CN113673777B - Desert succession prediction method under climate change condition - Google Patents
Desert succession prediction method under climate change condition Download PDFInfo
- Publication number
- CN113673777B CN113673777B CN202110992626.5A CN202110992626A CN113673777B CN 113673777 B CN113673777 B CN 113673777B CN 202110992626 A CN202110992626 A CN 202110992626A CN 113673777 B CN113673777 B CN 113673777B
- Authority
- CN
- China
- Prior art keywords
- desert
- distribution
- model
- data
- oasis
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 33
- 230000008859 change Effects 0.000 title claims abstract description 22
- 230000007704 transition Effects 0.000 claims abstract description 17
- 238000013507 mapping Methods 0.000 claims abstract description 8
- 238000012795 verification Methods 0.000 claims abstract description 4
- 238000000528 statistical test Methods 0.000 claims description 18
- 239000000203 mixture Substances 0.000 claims description 17
- 230000006870 function Effects 0.000 claims description 15
- 210000002569 neuron Anatomy 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000004088 simulation Methods 0.000 claims description 4
- 238000010276 construction Methods 0.000 claims description 3
- 238000000491 multivariate analysis Methods 0.000 claims description 3
- 239000002689 soil Substances 0.000 claims 1
- 230000008569 process Effects 0.000 abstract description 3
- 238000012360 testing method Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 238000010792 warming 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Tourism & Hospitality (AREA)
- Economics (AREA)
- Human Resources & Organizations (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Business, Economics & Management (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- Probability & Statistics with Applications (AREA)
- Life Sciences & Earth Sciences (AREA)
- Game Theory and Decision Science (AREA)
- Educational Administration (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Entrepreneurship & Innovation (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a desert succession prediction method under a climate change condition, which belongs to the technical field of desert control. Acquiring historical normalized vegetation index (NDVI) month scale grid point data, and obtaining three types of land types of a specific area, namely a desert, an oasis and a desert-oasis transition zone by a self-organizing mapping method; extracting historical rainfall data in the area range of the corresponding type; constructing a multi-model Bayesian discrimination set framework, and using the extracted historical rainfall data for verification of the framework; and predicting succession among desert, oasis and desert-oasis transition zones under the condition of future climate change by utilizing a trained multi-model Bayesian discrimination set framework. The invention can effectively process a large amount of uncertainty existing in rainfall prediction data of the global climate model by utilizing a multi-model Bayesian discrimination set framework, and objectively predicts desert succession of a selected area under the climate change condition from the statistical angle. Provides technical support for desert control.
Description
Technical Field
The invention belongs to the technical field of desert control, and particularly relates to a desert succession prediction method under a climate change condition.
Background
The desert expansion not only can deteriorate the ecological environment of the affected area and reduce the land bearing capacity, but also can have a profound effect on the social and economic life of the core area and the surrounding areas; global climate warming, on the other hand, causes extreme climate events to occur frequently; the drought risk in the surrounding areas of the desert is huge. Therefore, analyzing the response of the future desert to climate change is of great importance for sustainable development of the surrounding areas of the desert.
The existing desert succession prediction method under the climate change condition is mainly based on the rainfall simulation grid point data in the future, and the desert range of a specific area is identified according to the set threshold value; on the basis, the prediction of the expansion or reduction succession of the desert in the area is realized by utilizing rainfall data of one hundred years in the future predicted by the global climate model. However, the existing methods have the following disadvantages: (1) setting the threshold is too subjective; (2) lack of relevant definition of desert-oasis transition zone; (3) Uncertainty of the simulation result of the global climate model is difficult to process, and an ideal succession prediction effect cannot be obtained.
Disclosure of Invention
The invention aims to provide a desert succession prediction method under a climate change condition, which is characterized by comprising the following steps of:
step one: acquiring month scale grid point data of a normalized vegetation index of the last 10 years, and obtaining three types of land types of a specific area, namely a desert, an oasis and a desert-oasis transition zone by a self-organizing mapping method;
step two: acquiring month scale rainfall data simulated by a plurality of sets of global climate models, and extracting historical rainfall data in the area range of the corresponding type based on the three types of land types;
step three: fitting rainfall data statistical distribution of a corresponding type by using a Gaussian mixture model, wherein parameters related to the Gaussian mixture model comprise normal distribution weight coefficients, average values and variances;
step four: constructing a multi-model Bayesian distinguishing set framework based on the fitted Gaussian mixture distribution corresponding to each type, and using the extracted historical rainfall data for verification of the framework;
the multi-model Bayesian discriminant set framework is constructed based on Bayesian discriminant analysis, wherein the Bayesian discriminant analysis is to input new sample observation data under the condition of known sample classification, calculate posterior probabilities corresponding to all classes and judge the attribution class of the new sample by comparing the posterior probabilities; the method of calculating the posterior probability can be expressed as:
wherein p (pi) k I x) is the posterior probability corresponding to the kth class, p (x|pi) k ) For the prior probability corresponding to the kth class, p (pi k ) A likelihood function corresponding to the kth class;
the construction of the multi-model Bayesian discrimination set framework further comprises the following sub-steps:
step four, first: the rainfall data statistical distribution of each set of global climate model corresponding to the area range of three types of land types of desert, oasis and desert-oasis transition zone is known, namely likelihood function p (pi) in Bayesian discriminant analysis which can be obtained in the step three k ) The method comprises the steps of carrying out a first treatment on the surface of the Calculating a priori probability p (x|pi) k ) The calculation method comprises the following steps:
wherein n is k For the number of corresponding grids of rainfall data in the range of the area where various land types are located, n total Counting total grid points of rainfall data in a selected area;
step four, two: for each set of global climate models, the global climate model is determined by the prior probability p (x|pi k ) And likelihood function p (pi k ) Calculating posterior probability of various land types;
and step four, three: constructing posterior probability sets of various land types corresponding to a plurality of sets of global climate models to form lattice point data;
and step four: using a multivariate analysis of variance method, namely F statistical test based on Wilks criterion, wherein the F statistical test is one form of statistical test, and the statistical test quantity obeys F-distribution; corresponding to a certain lattice point, using posterior probability set samples of various land types to approximate the F statistical test quantity: if the difference between the classes is obvious, the lattice point is judged to be the class corresponding to the maximum posterior probability set; otherwise, the lattice point is judged to be a fourth class, namely a class which can not be objectively judged; where Λ is a Wilks value, which can be expressed as:
in SS (x) E Is the sum of squares of errors, SS H Is the sum of the total squares; the F statistical test quantity can be approximated by calculation of Wilks values:
wherein d is the dimension of the variable, n e And n f The number of samples for two subclasses;
step four, five: based on the land type judged by each grid point of the selected area, the extracted historical rainfall data is input, the land type obtained by Bayesian judgment is compared with the original land type, and the misjudgment rate is calculated to verify the effectiveness of the framework; the false positive rate calculation method can be expressed as:
E i ＝n i,s /n i,t
wherein E is i For the misjudgment rate of the ith land type, n i,s In the range of the region where the ith land type is positioned, the number of lattice points, n, of which the type obtained by Bayesian discrimination is identical with the original land type i,t The total grid point number in the area range of the ith land type is set;
step five: and inputting a plurality of sets of global climate model rainfall prediction data, and predicting succession among desert, oasis and desert-oasis transition zones under the condition of future climate change by utilizing a constructed and trained multi-model Bayesian discrimination set framework.
The self-organizing mapping method in the first step is to search an optimal reference vector set to classify an input data set, train a connection weight vector corresponding to each reference vector, and realize self-adapting adjustment of a network, thereby completing pattern classification. Wherein the weight vector change can be expressed as:
wherein eta is the learning rate, ti is the topological distance between adjacent neurons and winning neurons, t max Is the maximum topological distance between adjacent neurons and winning neurons, w is the weight vector, w old Weight vector corresponding to last state, d i Is a multidimensional data vector to be input.
Fitting the rainfall data statistical distribution of the corresponding type by using the Gaussian mixture model in the third step refers to effective simulation of the rainfall data statistical distribution corresponding to each land type by using the Gaussian mixture model; its density function can be expressed as:
in which θ= (Φ) 1 ,...,φ m ,μ 1 ,...,μ m ,σ 1 ,...,σ m )，f i Representing an ith normal distribution probability density function; phi i Weight coefficient, μ representing the i-th distribution i Mean value sigma representing ith distribution i Representing the variance of the ith distribution; i represents the number; m represents the final number of i.
The invention has the beneficial effects that: (1) The invention adopts the self-organizing mapping method to effectively simulate the land type of the selected area, and solves the technical problem that the conventional rainfall threshold selection is difficult to objectively select. In addition, the normalized vegetation indexes in the area range of the divided land types meet the objective facts of the corresponding land types, so that the accurate definition of the desert-oasis transition zone is realized; (2) The Gaussian mixture model can effectively fit the rainfall random distribution in the area range of each land type, and overcomes the defect that the conventional random distribution (such as normal distribution, poisson distribution and the like) cannot pass hypothesis test when the rainfall distribution is fit. (3) The random distribution fitted by the Gaussian mixture model is well matched with the experience distribution of the observed data; the fit of the random distribution of rainfall over the area of each land type passed the hypothesis test (B value > 0.05). (4) The multi-model Bayesian discrimination set framework can effectively process a large amount of uncertainty existing in rainfall prediction data of the global climate model, and objectively predict desert succession of a selected area under the climate change condition from a statistical angle. Provides technical support for desert control.
Drawings
FIG. 1 is a schematic flow chart of a desert succession prediction method.
Fig. 2 is a schematic diagram of a rainfall statistical distribution of a northern africa desert land.
Fig. 3 is a schematic diagram of rainfall statistical distribution of the northern africa desert-oasis transition zone land.
Fig. 4 is a schematic diagram of a statistical distribution of rainfall on northern africa land.
Detailed Description
The invention provides a desert succession prediction method under a climate change condition, and the invention is further described below with reference to the accompanying drawings.
As shown in fig. 1, the desert succession prediction method under the climate change condition provided by the invention specifically comprises the following steps:
step one: in the embodiment, the saharan desert in North Africa is taken as a selected area, grid point data of a normalized vegetation index month scale in 2011-2020 of the area is obtained, and three types of land types of the area, namely the saharan desert, the oasis and a desert-oasis transition zone are obtained through a self-organizing mapping method;
in this embodiment, the self-organizing mapping method refers to searching an optimal reference vector set to classify an input data set, and training a connection weight vector corresponding to each reference vector to realize self-adapting adjustment of a network, thereby completing pattern classification. Wherein the weight vector change can be expressed as:
wherein eta is learning rate and t i Is the topological distance t between adjacent neurons and winning neurons max Is the maximum topological distance between adjacent neurons and winning neurons, w is the weight vector, w old Weight vector corresponding to last state, d i Is a multidimensional data vector to be input; i represents the number; m represents the final number of i.
The self-organizing mapping method in the first step can effectively simulate the land type of the selected area, and solves the technical problem that the conventional rainfall threshold selection is difficult to objectively select. In addition, the normalized vegetation indexes in the area range of the divided land types all accord with objective facts of the corresponding land types, and the accurate definition of the desert-oasis transition zone is realized, as shown in table 1.
Table 1 normalized Vegetation index and Bayesian discriminant analysis misjudgment Rate
Table 1 shows normalized vegetation indexes corresponding to the three land types obtained by clustering by a self-organizing map method, and the corresponding values meet objective facts from the annual average, the dry average and the wet average, so that the accurate definition of the desert-oasis transition zone is realized;
step two: in the embodiment, month scale rainfall data simulated by 13 sets of global climate models are obtained, and based on the three types of land types, historical rainfall data in the area range where the corresponding types are located are extracted;
step three: fitting rainfall data statistical distribution of a corresponding type by using a Gaussian mixture model, wherein parameters related to the Gaussian mixture model comprise normal distribution weight coefficients, average values and variances;
in this example, the statistical distribution of rainfall in the regions of the saharan desert, oasis, and desert-oasis transition zone in north africa was fitted separately using a mixed gaussian model. Fig. 2-4 show the corresponding statistical distribution of rainfall for the three land types in north africa, respectively. As shown, the random distribution fitted by the gaussian mixture model is well-matched with the empirical distribution of the observed data. The model density function can be expressed as:
in which θ= (Φ) 1 ,...,φ m ,μ 1 ,...,μ m ,σ 1 ,...,σ m )，f i Representing an ith normal distribution probability density function; phi i Weight coefficient, μ representing the i-th distribution i Mean value sigma representing ith distribution i Representing the variance of the ith distribution; i represents the number; m represents the final number of i.
The Gaussian mixture model in the third step can effectively fit the rainfall random distribution in the area range of each land type, and overcomes the defect that the conventional random distribution (such as normal distribution, poisson distribution and the like) cannot pass the hypothesis test when the rainfall distribution is fitted. As shown in fig. 2-4, the random distribution fitted by the mixed gaussian model is well matched with the empirical distribution of the observed data; the fit of the random distribution of rainfall over the area of each land type passed the hypothesis test (B value > 0.05).
Step four: constructing a multi-model Bayesian distinguishing set framework based on the fitted Gaussian mixture distribution corresponding to each type, and using the extracted historical rainfall data for verification of the framework;
in this embodiment, first, the multi-model bayesian discriminant set framework is constructed in the fourth step based on bayesian discriminant analysis. The bayesian discriminant analysis is to input new sample observation data under the condition of known sample classification, calculate posterior probabilities corresponding to each category, and judge the attribution category of the new sample by comparing the posterior probabilities. The method of calculating the posterior probability can be expressed as:
wherein p (pi) k |x) is the posterior probability of the kth class distribution, p (x|pi) k ) For the prior probability of the kth class distribution, p (pi k ) Likelihood functions for a k-th class distribution; the detailed multi-model bayesian discriminant collection framework construction may further include the sub-steps of:
step four, first: the rainfall data statistical distribution (which can be obtained in the step three) of each set of global climate model corresponding to the region range of three types of land types of North Africa desert, oasis and desert-oasis transition zone is known, namely likelihood function p (pi) in Bayesian discriminant analysis k ). Calculating a priori probability p (x|pi) k ) The calculation method comprises the following steps:
wherein n is k For the number of corresponding grids of rainfall data in the range of the area where various land types are located, n total Total grid points for rainfall data in northern africa area.
Step four, two: for each set of global climate models, the global climate model is determined by the prior probability p (x|pi k ) And likelihood function p (pi k ) Calculating posterior probability of various land types;
and step four, three: constructing posterior probability sets of various land types corresponding to a plurality of sets of global climate models to form lattice point data;
and step four: using a multivariate analysis of variance method, namely F statistical test based on Wilks criterion, wherein the F statistical test is one form of statistical test, and the statistical test quantity obeys F-distribution; corresponding to a certain lattice point, using posterior probability set samples of various land types to approximate the F statistical test quantity: if the difference between the classes is obvious, the lattice point is judged to be the class corresponding to the maximum posterior probability set; otherwise, the lattice point is judged to be a fourth class, namely a class which can not be objectively judged; where Λ is a Wilks value, which can be expressed as:
in SS (x) E Is the sum of squares of errors, SS H Is the sum of the total squares; the F statistical test quantity can be approximated by calculation of Wilks values:
wherein d is the dimension of the variable, n e And n f The number of samples for two subclasses;
step four, five: based on the land type determined by each grid point in the northern Africa area, the extracted historical rainfall data is input, the land type determined by Bayesian determination is compared with the original land type, and the misjudgment rate is calculated to verify the validity of the framework. The false positive rate calculation method can be expressed as:
E i ＝n i,s /n i,t
wherein E is i For the misjudgment rate of the ith land type, n i,s In the range of the region where the ith land type is positioned, the number of lattice points, n, of which the type obtained by Bayesian discrimination is identical with the original land type i,t The total grid number in the range of the area where the ith land type is located.
In the embodiment, the misjudgment rate corresponding to each land type is shown in table 1 and is within an acceptable range.
Step five: 13 sets of global climate model rainfall prediction data are input, and the succession among the northern African saharan desert, oasis and desert-oasis transition zones under the future climate change condition is predicted by utilizing a constructed and trained multi-model Bayesian discrimination set framework. The invention predicts the desert succession of the selected area under the climate change condition more objectively from the statistical angle. Providing a technical basis for processing uncertainty of the global climate model; provides technical support for desert control.
Claims (3)
1. The desert succession prediction method under the climate change condition is characterized by comprising the following steps of:
step one: acquiring month scale grid point data of a normalized vegetation index of the last 10 years, and obtaining three types of land types of a specific area, namely a desert, an oasis and a desert-oasis transition zone by a self-organizing mapping method;
step two: acquiring month scale rainfall data simulated by a plurality of sets of global climate models, and extracting historical rainfall data in the area range of the corresponding type based on the three types of land types;
step three: fitting rainfall data statistical distribution of a corresponding type by using a Gaussian mixture model, wherein parameters related to the Gaussian mixture model comprise normal distribution weight coefficients, average values and variances;
step four: constructing a multi-model Bayesian distinguishing set framework based on the fitted Gaussian mixture distribution corresponding to each type, and using the extracted historical rainfall data for verification of the framework;
the multi-model Bayesian discriminant set framework is constructed based on Bayesian discriminant analysis, wherein the Bayesian discriminant analysis is to input new sample observation data under the condition of known sample classification, calculate posterior probabilities corresponding to all classes and judge the attribution class of the new sample by comparing the posterior probabilities; the method of calculating the posterior probability can be expressed as:
wherein p (pi) k I x) is the posterior probability corresponding to the kth class, p (x|pi) k ) For the prior probability corresponding to the kth class, p (pi k ) A likelihood function corresponding to the kth class;
the construction of the multi-model Bayesian discrimination set framework further comprises the following sub-steps:
step four, first: the areas of the three types of soil types of desert, oasis and desert-oasis transition zone are known to be correspondingStatistical distribution of rainfall data of each set of global climate model, namely likelihood function p (pi) in Bayesian discriminant analysis which can be obtained in step three k ) The method comprises the steps of carrying out a first treatment on the surface of the Calculating a priori probability p (x|pi) k ) The calculation method comprises the following steps:
wherein n is k For the number of corresponding grids of rainfall data in the range of the area where various land types are located, n total Counting total grid points of rainfall data in a selected area;
step four, two: for each set of global climate models, the global climate model is determined by the prior probability p (x|pi k ) And likelihood function p (pi k ) Calculating posterior probability of various land types;
and step four, three: constructing posterior probability sets of various land types corresponding to a plurality of sets of global climate models to form lattice point data;
and step four: using a multivariate analysis of variance method, namely F statistical test based on Wilks criterion, wherein the F statistical test is one form of statistical test, and the statistical test quantity obeys F-distribution; corresponding to a certain lattice point, using posterior probability set samples of various land types to approximate the F statistical test quantity: if the difference between the classes is obvious, the lattice point is judged to be the class corresponding to the maximum posterior probability set; otherwise, the lattice point is judged to be a fourth class, namely a class which can not be objectively judged; where Λ is a Wilks value, which can be expressed as:
in SS (x) E Is the sum of squares of errors, SS H Is the sum of the total squares; the F statistical test quantity can be approximated by calculation of Wilks values:
wherein d is the dimension of the variable, n e And n f The number of samples for two subclasses;
step four, five: based on the land type judged by each grid point of the selected area, the extracted historical rainfall data is input, the land type obtained by Bayesian judgment is compared with the original land type, and the misjudgment rate is calculated to verify the effectiveness of the framework; the false positive rate calculation method can be expressed as:
E i ＝n i,s /n i,t
wherein E is i For the misjudgment rate of the ith land type, n i,s In the range of the region where the ith land type is positioned, the number of lattice points, n, of which the type obtained by Bayesian discrimination is identical with the original land type i,t The total grid point number in the area range of the ith land type is set;
step five: and inputting a plurality of sets of global climate model rainfall prediction data, and predicting succession among desert, oasis and desert-oasis transition zones under the condition of future climate change by utilizing a constructed and trained multi-model Bayesian discrimination set framework.
2. The method for predicting desert succession under climate change condition according to claim 1, wherein the self-organizing map in the first step is to search an optimal reference vector set to classify an input data set, and train a connection weight vector corresponding to each reference vector to realize self-adapting adjustment of a network, thereby completing pattern classification, wherein the weight vector change can be expressed as:
wherein eta is learning rate and t i Is the topological distance t between adjacent neurons and winning neurons max Is the maximum topological distance between adjacent neurons and winning neurons, w is the weight vector, w old Weight vector corresponding to last state, d i For multidimensional data vector to be input, table iThe number is shown.
3. The desert succession prediction method under the climate change condition according to claim 1, wherein the step three is characterized in that the fit of the Gaussian mixture model to the rainfall data statistical distribution of the corresponding type refers to the realization of effective simulation of the rainfall data statistical distribution corresponding to each land type by utilizing the Gaussian mixture model; its density function can be expressed as:
in which θ= (Φ) 1 ,...,φ m ,μ 1 ,...,μ m ,σ 1 ,...,σ m )，f i Representing an ith normal distribution probability density function; phi i Weight coefficient, μ representing the i-th distribution i Mean value sigma representing ith distribution i Representing the variance of the ith distribution; i represents the number; m represents the final number of i.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110992626.5A CN113673777B (en) | 2021-08-27 | 2021-08-27 | Desert succession prediction method under climate change condition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110992626.5A CN113673777B (en) | 2021-08-27 | 2021-08-27 | Desert succession prediction method under climate change condition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113673777A CN113673777A (en) | 2021-11-19 |
| CN113673777B true CN113673777B (en) | 2024-02-02 |
Family
ID=78546734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110992626.5A Active CN113673777B (en) | 2021-08-27 | 2021-08-27 | Desert succession prediction method under climate change condition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113673777B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116882325B (en) * | 2023-09-07 | 2024-01-26 | 长江三峡集团实业发展(北京)有限公司 | Evaluation method and device for pollutant transportation process in desert area |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108761574A (en) * | 2018-05-07 | 2018-11-06 | 中国电建集团北京勘测设计研究院有限公司 | Rainfall evaluation method based on Multi-source Information Fusion |
| CN112326574A (en) * | 2020-11-04 | 2021-02-05 | 暨南大学 | Spectrum wavelength selection method based on Bayesian classification |
-
2021
- 2021-08-27 CN CN202110992626.5A patent/CN113673777B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108761574A (en) * | 2018-05-07 | 2018-11-06 | 中国电建集团北京勘测设计研究院有限公司 | Rainfall evaluation method based on Multi-source Information Fusion |
| CN112326574A (en) * | 2020-11-04 | 2021-02-05 | 暨南大学 | Spectrum wavelength selection method based on Bayesian classification |
Non-Patent Citations (5)
| Title |
|---|
| 买买提依明・吾买尔 ; 阿布都热西提・阿布都外力 * |
| 沙漠扩散预测模型的研究;马衣拉・木沙江;阿布都热西提・阿布都外力;;新疆农业科学(05);190-195 * |
| 甘肃环县沙漠化时空演变及风险性预测分析;丁文广;王丽娜;耿怡颖;;兰州大学学报(自然科学版)(01);48-55 * |
| 绿洲―沙漠交错带沙漠扩散预测模型――以新疆民丰县为例;买买提依明・吾买尔;阿布都热西提・阿布都外力;;新疆农业科学(06);240-244 * |
| 马衣拉・木沙江 ; 阿布都热西提・阿布都外力 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113673777A (en) | 2021-11-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sauquet et al. | Comparison of catchment grouping methods for flow duration curve estimation at ungauged sites in France | |
| Bardossy et al. | Fuzzy rule-based downscaling of precipitation | |
| CN115270965B (en) | Power distribution network line fault prediction method and device | |
| Shrestha et al. | Data‐driven approaches for estimating uncertainty in rainfall‐runoff modelling | |
| CN109840873A (en) | A kind of Cross Some Region Without Data Hydro-Model Parameter Calibration Technology fields method based on machine learning | |
| CN111665575B (en) | Medium-and-long-term rainfall grading coupling forecasting method and system based on statistical power | |
| Li et al. | Real-time flood forecast using the coupling support vector machine and data assimilation method | |
| CN106778838A (en) | A kind of method for predicting air quality | |
| Valverde et al. | Neural network and fuzzy logic statistical downscaling of atmospheric circulation-type specific weather pattern for rainfall forecasting | |
| Zhao et al. | The application of BP neural networks to analysis the national vulnerability. | |
| CN111860576A (en) | Endometrium tumor classification labeling method based on random forest | |
| CN110942182A (en) | Method for establishing typhoon prediction model based on support vector regression | |
| CN113673777B (en) | Desert succession prediction method under climate change condition | |
| CN113887635B (en) | Basin similarity classification method and classification device | |
| CN117808214A (en) | Hydraulic engineering data analysis system | |
| CN117540303A (en) | Landslide susceptibility assessment method and system based on cross semi-supervised machine learning algorithm | |
| CN113377750A (en) | Hydrological data cleaning method and system | |
| Lu et al. | Uncertainty quantification of machine learning models to improve streamflow prediction under changing climate and environmental conditions | |
| Verellen et al. | Using data-driven models to estimate the energy use of buildings based on a building stock model | |
| AhmadEbrahimpour et al. | Application of global precipitation dataset for drought monitoring and forecasting over the Lake Urmia basin with the GA-SVR model | |
| CN105279308A (en) | Oceanic whitecap coverage algorithm based on successive data rejection | |
| CN117933316B (en) | Groundwater level probability forecasting method based on interpretable Bayesian convolution network | |
| Tsang et al. | Region of influence method improves macroinvertebrate predictive models in Maryland | |
| CN103955953A (en) | Terrain collaborative variable selection method for digital soil cartography | |
| Shrestha et al. | Assessing model prediction limits using fuzzy clustering and machine learning |
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 |