data into, it, can be with after carrying out neural metwork training with data to training
Obtain a model.It is verified finally, verifying data set used is input in the model, verifies the accuracy of model,
The accuracy of verify data is indicated using the related coefficient between forecast data and predictand y.Use BP neural network model
Model of the accuracy of the data forecast 72%, than using conventional model (0.37), after being improved to conventional model
(0.53) it is improved to some extent in accuracy.
The present invention also provides it is a kind of forecast tropical cyclone generate frequency device, the device described below can with it is upper
The method of text description corresponds to each other reference.Referring to FIG. 4, Fig. 4 provides a kind of forecast tropical cyclone life by the embodiment of the present invention
At the schematic diagram of the device of frequency;In conjunction with Fig. 4, which includes:
Regression analysis module, for the characteristic value and tropical cyclone of the variable there may be influence will to be generated on tropical cyclone
The generation frequency of generation carries out stepwise regression analysis, and examines, if upchecked, retains variable;
Neural network module, the generation frequency conduct for using the variable of reservation to generate as characteristic value, tropical cyclone
Predictand is forecast using the method for BP neural network.
On the basis of the above embodiments, optionally, the regression analysis module, for possibility will to be generated to tropical cyclone
The generation frequency that the characteristic value and tropical cyclone of the variable having an impact generate carries out stepwise regression analysis, and examines, if inspection
It tests and passes through, then retain variable, include:
It selects and maximum variable is contributed to predictand, be added in regression equation;
Continue to find out the maximum variable of contribution in remaining variable, be added in regression equation;
It tests to variable, if passed through, it is determined that the variable is added;If not over being added just now
Variable is rejected;
It repeats the above steps to remaining variable, until there is no factor rejecting, is also introduced without the factor.
On the basis of the above embodiments, optionally, the neural network module, for using the variable of reservation as spy
The generation frequency that value indicative, tropical cyclone generate is forecast as predictand using the method for BP neural network;
The characteristic value of the variable is in the input layer of BP neural network;
The generation frequency of tropical cyclone is in output layer;
Mapping relations between the characteristic value of variable and the generation frequency of tropical cyclone are in hidden layer.
On the basis of the above embodiments, optionally, the method for inspection is that F is examined.
The present invention also provides a kind of equipment that forecast tropical cyclone generates frequency, referring to FIG. 5, Fig. 5 is that the present invention is real
Apply the schematic diagram that example provides a kind of equipment of forecast tropical cyclone generation frequency.With reference to Fig. 5, which includes:
Memory, for storing computer program;
Processor realizes the method that any of the above-described forecast tropical cyclone generates frequency when for executing the computer program
The step of.
The embodiment of the above method is please referred to for the introduction of equipment provided by the present invention, the present invention does not do superfluous herein
It states.
The present invention also provides a kind of computer readable storage medium, calculating is stored on the computer readable storage medium
Machine program, the computer program realize that any of the above-described forecast tropical cyclone generates the step of the method for frequency when being executed by processor
Suddenly.
The computer readable storage medium may include: USB flash disk, mobile hard disk, read-only memory (Read-Only
Memory, ROM), random access memory (Random Access Memory, RAM), magnetic or disk etc. is various to deposit
Store up the medium of program code.
Above method embodiment is please referred to for the introduction of computer readable storage medium provided by the present invention, the present invention
This will not be repeated here.
Each embodiment is described in a progressive manner in specification, the highlights of each of the examples are with other realities
The difference of example is applied, the same or similar parts in each embodiment may refer to each other.For device disclosed in embodiment, set
For standby and computer readable storage medium, since it is corresponded to the methods disclosed in the examples, so the comparison of description is simple
Single, reference may be made to the description of the method.
Professional further appreciates that, unit described in conjunction with the examples disclosed in the embodiments of the present disclosure
And algorithm steps, can be realized with electronic hardware, computer software, or a combination of the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is implemented in hardware or software actually, the specific application and design constraint depending on technical solution.Profession
Technical staff can use different methods to achieve the described function each specific application, but this realization is not answered
Think beyond the scope of this invention.
The step of method described in conjunction with the examples disclosed in this document or algorithm, can directly be held with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable programmable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
Computer management method provided by the present invention, relevant device and computer readable storage medium are carried out above
It is discussed in detail.Used herein a specific example illustrates the principle and implementation of the invention, above embodiments
Explanation be merely used to help understand method and its core concept of the invention.It should be pointed out that for the common of the art
, without departing from the principle of the present invention, can be with several improvements and modifications are made to the present invention for technical staff, these
Improvement and modification also fall into the protection scope of the claims in the present invention.
Claims (10)
1. forecasting the method that tropical cyclone generates frequency, characterized in that the variable there may be influence will be generated on tropical cyclone
Characteristic value and the generation frequency that generates of tropical cyclone carry out stepwise regression analysis, and examine, if upchecked, retain change
Amount;
The variable of reservation is used, as predictand, to use neural network as characteristic value, the generation frequency of tropical cyclone generation
Method is forecast.
2. the method that forecast tropical cyclone according to claim 1 generates frequency, characterized in that will be generated to tropical cyclone
The generation frequency generated there may be the characteristic value of the variable of influence and tropical cyclone carries out stepwise regression analysis, and examines, such as
Fruit is upchecked, then retains variable, includes:
It selects and maximum variable is contributed to predictand, be added in regression equation;
Continue to find out the maximum variable of contribution in remaining variable, be added in regression equation;
It tests to variable, if passed through, it is determined that the variable is added;If not over the variable being added just now
It rejects;
It repeats the above steps to remaining variable, until there is no factor rejecting, is also introduced without the factor.
3. the method that forecast tropical cyclone according to claim 1 generates frequency, characterized in that made using the variable retained
Value, the generation frequency of tropical cyclone generation are characterized as predictand, is forecast using the method for BP neural network;
The characteristic value of the variable is in the input layer of BP neural network;
The generation frequency of tropical cyclone is in output layer;
Mapping relations between the characteristic value of variable and the generation frequency of tropical cyclone are in hidden layer.
4. forecasting that tropical cyclone generates the model of frequency, characterized in that the tropical gas of any forecast according to claim 1 to 3
The method that rotation generates frequency obtains.
5. forecasting that tropical cyclone generates the device of frequency, characterized in that include:
Regression analysis module, characteristic value and tropical cyclone for that will generate the variable there may be influence on tropical cyclone generate
Generation frequency carry out stepwise regression analysis, and examine, if upchecked, retain variable;
Neural network module, the generation frequency for using the variable of reservation to generate as characteristic value, tropical cyclone are used as forecast
Amount, is forecast using the method for BP neural network.
6. the device that forecast tropical cyclone according to claim 5 generates frequency, characterized in that
The regression analysis module, for the characteristic value and tropical cyclone of the variable there may be influence will to be generated on tropical cyclone
The generation frequency of generation carries out stepwise regression analysis, and examines, if upchecked, retains variable, includes:
It selects and maximum variable is contributed to predictand, be added in regression equation;
Continue to find out the maximum variable of contribution in remaining variable, be added in regression equation;
It tests to variable, if passed through, it is determined that the variable is added;If not over the variable being added just now
It rejects;
It repeats the above steps to remaining variable, until there is no factor rejecting, is also introduced without the factor.
7. the device that forecast tropical cyclone according to claim 5 generates frequency, characterized in that
The neural network module, the generation frequency conduct for using the variable of reservation to generate as characteristic value, tropical cyclone
Predictand is forecast using the method for BP neural network;
The characteristic value of the variable is in the input layer of BP neural network;
The generation frequency of tropical cyclone is in output layer;
Mapping relations between the characteristic value of variable and the generation frequency of tropical cyclone are in hidden layer.
8. the device that forecast tropical cyclone according to claim 6 generates frequency, characterized in that the method for inspection is F
It examines.
9. a kind of equipment that forecast tropical cyclone generates frequency, characterized in that include:
Memory, for storing computer program;
Processor realizes such as Claims 1-4 described in any item forecast tropical cyclones when for executing the computer program
The step of generating the method for frequency.
10. a kind of computer readable storage medium, characterized in that the computer-readable recording medium storage has computer journey
Sequence realizes that the described in any item forecast tropical cyclones of Claims 1-4 such as generate when the computer program is executed by processor
The step of method of frequency.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910240549.0A CN109960882A (en) | 2019-03-28 | 2019-03-28 | Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910240549.0A CN109960882A (en) | 2019-03-28 | 2019-03-28 | Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109960882A true CN109960882A (en) | 2019-07-02 |
Family
ID=67025198
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910240549.0A Withdrawn CN109960882A (en) | 2019-03-28 | 2019-03-28 | Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109960882A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111338005A (en) * | 2020-03-02 | 2020-06-26 | 中国人民解放军国防科技大学 | Prediction method for generation frequency of tropical cyclone in northwest Pacific ocean on monthly scale |
CN112596127A (en) * | 2020-11-25 | 2021-04-02 | 中国人民解放军国防科技大学 | Novel method for calculating typhoon potential generation index |
-
2019
- 2019-03-28 CN CN201910240549.0A patent/CN109960882A/en not_active Withdrawn
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111338005A (en) * | 2020-03-02 | 2020-06-26 | 中国人民解放军国防科技大学 | Prediction method for generation frequency of tropical cyclone in northwest Pacific ocean on monthly scale |
CN111338005B (en) * | 2020-03-02 | 2021-07-30 | 中国人民解放军国防科技大学 | Prediction method for generation frequency of tropical cyclone in northwest Pacific ocean on monthly scale |
CN112596127A (en) * | 2020-11-25 | 2021-04-02 | 中国人民解放军国防科技大学 | Novel method for calculating typhoon potential generation index |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Quinn et al. | Exploring how changing monsoonal dynamics and human pressures challenge multireservoir management for flood protection, hydropower production, and agricultural water supply | |
Pawson et al. | The GCM-Reality Intercomparison Project for SPARC (GRIPS): Scientific issues and initial results | |
Neggers | A dual mass flux framework for boundary layer convection. Part II: Clouds | |
Hung et al. | An artificial neural network model for rainfall forecasting in Bangkok, Thailand | |
Liu et al. | Evaluation of two statistical downscaling models for daily precipitation over an arid basin in China | |
Thirel et al. | A past discharges assimilation system for ensemble streamflow forecasts over France–Part 1: Description and validation of the assimilation system | |
Weyn et al. | The dependence of the predictability of mesoscale convective systems on the horizontal scale and amplitude of initial errors in idealized simulations | |
US9262723B2 (en) | Predicting climate data using climate attractors derived from a global climate model | |
Lloyd | I—Elisabeth A. Lloyd: Varieties of support and confirmation of climate models | |
Hartmann et al. | Response of baroclinic life cycles to barotropic shear | |
Khan et al. | Assessing hydrologic impact of climate change with uncertainty estimates: Bayesian neural network approach | |
Neggers et al. | Power-law scaling in the internal variability of cumulus cloud size distributions due to subsampling and spatial organization | |
Chu et al. | Sensitivities of modelling storm surge to bottom friction, wind drag coefficient, and meteorological product in the East China Sea | |
Ukkonen et al. | Evaluation of thunderstorm predictors for Finland using reanalyses and neural networks | |
Yoo et al. | Analysis of intraseasonal and interannual variability of the Asian summer monsoon using a hidden Markov model | |
Juza et al. | From basin to sub-basin scale assessment and intercomparison of numerical simulations in the Western Mediterranean Sea | |
CN109960882A (en) | Forecast that tropical cyclone generates method, apparatus, equipment and the storage medium of frequency | |
Lai et al. | Conceptual Sim-Heuristic optimization algorithm to evaluate the climate impact on reservoir operations | |
CN106679761A (en) | Method and system for forecasting comprehensive water levels of sea areas in real time | |
Hassanian et al. | Iceland wind farm assessment case study and development: An empirical data from wind and wind turbine | |
Paquin et al. | Modelled variations of deep convection in the Irminger Sea during 2003–10 | |
Ramesh et al. | Forecasting Indian Monsoon Rainfall Including Within Year Seasonal Variability | |
Rasp | Statistical methods and machine learning in weather and climate modeling | |
CN111428420B (en) | Method and device for predicting sea surface flow velocity, computer equipment and storage medium | |
Zhang et al. | Impact of enthalpy-based ensemble filtering sea ice data assimilation on decadal predictions: Simulation with a conceptual pycnocline prediction model |
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 | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20190702 |
|
WW01 | Invention patent application withdrawn after publication |