Grassland grazing intensity estimation method based on BMogLSTM-KAN model and related device

Abstract

The invention discloses a grassland grazing intensity estimation method based on a BMogLSTM-KAN model and a related device, and relates to the technical field of grassland management.The method comprises the steps that an overground biomass inversion result, remote sensing data, topographic data, meteorological data and grazing activity factors of an area to be estimated are obtained, and then the grazing intensity of the area to be estimated is estimated based on a trained grassland grazing intensity estimation model; according to the data of the to-be-estimated area, obtaining a grazing intensity estimation result of the to-be-estimated area; specifically, the grassland grazing intensity estimation model is constructed according to a BMogLSTM-KAN model obtained by improving a MogLSTM model based on a batch attention module and a KAN mechanism. The improved model is utilized to fully learn the relationship among the above-ground biomass inversion result, the remote sensing data, the topographic data, the meteorological data, the grazing activity factor and the grazing intensity, and can be used for accurately estimating the grassland grazing intensity.

Description

Technical Field

[0001] The present application relates to the field of grassland management technology, and in particular to a grassland grazing intensity estimation method and related devices based on a BMogLSTM-KAN model. Background Art

[0002] Grassland is one of the most widely distributed ecosystems on Earth, covering about 40% of the land surface area. Grassland ecosystems are an important part of natural ecosystems, providing humans with a wide range of biodiversity and ecosystem services, and playing an important role in pasture feed, water conservation, wind and sand fixation, climate regulation, soil and water conservation, and food security. Grazing is the main way to use grasslands, and its output accounts for about 30% of the global meat supply. With the development of grassland animal husbandry, overgrazing has occurred in some parts of Inner Mongolia, and the number of livestock such as cattle and sheep far exceeds the carrying capacity of grasslands, and the contradiction between grasslands and livestock has become increasingly prominent. Grassland degradation caused by grazing not only affects the sustainable development of animal husbandry, but also leads to the weakening of ecosystem functions, posing a huge threat to local and surrounding ecological security. Therefore, a scientific understanding of grazing intensity is crucial to the sustainable development of grassland ecosystems.

[0003] Grazing intensity (GI) refers to the intensity of grazing utilization, that is, the number of grazing livestock per unit area within a certain period of time. It is an important indicator reflecting the degree of grassland utilization. Moderate grazing can improve grassland productivity, biodiversity and ecosystem stability. Excessive grazing will have a negative impact on biodiversity and service functions such as carbon, nitrogen and nutrient cycles, leading to ecological problems such as grassland degradation. Quantifying grassland grazing intensity is of great significance for adjusting and improving grassland ecosystem management, improving grassland ecological quality and ensuring China's ecological security. How to accurately estimate grassland grazing intensity is an urgent problem to be solved. Summary of the invention

[0004] The purpose of this application is to provide a grassland grazing intensity estimation method and related devices based on the BMogLSTM-KAN model, which can accurately estimate the grassland grazing intensity.

[0005] To achieve the above objectives, this application provides the following solutions:

[0006] In a first aspect, the present application provides a grassland grazing intensity estimation method based on a BMogLSTM-KAN model, comprising the following steps:

[0007] Obtain the inversion results of aboveground biomass, remote sensing data, topographic data, meteorological data and grazing activity factors of the area to be estimated.

[0008] Based on the trained grassland grazing intensity estimation model, the grazing intensity estimation result of the area to be estimated is obtained according to the aboveground biomass inversion result of the area to be estimated, the remote sensing data, the terrain data, the meteorological data and the grazing activity factor; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

[0009] Optionally, before obtaining the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model further includes the following steps:

[0010] A grazing intensity estimation sample data set is constructed; the grazing intensity estimation sample data set includes a plurality of grazing intensity estimation samples; the grazing intensity estimation samples include aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors as independent variables, and grassland grazing intensity as a dependent variable.

[0011] Based on the BMogLSTM-KAN model, an initial grassland grazing intensity estimation model was constructed.

[0012] Based on the grazing intensity estimation sample data set, the initial grassland grazing intensity estimation model is trained to obtain a trained grassland grazing intensity estimation model.

[0013] Optionally, constructing a sample data set for grazing intensity estimation specifically includes the following steps:

[0014] Field investigation and record are conducted in the study area to obtain grassland grazing conditions in the study area; the grassland grazing conditions include the number of grazing livestock, the type of grazing livestock, the pasture area and the grazing time; the study area is a grassland area of ​​the same type as the area to be estimated.

[0015] Based on the grassland grazing conditions in the study area, the grassland grazing intensity in the study area was calculated.

[0016] Field measurements were carried out in the study area to obtain the inversion results of the aboveground biomass in the study area.

[0017] Remote sensing data, topographic data, meteorological data and grazing activity factors of the study area were obtained.

[0018] The aboveground biomass inversion results, remote sensing data, topographic data, meteorological data and grazing activity factors of the study area were taken as independent variables, and the grassland grazing conditions in the study area were taken as dependent variables to obtain a sample of grazing intensity estimation.

[0019] The above process is repeated in several study areas to obtain a grazing intensity estimation sample data set including several grazing intensity estimation samples.

[0020] Optionally, the grassland grazing intensity in the study area during the grazing period can be calculated according to the following formula:

[0021]

[0022] Among them, GI is the grassland grazing intensity in the study area, SU is the sheep unit value in the study area, and the sheep unit is based on sheep, and different grazing livestock species in the study area are converted into sheep units, and GA is the pasture area.

[0023] Optionally, the BMogLSTM-KAN model includes a batch attention module, a MogLSTM model and a KAN mechanism; the batch attention module includes multi-head attention, layer normalization and a feedforward network; the MogLSTM model is an improved model obtained by adding an interactive mechanism to the traditional LSTM model.

[0024] Optionally, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model further includes the following steps:

[0025] According to the grazing intensity estimation results of the area to be estimated, a spatial distribution map of grazing intensity in the area to be estimated is drawn to determine the overall trend of grazing intensity in the area to be estimated.

[0026] In the second aspect, the present application provides a grassland grazing intensity estimation system based on the BMogLSTM-KAN model, comprising:

[0027] The regional data acquisition module is used to obtain the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated.

[0028] The regional grazing intensity estimation module is used to obtain the grazing intensity estimation result of the area to be estimated based on the trained grassland grazing intensity estimation model, according to the aboveground biomass inversion result of the area to be estimated, the remote sensing data, the terrain data, the meteorological data and the grazing activity factor; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

[0029] In a third aspect, the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the grassland grazing intensity estimation method based on the BMogLSTM-KAN model described above.

[0030] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the grassland grazing intensity estimation method based on the BMogLSTM-KAN model described above.

[0031] In a fifth aspect, the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the steps of the grassland grazing intensity estimation method based on the BMogLSTM-KAN model described above.

[0032] According to the specific embodiments provided in this application, this application discloses the following technical effects:

[0033] The present application provides a grassland grazing intensity estimation method and related devices based on the BMogLSTM-KAN model. In the method, the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated are obtained, and then based on the trained grassland grazing intensity estimation model, the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated are obtained to obtain the grazing intensity estimation results of the area to be estimated; specifically, the grassland grazing intensity estimation model of the present application is a BMogLSTM-KAN model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism The constructed model, this fusion not only retains the high efficiency performance of MogLSTM in sequence learning, but also strengthens the model's ability to learn the relationship between different samples in a small batch through the batch attention mechanism. In addition, the introduction of the KAN mechanism also provides a new activation function expression for the model, so that the weight parameters themselves are also involved in the learning process, further improving the model's ability to approximate complex functions. The improved model is used to fully learn the relationship between aboveground biomass inversion results, remote sensing data, terrain data, meteorological data, grazing activity factors and grazing intensity, and can accurately estimate grassland grazing intensity, which provides an important reference for scientific protection and sustainable management of grasslands. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0035] Figure 1 A flowchart of a grassland grazing intensity estimation method based on the BMgLSTM-KAN model provided in one embodiment of the present application.

[0036] Figure 2 A flowchart of constructing and training a grassland grazing intensity estimation model in a grassland grazing intensity estimation method based on a BMogLSTM-KAN model provided in an embodiment of the present application.

[0037] Figure 3 A schematic diagram of the structure of a traditional LSTM model in a grassland grazing intensity estimation method based on a BMogLSTM-KAN model provided in an embodiment of the present application.

[0038] Figure 4 A schematic diagram of the structure of a MogLSTM model used in a grassland grazing intensity estimation method based on a BMogLSTM-KAN model provided in an embodiment of the present application.

[0039] Figure 5 A schematic diagram of the structure of the BMgLSTM-KAN model in a grassland grazing intensity estimation method based on the BMgLSTM-KAN model provided in one embodiment of the present application.

[0040] Figure 6 A schematic diagram of the structure of a batch attention module in a grassland grazing intensity estimation method based on a BMogLSTM-KAN model provided in one embodiment of the present application.

[0041] Figure 7 A schematic diagram of the network structure of the KAN mechanism in a grassland grazing intensity estimation method based on the BMgLSTM-KAN model provided in one embodiment of the present application.

[0042] Figure 8 A schematic diagram of the results of residual analysis of a BMogLSTM-KAN model in a grassland grazing intensity estimation method based on the BMogLSTM-KAN model provided in one embodiment of the present application.

[0043] Fig. 9 A box plot of the regression performance of three algorithms on the same data set in a grassland grazing intensity estimation method based on a BMgLSTM-KAN model provided in one embodiment of the present application.

[0044] Fig.10 A schematic diagram of the fitting effect of the evaluation indicators of the BMogLSTM-KAN model in a grassland grazing intensity estimation method based on the BMogLSTM-KAN model provided in one embodiment of the present application.

[0045] Fig.11 A spatial distribution map of grazing intensity of typical grasslands in the temperate zone of Inner Mongolia in a grassland grazing intensity estimation method based on the BMogLSTM-KAN model provided in one embodiment of the present application.

[0046] Fig.12 A schematic diagram of the functional modules of a grassland grazing intensity estimation system based on the BMogLSTM-KAN model provided in one embodiment of the present application.

[0047] Fig.13 A schematic diagram of the structure of a computer device provided in one embodiment of the present application. DETAILED DESCRIPTION

[0048] 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 described embodiments 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 ordinary technicians in this field without creative work are within the scope of protection of this application.

[0049] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0050] The embodiment of the present application provides a method for estimating grassland grazing intensity based on a BMogLSTM-KAN model. The method can be executed by a computer device, specifically, can be executed by a computer device such as a terminal or a server alone, or can be executed by a terminal and a server together, including the following steps:

[0051] A1. Obtain the inversion results of aboveground biomass, remote sensing data, topographic data, meteorological data and grazing activity factors of the area to be estimated.

[0052] A2. Based on the trained grassland grazing intensity estimation model, the grazing intensity estimation result of the area to be estimated is obtained according to the inversion results of the aboveground biomass, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

[0053] In an exemplary embodiment, before step A1, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model further includes a process of constructing and training a grassland grazing intensity estimation model, such as Figure 2 As shown, the process includes the following steps:

[0054] B1. Construct a grazing intensity estimation sample data set; the grazing intensity estimation sample data set includes a number of grazing intensity estimation samples; the grazing intensity estimation samples include aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors as independent variables, and grassland grazing intensity as a dependent variable. In this embodiment, step B1 constructs a grazing intensity estimation sample data set, specifically including the following steps:

[0055] B11. Conduct field surveys and records in the study area to obtain grassland grazing conditions in the study area; the grassland grazing conditions include the number of grazing livestock, the types of grazing livestock, the pasture area and the grazing time; the study area is a grassland area of ​​the same type as the area to be estimated.

[0056] B12. Calculate the grassland grazing intensity of the study area according to the grassland grazing conditions in the study area. In this embodiment, the grassland grazing intensity of the study area during the grazing period is calculated according to the following formula:

[0057]

[0058] Among them, GI is the grassland grazing intensity in the study area, SU is the sheep unit value in the study area, and the sheep unit is based on sheep, and different grazing livestock species in the study area are converted into sheep units, and GA is the pasture area.

[0059] B13. Conduct field measurements in the study area and obtain the inversion results of the aboveground biomass in the study area.

[0060] B14. Obtain remote sensing data, topographic data, meteorological data and grazing activity factors of the study area.

[0061] B15. Take the aboveground biomass inversion results, remote sensing data, topographic data, meteorological data and grazing activity factors of the study area as independent variables, and the grassland grazing conditions in the study area as the dependent variable to obtain a grazing intensity estimation sample.

[0062] B16. Repeat the above steps B11 to B15 in several study areas to obtain a grazing intensity estimation sample data set including several grazing intensity estimation samples.

[0063] B2. Based on the BMogLSTM-KAN model, an initial grassland grazing intensity estimation model is constructed. The BMogLSTM-KAN model includes a batch attention module, a MogLSTM model and a KAN mechanism; the batch attention module includes multi-head attention, layer normalization and a feedforward network; the MogLSTM model is an improved model obtained by adding an interactive mechanism to the traditional LSTM model.

[0064] B3. Based on the grazing intensity estimation sample data set, the initial grassland grazing intensity estimation model is trained to obtain a trained grassland grazing intensity estimation model.

[0065] In order to facilitate understanding of the overall trend of grazing intensity in the area to be estimated, the grassland grazing intensity estimation method based on the BMgLSTM-KAN model provided in this embodiment further includes the following steps:

[0066] A3. Based on the grazing intensity estimation results of the area to be estimated, a spatial distribution map of grazing intensity in the area to be estimated is drawn to determine the overall trend of grazing intensity in the area to be estimated.

[0067] The following is a specific example to illustrate the effectiveness of the grassland grazing intensity estimation method based on the BMogLSTM-KAN model provided by the above embodiment of the present application. First, a study area is selected. The study area in this example is located in the typical temperate grassland of Inner Mongolia, 43°03′ to 46°73′ north latitude, 113°46′ to 119°21′ east longitude, and an average altitude of 1076m. The region is dominated by a semi-arid climate, with an average annual temperature of 2.5 degrees Celsius and an average annual precipitation of about 271 mm, of which 80% is concentrated in the growing season from May to August. According to the Chinese Soil Classification System, the soil type in the region is sandy chestnut. The study area is dominated by typical grassland as the main ecosystem type, and the vegetation type is herbaceous plants. The dominant species include sand onion, Anemarrhena, Fescue, Stipa grandis, Leymus chinensis, etc., and the growing season is from April to August. The main grazing livestock species are Simmental cattle and big-tailed sheep.

[0068] After the study area is determined, the first step is to collect known data in the study area for training the deep learning model, including the following steps:

[0069] (1) AGB and GI data were obtained by field measurement in the study area:

[0070] Field data collection in the study area was carried out from July 1 to August 1, 2023, and two types of measured data, GI and grassland AGB, were collected. The GI data were investigated and recorded with the help of GPS toolbox, including the number and type of grazing livestock, pasture area, and grazing time.

[0071] Grassland grazing intensity refers to the number of livestock per unit area within a certain period of time. The measured grassland GI is calculated based on field data. In order to make GI comparable between different paddocks, the number of various livestock is uniformly converted into sheep units (SU) in this example. Among them, 1 sheep is equal to 1SU, 1 cow is equal to 4SU, and 1 horse is equal to 5SU. The calculation method of GI is as follows:

[0072]

[0073] GI is grazing intensity (head / ha); SU is sheep unit (head); GA is grazing area (ha). The measured GI data are used for training and testing of subsequent estimation models.

[0074] Regarding the sampling of above-ground biomass (AGB), we selected typical temperate grassland communities with flat terrain and relatively uniform growth conditions to set up sample plots (a total of 75 sample plots and 225 sample AGB data were collected), and the distance between sample plots (30m×30m) was not less than 5km. Three sample plots (1m×1m) were randomly selected from each sample plot, and the grassland AGB was cut at the ground level. The longitude and latitude of the sample plot and the sampling time were recorded. The samples were bagged and numbered and brought back to the laboratory to be dried in a 65℃ oven to constant weight to obtain the dry weight of AGB for each sample plot. Finally, the average value of AGB of the three sample plots was taken as the sample plot data to obtain the measured ground data of grassland AGB.

[0075] AGB can usually reflect the grass condition of pastures and is an important indicator for evaluating grassland GI. The inversion of grassland AGB is already very mature. This study used the random forest model to invert AGB based on field measured AGB data. The accuracy of the AGB inversion model is R 2 =0.94, RMSE =7.50. The AGB inversion result is used to study grassland grazing intensity as one of the important influencing factors of the grazing intensity estimation model.

[0076] (2) Obtain remote sensing data, topographic data, meteorological data and grazing activity factors of the study area through the Internet:

[0077] Sentinel-2 is composed of two complementary satellites (Sentinel-2A / B) launched by the European Space Agency (ESA). The multispectral instrument (MSI) on Sentinel-2 has multiple spectral bands from visible light to shortwave infrared (443-2190nm), high spatial resolution, and a revisit period of 5 days. The Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI) are the most widely used vegetation indices in vegetation monitoring, which can reflect the vegetation status of pastures and are negatively correlated with GI. In this example, the Sentinel-2MSI image product is called in the GEE platform, and the product is pre-processed by QA cloud removal, image fusion, and calculation of vegetation index.

[0078] The terrain elevation data (DEM) comes from the European Space Agency (ESA), with a resolution of 30 meters and a spatial reference of GCS_WGS_1984. The livestock grazing process varies with the terrain, and the gentler the slope, the easier it is to graze. ArcGIS software was used to process the DEM data to obtain slope and aspect data.

[0079] Meteorological data include temperature and precipitation data from the National Oceanic and Atmospheric Administration (NOAA). Temperature and precipitation affect grassland vegetation growth and have a certain correlation with grazing intensity. Using temperature and precipitation station data in conjunction with DEM data, Kriging interpolation of meteorological data was performed to obtain the monthly average temperature and monthly average precipitation data of the study area.

[0080] Considering that high-coverage grasslands near water sources are more convenient for herders to graze and livestock to drink water (livestock need water in addition to food during grazing), and at the same time, livestock grazing and foraging are strongly correlated with the location of settlements. Therefore, the distance to the river and the distance to the settlement are selected as grazing activity factors. The river data comes from Open StreetMap, and the settlement data comes from the National Tibetan Plateau Data Center.

[0081] Before using the known data obtained through the above steps, it is necessary to perform corresponding preprocessing on the data to generate GI measured data and AGB, NDVI, EVI, DEM, slope, aspect, monthly average temperature, monthly average precipitation, distance to rivers DTR, and distance to settlements DTS. The resolution and data format of each data are shown in Table 1.

[0082] Table 1 Resolution and data format of each data obtained in this example

[0083]

[0084] Next, we build the model and establish the BMogLSTM-KAN regression model, taking the GI data of the measured points as the dependent variable and the corresponding AGB, NDVI, EVI, climate, terrain, and grazing activity data as the independent variables as the input data of the model, adjusting the model parameters, and training and testing the model. The specific steps include:

[0085] (1) Construction and training of BMogLSTM-KAN model:

[0086] As a type of recurrent neural network (RNN), LSTM can capture long-term dependencies in data. LSTM can effectively avoid the gradient explosion or vanishing problem in RNN. Figure 3 shows the node connection of a typical LSTM unit, where c t-1 and h t-1 Represent the cell state and output of the previous time step respectively. In the LSTM model, there are three gates, including the forget gate f, the input gate i, the output gate o and the internal cell state c. The weight matrix, recursive weight and bias are respectively denoted as W1=[W f ,W i ,W c ,W o ],W2=[W hf ,W hi ,W hc ,W ho ] and b1=[b f ,b i ,b c ,b o ], then the feedforward process of LSTM at time t is described as:

[0087]

[0088] Among them, σ and tanh are activation functions, and ⊙ represents the product.

[0089] The Mogrifier LSTM model (MogLSTM for short) adds an interactive mechanism to the LSTM model, which allows the current state input unit to fully interact with the hidden unit, and significantly improves the performance in long sequence data processing. In the storage unit of the LSTM network, the previous hidden state h t-1 With the current input x t There is a lack of deep interaction between them, and they only have a simple interaction when entering the unit. Melis et al. believe that h t-1 and x t The mutual independence of the hidden state h may limit the generalization ability of the LSTM network. Melis et al. improved the LSTM model and introduced a gating mechanism called "interaction gate". This mechanism ensures that the hidden state ht-1 and input x t Sufficient interaction is performed before entering the storage unit, thereby improving the learning efficiency of the model. Figure 4 The MogLSTM model with a 2-round interaction mechanism is shown. If the process of updating data in the LSTM model is expressed as a function:

[0090] LSTM(x t ,h t-1 ,c t-1 )=(h t ,c t ).

[0091] Then the mechanism of MogLSTM updating information can be expressed as:

[0092]

[0093] in, Represents x after the interaction gate t and h t-1 Output.

[0094] In this example, the MogLSTM model is built in Python 3.8. Furthermore, based on MogLSTM, the batch attention module (Batch Attention Module) and the weight activation function of KAN (Kolmogorov-Arnold Networks) are introduced, such as Figure 5 As shown in the figure. This fusion not only retains the high efficiency of MogLSTM in sequence learning, but also strengthens the model's ability to learn the relationship between different samples in a small batch through the batch attention mechanism. Specifically, the batch attention module allows the model to not only focus on the characteristics of a single sample when processing data, but also capture the relationship between samples, thereby achieving a more comprehensive feature representation. In addition, the introduction of KAN also provides the model with a new activation function (spline function as a weight activation function) expression, so that the weight parameters themselves are also involved in the learning process, further improving the model's ability to approximate complex functions. This BMogLSTM-KAN, which combines batch attention and weight activation functions, is not only innovative in theory, but also shows excellent performance in practical applications.

[0095] 1) Batch Attention Module

[0096] During the information propagation process, MogLSTM ignores the possible information propagation between different training samples in each small batch. The relationships between different data samples are diverse and complex. Effectively extracting the relationships between samples is crucial to improving the learning efficiency of the model. In the prior art, a simple and effective batch attention module is proposed to promote robust representation learning by utilizing the ability of the Transformer encoder to automatically learn sample relationships. The batch attention module consists of multi-head attention (Multi-Head), layer normalization (LayerNorm) and feedforward network (FFN). In the batch attention module, the Transformer encoder is used to process the batch dimension of the input data. The specific process is as follows: Figure 6 Assume is the sequence of input features, where B is the length of the sequence, i.e., the batch size, and d model is the dimension of the feature.

[0097] The batch attention module framework is specifically manifested in that the Transformer network is introduced and the structural advantages of the Transformer network are used to mine the relationship between samples. Through the proposed batch attention module, we can pass messages along different samples within each mini-batch. From an optimization point of view, all samples can contribute to the learning of any sample. Therefore, the batch attention module implicitly enriches the current training data based on the samples of the entire mini-batch. In this way, the attention mechanism in the batch attention module becomes the cross-attention between different samples in each mini-batch.

[0098] 2) KAN Mechanism

[0099] Inspired by the Kolmogorov-Arnold representation theorem, a neural network structure Kolmogorov-Arnold Networks (KANs) is proposed in the prior art as an alternative to multilayer perceptrons (MLPs). The core idea of ​​KANs is to replace the fixed activation functions at the nodes (neurons) in traditional neural networks with learnable activation functions on the edges (weights), and place these functions on the edges (weights) of the network instead of on the nodes (neurons) in traditional MLPs. KAN does not have a linear weight matrix, and each of its weight parameters is replaced by a learnable one-dimensional function parameterized as a spline function.

[0100] Because KAN has MLP outside and spline curve inside, KAN can not only learn the combined structure (external degrees of freedom), but also approximate the single variable function (internal degrees of freedom) well and learn functions more accurately, such as Figure 7As shown, KAN replaces fixed activation functions with learnable activation functions (“weights”) on the edges. The unique framework of LAN enables KAN to approximate nonlinear functions more effectively while keeping the network size smaller. KAN addresses these limitations by incorporating learnable spline functions on each weight, providing greater flexibility and accuracy in learning nonlinear relationships. Compared to MLP, this unique architecture provides faster neural scaling laws and requires fewer parameters.

[0101] The Kolmogorov-Arnold theorem is specifically expressed as follows: It can be represented by a finite combination of single variable continuous functions and additive binary operations. The Kolmogorov-Arnold theorem is defined by KANs of shape [n,2n+1,1]. Here, a general deeper KAN architecture is defined using [n1,...,nL+1], where L represents the number of layers of the KAN.

[0102]

[0103] All operations in the formula are differentiable, so KAN can be trained through back propagation. The BMogLSTM-KAN model is built in Python 3.8 environment. The model training process and GI estimation process are consistent with the MogLSTM model.

[0104] (2) Model performance evaluation and comparison:

[0105] In this case, the coefficient of determination (R 2 ), root mean square error (RMSE) and mean absolute error (MAE) were used as evaluation indicators for the accuracy of grassland grazing intensity estimation models. 2 It reflects the fitting ability between the estimated value and the measured value, RMSE reflects the degree of dispersion between the estimated value and the measured value, and MAE reflects the average level of the difference between the estimated value and the measured value. The unit of RMSE is head / ha -1 , which is the same as the unit of the measured value of grassland grazing intensity. The calculation formulas are as follows:

[0106]

[0107] Where n is the number of samples, y i represents the measured value of GI, is the average value of GI, is an estimate of GI.

[0108] In this example, the MogLSTM model was further improved to develop the BMogLSTM-KAN model. A residual analysis was performed on the BMogLSTM-KAN model. During the analysis, a total of 197 sample points were used, of which 80% of the data was divided into training sets and the remaining 20% ​​was used as test sets. Figure 8 As shown in the figure, the distance between the measured value and the estimated value represents the residual, and the larger the distance between the two points, the larger the residual. The improved BMogLSTM-KAN performs well in estimating grazing intensity. The residual of BMogLSTM-KAN is the smallest overall, and the estimated GI curve is highly consistent with the actual GI measurement value, showing excellent fitting performance.

[0109] In order to verify the improvement effect of the BMogLSTM-KAN model in this example, the fitting effects of BMogLSTM, MogLSTM-KAN, and BMogLSTM-KAN on the same data set are compared. In order to further evaluate the stability of the model and reduce random fluctuations during training, we used a 10-cycle training method to evaluate the model performance. The performance of the model on the test set was recorded, including R 2 , RMSE, MAE and loss value. By comparing the regression losses of different algorithms, it can be found that the average regression loss of the BMogLSTM-KAN algorithm is the lowest, specifically BMogLSTM-KAN(0.80)<BMogLSTM(1.17)<MogLSTM-KAN(1.32). Fig. 9 The regression performance box plots of the three algorithms on the same data set are shown. As shown in the figure, the upper and lower limits of the four indicators of BMogLSTM and MogLSTM-KAN are far apart, and the upper and lower limits of RMSE are more than 0.55, indicating that the model is not stable enough. The four indicators of the BMogLSTM-KAN algorithm are evenly distributed, and the gap between the quarter point and the three quarter point is relatively small, so the stability is better.

[0110] Through the comparison between different models, in this example, the BMogLSTM-KAN model was used to complete the grassland grazing intensity estimation task. The input factors for model establishment include 10 indicators, including aboveground biomass, vegetation index, meteorological factors, terrain factors and grazing activity factors, specifically AGB, NDVI, EVI, monthly average temperature, monthly average precipitation, DEM, slope, slope aspect, DTR and DTS. The input data is divided into training data set and test data set according to 8:2, and R 2 , RMSE and MAE are used as evaluation indicators for the accuracy of grazing intensity estimation models, such as Fig.10As shown in the figure, the BMogLSTM-KAN model with added batch attention module and KAN mechanism has the best fitting effect, with a determination coefficient of 0.93 and a root mean square error of 0.68.

[0111] Finally, the trained grassland grazing intensity estimation model obtained in the above steps is used to obtain the grazing intensity estimation results of the estimated area based on the aboveground biomass inversion results AGB, remote sensing data (NDVI, EVI), terrain data (DEM, slope, slope aspect), meteorological data (monthly average temperature, monthly average precipitation) and grazing activity factors (DTR, DTS). Fig.11 The figure shows the estimation results of grazing intensity of typical grasslands in temperate zone of Inner Mongolia by BMogLSTM-KAN model. The overall trend of the spatial distribution map of grazing intensity of typical grasslands in temperate zone of Inner Mongolia is as follows: GI gradually increases from northeast to southwest of the study area. The GI estimated by BMogLSTM-KAN ranges from 0 to 8.51 head / ha. In addition, the GI estimated by BMogLSTM-KAN model shows that the areas with high GI are greater than those with low GI.

[0112] The results of the estimation of grassland grazing intensity using the BMogLSTM-KAN model were statistically analyzed according to the county (banner) boundaries. As shown in Table 2, the maximum grazing intensity of each county (banner) in the typical temperate grassland of Inner Mongolia exceeded 8.2 head / ha, indicating that overgrazing existed in all counties. In particular, the mean grazing intensity of Xilinhot City and Abaga Banner was 5.34 head / ha and 5.92 head / ha, respectively, with a small standard deviation, reflecting that the grazing intensity in these two regions was not only high, but also widely distributed.

[0113] Table 2 Grazing intensity in each county (banner) of typical temperate grassland in Inner Mongolia

[0114] name Minimum Maximum average value Standard Deviation East Ujimchin Banner 0 8.337016106 4.45583 2.2152212 West Ujimchin Banner 0 8.289620399 3.981979 2.441014 Abaga Banner 0 8.505903244 5.921941 1.5667299 Xilinhot City 0 8.301514626 5.339965 1.909221 Linxi County 0 8.202166557 2.361669 2.6007649 Keshiketeng Banner 0 8.341110229 2.837717 2.1454055

[0115] In this case, the improved BMogLSTM-KAN model is used to estimate the grazing intensity of typical grasslands in the temperate zone of Inner Mongolia based on measured grazing intensity data and grazing influencing factors (AGB, NDVI, EVI, monthly average temperature, monthly average precipitation, DEM, slope, aspect, DTR, DTS), quantitatively study grassland grazing intensity, and achieve the estimation of grazing intensity of typical grasslands in the temperate zone of Inner Mongolia. We analyzed the performance of the model in grazing intensity estimation and found that the deep learning model BMogLSTM-KAN showed good results in estimating GI, R 2=0.93, RMSE=0.68. This model is based on the MogLSTM model and is optimized and improved by introducing advanced batch attention modules and KAN mechanisms. In the comparative evaluation of the model's optimization process, the model demonstrated the accuracy of its estimation and the robustness of its performance. Finally, 10 key grazing influencing factors were used as input data, and 4 models were used to estimate the grazing intensity of typical temperate grasslands in Inner Mongolia. The spatial distribution map of grazing intensity of typical temperate grasslands in Inner Mongolia was depicted, and the distribution of grazing intensity in the study area was quantitatively evaluated. The research results can provide an important reference for the scientific protection and sustainable management of grasslands in Inner Mongolia.

[0116] Based on the same inventive concept, the embodiment of the present application also provides a system for implementing the above-mentioned method for estimating grassland grazing intensity based on the BMogLSTM-KAN model. The implementation scheme for solving the problem provided by the system is similar to the implementation scheme recorded in the above-mentioned method, so the specific limitations in one or more system embodiments provided below can refer to the limitations of the method for estimating grassland grazing intensity based on the BMogLSTM-KAN model above, which will not be repeated here.

[0117] In an exemplary embodiment, Fig.12 As shown in the figure, a grassland grazing intensity estimation system based on the BMogLSTM-KAN model is provided, including the following modules:

[0118] The regional data acquisition module is used to obtain the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated.

[0119] The regional grazing intensity estimation module is used to obtain the grazing intensity estimation result of the area to be estimated based on the trained grassland grazing intensity estimation model, according to the aboveground biomass inversion result of the area to be estimated, the remote sensing data, the terrain data, the meteorological data and the grazing activity factor; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

[0120] certainly, Fig.12 The architecture shown is only exemplary and can be omitted according to actual needs when implementing different functions. Fig.12 One or at least two components of the system shown.

[0121] In an exemplary embodiment, a computer device is provided. The computer device may be a server or a terminal. The internal structure diagram thereof may be as follows: Fig.13As shown. The computer device includes a processor, a memory, an input / output interface (Input / Output, referred to as I / O) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through a system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The input / output interface of the computer device is used to exchange information between the processor and an external device. The communication interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a grassland grazing intensity estimation method based on the BMogLSTM-KAN model provided in the above method embodiment can be implemented.

[0122] Those skilled in the art will understand that Fig.13 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0123] In an exemplary embodiment, a computer device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0124] In an exemplary embodiment, a computer-readable storage medium is provided, storing a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0125] In an exemplary embodiment, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the steps in the above method embodiments are implemented.

[0126] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant regulations.

[0127] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM may be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0128] The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. The non-relational database may include a distributed database based on blockchain, etc., but is not limited thereto. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but is not limited thereto.

[0129] The technical features of the above embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0130] This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. At the same time, for those skilled in the art, according to the ideas of this 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 limiting this application.

Claims

1. A grassland grazing intensity estimation method based on BMogLSTM-KAN model, characterized in that: include: Obtain the inversion results of aboveground biomass, remote sensing data, topographic data, meteorological data and grazing activity factors of the area to be estimated; Based on the trained grassland grazing intensity estimation model, the grazing intensity estimation result of the area to be estimated is obtained according to the aboveground biomass inversion result of the area to be estimated, the remote sensing data, the terrain data, the meteorological data and the grazing activity factor; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

2. The method for estimating grassland grazing intensity based on the BMogLSTM-KAN model according to claim 1, characterized in that: Before obtaining the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model also includes: Constructing a grazing intensity estimation sample data set; the grazing intensity estimation sample data set includes a plurality of grazing intensity estimation samples; the grazing intensity estimation samples include aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors as independent variables, and grassland grazing intensity as a dependent variable; Based on the BMogLSTM-KAN model, an initial grassland grazing intensity estimation model was constructed; Based on the grazing intensity estimation sample data set, the initial grassland grazing intensity estimation model is trained to obtain a trained grassland grazing intensity estimation model.

3. The grassland grazing intensity estimation method based on the BMogLSTM-KAN model according to claim 2, characterized in that: Construct a sample data set for grazing intensity estimation, including: Conduct field surveys and records in the study area to obtain grassland grazing conditions in the study area; the grassland grazing conditions include the number of grazing livestock, the type of grazing livestock, the pasture area and the grazing time; the study area is a grassland area of ​​the same type as the area to be estimated; According to the grassland grazing conditions in the study area, the grassland grazing intensity in the study area was calculated; Field measurements were conducted in the study area to obtain the inversion results of aboveground biomass in the study area; Obtain remote sensing data, topographic data, meteorological data and grazing activity factors in the study area; The aboveground biomass inversion results, remote sensing data, topographic data, meteorological data and grazing activity factors of the study area were used as independent variables, and the grassland grazing situation in the study area was used as the dependent variable to obtain a sample of grazing intensity estimation; The above process is repeated in several study areas to obtain a grazing intensity estimation sample data set including several grazing intensity estimation samples.

4. The method for estimating grassland grazing intensity based on the BMogLSTM-KAN model according to claim 3, characterized in that: The grassland grazing intensity in the study area during the grazing period was calculated according to the following formula: Among them, GI is the grassland grazing intensity in the study area, SU is the sheep unit value in the study area, and the sheep unit is based on sheep, and different grazing livestock species in the study area are converted into sheep units, and GA is the pasture area.

5. The method for estimating grassland grazing intensity based on the BMogLSTM-KAN model according to claim 1, characterized in that: The BMogLSTM-KAN model includes a batch attention module, a MogLSTM model and a KAN mechanism; the batch attention module includes multi-head attention, layer normalization and a feedforward network; the MogLSTM model is an improved model obtained by adding an interactive mechanism to the traditional LSTM model.

6. The method for estimating grassland grazing intensity based on the BMogLSTM-KAN model according to claim 1, characterized in that: The grassland grazing intensity estimation method based on the BMgLSTM-KAN model also includes: According to the grazing intensity estimation results of the area to be estimated, a spatial distribution map of grazing intensity in the area to be estimated is drawn to determine the overall trend of grazing intensity in the area to be estimated.

7. A grassland grazing intensity estimation system based on BMogLSTM-KAN model, characterized in that: include: The regional data acquisition module is used to obtain the aboveground biomass inversion results, remote sensing data, terrain data, meteorological data and grazing activity factors of the area to be estimated; The regional grazing intensity estimation module is used to obtain the grazing intensity estimation result of the area to be estimated based on the trained grassland grazing intensity estimation model, according to the aboveground biomass inversion result of the area to be estimated, the remote sensing data, the terrain data, the meteorological data and the grazing activity factor; the grassland grazing intensity estimation model is a model constructed based on the BMogLSTM-KAN model; the BMogLSTM-KAN model is a model obtained by improving the MogLSTM model based on the batch attention module and the KAN mechanism.

8. A computer device comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the grassland grazing intensity estimation method based on the BMogLSTM-KAN model as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model described in any one of claims 1 to 6 is implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the grassland grazing intensity estimation method based on the BMogLSTM-KAN model described in any one of claims 1 to 6 is implemented.

Images