CN108428007A - A kind of recognition methods of Driving forces of land use change, system and device - Google Patents
A kind of recognition methods of Driving forces of land use change, system and device Download PDFInfo
- Publication number
- CN108428007A CN108428007A CN201810120621.1A CN201810120621A CN108428007A CN 108428007 A CN108428007 A CN 108428007A CN 201810120621 A CN201810120621 A CN 201810120621A CN 108428007 A CN108428007 A CN 108428007A
- Authority
- CN
- China
- Prior art keywords
- land use
- data
- use change
- identification model
- driven factor
- 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
- 230000008859 change Effects 0.000 title claims abstract description 67
- 238000000034 method Methods 0.000 title claims abstract description 55
- 238000005457 optimization Methods 0.000 claims abstract description 12
- 230000006870 function Effects 0.000 claims description 15
- 238000010219 correlation analysis Methods 0.000 claims description 6
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 238000013439 planning Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000000265 homogenisation Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 5
- 235000013372 meat Nutrition 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 4
- 230000003247 decreasing effect Effects 0.000 abstract description 2
- 238000010276 construction Methods 0.000 description 8
- 230000000875 corresponding effect Effects 0.000 description 7
- 238000011160 research Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 239000000284 extract Substances 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000002596 correlated effect Effects 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 230000007423 decrease Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000002452 interceptive effect Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 239000002689 soil Substances 0.000 description 2
- 244000025254 Cannabis sativa Species 0.000 description 1
- VAYOSLLFUXYJDT-RDTXWAMCSA-N Lysergic acid diethylamide Chemical compound C1=CC(C=2[C@H](N(C)C[C@@H](C=2)C(=O)N(CC)CC)C2)=C3C2=CNC3=C1 VAYOSLLFUXYJDT-RDTXWAMCSA-N 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 238000013316 zoning Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/26—Government or public services
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of recognition methods of Driving forces of land use change, system and devices, and this approach includes the following steps：Obtain the data of the driven factor in survey region；According to the data of the driven factor of acquisition, M iteration optimization is carried out to prediction model using stochastic gradient lift method, obtains driving force identification model；The contribution rate of the driven factor of different land use variation is calculated according to driving force identification model, the system and device are used to execute the recognition methods of Driving forces of land use change.The present invention is iterated optimization using stochastic gradient lift method to prediction model, and stochastic gradient lift method Ensemble classifier tree and lift method, the deviation of prediction model can be steadily decreasing by iteration, to obtain high-precision driving force identification model.The present invention can be widely applied to environment modeling techniques field.
Description
Technical field
The present invention relates to environment modeling techniques field, especially a kind of recognition methods of Driving forces of land use change is
System and device.
Background technology
Land use change survey modeling is the emphasis of soil scientific research, and the research for land use change survey mechanism is soil
It is the research core of land change problem using the basis of variation model.Thus land use change survey attribution is particularly important, is
It inquires into land use driving mechanism, support urban planning with policy making and assessment land use to eco-environmental impact
Important tool.
Land use change survey attribution model is variation and the various shadows by analyzing land use time and spatial framework
The factor and its mode of action for ringing land use change survey, disclose the feature of land use change survey and illustrate land use change survey
Process.Land use change survey attribution model mainly uses empirical model, the model to show variable using linear correlation at present
Between relationship mathematical equation, using land use change survey class label as dependent variable, driving force factors form independent variable, lead to
Regression equation coefficient value is crossed to reflect the relationship of land use change survey, this method is simple and practical, and autgmentability is strong, can carry out single
Simulation.However in complicated man-land territorial system, between numerous societies, economy, technology and natural environmental condition
Interaction is not a kind of simple linear relationship.The cause and effect normal form Land in Regional Land attribute beyond expression of words that linear model obtains
The nonlinear change phenomenon influenced by critical value, mutation or enchancement factor, therefore the simulation precision of model is relatively low.
Invention content
In order to solve the above technical problems, it is an object of the invention to：A kind of land use change survey that simulation precision is high is provided
Recognition methods, system and the device of driving force.
The first technical solution for being taken of the present invention is：
A kind of recognition methods of Driving forces of land use change, includes the following steps：
Obtain the data of the driven factor in survey region；
According to the data of the driven factor of acquisition, M iteration optimization is carried out to prediction model using stochastic gradient lift method,
Obtain driving force identification model；
The contribution rate of the driven factor of different land use variation is calculated according to driving force identification model.
Further, the expression formula of the driving force identification model is：
Wherein, FM(x) prediction model after M iteration optimization is represented, i.e. driving force identification model, M represents iteration
Number, J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, CjmRepresenting makes the loss letter of prediction model
The corresponding output valve of leaf node that number minimizes, I representative samples.
Further, further comprising the steps of：
Resolution ratio homogenization processing is carried out to the data of the driven factor got.
Further, further comprising the steps of：
Utilize the correlativity between partial Correlation Analysis land use change survey and single driven factor.
Further, further comprising the steps of：
The degree of fitting of driving force identification model is verified, and uses ten foldings crosscheck method verification driving force identification model
Precision.
Further, the data of the driven factor include physical geography data and socioeconomic data, wherein：
The physical geography data include at least one of landform, the hydrology and meteorological data；
The socioeconomic data include land use pattern figure, land use planning figure, location information, land policy,
Gross national product, GNP per capita, social retail product gross sales amount, primary industry gross national product, the second production
Industry gross national product, Gross National Product of Tertiary Industry, per capita output of grain, production of meat, investment in fixed assets are total per capita
Volume, Engel coefficient, gini index, town dweller per capita salary, cottar per capita salary, total population density and people from cities and towns
At least one of mouth density.
Further, the contribution rate of the driven factor that different land use variation is calculated according to driving force identification model,
The step for be specially：
According to driving force identification model carry out weight calculation, extract the weighted score of driven factor, to obtain driving because
The contribution rate of son.
Second of technical solution being taken of the present invention be：
A kind of identifying system of Driving forces of land use change, including：
Acquisition module, the data for obtaining the driven factor in survey region；
Iteration module, for the data according to the driven factor of acquisition, using stochastic gradient lift method to prediction model into
M iteration optimization of row, obtains driving force identification model；
Computing module, the contribution of the driven factor for calculating different land use variation according to driving force identification model
Rate.
Further, the expression formula of the driving force identification model is：
Wherein, M represents the number of iteration, and J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, Cjm
Representative makes the corresponding output valve of leaf node that the loss function of prediction model minimizes, I representative samples.
The third technical solution for being taken of the present invention is：
A kind of identification device of Driving forces of land use change, including：
Memory, for storing program；
Processor, the method for loading described program to execute Driving forces of land use change.
The beneficial effects of the invention are as follows：The present invention is iterated optimization using stochastic gradient lift method to prediction model,
And stochastic gradient lift method Ensemble classifier tree and lift method, the deviation of prediction model is enough steadily decreasing by iteration, to
Obtain high-precision driving force identification model.
Description of the drawings
Fig. 1 is a kind of flow chart of the recognition methods of Driving forces of land use change of the present invention；
Fig. 2 is a kind of flow chart of the recognition methods of the Driving forces of land use change of specific embodiment of the present invention；
Fig. 3 is the statistical chart of certain province farmland variation and its driven factor contribution rate in a kind of specific embodiment of the present invention；
Fig. 4 is the statistical chart for the Main Driving Factors contribution rate that certain in a kind of specific embodiment of the present invention saves construction land.
Specific implementation mode
Referring to Fig.1, a kind of recognition methods of Driving forces of land use change, includes the following steps：
Obtain the data of the driven factor in survey region；
According to the data of the driven factor of acquisition, M iteration optimization is carried out to prediction model using stochastic gradient lift method,
Obtain driving force identification model；
The contribution rate of the driven factor of different land use variation is calculated according to driving force identification model.
Further, the expression formula of the driving force identification model is：
Wherein, FM(x) prediction model after M iteration optimization is represented, i.e. driving force identification model, M represents iteration
Number, J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, CjmRepresenting makes the loss letter of prediction model
The corresponding output valve of leaf node that number minimizes, I representative samples.Wherein, prediction model is with F (x), the expression of F (x),
Those skilled in the art can be flexibly arranged according to actual needs.Such as F (x)=A1x+A2x+A3x+……+Anx；F (x)=W0+
W1x1+W2x2+…+Wnxn。
It is further used as preferred embodiment, it is further comprising the steps of：
Resolution ratio homogenization processing is carried out to the data of the driven factor got.
It is further used as preferred embodiment, it is further comprising the steps of：
Utilize the correlativity between partial Correlation Analysis land use change survey and single driven factor.
It is further used as preferred embodiment, it is further comprising the steps of：
The degree of fitting of driving force identification model is verified, and uses ten foldings crosscheck method verification driving force identification model
Precision.
It is further used as preferred embodiment, the data of the driven factor include physical geography data and social economy
Data, wherein：
The physical geography data include at least one of landform, the hydrology and meteorological data；
The socioeconomic data include land use pattern figure, land use planning figure, location information, land policy,
Gross national product, GNP per capita, social retail product gross sales amount, primary industry gross national product, the second production
Industry gross national product, Gross National Product of Tertiary Industry, per capita output of grain, production of meat, investment in fixed assets are total per capita
Volume, Engel coefficient, gini index, town dweller per capita salary, cottar per capita salary, total population density and people from cities and towns
At least one of mouth density.
It is further used as preferred embodiment, it is described that different land use variation is calculated according to driving force identification model
The contribution rate of driven factor, the step for be specially：
According to driving force identification model carry out weight calculation, extract the weighted score of driven factor, to obtain driving because
The contribution rate of son.
A kind of identifying system of Driving forces of land use change corresponding to method with Fig. 1, including：
Acquisition module, the data for obtaining the driven factor in survey region；
Iteration module, for the data according to the driven factor of acquisition, using stochastic gradient lift method to prediction model into
M iteration optimization of row, obtains driving force identification model；
Computing module, the contribution of the driven factor for calculating different land use variation according to driving force identification model
Rate.
It is further used as preferred embodiment, the expression formula of the driving force identification model is：
Wherein, M represents the number of iteration, and J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, Cjm
Representative makes the corresponding output valve of leaf node that the loss function of prediction model minimizes, I representative samples.
A kind of identification device of Driving forces of land use change corresponding to method with Fig. 1, including：
Memory, for storing program；
Processor, the method for loading described program to execute Driving forces of land use change.
The present invention is further detailed with specific embodiment with reference to the accompanying drawings of the specification.
With reference to Fig. 2, present embodiment discloses a kind of recognition methods of Driving forces of land use change, this method includes following
Step：
S1, obtain survey region in driven factor data.
Survey region is determined first, collects the side such as history land use change survey, politics, economy, population, nature in research area
Face counts and Space Elements data, establishes land use geographic information data collection in survey region, the geographic information data collection
Including two class data of physical geography and social economy：Physical geography data include：Landform, the hydrology, meteorological data；Social economy
Data include land use pattern figure, land use planning figure, location information, land policy, gross national product, per capita state
People's total output value, social retail product gross sales amount, primary industry gross national product, secondary industry gross national product, third
Industry gross national product, per capita output of grain, per capita production of meat, gross fixed assets investment, Engel coefficient, Geordie
The data such as index, town dweller per capita salary, cottar per capita salary, total population density, urban population density.
S2, resolution ratio homogenization processing is carried out to the data of the driven factor got.For example, the temperature in meteorological factor
It is usually raster data or based on the measured value of meteorological site with rainfall data, land use data is based on grid
Or plot, gross national product, GNP per capita and social retail product gross sales amount data be based on
Town, county, city or bigger regional scale, total population density and urban population density are generally based on the data of grid, fixed
Assets investment total value, Engel coefficient, gini index and town dweller per capita salary are generally from the yearbook number based on county domain
According to these data resolutions have very big difference：Based on meteorological site, based on grid, plot or data based on the cities Xian He
It needs to carry out the time and space scale unification carries out land use change survey classification, Reasons again.Data prediction is to realize to differentiate
Rate uniforms, and generally uses spatial interpolation methods to the data based on meteorological site；To be based on grid, plot or based on county and
The data in city use using the maximum region method average as standard area zoning (such as：City>County>Town>Plot is with city
Standard area calculates the average value in all counties in a city).Different resolution time scale data are also to use with maximum time
Scale is the standard time to calculate method average in the period, such as year>Season>Month>It, calculates by the standard time of year
The average value of all seasons in 1 year.The driven factor data of time and spatial resolution homogenization are thus obtained.Due to
The presence of exceptional value can have spatial variability function significant impact, this research to identify that exceptional value, i.e. sample are flat using domain method
Mean value adds and subtracts 3 times of standard deviations, and the data other than this section are set to exceptional value, then uses respectively normal minimum and maximum
Value replaces exceptional value to control the quality of data.By this step, it is capable of the precision of lift scheme and the generation effect of model
Rate.
It is excellent to carry out M iteration using stochastic gradient lift method to prediction model for S3, the data according to the driven factor of acquisition
Change, obtains driving force identification model.Wherein the step is specific as follows：
In the present embodiment, prediction model is indicated with F (x), then calculates minimum loss letter of the sample point at F (x)
Number, algorithm one new regression tree of each grey iterative generation, for N number of sample pointThe expression of loss function can be led
Formula is： L(yi,F(xi)), the number of iteration, m={ 1,2,3 ... ..., M }, the recurrence obtained after M iteration are indicated with M
Anticipation function is with FM(x) it indicates, i.e. driving force identification model.
First, regression tree is initializedEstimation one makes loss function pole
The long numerical value of smallization, at this time the regression tree be only there are one put tree, then iteration establish M regression tree：The estimation of residual error
Value, sees below expression formula.
From m=1 to M (i.e. first layer recycles), i=1 to N (i.e. the second layer recycles), counting loss function exists in negative gradient
The value of "current" model, and using it as the estimated value of residual error：The estimation value expression of the residual error is：
Wherein, γimIndicate residual error estimated value, L (yi,F(xi)) indicate that loss function, F (x) indicate anticipation function, pass through
The regression forecasting function obtained after M-1 iteration is with FM-1(x) it indicates.It is learnt by gradient descent method, the negative gradient of loss function L
In "current" model F (xi) value when, L will most decline soon, that is, optimal model process.
For γimThe regression tree of fitting obtains the leaf node region R of the m treejm, wherein J is the quantity of leaf node, j
=1,2,3 ... ..., J.
J=1 to J (two layers of cycle), calculates：
Wherein, CjmRepresentative makes the corresponding output valve of leaf node that the loss function of prediction model minimizes, to j=1,2 ...
J, using linear search, estimates that each leaf node corresponds to the output valve of class, makes loss letter that is, to each leaf node on the tree
Number minimizes, and updates FM(x)；
It obtains
Final model can be expressed as：
Wherein, M represents the number of iteration, and J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, I generations
Table sample sheet.
S4, weight calculation is carried out according to driving force identification model, extracts the weighted score of driven factor, to obtain driving
The contribution rate of the factor.
In step s 4, by establishing model new one by one on the gradient direction of residual error reduction, according to loss function
The weight calculation of minimum final mask extracts weighted score, obtains the driven factor contribution of different land use classification variation
Rate；Accurately screen and identify using established optimal accuracy model the crucial driving of different land use classification variation because
Son；Then according to the contribution rate of feature land used status and its crucial driven factor, this feature land used status is determined
Crucial driven factor type and its influence degree.
S5, the correlativity between partial Correlation Analysis land use change survey and single driven factor is utilized.Partial correlation is to arrange
In addition to analyzing the correlativity between land use change survey and single driven factor under the influence of its dependent variable, iterative method point is utilized
Analyse the non-linear relation of each driven factor and feature land used status variation, most significantly correlated one group of driven factor with
Reciprocation between the variation of feature land used status.
S6, the degree of fitting for verifying driving force identification model, and identify mould using ten foldings crosscheck method verification driving force
The precision of type.In this step, reliable and stable model in order to obtain, cross-checks estimation stochastic gradient using ten foldings and is promoted
The arithmetic accuracy of model.Models fitting goodness is measured with Pesudo R square.The value of Pesudo R square illustrates mould closer to 1
The fitting degree of type is better；Conversely, the value of Pesudo R square is smaller, illustrate that the fitting degree of model is poorer.
The present embodiment combines the land use change survey made a living to be studied.
Data source：1. land use data using 12 scapes covers the remotely-sensed data that certain saves landsat satellites, (resolution ratio is
30m), the second stage of totally 24 scape remote sensing images of nineteen ninety-five and 2005, by image preprocessing, geographic registration, radiant correction, image
After the remote sensing image processings processes such as splicing, exercise supervision classification, and interpretation obtains six class land use patterns：Farmland, forest, grass
Ground, water body, bare area and construction land.Therefore, the land use data of nineteen ninety-five and 2005 is obtained.2. land use
Change driven factor data：Nineteen ninety-five and 2005, physiographic factor data include：Temperature (SAT) and height above sea level (ELE).Society
Economic factor data includes：Land use planning, land policy (policy), gross national product (GDP), it is national per capita
Total output value (GDPC), social retail product gross sales amount (TRSCG), primary industry gross national product (GDPP), secondary industry
Gross national product (GDPS), Gross National Product of Tertiary Industry (GDPT), Engel coefficient (EC), cottar's per capita income
(TIRI), total population density (TPOP).
Then according to land use and driven factor data, the land use change survey rate of transform and driven factor variation are calculated
Rate.Dependent variable of the land use change survey rate of transform as stochastic gradient lift scheme, driven factor change rate are independent variable.Its
The middle land use change survey rate of transform and driven factor change rate are calculated by EXCEL, and stochastic gradient lift scheme is simulated by R
The program of language editor is completed.
The analysis result of the present embodiment is as follows：
Fig. 3 shows that the Main Driving Factors that farmland changes between certain is saved 1995 to 2005 years are land policy (policy) and township
Village's per capita income (TIRI), contribution rate are respectively 42.9% and 39.7%, and contribution rate of accumulative total reaches 82.6%.It is other naturally
Reason and social ecnomicfactors contribution rate are relatively low.Social ecnomicfactors are the key that Jiangxi Province farmland variation driven factors.Partial correlation
Analysis shows farmland variation is in just with land policy, rural per capita income, the density of population and primary industry gross national product
Correlation, with the increase of land policy, rural per capita income, the density of population and primary industry gross national product, farmland variation
First constant be further added by is presented finally to keep constant.
Fig. 4 show the Main Driving Factors of construction land change between certain is saved 1995 to 2005 years be Engel coefficient (EC),
Gross National Product of Tertiary Industry (GDPT), gross national product (GDP) and rural per capita income (TIRI), contribution rate difference
It is 40.1%, 10.9%, 10.5% and 7.7%, contribution rate of accumulative total reaches 69.2%.Other physical geographys and social economy because
Sub- contribution rate is relatively low.Social ecnomicfactors are the key that Jiangxi Province's construction land change driven factors.Partial Correlation Analysis shows
Construction land change and Engel coefficient and secondary industry gross national product are negatively correlated, with Engel coefficient and second
The increase of industry gross national product, construction land change are presented first constant decline again and finally keep constant；Construction land change
It is proportionate with gross national product and rural per capita income, with the increase of gross national product and rural per capita income, agriculture
Field variation is presented first constant be further added by and finally keeps constant.
It finally tests to model, examines the fitting essence for showing that farmland and construction land stochastic gradient promote regression model
Degree is higher, and pseudo R square are respectively 0.82 and 0.89.
In conclusion the present invention has the following advantages：
(1) insensitive to multicollinearity, it is as a result more steady to missing data and outlier, it can efficiently handle not
The effect of the land use change survey driven factor of same type and up to thousands of potential driven factors.
(2) pass of land use change survey and its driven factor can be fitted well by not needing complicated model hypothesis
System.
(3) have analysis land use change survey and the interactive ability of its driven factor, overcome traditional land use
Change attribution method because interactive complexity in other models through commonly overlooked defect.
(4) it is not likely to produce the overfitting to data.
Therefore, the driven factor and its influence degree of stochastic gradient lift scheme identification land use change survey, essence can be applied
It really screens the crucial driven factor of land use change survey and quantifies and calculate its contribution rate, and the reciprocation between analysis factor.
Model method based on stochastic gradient boosting algorithm is that land use change survey attribution research opens completely new visual angle.
For the step number in embodiment, it is arranged only for the purposes of illustrating explanation, the sequence between step is not
Any restriction is done, the execution sequence of each step in embodiment can be adapted to according to the understanding of those skilled in the art
Property adjustment.
It is to be illustrated to the preferable implementation of the present invention, but the present invention is not limited to the embodiment above, it is ripe
Various equivalent variations or replacement can also be made under the premise of without prejudice to spirit of that invention by knowing those skilled in the art,
These equivalent deformations or replacement are all contained in the application claim limited range.
Claims (10)
1. a kind of recognition methods of Driving forces of land use change, it is characterised in that：Include the following steps：
Obtain the data of the driven factor in survey region；
According to the data of the driven factor of acquisition, optimization is iterated to prediction model using stochastic gradient lift method, is driven
Dynamic identification model；
The contribution rate of the driven factor of different land use variation is calculated according to driving force identification model.
2. a kind of recognition methods of Driving forces of land use change according to claim 1, it is characterised in that：The driving
The expression formula of power identification model is：
Wherein, FM(x) prediction model after M iteration optimization is represented, M represents the number of iteration, and J represents the number of leaf node
Amount, RjmRepresent the leaf node region of the m tree, CjmIt is corresponding to represent the leaf node for making the loss function of prediction model minimize
Output valve, I representative samples.
3. a kind of recognition methods of Driving forces of land use change according to claim 1, which is characterized in that further include with
Lower step：
Resolution ratio homogenization processing is carried out to the data of the driven factor got.
4. a kind of recognition methods of Driving forces of land use change according to claim 1, which is characterized in that further include with
Lower step：
Utilize the correlativity between partial Correlation Analysis land use change survey and single driven factor.
5. a kind of recognition methods of Driving forces of land use change according to claim 1, which is characterized in that further include with
Lower step：
Verify the degree of fitting of driving force identification model, and the essence using ten foldings crosscheck method verification driving force identification model
Degree.
6. a kind of recognition methods of Driving forces of land use change according to claim 1, it is characterised in that：The driving
The data of the factor include physical geography data and socioeconomic data, wherein：
The physical geography data include at least one of landform, the hydrology and meteorological data；
The socioeconomic data includes land use pattern figure, land use planning figure, location information, land policy, its people
Total output value, GNP per capita, social retail product gross sales amount, primary industry gross national product, secondary industry state
People's total output value, Gross National Product of Tertiary Industry, per capita output of grain, per capita production of meat, gross fixed assets investment, grace
Ge Er coefficients, gini index, town dweller per capita salary, cottar per capita salary, total population density and urban population density
At least one of.
7. a kind of recognition methods of Driving forces of land use change according to claim 1, it is characterised in that：The basis
Driving force identification model calculate different land use variation driven factor contribution rate, the step for be specially：
Weight calculation is carried out according to driving force identification model, the weighted score of driven factor is extracted, to obtain driven factor
Contribution rate.
8. a kind of identifying system of Driving forces of land use change, it is characterised in that：Including：
Acquisition module, the data for obtaining the driven factor in survey region；
Iteration module is used for the data of the driven factor according to acquisition, is changed to prediction model using stochastic gradient lift method
Generation optimization, obtains driving force identification model；
Computing module, the contribution rate of the driven factor for calculating different land use variation according to driving force identification model.
9. a kind of identifying system of Driving forces of land use change according to claim 8, it is characterised in that：The driving
The expression formula of power identification model is：
Wherein, M represents the number of iteration, and J represents the quantity of leaf node, RjmRepresent the leaf node region of the m tree, CjmRepresentative makes
The corresponding output valve of leaf node that the loss function of prediction model minimizes, I representative samples.
10. a kind of identification device of Driving forces of land use change, it is characterised in that：Including：
Memory, for storing program；
Processor, the method for loading described program to execute Driving forces of land use change as described in claim 1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810120621.1A CN108428007A (en) | 2018-02-07 | 2018-02-07 | A kind of recognition methods of Driving forces of land use change, system and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810120621.1A CN108428007A (en) | 2018-02-07 | 2018-02-07 | A kind of recognition methods of Driving forces of land use change, system and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108428007A true CN108428007A (en) | 2018-08-21 |
Family
ID=63156781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810120621.1A Pending CN108428007A (en) | 2018-02-07 | 2018-02-07 | A kind of recognition methods of Driving forces of land use change, system and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108428007A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1987906A (en) * | 2005-12-22 | 2007-06-27 | 中国煤炭地质总局航测遥感局 | Method for dynamicaly predicting land use change |
| CN102082433A (en) * | 2010-11-29 | 2011-06-01 | 沈阳工业大学 | Predicting device and method of voltage stability of wind power parallel network system |
| CN102646164A (en) * | 2012-02-28 | 2012-08-22 | 黄波 | Land use change modeling method and system implemented in combination with spatial filtering |
| KR101703972B1 (en) * | 2015-09-30 | 2017-02-22 | 한국지질자원연구원 | System and method for predicting groundwater potential area using spatial information |
| US20170091641A1 (en) * | 2015-09-30 | 2017-03-30 | Tata Consultancy Services Limited | Methods and systems for optimizing hidden markov model based land change prediction |
| CN107220732A (en) * | 2017-05-31 | 2017-09-29 | 福州大学 | A kind of power failure complaint risk Forecasting Methodology based on gradient boosted tree |
| CN107563539A (en) * | 2017-07-24 | 2018-01-09 | 佛山市顺德区中山大学研究院 | Short-term and long-medium term power load forecasting method based on machine learning model |
-
2018
- 2018-02-07 CN CN201810120621.1A patent/CN108428007A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1987906A (en) * | 2005-12-22 | 2007-06-27 | 中国煤炭地质总局航测遥感局 | Method for dynamicaly predicting land use change |
| CN102082433A (en) * | 2010-11-29 | 2011-06-01 | 沈阳工业大学 | Predicting device and method of voltage stability of wind power parallel network system |
| CN102646164A (en) * | 2012-02-28 | 2012-08-22 | 黄波 | Land use change modeling method and system implemented in combination with spatial filtering |
| KR101703972B1 (en) * | 2015-09-30 | 2017-02-22 | 한국지질자원연구원 | System and method for predicting groundwater potential area using spatial information |
| US20170091641A1 (en) * | 2015-09-30 | 2017-03-30 | Tata Consultancy Services Limited | Methods and systems for optimizing hidden markov model based land change prediction |
| CN107220732A (en) * | 2017-05-31 | 2017-09-29 | 福州大学 | A kind of power failure complaint risk Forecasting Methodology based on gradient boosted tree |
| CN107563539A (en) * | 2017-07-24 | 2018-01-09 | 佛山市顺德区中山大学研究院 | Short-term and long-medium term power load forecasting method based on machine learning model |
Non-Patent Citations (1)
| Title |
|---|
| 周双: "丘陵山区农田土壤有机碳密度影响因素与制图研究", 《中国优秀硕士学位论文全文数据库农业科技辑》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Zhang et al. | A feature selection and multi-model fusion-based approach of predicting air quality | |
| Gigović et al. | Testing a new ensemble model based on SVM and random forest in forest fire susceptibility assessment and its mapping in Serbia’s Tara National Park | |
| Khoshnood Motlagh et al. | Analysis and prediction of land cover changes using the land change modeler (LCM) in a semiarid river basin, Iran | |
| Quan et al. | GIS-based landslide susceptibility mapping using analytic hierarchy process and artificial neural network in Jeju (Korea) | |
| CN112200358B (en) | Landslide level prediction method, landslide level prediction device, landslide level prediction equipment and storage medium | |
| Wong et al. | High-resolution calculation of the urban vegetation fraction in the Pearl River Delta from the Sentinel-2 NDVI for urban climate model parameterization | |
| Ma et al. | Evaluating carbon fluxes of global forest ecosystems by using an individual tree-based model FORCCHN | |
| CN102646164A (en) | Land use change modeling method and system implemented in combination with spatial filtering | |
| CN110956412A (en) | Flood dynamic assessment method, device, medium and equipment based on real-scene model | |
| Alahmadi et al. | A comparison of small-area population estimation techniques using built-area and height data, Riyadh, Saudi Arabia | |
| Ohana-Levi et al. | Using satellite thermal-based evapotranspiration time series for defining management zones and spatial association to local attributes in a vineyard | |
| Zoraghein et al. | A spatial population downscaling model for integrated human-environment analysis in the United States | |
| CN108763673A (en) | The Driving forces of land use change screening technique and device returned based on LASSO | |
| Salman Aal-shamkhi et al. | Extraction and modeling of urban sprawl development in Karbala City using VHR satellite imagery | |
| Jamshidi et al. | Disaggregation of conventional soil map by generating multi realizations of soil class distribution (case study: Saadat Shahr plain, Iran) | |
| CN112200355B (en) | Landslide prediction method, landslide prediction device, landslide prediction equipment and storage medium | |
| Hales et al. | Bias correcting discharge simulations from the GEOGloWS global hydrologic model | |
| Momm et al. | Automated watershed subdivision for simulations using multi-objective optimization | |
| CN117290684B (en) | Transformer-based high-temperature drought weather early warning method and electronic equipment | |
| CN115829163B (en) | Multi-mode integration-based runoff prediction method and system for middle and lower reaches of Yangtze river | |
| CN113239112B (en) | Third-production growth amount visualization method and device based on GIS system | |
| CN108428007A (en) | A kind of recognition methods of Driving forces of land use change, system and device | |
| Irwin | Assessment of the regionalization of precipitation in two Canadian climate regions: A fuzzy clustering approach | |
| CN113254554B (en) | Urban block heat island modeling method and system based on map capturing and cluster learning | |
| CN116011319A (en) | Urban expansion simulation method based on driving factor analysis |
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 | ||
| CB02 | Change of applicant information |
Address after: Tianhe District Tianyuan road Guangzhou City, Guangdong province 510650 No. 808 Applicant after: Institute of ecological environment and soil, Guangdong Academy of Sciences Address before: Guangzhou City, Guangdong province 510650 Tianyuan Road No. 808 Applicant before: GUANGDONG INSTITUTE OF ECO-ENVIRONMENT AND SOIL SCIENCES |
|
| CB02 | Change of applicant information | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180821 |
|
| RJ01 | Rejection of invention patent application after publication |