CN107219566A - Cloud prediction and forecasting procedure based on GM (1,1) gray model - Google Patents
Cloud prediction and forecasting procedure based on GM (1,1) gray model Download PDFInfo
- Publication number
- CN107219566A CN107219566A CN201710300151.2A CN201710300151A CN107219566A CN 107219566 A CN107219566 A CN 107219566A CN 201710300151 A CN201710300151 A CN 201710300151A CN 107219566 A CN107219566 A CN 107219566A
- Authority
- CN
- China
- Prior art keywords
- cloud
- prediction
- model
- data
- predicted
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
Landscapes
- Environmental & Geological Engineering (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Atmospheric Sciences (AREA)
- Biodiversity & Conservation Biology (AREA)
- Ecology (AREA)
- Environmental Sciences (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Cloud prediction and forecasting procedure based on the variable Grey models GM (1,1) of single order one, belong to concrete application of the Computerized Information Processing Tech in atmospheric science research field.It is characterized in that：It regard original satellite cloud atlas (remote sensing data) or satellite remote sensing product as data source, historical summary is closed on as data set using a small amount of, in data set with it is each when time fixed position point on data configuration time series set up grey forecasting model, growth factor and extinction factor pair model output result is set to be regulated and controled, predict the numerical value on the location point in following development trend by Models computed, predicted value combination in all fixed position points is output as new product collection, it is last that the state partly or wholly for parsing target information is concentrated in the product for representing to-be, realize the generation to the cloud of future time instance, development, it is mobile, merge, the prediction and forecast of the evolution process such as extinction.
Description
Technical field
The invention belongs to concrete application of the Computerized Information Processing Tech in atmospheric science research field.More particularly to one
Plant cloud prediction and forecasting procedure based on GM (1,1) grey forecasting model.
Background technology
Cloud is the visible polymer of steam and the nuclei of condensation in atmosphere, is one of very important meteorological element, is also table
Levy an important factor of Earth Atmosphere System behavior and geophysics state.In the prediction methods of cloud, atmospheric science
In with Atmosphere System develop in because although the cloud Forecasting Methodology based on numerical weather forecast (numerical model) at foothold has
Sufficient dynamics, thermodynamics and physics theory support, but predict the outcome and release with sufficiently complex, is settled finally in practicality by one
Limit；Although the external presentation moved with cloud cluster is characterized as that the cloud Forecasting Methodology based on linear extrapolation of point of observation is simple, exist
Predict timeliness it is short, the problem of predictablity rate is low, and can only predict existing cloud cluster barycenter (or center) situation of movement and nothing
The new life of method prediction cloud and the dissipation of cloud.
The content of the invention
There is provided a kind of new with physical significance for deficiency of the invention for the prediction of both the above cloud and forecasting procedure
Prediction and forecasting procedure with the cloud of convenient practicality.
The present invention is achieved by the following technical solutions：Cloud prediction and forecast of the one kind based on GM (1,1) gray model
Method, it is characterised in that include following key step：
(1) original satellite cloud atlas (remote sensing data) or satellite remote sensing product are reported into point " 0 moment " as data source using rising
And its n a small amount of constant duration before closes on historical summary as initial operational data collection；
(2) operational data concentrate with it is each when secondary datum plane in fixed position point on data configuration time series,
GM (1,1) grey forecasting model is set up, model is resolved, passes through the statistical nature of initial operational data collection, the model modeled every time
Development coefficient and set growth factor, extinction factor pair model output result are revised and regulated and controled, and export the location point
On prediction numerical value, regard this numerical value as the predicted value at this position of following subsequent time；
(3) the predicted value combination in all fixed position points that will be calculated through previous step has been output as report point " when 0
Carve " the predicted value plane at the 1st moment afterwards, and this predicted value panel data is added to operational data concentrated, sequentially in time
Arrangement, updates operational data collection, repeats (2) step process, using the method for recursion, is sequentially output the 2nd, the 3rd to the m moment
Predicted value plane, build prediction product collection.Represent to-be prediction product concentrate parsing target information part or
Overall state, realizes prediction to evolution process such as generation, development, movement, merging, the extinctions of cloud of future time instance and in advance
Report.
It is of the invention with existing cloud prediction and forecasting procedure compared with, have an advantageous effect in that：The life of cloud is taken into full account
Disappear the nonlinear and nonstationary feature of differentiation, and overall prediction is implemented using first superposition background, then from prediction product parsing and point
The way of target development evolvement trend is analysed, the office of general cloud cluster barycenter (or center) linear extrapolation prediction methods is overcome
Limit, can predict movement and the in-house development situation of cloud cluster, sensitively can catch and predict again the newborn information of outside cloud.Together
When, present invention, avoiding the huge computing of the cloud prediction methods based on numerical weather forecast (numerical model), it is to avoid multiple
Process is used in miscellaneous releasing.Empirical tests have higher accuracy and practicality in the short-term prediction forecast application of cloud and cloud cluster.
Brief description of the drawings
Fig. 1 is based on the cloud prediction and forecasting procedure flow chart of GM (1,1) gray model
Fig. 2 is based on prediction and the value of forecasting figure explanation in the MCSs cloud covered areas domain of GM (1,1) gray model：Being one makes
It is pre- with mesoscale convective system (Mesoscale Convective Systems, MCSs) cloud system of GM (1,1) gray model
Survey the image conversion displaying of result.Used in example FY2G meteorological satellites zebra time 03 on May 28th, 2016,04,05,06,
When 07 totally 5 when time Southern Hemisphere mid low latitude region overhead satellite visual field in 65*93 pixel coverages size (ground about 325*
L1 level vapor channels Value of Remote Sensing Data 465km) is source data, to it is following 5 when time (during 09-13) cloud evolution enter
Prediction is gone.The 2nd, 4,6 row bianry images are the edge of the MCSs cloud targets parsed in figure；
Fig. 3 is based on the forecast accuracy curve map explanation in the MCSs cloud covered areas domain of GM (1,1) gray model：It is to use
Forecast accuracy in the MCSs cloud covered areas domain prediction example of GM (1,1) gray model is with the forecast increased change feelings of duration
Condition.The predictablity rate index used is：POD=n is compared in detectionSuccess/(nSuccess+nFail to report), FAR=n is compared in falseFalse/(nSuccess+nFalse),
Critical success index CSI=nSuccess/(nSuccess+nFalse+nFail to report), fail to report and compare MAR=nFail to report/(nSuccess+nFail to report), wherein, when prediction and fact
When being consistent referred to as " success "；When prediction meets standard less than standard, and fact, referred to as " fail to report "；When predictor standardization,
And it is live when being less than standard, referred to as " make a false report ".Shown in figure, the MCSs cloud covered areas domain of GM (1,1) gray model is used
1-2 hours predictablity rates are higher, it is relatively low with false alert rate to fail to report, and the error of forecast result gradually increases with forecast duration.
Embodiment
Below in conjunction with the accompanying drawings 1 and example, embodiments of the present invention are done with the detailed description of idiographic flow, effect of the invention
It can will become more apparent：
1. the area-of-interest of 65*93 pixel sizes is chosen in L1 grades of vapor channel Value of Remote Sensing Data, with 2016 5
Data information when during month zebra time 03,04,05,06,07 on the 28th 5 in the region time as initial operational data collection BT,
It is respectively labeled as BT_4, BT_3, BT_2, BT_1, BT_0；
2. extracting the numerical value at each datum plane position (i, j) place in data set sequentially in time, sequence is arranged as：
x(0)(1), x(0)(2), x(0)(3), x(0)(4), x(0)(5), referred to as original series, one-accumulate, generation are implemented by this sequence
Sequence：x(1)(1), x(1)(2), x(1)(3), x(1)(4), x(1)(5), the two meets formula relational expression：And meet GM (1,1) grey forecasting model equation：By the mould
The least-squares estimation for the parameter vector that the development coefficient a of type and grey control action amount b are constituted meets formula：Wherein：
It is so as to the response of trying to achieve GM (1,1) grey forecasting model：Enter
And discretization forecasting sequence can be solved
During in this example, this is calculated, n=5.By above method, by what is solvedRevised using following formula：
In formula, ai,jFor position (i, j) place model development coefficient, x(0)(n) it is the tail-end value of modeling original series, α is extinction
The factor, β be growth factor (take α=0.85, β=1.02) in this example,For initial operational data collection
Minimum,For initial operational data maximum of a set.
Will be revisedNumerical value as next moment at plan-position (i, j) place predicted value.Count successively
Calculate the predicted value that simultaneously output data concentrates next moment at all plan-positions.
3. the predicted value at all positions to be output as to the predicted value plane PP_1 of subsequent time according to position grouping, update
Operational data collection is { BT_4, BT_3, BT_2, BT_1, BT_0, PP_1 }, repeats (2) step process, and recursion resolves output
PP_2, PP_3 ..., untill PP_m (m=5 in this example), build prediction product collection for PP_1, PP_2, PP_3, PP_4,
PP_5}.That is PP_1, PP_2, PP_3, PP_4, PP_5 correspond to zebra time 08,09,10,11,12 on May 28th, 2016 respectively
When area-of-interest prediction L1 level vapor channel remotely-sensed datas.Parsing mesoscale convective system is concentrated in prediction product
(Mesoscale Convective Systems, MCSs) cloud system, realize -5 hours 1 hour MCSs clouds overlay area develop into
The prediction and forecast of journey.
The example explanation present invention have it is required historical summary is few, operand is small, computing take it is short, implement it is convenient, can be pre-
The movement and in-house development situation for surveying cloud cluster sensitively can catch and predict again the newborn information of outside cloud, while having air concurrently
The characteristics of environment field physical significance, there is very high accuracy using the Short-term Forecast product of the inventive method.
Examples detailed above is used for illustrating the present invention, rather than limits the invention, in the spirit and power of the present invention
In the protection domain that profit is required, any modifications and changes made to the present invention both fall within protection scope of the present invention.
Claims (5)
1. cloud prediction and forecasting procedure of the one kind based on GM (1,1) gray model, it is characterised in that include following key step：
(1) using original satellite cloud atlas (remote sensing data) or satellite remote sensing product as data source, using rise report point " 0 moment " and its
N a small amount of constant duration before closes on historical summary as initial operational data collection；
(2) operational data concentrate with it is each when time datum plane in fixed position point on data configuration time series, foundation
GM (1,1) grey forecasting model, resolves model, passes through the statistical nature of initial operational data collection, the model development modeled every time
Coefficient and set growth factor, extinction factor pair model output result are revised and regulated and controled, and are exported on the location point
Numerical value is predicted, this numerical value is regard as the predicted value at this position of following subsequent time；
(3) the predicted value combination in all fixed position points that will be calculated through previous step has been output as report point " 0 moment "
The predicted value plane at the 1st moment, and this predicted value panel data is added into operational data concentration afterwards, is arranged sequentially in time
Row, update operational data collection, repeat (2) step process, using the method for recursion, are sequentially output the 2nd, the 3rd to the m moment
Predicted value plane, builds prediction product collection.The local or whole of parsing target information is concentrated in the prediction product for representing to-be
The state of body, realizes the prediction and forecast to evolution process such as generation, development, movement, merging, the extinctions of cloud of future time instance.
2. the cloud based on GM (1,1) gray model is predicted and forecasting procedure according to claim 1, it is characterised in that：Use
Data composition time series in fixed position point in data set in each datum plane is modeled, each data in data set
The method that all location points in plane are slided by position is traveled through.
3. the cloud based on GM (1,1) gray model is predicted and forecasting procedure according to claim 1, it is characterised in that：Use
The Forecasting Methodology of GM (1,1) gray model.
4. the cloud based on GM (1,1) gray model is predicted and forecasting procedure according to claim 1, it is characterised in that：Use
The statistical nature of initial operational data collection, the model development coefficient modeled every time and set growth factor, extinction factor pair
Model output result is revised and regulated and controled.
5. the cloud based on GM (1,1) gray model is predicted and forecasting procedure according to claim 1, it is characterised in that：Pass through
The state partly or wholly of parsing target information is concentrated in the prediction product for being superimposed ambient field information, realization is drilled target
The prediction and forecast of change process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710300151.2A CN107219566B (en) | 2017-05-02 | 2017-05-02 | Cloud prediction and forecasting procedure based on GM (1,1) gray model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710300151.2A CN107219566B (en) | 2017-05-02 | 2017-05-02 | Cloud prediction and forecasting procedure based on GM (1,1) gray model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107219566A true CN107219566A (en) | 2017-09-29 |
| CN107219566B CN107219566B (en) | 2019-05-07 |
Family
ID=59943765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710300151.2A Active CN107219566B (en) | 2017-05-02 | 2017-05-02 | Cloud prediction and forecasting procedure based on GM (1,1) gray model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107219566B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102833326A (en) * | 2012-08-15 | 2012-12-19 | 广东工业大学 | Grey prediction-based cloud storage load balancing method |
| CN104216334A (en) * | 2014-09-16 | 2014-12-17 | 北京工业大学 | Selection optimization method of temperature measurement point combination for positioning errors of numerically-controlled machine tool under thermal effect |
| CN104933469A (en) * | 2015-03-02 | 2015-09-23 | 华北电力大学（保定） | Short-term wind speed forecasting method based on grey generating perturbation model |
| CN105488730A (en) * | 2015-12-30 | 2016-04-13 | 郑州光力科技股份有限公司 | Fire risk monitoring method of easy firing point at goaf |
| CN106597574A (en) * | 2016-12-30 | 2017-04-26 | 重庆邮电大学 | Weather temperature prediction method and device based on time-varying cloud model |
-
2017
- 2017-05-02 CN CN201710300151.2A patent/CN107219566B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102833326A (en) * | 2012-08-15 | 2012-12-19 | 广东工业大学 | Grey prediction-based cloud storage load balancing method |
| CN104216334A (en) * | 2014-09-16 | 2014-12-17 | 北京工业大学 | Selection optimization method of temperature measurement point combination for positioning errors of numerically-controlled machine tool under thermal effect |
| CN104933469A (en) * | 2015-03-02 | 2015-09-23 | 华北电力大学（保定） | Short-term wind speed forecasting method based on grey generating perturbation model |
| CN105488730A (en) * | 2015-12-30 | 2016-04-13 | 郑州光力科技股份有限公司 | Fire risk monitoring method of easy firing point at goaf |
| CN106597574A (en) * | 2016-12-30 | 2017-04-26 | 重庆邮电大学 | Weather temperature prediction method and device based on time-varying cloud model |
Non-Patent Citations (2)
| Title |
|---|
| 李祚泳: ""用灰色GM(1，1)模型作异常降水预测"", 《四川大学学报（自然科学版）》 * |
| 王志强: ""灰色模型在环渤海地区降水过程预报中的应用"", 《干旱气象》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107219566B (en) | 2019-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ferreira et al. | Short-term forecast of wind speed through mathematical models | |
| Lalic et al. | An empirical relation describing leaf-area density inside the forest for environmental modeling | |
| CN107703564A (en) | A kind of precipitation predicting method, system and electronic equipment | |
| Jing et al. | An environment‐dependent probabilistic tropical cyclone model | |
| US10545263B2 (en) | Systems and methods for generating computer-based representations of probabilities of precipitation occurrences and intensities | |
| Douville | Relevance of soil moisture for seasonal atmospheric predictions: is it an initial value problem? | |
| CN107179566A (en) | The self study modification method and system of a kind of district weather forecasting | |
| Tiwari et al. | The role of land surface schemes in the regional climate model (RegCM) for seasonal scale simulations over Western Himalaya | |
| US20100100328A1 (en) | System and Method for Generating a Cloud Type and Coverage Prediction Database | |
| Zhang et al. | HWRF based ensemble prediction system using perturbations from GEFS and stochastic convective trigger function | |
| CN114493052B (en) | Multi-model fusion self-adaptive new energy power prediction method and system | |
| Lin et al. | Forecasts of hurricanes using large-ensemble outputs | |
| Arcomano et al. | A hybrid atmospheric model incorporating machine learning can capture dynamical processes not captured by its physics‐based component | |
| Zhang et al. | The role of radiative interactions in tropical cyclone development under realistic boundary conditions | |
| Wen et al. | Applying an artificial neural network to simulate and predict Chinese fir (Cunninghamia lanceolata) plantation carbon flux in subtropical China | |
| van Oel et al. | Application of multi-agent simulation to evaluate the influence of reservoir operation strategies on the distribution of water availability in the semi-arid Jaguaribe basin, Brazil | |
| Routray et al. | Influence of ASCAT soil moisture on prediction of track and intensity of landfall tropical cyclones | |
| CN107180227A (en) | Cloud prediction and forecasting procedure based on DTMGM+ gray models | |
| Ding et al. | Impact of Convection‐Permitting and Model Resolution on the Simulation of Mesoscale Convective System Properties Over East Asia | |
| Tan et al. | Projected changes of typhoon intensity in a regional climate model: Development of a machine learning bias correction scheme | |
| Sateesh et al. | Numerical simulation of tropical cyclone thane: role of boundary layer and surface drag parameterization schemes | |
| Rousseau-Rizzi et al. | A thermodynamic perspective on steady-state tropical cyclones | |
| CN107219566A (en) | Cloud prediction and forecasting procedure based on GM (1,1) gray model | |
| Sun et al. | Nonstationarities and at-site probabilistic forecasts of seasonal precipitation in the East River Basin, China | |
| Loi et al. | Prediction of tropical cyclogenesis based on machine learning methods and its SHAP interpretation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |