Soil humidity estimation model training method and device, electronic equipment and storage medium

Abstract

The invention discloses a soil humidity estimation model training method and device, electronic equipment and a computer readable storage medium, and relates to the field of agriculture and forestry. The soil humidity estimation model training method comprises the following steps: acquiring recorded humidity data of a recording area, determining a vegetation time-lag duration of influence of a normalized vegetation index on soil humidity change according to the recorded humidity data, and determining a rainfall time-lag duration of influence of rainfall on the soil humidity change according to the recorded humidity data, determining a target normalized vegetation index and a target precipitation corresponding to the soil humidity; and constructing a model training data set according to the corresponding relationship among the soil humidity set, the target normalized vegetation index and the target precipitation, and performing model training on the soil humidity estimation model according to the model training data set. The technical problem of low precision of soil humidity estimation in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of agriculture and forestry, and specifically to a soil moisture estimation model training method and device, electronic equipment and computer-readable storage medium. Background Art

[0002] In recent years, the intensification of global climate change and the advancement of ecological restoration projects have made the role of soil moisture in regional water cycles, energy balance, and carbon cycles increasingly important. In semi-arid and arid regions, in particular, deep soils provide long-term moisture support for vegetation and influence ecosystem stability.

[0003] However, existing soil moisture monitoring methods have limitations. Fixed-point probe monitoring is only applicable to small scales, while remote sensing technology, limited by vegetation cover and signal penetration, can only capture surface soil moisture, making it difficult to meet the high spatiotemporal resolution required for monitoring deep soil moisture. Current deep soil moisture inversion methods rely primarily on environmental factors such as precipitation, topography, and land use, combined with machine learning models for prediction. However, while precipitation affects surface soil moisture rapidly, deep soil moisture often has a long delayed response. Existing research has paid little attention to the delayed response characteristics of precipitation at different soil depths, which affects the accuracy of soil moisture estimation. Summary of the Invention

[0004] In view of this, it is necessary to provide a soil moisture estimation model training method and device, electronic device and computer-readable storage medium to solve the technical problem of low accuracy of soil moisture estimation in the prior art.

[0005] In order to solve the above technical problems, in the first aspect, the present application provides a soil moisture estimation model training method, including: obtaining recorded humidity data of a recorded area, the recorded humidity data including the normalized vegetation index, precipitation and soil moisture set of the recorded area at multiple different time points, the soil moisture set including soil moisture at several soil depths; for any soil depth, determining the vegetation lag duration of the normalized vegetation index affecting the soil moisture change according to the recorded humidity data, and determining the precipitation lag duration of the precipitation affecting the soil moisture change according to the recorded humidity data; for any soil moisture, determining the normalized vegetation index corresponding to the time point delayed by the vegetation lag duration in the recorded humidity data as the target normalized vegetation index corresponding to the soil moisture, and determining the precipitation corresponding to the time point advanced by the precipitation lag duration in the recorded humidity data as the target precipitation corresponding to the soil moisture; constructing a model training data set according to the correspondence between the soil moisture set, the target normalized vegetation index and the target precipitation, and performing model training on the soil moisture estimation model based on the model training data set.

[0006] In an optional embodiment, obtaining recorded humidity data of a recorded area further includes: obtaining slope data, potential evapotranspiration data, and surface temperature data of the recorded area; for any soil depth, determining a target parameter having a correlation with soil moisture greater than a preset correlation threshold, the target parameter being one or more of the slope data, the potential evapotranspiration data, and the surface temperature data; constructing a model training data set based on the correspondence between the soil moisture set, the target normalized vegetation index, and the target precipitation, including: constructing a model training data set based on the correspondence between the soil moisture set, the target normalized vegetation index, the target precipitation, and the target parameters.

[0007] In an optional embodiment, determining a target parameter having a correlation with soil moisture greater than a preset correlation threshold includes: constructing a correlation matrix between slope data, potential evapotranspiration data, and surface temperature data and soil moisture, and determining the target parameter based on the correlation matrix.

[0008] In an optional embodiment, determining the vegetation lag duration of the influence of the normalized vegetation index on the soil moisture change based on the recorded humidity data includes: constructing a cross-correlation coefficient formula, substituting the normalized vegetation index and the recorded humidity data into the cross-correlation coefficient formula, and calculating the vegetation lag duration; determining the precipitation lag duration of the influence of precipitation on the soil moisture change based on the recorded humidity data includes: substituting the precipitation and the recorded humidity data into the cross-correlation coefficient formula, and calculating the precipitation lag duration; the cross-correlation coefficient formula includes: ; in, is the total number of time points in the recorded humidity data, When the normalized vegetation index is the vegetation lag time, When the precipitation is the precipitation lag time, To record humidity data, and They are and The average value of .

[0009] In an optional embodiment, the soil moisture estimation model training method further includes: establishing a regression model, and performing accuracy verification on the vegetation time lag and the precipitation time lag according to the recorded humidity data based on the regression model.

[0010] In an optional embodiment, obtaining recorded humidity data of a recorded area further includes: obtaining land use data of the recorded area, and recording the land type of the recorded area according to the land use data; after training the soil moisture estimation model according to the model training data set, the soil moisture estimation model training method further includes: labeling the trained soil moisture estimation model based on the land type.

[0011] In an optional embodiment, a model training data set is constructed based on the correspondence between a soil moisture set, a target normalized vegetation index, and a target precipitation, including: for any first soil moisture in the soil moisture set, determining the relevant soil moisture corresponding to the first soil moisture, determining the adjacent soil correlation coefficient between the first soil moisture and the relevant soil moisture, the relevant soil moisture being the soil moisture corresponding to a second soil depth whose difference from a first soil depth corresponding to the first soil moisture is less than a preset difference threshold; and constructing a model training data set based on the correspondence between the soil moisture set, the target normalized vegetation index, the target precipitation, and the adjacent soil correlation coefficient.

[0012] In the second aspect, an embodiment of the present application provides a soil moisture estimation model training device, including: a data acquisition module, the data acquisition module is used to acquire recorded humidity data of a recorded area, the recorded humidity data includes a normalized vegetation index, precipitation and soil humidity set of the recorded area at multiple different time points, and the soil humidity set includes soil humidity at several soil depths; a time lag determination module, the time lag determination module is used to determine, for any soil depth, the vegetation time lag duration of the normalized vegetation index affecting the soil humidity change based on the recorded humidity data, and determine the precipitation affecting the soil humidity change based on the recorded humidity data. The precipitation lag time length is 200 milliseconds; the data reorganization module is used to determine, for any soil moisture, the normalized vegetation index corresponding to the time point of the delayed vegetation lag time length in the recorded humidity data as the target normalized vegetation index corresponding to the soil moisture, and determine the precipitation corresponding to the time point of the advanced precipitation lag time length in the recorded humidity data as the target precipitation corresponding to the soil moisture, and construct a model training data set according to the correspondence between the soil moisture set, the target normalized vegetation index, and the target precipitation; the model training module is used to train the soil moisture estimation model according to the model training data set.

[0013] In a third aspect, an embodiment of the present application provides an electronic device, including a memory and a processor, wherein the memory is used to store programs; the processor is coupled to the memory and is used to execute the programs stored in the memory to implement the soil moisture estimation model training method as described above.

[0014] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium for storing a computer-readable program or instruction, which, when executed by a processor, can implement the soil moisture estimation model training method as described above.

[0015] Beneficial effects of this application: Compared with the related art, in the embodiment of the present application, when constructing the training data of the soil moisture estimation model, the vegetation lag time of the normalized vegetation index affecting the soil moisture change is determined based on the recorded humidity data. Since vegetation absorbs water during its growth, the normalized vegetation index has a delay in the change of soil moisture. Subsequently, the normalized vegetation index corresponding to the time point with the delayed vegetation lag time in the recorded humidity data is used as the target normalized vegetation index corresponding to the soil moisture, that is, the normalized vegetation index with the greatest correlation with the soil moisture change is determined as the target normalized vegetation index; similarly, the precipitation lag time of the precipitation affecting the soil moisture change is determined based on the recorded humidity data. Since surface precipitation gradually penetrates downward, Precipitation has an advance effect on soil moisture changes. Subsequently, the precipitation corresponding to the time point with the advance precipitation lag in the humidity data will be recorded as the target precipitation corresponding to the soil moisture, that is, the precipitation with the greatest correlation with soil moisture changes will be determined as the target precipitation; then, a model training data set is constructed based on the correspondence between the soil moisture set, the target normalized vegetation index, and the target precipitation to improve the correlation between the normalized vegetation index and precipitation in the training data and soil moisture changes. Using training data with stronger correlation to form a model training data set to train the soil moisture estimation model can improve the training effect of the soil moisture estimation model, thereby improving the accuracy of the soil moisture estimation model's estimation results. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0017] Figure 1 A flow chart of the soil moisture estimation model training method provided in an embodiment of the present application; Figure 2 A schematic diagram of a process for training a soil moisture estimation model in the soil moisture estimation model training method provided in an embodiment of the present application; Figure 3 This is a schematic diagram of the structure of the soil moisture estimation model training device provided in an embodiment of the present application; Figure 4A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the embodiments described are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0019] In the description of the embodiments of this application, unless otherwise specified, "plurality" means two or more. "And / or" describes the association relationship between associated objects, indicating that three relationships can exist. For example, "A and / or B" can mean: A exists alone, A and B exist simultaneously, or B exists alone.

[0020] The terms "first," "second," and so on, used in the embodiments of this application are for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, technical features designated as "first" or "second" may explicitly or implicitly include at least one such feature.

[0021] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0022] The present application provides a soil moisture estimation model training method and device, an electronic device, and a computer-readable storage medium, which are described below respectively.

[0023] Please refer to Figure 1 The present application provides a soil moisture estimation model training method, which is applied to training a soil moisture estimation model. After the model is completed, the soil moisture estimation model can estimate soil moisture based on provided data related to soil moisture. The soil moisture estimation model training method provided in this embodiment specifically includes the following steps: Step S101: Acquire recorded humidity data of a recording area.

[0024] Step S102: For any soil depth, determine the vegetation lag duration of the normalized vegetation index affecting soil moisture changes based on the recorded moisture data.

[0025] Step S103: For any soil depth, determine the precipitation time lag duration for the effect of precipitation on soil moisture changes based on the recorded moisture data.

[0026] Step S104: For any soil moisture, determine the normalized vegetation index corresponding to the time point delayed by the vegetation time lag in the recorded moisture data as the target normalized vegetation index corresponding to the soil moisture.

[0027] Step S105: For any soil moisture, determine the precipitation corresponding to the time point of the advance precipitation lag time in the recorded humidity data as the target precipitation corresponding to the soil moisture.

[0028] Step S106: constructing a model training data set according to the corresponding relationship between the soil moisture set, the target normalized difference vegetation index, and the target precipitation.

[0029] Step S107: performing model training on the soil moisture estimation model according to the model training data set.

[0030] Compared with the related art, in the embodiment of the present application, when constructing the training data of the soil moisture estimation model, the vegetation lag time of the normalized vegetation index affecting the soil moisture change is determined based on the recorded humidity data. Since vegetation absorbs water during its growth, the normalized vegetation index has a delay in the change of soil moisture. Subsequently, the normalized vegetation index corresponding to the time point with the delayed vegetation lag time in the recorded humidity data is used as the target normalized vegetation index corresponding to the soil moisture, that is, the normalized vegetation index with the greatest correlation with the soil moisture change is determined as the target normalized vegetation index; similarly, the precipitation lag time of the precipitation affecting the soil moisture change is determined based on the recorded humidity data. Since surface precipitation gradually penetrates downward, Precipitation has an advance effect on soil moisture changes. Subsequently, the precipitation corresponding to the time point with the advance precipitation lag in the humidity data will be recorded as the target precipitation corresponding to the soil moisture, that is, the precipitation with the greatest correlation with soil moisture changes will be determined as the target precipitation; then, a model training data set is constructed based on the correspondence between the soil moisture set, the target normalized vegetation index, and the target precipitation to improve the correlation between the normalized vegetation index and precipitation in the training data and soil moisture changes. Using training data with stronger correlation to form a model training data set to train the soil moisture estimation model can improve the training effect of the soil moisture estimation model, thereby improving the accuracy of the soil moisture estimation model's estimation results.

[0031] In step S101, the recording area can be any area that can provide soil moisture-related data. Obtaining recorded moisture data for the recording area can specifically include obtaining a set of normalized vegetation indexes, precipitation, and soil moisture for the recording area at multiple different time points. The soil moisture set includes soil moisture at multiple soil depths. Specifically, soil moisture at multiple different time points for each soil depth in the recording area, as well as normalized vegetation indexes and precipitation for the recording area at multiple different time points, are obtained. These data are then mapped based on the time points to form recorded moisture data.

[0032] Specifically, the normalized vegetation index, precipitation, and soil moisture sets of a recording area at multiple different time points can be obtained through data from public channels. For example, the normalized vegetation index, precipitation, and soil moisture sets of a recording area at multiple different time points can be obtained directly from the data disclosed in the Global Land Data Assimilation System GLDAS or ESA CCI (ClimateChange Initiative of the European Space Agency (ESA)) as recorded humidity data.

[0033] The Global Land Data Assimilation System (GLDAS) contains 36 land surface fields from January 2000 to the present, covering a wide range of meteorological elements such as temperature, humidity, wind speed, precipitation, radiation flux, soil moisture, and soil temperature. This provides rich data support for research and applications in meteorology, agriculture, transportation, energy, urban management, and other fields. The ESA CCI project encompasses multiple aspects, such as the development of datasets on land cover, fires, and soil moisture.

[0034] Furthermore, in some embodiments of the present application, the recorded humidity data can be obtained only from the Global Land Data Assimilation System GLDAS, or only from the ESA CCI, or the humidity-related data of the recorded area can be obtained from both the Global Land Data Assimilation System GLDAS and the ESA CCI at the same time, and the two can be mutually verified, the missing data can be supplemented, and the invalid data can be removed to serve as the recorded humidity data, thereby improving the reliability and accuracy of the recorded humidity data.

[0035] Specifically, mutual verification of humidity-related data of the recorded area obtained from GLDAS and ESA CCI may include: After obtaining soil moisture data from the official ESA CCI and GLDAS websites, the two soil moisture data were temporally and spatially matched to ensure consistent timestamps and spatial resolution. Further, metrics such as root mean square error (RMSE), mean absolute error (MAE), correlation coefficient (R), and bias (Bias) were used to assess the consistency of soil moisture data obtained by GLDAS-2.1 and ESA CCI. Long-term trend analysis of soil moisture data obtained by GLDAS-2.1 and ESA CCI was then performed, such as using the Mann-Kendall trend test. The correlation coefficient (R), bias (Bias), Mann-Kendall statistic (S), variance (Var(S)), and standardized test statistic (Z) were calculated using the following formulas: ; ; ; ; ; in, and is the data point in the time series, m is the number of groups of repeated values ​​in the data, is the number of repeated values ​​in the kth group, and sgn is the sign function, defined as follows: .

[0036] In step S102, the vegetation lag duration of the influence of the normalized vegetation index on the soil moisture change is determined based on the recorded humidity data. Specifically, a cross-correlation coefficient formula is constructed, and the normalized vegetation index and the recorded humidity data are substituted into the cross-correlation coefficient formula to calculate the vegetation lag duration.

[0037] The mutual correlation coefficient formula includes: ; in, is the total number of time points in the recorded humidity data, is the normalized difference vegetation index, is the vegetation lag time, To record humidity data, and They are and The average value of .

[0038] Similarly, in the embodiment of the present application, the precipitation lag duration of the influence of precipitation on soil moisture change can be determined based on the recorded humidity data in step S103 by substituting the precipitation and recorded humidity data into the correlation coefficient formula to calculate the precipitation lag duration. In this case, the above correlation coefficient formula is is the precipitation, is the precipitation lag time.

[0039] Furthermore, in some embodiments of the present application, after determining the vegetation lag duration and the precipitation lag duration in steps S102 and S103, a step of verifying the accuracy of the vegetation lag duration and the precipitation lag duration may also be included. Specifically, a regression model may be established, and the accuracy of the vegetation lag duration and the precipitation lag duration may be verified based on the recorded humidity data based on the regression model. The regression model may specifically be a Granger causality test model, and the specific formula is as follows: ; ; Where, and Represent the time series, namely the vegetation lag time and the precipitation lag time, 、 、 and is the regression coefficient, and is the residual term that is considered uncorrelated, and is the maximum lag order.

[0040] Furthermore, in step S104, the normalized vegetation index corresponding to the time point delayed by the vegetation lag time in the recorded humidity data is determined as the target normalized vegetation index corresponding to the soil moisture. Specifically, for any soil moisture a1, it corresponds to a time point t1 and a normalized vegetation index b1 in the recorded humidity data. By delaying this time point t1 by the vegetation lag time t3, the time point t2=t3+t1 can be obtained. The normalized vegetation index b2 corresponding to the time point t2 in the recorded humidity data is used as the target normalized vegetation index corresponding to the soil moisture a1, that is, the normalized vegetation index b2 is used to replace the normalized vegetation index b1 corresponding to the soil moisture a1.

[0041] Similarly, in step S105, the precipitation corresponding to the time point in advance of the precipitation lag time in the recorded humidity data is determined as the target precipitation corresponding to the soil humidity. Specifically, for any soil humidity a1, it corresponds to a time point t1 and a precipitation c1 in the recorded humidity data. This time point t1 is advanced by the precipitation lag time t4 to obtain the time point t5=t4+t1. The precipitation c2 corresponding to the time point t5 in the recorded humidity data is used as the target precipitation corresponding to the soil humidity a1, that is, the precipitation c1 is replaced by the precipitation c2 to correspond to the soil humidity a1.

[0042] For further information, please refer to Figure 2 In some embodiments of the present application, the step S107 of training the soil moisture estimation model according to the model training data set may specifically include the following steps: Step S201: The model training data set is used as the prediction variable data set, and the soil moisture in the humidity data is recorded as the target variable data set.

[0043] Step S202: Standardize the prediction variable dataset and the target variable dataset to obtain standardized data.

[0044] Step S203: All standardized data are randomly divided into two groups, namely a training set and a test set.

[0045] Step S204: performing model training on the soil moisture estimation model based on the training set and the test set.

[0046] In this step, the soil moisture estimation model may be trained using a random forest model. Further, the training of the random forest model may include: The test set is fed into the random forest model as input. The model then extracts correlation features from the input parameters using a model operation algorithm such as GridSearch or Bayesian Optimization, combined with 10-fold cross-validation. Finally, the accuracy of the random forest model's soil moisture estimation can be verified. For example, the root mean square error (RMSE), relative root mean square error (RRMSE), and goodness of fit (R²) can be used as evaluation metrics to assess its accuracy. The specific verification process calculation formula is as follows: ; ; ; in is the actual value, is the predicted value, is the mean of the actual values.

[0047] Furthermore, in some embodiments of the present application, obtaining recorded humidity data of the recorded area in step S101 may also include: obtaining slope data, potential evapotranspiration data, and surface temperature data of the recorded area, and selecting a target parameter for each soil moisture whose correlation is greater than a preset correlation threshold.

[0048] Specifically, selecting the target parameter from the slope data, potential evapotranspiration data, and surface temperature data can include constructing a relationship matrix between the slope data, potential evapotranspiration data, surface temperature data, and soil moisture, determining the target parameter based on the relationship matrix, and then adding the target parameter to the model training dataset to train the soil moisture estimation model.

[0049] Specifically, the correlation matrix can be constructed by calculating the correlation coefficients between the slope data, potential evapotranspiration data, and surface temperature data and soil moisture, respectively, and then constructing the correlation matrix based on all the correlation coefficients. Each correlation coefficient in the correlation matrix is ​​compared with a preset correlation threshold, and the slope data and / or potential evapotranspiration data and / or surface temperature data having a correlation coefficient greater than the preset correlation threshold is determined as the target parameter.

[0050] Furthermore, in some embodiments of the present application, the model training dataset may also include soil moisture of adjacent soil layers. Specifically, for any first soil moisture in the soil moisture set, a related soil moisture corresponding to the first soil moisture is determined, and a soil adjacent correlation coefficient between the first soil moisture and the related soil moisture is determined, where the related soil moisture is the soil moisture corresponding to a second soil depth where the difference between the first soil depth corresponding to the first soil moisture is less than a preset difference threshold. A model training dataset is constructed based on the corresponding relationship between the soil moisture set, the target normalized difference vegetation index, the target precipitation, and the soil adjacent correlation coefficient. Specifically, for the first soil moisture a1, it corresponds to a first soil depth d1 in the recorded moisture data. The recorded moisture data also includes second soil depths d2, d3, d4..., whose soil depth difference with the first soil depth d1 is less than a preset difference threshold. Each second soil depth d2, d3, d4... corresponds to a number of soil moistures a2, a3, a4... in the recorded moisture data. These soil moistures a2, a3, a4... are used as relevant soil moistures corresponding to the first soil moisture a1. The adjacent soil correlation coefficient between the relevant soil moisture and the first soil moisture a1 is calculated, and the adjacent soil correlation coefficient and the first soil moisture a1 are added to the model training data set accordingly.

[0051] Furthermore, in some embodiments of the present application, land use data of the recorded area can also be obtained, and the land type of the recorded area can be determined based on the land use data. Land types can include, for example, woodland, grassland, shrubs, etc. After the soil moisture estimation model is trained based on the model training data set, the trained soil moisture estimation model can also be labeled based on the land type. The trained soil moisture estimation model is labeled based on the land type, that is, the model training data set of the soil moisture estimation model is constructed separately for different land use conditions, so that the soil moisture estimation model is more consistent with the actual land use conditions, thereby improving the soil moisture estimation accuracy of the soil moisture estimation model.

[0052] Please refer to Figure 3 , the embodiment of the present application also provides a soil moisture estimation model training device, comprising: The data acquisition module 100 is used to acquire recorded humidity data of a recording area, wherein the recorded humidity data includes a normalized vegetation index, precipitation, and soil moisture set of the recording area at multiple different time points, wherein the soil moisture set includes soil moisture at multiple soil depths; a time lag determination module 200 for determining, for any soil depth, a vegetation time lag duration for the effect of the normalized vegetation index on soil moisture changes based on the recorded moisture data, and a precipitation time lag duration for the effect of precipitation on soil moisture changes based on the recorded moisture data; The data reorganization module 300 is configured to, for any soil moisture, determine a normalized vegetation index corresponding to a time point after a vegetation lag in the recorded moisture data as a target normalized vegetation index corresponding to the soil moisture, determine a precipitation amount corresponding to a time point before a precipitation lag in the recorded moisture data as a target precipitation amount corresponding to the soil moisture, and construct a model training data set based on the corresponding relationship between the soil moisture set, the target normalized vegetation index, and the target precipitation amount; The model training module 400 is used to perform model training on the soil moisture estimation model according to the model training data set.

[0053] Please refer to Figure 4 The embodiment of the present application also provides an electronic device. The electronic device includes a processor 101 and a memory 102. Figure 4 Only some of the components of the electronic device are shown, but it should be understood that it is not required to implement all of the shown components, and more or fewer components may be implemented instead.

[0054] In some embodiments, the processor 101 may be a central processing unit (CPU), a microprocessor, or other data processing chip, configured to execute program codes or process data stored in the memory 102 , such as the magnetic resonance image optimization method of the present invention.

[0055] In some embodiments, processor 101 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, processor 101 may be local or remote. In some embodiments, processor 101 may be implemented on a cloud platform. In one embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an on-premises cloud, a multi-cloud, or any combination thereof.

[0056] In some embodiments, the memory 102 may be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device. In other embodiments, the memory 102 may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device.

[0057] Furthermore, the memory 102 may include both an internal storage unit of the electronic device and an external storage device. The memory 102 is used to store application software installed in the electronic device and various data.

[0058] Furthermore, the embodiments of the present invention do not specifically limit the types of electronic devices mentioned, and the electronic devices may be portable electronic devices such as mobile phones, tablet computers, personal digital assistants (PDAs), wearable devices, and laptop computers. Exemplary embodiments of portable electronic devices include but are not limited to portable electronic devices equipped with iOS, Android, Microsoft, or other operating systems. The above-mentioned portable electronic devices may also be other portable electronic devices, such as laptop computers with touch-sensitive surfaces (e.g., touch panels). It should also be understood that in some other embodiments of the present invention, the electronic device may not be a portable electronic device, but a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0059] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, the steps or functions in the magnetic resonance image optimization method provided by the above-mentioned method embodiments can be implemented.

[0060] Those skilled in the art will appreciate that all or part of the process flow of the above-described method embodiment can be implemented by instructing related hardware (such as a processor, controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a magnetic disk, an optical disk, a read-only memory, or a random access memory.

[0061] The above is a detailed introduction to the ship communication method, device and storage medium provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, based on the ideas of the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A soil moisture estimation model training method, characterized in that: include: Acquire recorded humidity data of the recording area, the recorded humidity data including normalized vegetation index, precipitation and soil moisture set of the recording area at multiple different time points, the soil moisture set including soil moisture at multiple soil depths; For any soil depth, the vegetation lag duration of the effect of the normalized vegetation index on soil moisture changes is determined based on the recorded moisture data, and the precipitation lag duration of the effect of precipitation on soil moisture changes is determined based on the recorded moisture data; For any soil moisture, the normalized vegetation index corresponding to the time point with a delayed vegetation lag in the recorded humidity data is determined as the target normalized vegetation index corresponding to the soil moisture, and the precipitation corresponding to the time point with a leading precipitation lag in the recorded humidity data is determined as the target precipitation corresponding to the soil moisture; A model training data set is constructed according to the correspondence between the soil moisture set, the target normalized difference vegetation index, and the target precipitation, and the soil moisture estimation model is trained according to the model training data set.

2. The soil moisture estimation model training method according to claim 1, characterized in that: Get the recorded humidity data of the recorded area, including: Obtain slope data, potential evapotranspiration data, and surface temperature data for the recording area; For any soil depth, determining a target parameter having a correlation with soil moisture greater than a preset correlation threshold, the target parameter being one or more of slope data, potential evapotranspiration data, and surface temperature data; The model training dataset is constructed based on the correspondence between the soil moisture set, the target normalized difference vegetation index, and the target precipitation, including: A model training dataset is constructed based on the soil moisture set, target normalized difference vegetation index, target precipitation and the correspondence between target parameters.

3. The soil moisture estimation model training method according to claim 2, characterized in that: Identify target parameters whose correlation with soil moisture is greater than a preset correlation threshold, including: The correlation matrix among slope data, potential evapotranspiration data, surface temperature data and soil moisture is constructed, and the target parameters are determined according to the correlation matrix.

4. The soil moisture estimation model training method according to claim 1, characterized in that: Determining a vegetation time lag duration of the influence of the normalized vegetation index on soil moisture changes based on the recorded humidity data, including: constructing a cross-correlation coefficient formula, substituting the normalized vegetation index and the recorded humidity data into the cross-correlation coefficient formula, and calculating the vegetation time lag duration; Determining a precipitation lag duration of the effect of precipitation on soil moisture changes based on recorded humidity data, including: substituting precipitation and recorded humidity data into a correlation coefficient formula to calculate the precipitation lag duration; The cross-correlation coefficient formulas include: ； in, is the total number of time points in the recorded humidity data, When the normalized vegetation index is the vegetation lag time, When the precipitation is the precipitation lag time, To record humidity data, and They are and The average value of .

5. The soil moisture estimation model training method according to claim 4, characterized in that: The soil moisture estimation model training method also includes: A regression model was established, and the accuracy of vegetation time lag and precipitation time lag was verified based on the recorded humidity data.

6. The soil moisture estimation model training method according to claim 1, characterized in that: Get the recorded humidity data of the recorded area, including: Obtaining land use data of the recorded area, and recording the land type of the area according to the land use data; After the soil moisture estimation model is trained according to the model training data set, the soil moisture estimation model training method further includes: Label the trained soil moisture estimation model based on land type.

7. The soil moisture estimation model training method according to claim 1, characterized in that: The model training dataset is constructed based on the correspondence between the soil moisture set, the target normalized difference vegetation index, and the target precipitation, including: For any first soil moisture in the soil moisture set, determining a correlated soil moisture corresponding to the first soil moisture, and determining a soil correlation coefficient between the first soil moisture and the correlated soil moisture, where the correlated soil moisture is a soil moisture corresponding to a second soil depth at which a difference from a first soil depth corresponding to the first soil moisture is less than a preset difference threshold; The model training dataset is constructed based on the correspondence between the soil moisture set, the target normalized difference vegetation index, the target precipitation and the adjacent soil correlation coefficient.

8. A soil moisture estimation model training device, characterized in that: include: A data acquisition module is used to acquire recorded humidity data of the recording area, the recorded humidity data including normalized vegetation index, precipitation and soil humidity set of the recording area at multiple different time points, the soil humidity set including soil humidity at multiple soil depths; a time lag determination module, for any soil depth, for determining a vegetation time lag duration of the effect of the normalized vegetation index on soil moisture changes based on the recorded moisture data, and for determining a precipitation time lag duration of the effect of precipitation on soil moisture changes based on the recorded moisture data; A data reorganization module, for any soil moisture, is used to determine the normalized vegetation index corresponding to the time point with a delayed vegetation time lag in the recorded humidity data as the target normalized vegetation index corresponding to the soil moisture, determine the precipitation corresponding to the time point with an advanced precipitation time lag in the recorded humidity data as the target precipitation corresponding to the soil moisture, and construct a model training data set based on the correspondence between the soil moisture set, the target normalized vegetation index, and the target precipitation; The model training module is used to train the soil moisture estimation model based on the model training dataset.

9. An electronic device, characterized in that: comprising a memory and a processor, wherein, Memory, used to store programs; A processor is coupled to the memory and is configured to execute a program stored in the memory to implement the soil moisture estimation model training method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the soil moisture estimation model training method according to any one of claims 1 to 7.

Images