CN-121998516-A - Method and device for constructing soil suitable-tillage prediction model and evaluating soil suitable-tillage

CN121998516ACN 121998516 ACN121998516 ACN 121998516ACN-121998516-A

Abstract

The invention relates to the technical field of ecological suitability evaluation, and discloses a method and a device for constructing a soil suitable-tillage prediction model and evaluating the soil suitable-tillage, wherein the method for constructing the soil suitable-tillage prediction model comprises the steps of obtaining target soil suitable-tillage evaluation result grid data, target evaluation index grid data and a soil suitable-tillage index system of a target area, wherein the soil suitable-tillage index system comprises a plurality of evaluation dimensions, and each evaluation dimension comprises a plurality of evaluation indexes; the method comprises the steps of constructing a Bayesian network structure based on a land suitable-tillage index system, converting target land suitable-tillage evaluation result grid data and target evaluation index grid data into a sample data table, inputting the sample data table into the Bayesian network structure for parameter learning, and obtaining a land suitable-tillage prediction model, wherein each root node in the Bayesian network structure is a plurality of evaluation indexes in the land suitable-tillage index system, each middle node is an evaluation dimension, and each leaf node is a land suitable-tillage evaluation result.

Inventors

YAO MINGLEI
CHEN YUYANG
LI SONGCHEN
YU TAO
CAO YAN

Assignees

三峡环境科技有限公司
中国长江三峡集团有限公司

Dates

Publication Date: 20260508
Application Date: 20260407

Claims (10)

1. A method for constructing a soil suitable-tillage prediction model is characterized by comprising the following steps: Acquiring target land suitable-tillage evaluation result grid data, target evaluation index grid data and a land suitable-tillage index system of a target area, wherein the land suitable-tillage index system comprises a plurality of evaluation dimensions, and each evaluation dimension comprises a plurality of evaluation indexes; Constructing a Bayesian network structure based on the soil suitable-tillage index system, wherein each root node in the Bayesian network structure is a plurality of evaluation indexes in the soil suitable-tillage index system, the middle node is an evaluation dimension, and the leaf nodes are soil suitable-tillage evaluation results; And converting the target land suitable-tillage evaluation result grid data and the target evaluation index grid data into a sample data table, inputting the sample data table into a Bayesian network structure for parameter learning, and obtaining a land suitable-tillage prediction model.
2. The method according to claim 1, wherein the target land suitable tillability evaluation result raster data and the target evaluation index raster data are obtained by: Acquiring land suitable-tillage evaluation result grid data and evaluation index grid data of a target area; Carrying out coordinate unification processing on the land suitable-tillage evaluation result raster data and the evaluation index raster data to obtain processed land suitable-tillage evaluation result raster data and evaluation index raster data; And carrying out space scale unification processing on the processed land suitable-tillage evaluation result raster data and the evaluation index raster data to obtain target land suitable-tillage evaluation result raster data and target evaluation index raster data.
3. A method for evaluating soil suitability for tillage, the method comprising: Acquiring a post-verification data set of the field to be evaluated, wherein the post-verification data set is determined according to a field experience data set of the field to be evaluated, and the field experience data set is used for representing experience values of a plurality of evaluation indexes of the field to be evaluated; Inputting the post-verification data set into a pre-constructed land suitable-tillage level prediction model, so that the land suitable-tillage level prediction model calculates land suitable-tillage probability distribution and confidence interval of the field to be evaluated, wherein the land suitable-tillage level prediction model is constructed according to the land suitable-tillage prediction model construction method according to claim 1 or 2; And determining the evaluation result of the soil suitable-tillage property of the field to be evaluated based on the soil suitable-tillage probability distribution of the field to be evaluated and the confidence interval.
4. A method according to claim 3, characterized in that the method further comprises: If the suitable-tillage grade of the field to be evaluated is lower than the preset grade based on the suitable-tillage evaluation result of the field to be evaluated, obtaining the soil suitable-tillage probability distribution of each evaluation dimension, wherein the soil suitable-tillage probability distribution of each evaluation dimension is calculated by using a soil suitable-tillage grade prediction model; Determining suitability grades of the corresponding evaluation dimensions based on the land suitability probability distribution of each evaluation dimension; A structured management text is generated based on the suitability level for each evaluation dimension.
5. The method of claim 4, wherein after the step of generating structured management text based on suitability levels for each evaluation dimension, the method further comprises: And sending the structured management text of the field to be evaluated and the evaluation result of the suitable ploughability of the field to be evaluated to a display terminal for display.
6. A land tillability prediction model construction device, the device comprising: the first acquisition module is used for acquiring target land suitable-tillage evaluation result grid data, target evaluation index grid data and a land suitable-tillage index system of a target area, wherein the land suitable-tillage index system comprises a plurality of evaluation dimensions, and each evaluation dimension comprises a plurality of evaluation indexes; the construction module is used for constructing a Bayesian network structure based on the soil suitable-tillage index system, each root node in the Bayesian network structure is a plurality of evaluation indexes in the soil suitable-tillage index system, the middle node is an evaluation dimension, and the leaf node is a soil suitable-tillage evaluation result; the first determining module is used for converting the target land suitable-tillage evaluation result grid data and the target evaluation index grid data into a sample data table, inputting the sample data table into a Bayesian network structure for parameter learning, and obtaining a land suitable-tillage prediction model.
7. A land tillability evaluation device, the device comprising: The second acquisition module is used for acquiring a post-verification data set of the field to be evaluated, wherein the post-verification data set is determined according to a field experience data set of the field to be evaluated, and the field experience data set is used for representing experience values of a plurality of evaluation indexes of the field to be evaluated; The calculation module is used for inputting the post-verification data set into a pre-constructed land suitable-tillage level prediction model so that the land suitable-tillage level prediction model calculates land suitable-tillage probability distribution and confidence intervals of the field to be evaluated, and the land suitable-tillage level prediction model is constructed according to the land suitable-tillage prediction model construction method according to claim 1 or 2; And the second determining module is used for determining the suitable-tillage evaluation result of the field to be evaluated based on the soil suitable-tillage probability distribution of the field to be evaluated and the confidence interval.
8. An electronic device, comprising: A memory and a processor, the memory and the processor being communicatively connected to each other, the memory having stored therein computer instructions, the processor executing the computer instructions to perform the method of constructing a soil suitability prediction model as claimed in any one of claims 1 or 2, or to perform the method of evaluating soil suitability as claimed in any one of claims 3 to 5.
9. A computer-readable storage medium having stored thereon computer instructions for causing a computer to execute the soil suitability prediction model construction method according to any one of claims 1 or 2 or to execute the soil suitability assessment method according to any one of claims 3 to 5.
10. A computer program product comprising computer instructions for causing a computer to perform the method of construction of a soil suitability prediction model as claimed in any one of claims 1 or 2, or to perform the method of assessment of soil suitability as claimed in any one of claims 3 to 5.

Description

Method and device for constructing soil suitable-tillage prediction model and evaluating soil suitable-tillage Technical Field The invention relates to the technical field of ecological suitability evaluation, in particular to a method and a device for constructing a land suitable-tillage prediction model and evaluating the land suitable-tillage. Background The soil suitable-tillage evaluation is a core link of agricultural resource management, and can help farmers accurately judge the suitable-tillage potential and limiting factors of the self-farm block. However, in the related art, macro scale is used as a core, and the evaluation result of the soil suitable for the tillage is output, so that decision support is mainly provided for management departments or planning institutions to compile land utilization plans. The evaluation model design is oriented to regional planning, the spatial scale of the evaluation index is macroscopic, and the heterogeneity difference of the microscopic scale is not considered, so that the evaluation result is disjointed with the microscopic demand of the farmer, and the farmer cannot be directly guided to develop the reserve farmland resources on the field-level microscopic scale. Disclosure of Invention The invention provides a method and a device for constructing a soil suitable-tillage prediction model and evaluating the soil suitable-tillage, which are used for solving the problems that in the related technology, the evaluation model design is oriented to regional planning, the spatial scale of an evaluation index is macroscopic, the heterogeneity difference of a microscopic scale is not considered, so that an evaluation result is disjointed with the microscopic demand of a farmer, and the farmer cannot be directly guided to develop the reserve tillage resources on the field-level microscopic scale. The invention provides a construction method of a soil suitability prediction model, which comprises the steps of obtaining target soil suitability evaluation result grid data, target evaluation index grid data and a soil suitability index system of a target area, wherein the soil suitability index system comprises a plurality of evaluation dimensions, each evaluation dimension comprises a plurality of evaluation indexes, constructing a Bayesian network structure based on the soil suitability index system, wherein each node in the Bayesian network structure is a plurality of evaluation indexes in the soil suitability index system, each intermediate node is an evaluation dimension, each leaf node is a soil suitability evaluation result, converting the target soil suitability evaluation result grid data and the target evaluation index grid data into a sample data table, and inputting the sample data table into the Bayesian network structure for parameter learning to obtain the soil suitability prediction model. According to the method for constructing the soil suitable-tillage prediction model, provided by the invention, the Bayesian network structure with the evaluation index as the root node, the evaluation dimension as the intermediate node and the suitable-tillage evaluation result as the leaf node is constructed, and the macro grid data is utilized to complete parameter learning, so that the problems of macro evaluation dimension and disjoint with the microscopic demands of farmers in the related technology are effectively solved. Compared with the traditional deterministic model for regional planning, the Bayesian network disclosed by the invention can be used for carrying out probability fusion on priori knowledge of macroscopic grid data and posterior data of farm user field experience, so that the statistical rule of a regional layer is reserved, the field-level heterogeneity difference is fully considered, and the evaluation result is converted from the macroscopic planning basis of a service management department into a microscopic decision support capable of directly guiding the farm user to prepare the farmland resource development. Meanwhile, the Bayesian network naturally supports incomplete data input, and the uncertainty is quantified through probability distribution and confidence intervals, so that the applicability and the robustness of the model in a scene that only partial empirical data can be provided by farmers are obviously improved. In an alternative embodiment, the target land suitable-tillage evaluation result grid data and the target evaluation index grid data are obtained by obtaining land suitable-tillage evaluation result grid data and evaluation index grid data of a target area, performing coordinate unification processing on the land suitable-tillage evaluation result grid data and the evaluation index grid data to obtain processed land suitable-tillage evaluation result grid data and evaluation index grid data, and performing spatial scale unification processing on the processed land suitable-tillage evaluation result grid data and the evalua