CN117056858A - Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data - Google Patents
Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data Download PDFInfo
- Publication number
- CN117056858A CN117056858A CN202310939608.XA CN202310939608A CN117056858A CN 117056858 A CN117056858 A CN 117056858A CN 202310939608 A CN202310939608 A CN 202310939608A CN 117056858 A CN117056858 A CN 117056858A
- Authority
- CN
- China
- Prior art keywords
- model
- attribute data
- soil attribute
- soil
- nonlinear
- 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
- 239000002689 soil Substances 0.000 title claims abstract description 120
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000010606 normalization Methods 0.000 claims abstract description 67
- 230000004927 fusion Effects 0.000 claims abstract description 38
- 230000007613 environmental effect Effects 0.000 claims abstract description 18
- 239000011159 matrix material Substances 0.000 claims description 31
- 238000012545 processing Methods 0.000 claims description 10
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000007499 fusion processing Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 238000007500 overflow downdraw method Methods 0.000 abstract description 4
- 230000006870 function Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 5
- 238000013528 artificial neural network Methods 0.000 description 5
- 229910052799 carbon Inorganic materials 0.000 description 5
- 238000012417 linear regression Methods 0.000 description 5
- 238000012549 training Methods 0.000 description 5
- 238000012795 verification Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a soil attribute space prediction method integrating multisource data and space autocorrelation thereof, which is used for collecting environmental variables of a target area; inputting the environment variables of the target area into a pre-trained linear relation model of soil attribute data and environment variables and a pre-trained nonlinear relation model of soil attribute data and environment variables respectively to obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value; and fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by using a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area. The advantages are that: the precision of the prediction result is greatly improved compared with the original fused data, and is better than that of the conventional linear and nonlinear fusion method; the invention can keep good prediction performance under various complex environments by considering the spatial autocorrelation of soil attributes.
Description
Technical Field
The invention relates to a soil attribute space prediction method for fusing multisource data and space autocorrelation thereof, and belongs to the technical field of digital soil mapping and data fusion.
Background
In current research on predicting basic soil attribute data such as soil organic carbon, the linear or nonlinear relationship between a plurality of environmental factors and predicted soil attributes is often described by constructing a plurality of prediction models. A common linear model is a multiple linear regression model (multiple linear regression, MLR) and a geographically weighted regression model (geographically weighted regression, GWR). Wherein GWR is an expansion of a common linear regression model, which follows the spatial autocorrelation proposed by the first law of geography and innovatively applies the idea of local regression. The weights of the data in the partial regression equation are calculated by introducing the spatial positions of the data, so that the parameters of the dependent variable and the explanatory variable of each position can be estimated under the consideration of the spatial weights of adjacent points. Because the soil property has spatial autocorrelation, the prediction effect of GWR is superior to that of a common linear regression model based on a least square method. And a series of models are derived by taking GWR as a core idea. The multiscale geographic weighted regression (multiscale geographically weighted regression, MGWR) method is expansion of GWR, is not limited by the fact that all modeling processes are in the same spatial scale, and has better adaptability in large-scale and environment complex areas. MGWRs take into account changes in the spatial scale of relationships by allocating a separate bandwidth for each relationship in the model, as opposed to standard GWR using the same bandwidth for all relationships between environmental variables and predicted variables in the model. Furthermore, unlike the weighted least squares algorithm used in the standard GWR model, MGWR uses a backward fitting algorithm. In terms of nonlinear models, the commonly used machine learning models include Random Forest (RF), support vector machines (support vector machine, SVM), artificial neural networks (artificial neural network, ANN), and the like. Nonlinear machine learning methods are generally superior to linear methods in training because the relationship between soil properties and environmental variables is complex and not linear in nature. Among these nonlinear methods, artificial neural networks are widely used due to their strong adaptive nonlinear fitting capability.
Although nonlinear methods such as artificial neural networks have higher accuracy in predicting basic soil property data such as soil organic carbon than ordinary linear methods, they still have drawbacks, one of which is a lack of consideration for spatial autocorrelation. This means that they only consider the relationship between the soil property and the environmental variable at that location, ignoring the influence of neighboring environmental variables, when predicting the soil property.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a soil attribute space prediction method integrating multisource data and space autocorrelation thereof.
In order to solve the technical problems, the invention provides a soil attribute space prediction method integrating multisource data and space autocorrelation thereof, which comprises the following steps:
collecting environment variables of a target area;
inputting the environment variables of the target area into a pre-trained linear relation model of soil attribute data and environment variables and a pre-trained nonlinear relation model of soil attribute data and environment variables respectively to obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by using a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
Furthermore, the linear relation model of the soil attribute data and the environment variable adopts an ANN model.
Furthermore, the nonlinear relation model of the soil attribute data and the environment variable adopts an MGWR model.
Further, the fusion process of the multi-scale geographic normalization weighted fusion model is as follows:
acquiring a variation of the multi-scale geographic normalization weighted fusion model and a target grid to be predicted;
dividing all grids into four quadrants on a plane by taking a target grid to be predicted as a center, and acquiring a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value of each grid layer in a range of a variation; all grids are grid data of a target area generated by map software, wherein the grid data consists of a plurality of grids;
acquiring a predetermined optimal normalization index and an optimal bandwidth of a corresponding linear model and an optimal normalization index and an optimal bandwidth of a corresponding linear non-nature model;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding linear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding linear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the linear model soil attribute data predictive values of all grid layers to obtain an independent variable matrix corresponding to the linear model; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the linear model to obtain a final value corresponding to the linear model;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding nonlinear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding nonlinear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the predicted values of the soil attribute data of the nonlinear models of all grid layers to obtain independent variable matrixes of the corresponding nonlinear models; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the nonlinear model to obtain a final value corresponding to the nonlinear model;
and adding the final value corresponding to the linear model and the final value corresponding to the nonlinear model to obtain the final predicted value of the soil attribute data of the target grid.
Further, the obtaining process of the optimal normalization index and the optimal bandwidth of the corresponding linear model and the optimal normalization index and the optimal bandwidth of the corresponding linear non-linear model specifically includes:
randomly generating a plurality of groups of normalization indexes and bandwidths corresponding to the linear models and normalization indexes and bandwidths corresponding to the nonlinear models in a preset range, wherein the sum of the normalization indexes of the linear models and the normalization indexes of the nonlinear models is 1;
obtaining an actual measurement value of soil attribute data of any grid in a target area;
taking the grid as a target grid, taking the normalization index and the bandwidth of each group of corresponding linear models and the normalization index and the bandwidth of corresponding nonlinear models as normalization indexes and bandwidths used in a multi-scale geographic normalization weighted fusion model, and calculating through the multi-scale geographic normalization weighted fusion model to obtain a plurality of final predicted values of soil attribute data corresponding to the grid;
and solving the mean square error of the final predicted value of the soil attribute data corresponding to the grid and the actual measured value of the soil attribute data of the grid, wherein the normalized index and the bandwidth of the corresponding linear model and the normalized index and the bandwidth of the corresponding nonlinear model of a group corresponding to the final predicted value of the soil attribute data with the minimum mean square error are the optimal normalized index and the optimal bandwidth of the corresponding linear model and the optimal normalized index and the optimal bandwidth of the corresponding linear nonlinear model.
Further, the expression of the multi-scale geographic normalization weighted fusion model is as follows:
wherein y is i ,X ij ,The independent variable matrix and the space weight matrix of the j-th fused variable obtained by searching the ith grid in the range of the variation range are respectively corresponding to the predicted value of the ith grid; h is the number of fused variables; p is p j ,bw j Are respectively toNormalized index and bandwidth of the jth fused variable; />The spatial weight matrix representing the jth fused variable is normalized to (0, p) j ) So that->The sum of all the spatial weight values in (a) is p j 。
A soil property spatial prediction apparatus that fuses multisource data and spatial autocorrelation thereof, comprising:
the acquisition module is used for acquiring environment variables of the target area;
the prediction module is used for respectively inputting the environmental variables of the target area into a pre-trained linear relation model of soil attribute data and environmental variables and a pre-trained nonlinear relation model of the soil attribute data and the environmental variables to respectively obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and the fusion module is used for fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by utilizing a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
A computer readable storage medium storing one or more programs, wherein the one or more programs comprise instructions, which when executed by a computing device, cause the computing device to perform any of the methods.
A computer device, comprising,
one or more processors, memory, and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for performing any of the methods.
The invention has the beneficial effects that: the fusion prediction result obtained by the method has higher precision than the original fused data, and is also superior to the conventional linear and nonlinear fusion methods. Meanwhile, the spatial autocorrelation of soil properties is considered in the fusion process, so that the method has the strongest adaptability to the whole environment of the prediction area, namely, the method can keep better prediction performance in various complex environments.
Drawings
FIG. 1 is a schematic flow chart of the present invention;
FIG. 2 is a schematic diagram of a model of the present invention;
FIG. 3 is a flow chart of an embodiment;
FIG. 4 is a schematic diagram of accuracy verification comparison;
FIG. 5 is a graph showing a comparison of spatial predictions of organic carbon reserves in soil using various methods.
Detailed Description
The invention is further described below with reference to the accompanying drawings. The following examples are only for more clearly illustrating the technical aspects of the present invention, and are not intended to limit the scope of the present invention.
Embodiment 1, as shown in fig. 1 and 3, this embodiment describes a soil attribute spatial prediction method for fusing multisource data and spatial autocorrelation thereof, which includes:
collecting environment variables of a target area;
inputting the environment variables of the target area into a pre-trained linear relation model of soil attribute data and environment variables and a pre-trained nonlinear relation model of soil attribute data and environment variables respectively to obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by using a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
The fusion process of the multi-scale geographic normalization weighted fusion model is as follows:
acquiring a variation of the multi-scale geographic normalization weighted fusion model and a target grid to be predicted;
dividing all grids into four quadrants on a plane by taking a target grid to be predicted as a center, and acquiring a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value of each grid layer in a range of a variation; all grids are raster data composed of a plurality of grids of a target area generated by using ArcMap.
The calculation of the variation is performed on ArcMap, i.e. the values of all grids predicted in the investigation region by the two models being fused (in this case the ANN model and the MGWR model). The calculation of the variation is based on a half-variant function. When the distance between sampling points increases, the semi-variation function reaches a relatively stable constant from an initial value, and the constant value is called a base station value; when the value of the half variation function reaches the base station value from the initial value, the interval distance of the sampling points is called a variation range. And taking the minimum value of the variation in the ANN model and the MGWR model as the calculation range of the fusion model, namely the variation obtained by the method.
The use of existing training methods for obtaining trained ANN models and MGWR models is not part of the present invention and is not further described herein.
As shown in fig. 2, the optimal normalization index and the optimal bandwidth of the corresponding linear model and the optimal normalization index and the optimal bandwidth of the corresponding linear non-nature model are obtained in advance;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding linear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding linear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the linear model soil attribute data predictive values of all grid layers to obtain an independent variable matrix corresponding to the linear model; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the linear model to obtain a final value corresponding to the linear model;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding nonlinear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding nonlinear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the predicted values of the soil attribute data of the nonlinear models of all grid layers to obtain independent variable matrixes of the corresponding nonlinear models; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the nonlinear model to obtain a final value corresponding to the nonlinear model;
and adding the final value corresponding to the linear model and the final value corresponding to the nonlinear model to obtain the final predicted value of the soil attribute data of the target grid.
The obtaining process of the optimal normalization index and the optimal bandwidth of the corresponding linear model and the optimal normalization index and the optimal bandwidth of the corresponding linear non-linear model specifically comprises the following steps:
randomly generating a plurality of groups of normalization indexes and bandwidths corresponding to the linear models and normalization indexes and bandwidths corresponding to the nonlinear models in a preset range, wherein the sum of the normalization indexes of the linear models and the normalization indexes of the nonlinear models is 1;
obtaining an actual measurement value of soil attribute data of any grid in a target area;
taking the grid as a target grid, taking the normalization index and the bandwidth of each group of corresponding linear models and the normalization index and the bandwidth of corresponding nonlinear models as normalization indexes and bandwidths used in a multi-scale geographic normalization weighted fusion model, and calculating through the multi-scale geographic normalization weighted fusion model to obtain a plurality of final predicted values of soil attribute data corresponding to the grid;
and solving the mean square error of the final predicted value of the soil attribute data corresponding to the grid and the actual measured value of the soil attribute data of the grid, wherein the normalized index and the bandwidth of the corresponding linear model and the normalized index and the bandwidth of the corresponding nonlinear model of a group corresponding to the final predicted value of the soil attribute data with the minimum mean square error are the optimal normalized index and the optimal bandwidth of the corresponding linear model and the optimal normalized index and the optimal bandwidth of the corresponding linear nonlinear model.
According to the method, different spatial weights are distributed for the fusion variables at different positions according to the distance between the fusion variables and the grid where the target fusion data are located. The range of the considered fused variables is determined by the minimum variance calculated by the half variance function (the smaller the distance between the sampling points within the variance, the greater the similarity. And, considering that different variables have different spatial scales, we combine the thought of MGWR to give different fused variables different bandwidths in fusion. Finally, after normalizing the spatial weights corresponding to the variables based on the specified normalization indexes respectively, weighting and aggregating to obtain a prediction result. It can be represented by the following formula:
wherein y is i ,X ij ,The independent variable matrix and the space weight matrix of the j-th fused variable obtained by searching the ith grid in the range of the variation range are respectively corresponding to the predicted value of the ith grid; h is the number of fused variables; p is p j ,bw j Normalized index and bandwidth corresponding to the j-th fused variable respectively; />The spatial weight matrix representing the jth fused variable is normalized to (0, p) j ) This makes +.>The sum of all the spatial weight values in (a) is p j . Furthermore, the->Each spatial weight in the spatial weight matrix is calculated based on a gaussian kernel function, as follows:
w＝exp(-D 2 /bw 2 )
where D is the distance from the corresponding mesh.
In this embodiment, taking the soil organic carbon reserves as an example, based on 952 soil profile actual measurement organic carbon reserves (762 is 80% profile data used for training, 190 is 20% profile data used for verification), and 9 environmental variables (precipitation, air temperature, multiscale ridge index, green light wave band, red light wave band, near infrared wave band and enhanced vegetation index), respectively constructing and utilizing training set data to train an ANN model and an MGWR model, respectively utilizing a conventional linear fusion method multiple linear regression MLR_F, a nonlinear fusion method artificial neural network ANN_F and our MGNW_F fusion ANN and MGWR prediction results, and respectively utilizing training set data to train three models. The prediction results of ANN, MGWR, MLR _ F, ANN _f and mgnw_f are respectively subjected to accuracy verification by using verification set data, and the results are sequentially shown in fig. 4, and it can be seen that mgnw_f is the lowest of the five models in terms of both the mean square error MSE and the mean absolute error MAE, and the correlation coefficient is the highest of the five models.
As shown in fig. 5, the predicted results of ANN, MGWR, MLR _ F, ANN _f and mgnw_f are obtained, where (a) is an ANN model, (b) is an MGWR model, (c) is mlr_f, (d) is ann_f, and (e) is mgnw_f. The prediction results of the fusion model can be well combined with the characteristics of the ANN model and the MGWR model, and a good fusion effect is achieved.
Embodiment 2, which is based on the same inventive concept as embodiment 1, presents a soil property spatial prediction apparatus that fuses multi-source data and spatial autocorrelation thereof, comprising:
the acquisition module is used for acquiring environment variables of the target area;
the prediction module is used for respectively inputting the environmental variables of the target area into a pre-trained linear relation model of soil attribute data and environmental variables and a pre-trained nonlinear relation model of the soil attribute data and the environmental variables to respectively obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and the fusion module is used for fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by utilizing a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
Embodiment 3, which is based on the same inventive concept as the other embodiments, introduces a computer-readable storage medium storing one or more programs, characterized in that the one or more programs include instructions, which when executed by a computing device, cause the computing device to perform any of the methods.
Embodiment 4, which is based on the same inventive concept as the other embodiments, presents a computer apparatus comprising,
one or more processors, memory, and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for performing any of the methods.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that modifications and variations could be made by those skilled in the art without departing from the technical principles of the present invention, and such modifications and variations should also be regarded as being within the scope of the invention.
Claims (9)
1. A soil attribute spatial prediction method integrating multisource data and spatial autocorrelation thereof is characterized by comprising the following steps:
collecting environment variables of a target area;
inputting the environment variables of the target area into a pre-trained linear relation model of soil attribute data and environment variables and a pre-trained nonlinear relation model of soil attribute data and environment variables respectively to obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by using a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
2. The method for spatial prediction of soil properties by fusion of multisource data and spatial autocorrelation thereof according to claim 1, wherein the linear relation model of the soil property data and environmental variables adopts an ANN model.
3. The method for spatial prediction of soil properties by fusion of multisource data and spatial autocorrelation thereof according to claim 1, wherein the nonlinear relation model of the soil property data and environmental variables adopts an MGWR model.
4. The method for spatial prediction of soil properties by fusion of multisource data and spatial autocorrelation thereof according to claim 1, wherein the fusion process of the multiscale geographic normalization weighted fusion model is as follows:
acquiring a variation of the multi-scale geographic normalization weighted fusion model and a target grid to be predicted;
dividing all grids into four quadrants on a plane by taking a target grid to be predicted as a center, and acquiring a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value of each grid layer in a range of a variation; all grids are grid data of a target area generated by map software, wherein the grid data consists of a plurality of grids;
acquiring a predetermined optimal normalization index and an optimal bandwidth of a corresponding linear model and an optimal normalization index and an optimal bandwidth of a corresponding linear non-nature model;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding linear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding linear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the linear model soil attribute data predictive values of all grid layers to obtain an independent variable matrix corresponding to the linear model; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the linear model to obtain a final value corresponding to the linear model;
calculating the space weights corresponding to different quadrants of each layer of relative target grids according to the optimal bandwidth of the corresponding nonlinear model, summing the space weights corresponding to different quadrants of each layer of relative target grids, averaging to obtain the average space weight of each quadrant, and carrying out normalization processing on the average space weight of each quadrant according to the optimal normalization index of the corresponding nonlinear model to obtain a normalized space weight matrix; calculating the average value of soil attribute data in the same quadrants of all grid layers according to the predicted values of the soil attribute data of the nonlinear models of all grid layers to obtain independent variable matrixes of the corresponding nonlinear models; multiplying and summing the normalized space weight matrix and the independent variable matrix corresponding to the nonlinear model to obtain a final value corresponding to the nonlinear model;
and adding the final value corresponding to the linear model and the final value corresponding to the nonlinear model to obtain the final predicted value of the soil attribute data of the target grid.
5. The method for spatial prediction of soil properties by fusion of multisource data and spatial autocorrelation thereof according to claim 4, wherein the process for obtaining the optimal normalization index and the optimal bandwidth of the corresponding linear model and the optimal normalization index and the optimal bandwidth of the corresponding linear non-linear model specifically comprises:
randomly generating a plurality of groups of normalization indexes and bandwidths corresponding to the linear models and normalization indexes and bandwidths corresponding to the nonlinear models in a preset range, wherein the sum of the normalization indexes of the linear models and the normalization indexes of the nonlinear models is 1;
obtaining an actual measurement value of soil attribute data of any grid in a target area;
taking the grid as a target grid, taking the normalization index and the bandwidth of each group of corresponding linear models and the normalization index and the bandwidth of corresponding nonlinear models as normalization indexes and bandwidths used in a multi-scale geographic normalization weighted fusion model, and calculating through the multi-scale geographic normalization weighted fusion model to obtain a plurality of final predicted values of soil attribute data corresponding to the grid;
and solving the mean square error of the final predicted value of the soil attribute data corresponding to the grid and the actual measured value of the soil attribute data of the grid, wherein the normalized index and the bandwidth of the corresponding linear model and the normalized index and the bandwidth of the corresponding nonlinear model of a group corresponding to the final predicted value of the soil attribute data with the minimum mean square error are the optimal normalized index and the optimal bandwidth of the corresponding linear model and the optimal normalized index and the optimal bandwidth of the corresponding linear nonlinear model.
6. The soil property spatial prediction method for fusing multisource data and spatial autocorrelation thereof according to claim 1, wherein the expression of the multiscale geographic normalization weighted fusion model is:
wherein y is i ,X ij ,Respectively, the predicted value corresponding to the ith grid and the ithSearching the independent variable matrix and the space weight matrix of the j-th fused variable by the grid in the range of the variation; h is the number of fused variables; p is p j ,bw j Normalized index and bandwidth corresponding to the j-th fused variable respectively; />The spatial weight matrix representing the jth fused variable is normalized to (0, p) j ) So that->The sum of all the spatial weight values in (a) is p j 。
7. A soil property spatial prediction apparatus that fuses multisource data and spatial autocorrelation thereof, comprising:
the acquisition module is used for acquiring environment variables of the target area;
the prediction module is used for respectively inputting the environmental variables of the target area into a pre-trained linear relation model of soil attribute data and environmental variables and a pre-trained nonlinear relation model of the soil attribute data and the environmental variables to respectively obtain a linear model soil attribute data predicted value and a nonlinear model soil attribute data predicted value;
and the fusion module is used for fusing the linear model soil attribute data predicted value and the nonlinear model soil attribute data predicted value by utilizing a multi-scale geographic normalization weighted fusion model to obtain a final predicted value of the soil attribute data of the target area.
8. A computer readable storage medium storing one or more programs, wherein the one or more programs comprise instructions, which when executed by a computing device, cause the computing device to perform any of the methods of claims 1-6.
9. A computer device, comprising,
one or more processors, memory, and one or more programs, wherein one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for performing any of the methods of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310939608.XA CN117056858A (en) | 2023-07-28 | 2023-07-28 | Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310939608.XA CN117056858A (en) | 2023-07-28 | 2023-07-28 | Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117056858A true CN117056858A (en) | 2023-11-14 |
Family
ID=88668385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310939608.XA Pending CN117056858A (en) | 2023-07-28 | 2023-07-28 | Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117056858A (en) |
-
2023
- 2023-07-28 CN CN202310939608.XA patent/CN117056858A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Kapoor et al. | A grey wolf optimizer based automatic clustering algorithm for satellite image segmentation | |
| CN114254561B (en) | Waterlogging prediction method, system and storage medium | |
| Jalalkamali | Using of hybrid fuzzy models to predict spatiotemporal groundwater quality parameters | |
| CN112512069B (en) | Network intelligent optimization method and device based on channel beam pattern | |
| GB2547816A (en) | Actually-measured marine environment data assimilation method based on sequence recursive spare filtering three-dimensional variation | |
| CN114501530B (en) | Method and device for determining antenna parameters based on deep reinforcement learning | |
| CN115223049B (en) | Knowledge distillation and quantification method for large model compression of electric power scene edge calculation | |
| Li et al. | A novel combined prediction model for monthly mean precipitation with error correction strategy | |
| CN104156943B (en) | Multi objective fuzzy cluster image change detection method based on non-dominant neighborhood immune algorithm | |
| CN114444378A (en) | Short-term power prediction method for regional wind power cluster | |
| CN112367130B (en) | Wireless network signal transmission intensity calculation method and computer storage medium | |
| CN113592144A (en) | Medium-and-long-term runoff probability forecasting method and system | |
| CN110689183A (en) | Cluster photovoltaic power probability prediction method, system, medium and electronic device | |
| CN115859805A (en) | Self-adaptive sequential test design method and device based on mixed point adding criterion | |
| CN112597870A (en) | Method for constructing land cover classification model | |
| CN110008439A (en) | Rainfall data Spatio-temporal modeling interpolation algorithm based on matrix decomposition | |
| CN114493052A (en) | Multi-model fusion self-adaptive new energy power prediction method and system | |
| CN114169502A (en) | Rainfall prediction method and device based on neural network and computer equipment | |
| CN115859765B (en) | Urban expansion prediction method, device, equipment and storage medium | |
| CN117131654A (en) | Target observation method based on nonlinear optimal disturbance of pre-analysis initial guess condition | |
| CN117056858A (en) | Soil attribute spatial prediction method integrating multisource data and spatial autocorrelation of multisource data | |
| CN115564989A (en) | Random forest algorithm for land use classification | |
| CN115526376A (en) | Multi-feature fusion generation countermeasure network ultra-short-term wind power prediction method | |
| CN115659165A (en) | Method, device, equipment and storage medium for constructing park load data sample | |
| CN115907131A (en) | Method and system for building electric heating load prediction model in northern area |
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 |