Monitoring, preventing and treating system for alpine grassland shrub formation and working method of monitoring, preventing and treating system

Abstract

The invention provides a monitoring and prevention system for alpine grassland shrubby and a working method thereof, and the system comprises a division and collection module which is used for carrying out the region delimiting of an alpine grassland, and collecting the image data and periodic data in the periodic time of the delimiting region through a satellite, an unmanned plane and manpower, the periodic data comprises periodic data acquisition time, a time interval and environment data of a delimited area at the time; wherein the environment data of the delimited region at the time comprises all parameters related to plant growth in the region, continuous data acquisition is carried out by selecting the region of the alpine grassland, the shrubby development situation of the alpine grassland can be visually known, and the evaluation result of the shrubby development situation of the alpine grassland can be obtained through the Bayesian network correlation evaluation model. The comprehensive benefits of the environment, the society and the economy under different management schemes can be predicted, quantitative prediction results can be provided for grassland management and policy making, and making of more scientific and reasonable management measures is helped.

Description

Technical Field

[0001] The invention relates to the technical field of alpine grassland monitoring and prevention, and in particular to a system for monitoring and preventing shrubbery in alpine grassland and a working method thereof. Background Art

[0002] In alpine grassland areas, the problem of shrubbery is becoming increasingly prominent, seriously threatening the ecological balance and health of grasslands. Although a variety of grassland ecological monitoring and management systems have been established, such as the "Grassland Ecological Health Intelligent Monitoring Network System", "Alpine Grassland Ecological Management System" and "Alpine Mountain Grassland Degradation Monitoring Platform", research on the dynamic monitoring system of grassland shrubbery is still insufficient.

[0003] However, most of the existing monitoring and prevention strategies are in the experimental and exploratory stage, lacking a mature system and effective management platform. Therefore, it is urgent to build a monitoring and prevention system and working methods for alpine grassland shrubbery that takes into account environmental protection, social welfare and economic development. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the deficiencies in the prior art, the present invention provides a system for monitoring and controlling shrubbery in alpine grasslands and a working method thereof, so as to solve the problems raised in the above-mentioned background technology.

[0006] (II) Technical solution

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a monitoring and prevention system for shrubbery in alpine grasslands, comprising:

[0008] The division and collection module is used to demarcate the alpine grassland area, and collect image data and periodic data within the period of the demarcated area through satellites, drones and manual work. The periodic data includes the time and time interval of periodic data collection and the environmental data of the demarcated area at that time;

[0009] The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area;

[0010] A data processing module, used for dividing and processing the image data and the periodic data collected by the division and collection module to obtain an image feature data set and a periodic feature data set;

[0011] A model management module, which constructs a prediction model based on the image feature data set and the period feature data set, and is also used to update the Bayesian network association evaluation model;

[0012] The decision-making and prevention module is used to input the real-time information of the alpine grassland into the Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

[0013] As a preferred embodiment of this embodiment, the data processing module divides and processes the image data and the periodic data collected by the division collection module to obtain an image feature data set and a periodic feature data set, including:

[0014] Data segmentation: image data and periodic data are segmented into several time periods based on the collected periodic time;

[0015] Data processing, deleting duplicate data and abnormal data in the segmented image data and periodic data, and renaming or defining the several image data and periodic data formed after segmentation to obtain several image feature data and periodic feature data.

[0016] As a preferred embodiment of this invention, the process of constructing the prediction model by the model management module includes:

[0017] Obtain image feature datasets and period feature datasets, and construct training sets;

[0018] Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data;

[0019] The target image data and the target period data are processed by using a neural network to generate an image-period feature map;

[0020] The image-period feature map is trained, and a prediction model is built based on the training results.

[0021] As a preferred embodiment of this embodiment, the processing of the target image data and the target period data by using a neural network to generate an image-period feature map includes:

[0022] The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data;

[0023] Based on a number of image feature data and period feature data obtained after segmentation, the mapping relationship model is adjusted and optimized to obtain an image-period feature mapping.

[0024] As a preferred embodiment of the present invention, the decision-making control module updates the Bayesian network association evaluation model and inputs the real-time information into the Bayesian network association evaluation model to obtain the control strategy of the alpine grassland, which includes:

[0025] The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model;

[0026] The real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

[0027] As a preferred embodiment of this invention, the real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland, which includes:

[0028] Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold;

[0029] The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor;

[0030] Determine the size of the impact factor and impact threshold;

[0031] If so, the periodic data of alpine grasslands are adjusted based on the differences between influencing factors and impact thresholds and multidimensional benefits;

[0032] If not, the alpine grassland will be maintained in its original state.

[0033] As a preferred embodiment of the present invention, the process of adjusting the periodic data of the alpine grassland based on the difference between the influencing factor and the influencing threshold and the multi-dimensional benefits includes:

[0034] Based on the image-period feature mapping and multi-dimensional benefits, the influence of the periodic features on the image features is determined, the weight of the influence of the periodic features on the image features under the multi-dimensional benefits condition is calculated, and the periodic data of the alpine grassland is adjusted based on the calculated weight;

[0035] The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is:

[0036] WC(x t )=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ

[0037] Where W is the weight coefficient, C(x t) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 It is the state quantity of the image feature data at time t-1.

[0038] As a preferred embodiment of the present invention, a method for monitoring and controlling the thicketization of alpine grasslands is provided, which adopts the above-mentioned system for monitoring and controlling the thicketization of alpine grasslands, and includes:

[0039] Delineate the alpine grassland area, collect image data and periodic data within the period of the delineated area through satellite, UAV and manual, the periodic data includes the time of periodic data collection, time interval and environmental data of the delineated area at that time;

[0040] The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area;

[0041] The collected image data and periodic data are divided and processed based on a preset time interval to obtain an image feature data set and a periodic feature data set;

[0042] Building a prediction model based on a graphic feature dataset and a period feature dataset, wherein the period feature dataset is an input of the prediction model and the image feature dataset is an output of the prediction model;

[0043] Update the Bayesian network association evaluation model, input the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtain the prevention and control strategy of the alpine grassland, and obtain the relationship data of the shrubbery trend of the alpine grassland;

[0044] Based on the alpine grassland shrubbery trend relationship data, a prevention and control management plan for the alpine grassland to be detected is obtained;

[0045] The process of building a predictive model includes:

[0046] Obtain image feature datasets and period feature datasets, and construct training sets;

[0047] Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data;

[0048] The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data;

[0049] The mapping relationship model is adjusted and optimized based on a number of image feature data and period feature data obtained after segmentation to obtain an image-period feature mapping;

[0050] The image-period feature map is trained, and a prediction model is built based on the training results.

[0051] As a preferred embodiment of the present invention, the process of updating the Bayesian network association evaluation model, inputting the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtaining the prevention and control strategy of the alpine grassland, and obtaining the relationship data of the shrubbery trend of the alpine grassland includes:

[0052] The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model;

[0053] Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold;

[0054] The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor;

[0055] Determine the size of the impact factor and impact threshold;

[0056] If so, the influence of the periodic features on the image features is determined based on the image-period feature mapping and the multidimensional benefit, the weight of the influence of the periodic features on the image features under the multidimensional benefit condition is determined, and the periodic data of the alpine grassland is adjusted based on the calculated weight;

[0057] The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is:

[0058] WC(x t )=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ

[0059] Where W is the weight coefficient, C(x t ) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 is the state quantity of the image feature data at time t-1;

[0060] If not, the alpine grassland will be maintained in its original state.

[0061] (III) Beneficial effects

[0062] The present invention provides a monitoring and control system for coping with alpine grassland shrubbery and a working method thereof, which have the following beneficial effects: by selecting an area of ​​alpine grassland for continuous data collection, the development trend of alpine grassland shrubbery can be intuitively understood; by processing the collected image data and periodic data, the correlation between alpine grassland shrubbery and periodic data can be determined; and a prediction model can be established based on the processed data; the prediction model can be trained based on an image feature data set and a periodic feature data set; the correlation between the image feature data set and the periodic feature data set in the prediction model is optimized using a Bayesian optimization algorithm, and a Bayesian network correlation evaluation model having a mapping relationship between the image feature data set and the periodic feature data set is obtained; Therefore, by simply collecting real-time information on alpine grasslands, it is possible to monitor the spatiotemporal dynamics of grassland shrub expansion in real time, promptly warn of potential problems, help managers quickly identify and respond to grassland ecological changes, and significantly improve monitoring accuracy, enhance the timeliness and effectiveness of grassland ecological protection. At the same time, through the Bayesian network association evaluation model, the comprehensive environmental, social and economic benefits under different management schemes can be predicted, which can provide quantitative prediction results for grassland management and policy formulation, help formulate more scientific and reasonable management measures, and by calculating the impact of these features on subsequent changes in image data, based on the changes in alpine grasslands and combined with multi-dimensional benefits, more accurate prevention and control plans for alpine grasslands can be formulated to achieve the best benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 A flow chart of a calibration method for installing a vehicle-mounted camera sensor according to the present invention;

[0064] Figure 2Schematic diagram of the structure of a vehicle camera calibration platform according to an embodiment of the present invention. DETAILED DESCRIPTION

[0065] The embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and cannot be understood as limiting the present invention.

[0066] The disclosure below provides many different embodiments or examples to realize different structures of the present invention. In order to simplify the disclosure of the present invention, the parts and settings of specific examples are described below. Of course, they are only examples, and the purpose is not to limit the present invention. In addition, the present invention can repeat reference numbers and / or reference letters in different examples, and this repetition is for the purpose of simplicity and clarity, which itself does not indicate the relationship between the various embodiments and / or settings discussed. In addition, the present invention provides various specific examples of processes and materials, but those of ordinary skill in the art can be aware of the application of other processes and / or the use of other materials.

[0067] like Figure 1 As shown, the embodiment of the present invention provides a system for monitoring and controlling shrubbery in alpine grasslands, including

[0068] The division and collection module is used to demarcate the alpine grassland area, and collect image data and periodic data within the period of the demarcated area through satellites, drones and manual work. The periodic data includes the time and time interval of periodic data collection and the environmental data of the demarcated area at that time;

[0069] The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area;

[0070] It should be noted that there are shrub areas in the alpine grassland in the selected area. The relevant parameters for plant growth are specifically the temperature, humidity, soil fertility, etc. of the area. The periodic data include field monitoring data such as vegetation, soil, meteorology, and hydrology.

[0071] A data processing module, used for dividing and processing the image data and the periodic data collected by the division and collection module to obtain an image feature data set and a periodic feature data set;

[0072] A model management module, which constructs a prediction model based on the image feature data set and the period feature data set, and is also used to update the Bayesian network association evaluation model;

[0073] The decision-making and prevention module is used to input the real-time information of the alpine grassland into the Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

[0074] Specifically, by selecting areas of alpine grassland for continuous data collection, we can intuitively understand the development trend of alpine grassland shrubbery. By processing the collected image data and periodic data (obtaining image feature data sets and periodic feature data sets after processing), we can determine the association between alpine grassland shrubbery and periodic data, and establish a prediction model based on these processed data. The prediction model can be trained based on the image feature data sets and the periodic feature data sets. The Bayesian optimization algorithm can be used to optimize the correlation between the image feature data sets and the periodic feature data sets in the prediction model, and a Bayesian network association evaluation model with a mapping relationship between the image feature data sets and the periodic feature data sets can be obtained. This can improve the prediction accuracy of the impact of changes in periodic data on image data, and thus obtain a more accurate alpine grassland shrubbery prevention and control plan.

[0075] This embodiment can monitor the spatiotemporal dynamics of grassland shrub expansion in real time simply by collecting instant alpine grassland information, promptly warn of potential problems, help managers quickly identify and respond to grassland ecological changes, and significantly improve monitoring accuracy, enhancing the timeliness and effectiveness of grassland ecological protection.

[0076] Among them, the cycle time of the demarcated area collected by satellite, drone and manual collection is at least one year, so as to have a deeper understanding of the relationship between image data and periodic data.

[0077] Furthermore, the data processing module divides and processes the image data and the periodic data collected by the division and collection module to obtain an image feature data set and a periodic feature data set, including:

[0078] Data segmentation: image data and periodic data are segmented into several time periods based on the collected periodic time;

[0079] When segmenting, the segmentation may be performed according to seasonal cycles or according to time periods corresponding to the growth habits of the plants.

[0080] Data processing, deleting duplicate data and abnormal data in the segmented image data and periodic data, and renaming or defining the several image data and periodic data formed after segmentation to obtain several image feature data and periodic feature data.

[0081] In the process of collecting data, since the time span is too large and the growth cycle of plants is not too slow, a large amount of duplicate data will eventually be collected. By dividing and processing the data, the data can be more standardized and easier to use.

[0082] Furthermore, the process of constructing the prediction model by the model management module includes:

[0083] Obtain image feature datasets and period feature datasets, and construct training sets;

[0084] Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data;

[0085] The target image data and the target period data are processed by using a neural network to generate an image-period feature map;

[0086] The image-period feature map is trained, and a prediction model is built based on the training results.

[0087] Specifically, the image feature dataset is a map of the changes in shrubbery in alpine grassland images within a period, which can intuitively show the changes in the alpine grassland shrubbery. Based on this, the data related to and changing in the alpine grassland shrubbery in the periodic feature dataset are extracted, and the inevitable connection between the two is analyzed through comparison, and then the mapping association between the two is determined through calculation.

[0088] Furthermore, the process of processing the target image data and the target period data by using a neural network to generate an image-period feature map includes:

[0089] The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data;

[0090] Based on a number of image feature data and period feature data obtained after segmentation, the mapping relationship model is adjusted and optimized to obtain an image-period feature mapping.

[0091] Specifically, the decision-making control module updates the Bayesian network association evaluation model and inputs the real-time information into the Bayesian network association evaluation model to obtain the control strategy of the alpine grassland, which includes:

[0092] The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model;

[0093] The real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

[0094] In this embodiment, the Bayesian network association evaluation model can be used to predict the comprehensive environmental, social and economic benefits under different management schemes, which can provide quantitative prediction results for grassland management and policy formulation, and help formulate more scientific and reasonable management measures.

[0095] It is understandable that most of the benefits are comprehensive benefits of the environment, society and economy.

[0096] Furthermore, the real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland, including:

[0097] Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold;

[0098] The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor;

[0099] Determine the size of the impact factor and impact threshold;

[0100] If so, the periodic data of alpine grasslands are adjusted based on the differences between influencing factors and impact thresholds and multidimensional benefits;

[0101] If not, the alpine grassland will be maintained in its original state.

[0102] Furthermore, the process of adjusting the periodic data of the alpine grassland based on the difference between the influencing factor and the influencing threshold and the multi-dimensional benefits includes:

[0103] Based on the image-period feature mapping and multi-dimensional benefits, the influence of the periodic features on the image features is determined, the weight of the influence of the periodic features on the image features under the multi-dimensional benefits condition is calculated, and the periodic data of the alpine grassland is adjusted based on the calculated weight;

[0104] The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is:

[0105] WC(x t )=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ

[0106] Where W is the weight coefficient, C(x t) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 It is the state quantity of the image feature data at time t-1.

[0107] Specifically, there are several types of periodic data, all of which will have an impact on the development of image data. By calculating the degree of influence of these features on the subsequent changes of image data (i.e., weight), we can formulate more accurate prevention and control plans for alpine grasslands based on the changes in alpine grasslands and combined with multi-dimensional benefits to achieve the best benefits.

[0108] like Figure 2 As shown, this embodiment also provides a working method for monitoring and controlling the thicketization of alpine grasslands, using the above-mentioned monitoring and controlling system for monitoring and controlling the thicketization of alpine grasslands, including:

[0109] Delineate the alpine grassland area, collect image data and periodic data within the period of the delineated area through satellite, UAV and manual, the periodic data includes the time of periodic data collection, time interval and environmental data of the delineated area at that time;

[0110] The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area;

[0111] The collected image data and periodic data are divided and processed based on a preset time interval to obtain an image feature data set and a periodic feature data set;

[0112] Building a prediction model based on a graphic feature dataset and a period feature dataset, wherein the period feature dataset is an input of the prediction model and the image feature dataset is an output of the prediction model;

[0113] Update the Bayesian network association evaluation model, input the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtain the prevention and control strategy of the alpine grassland, and obtain the relationship data of the shrubbery trend of the alpine grassland;

[0114] Based on the alpine grassland shrubbery trend relationship data, a prevention and control management plan for the alpine grassland to be detected is obtained;

[0115] The process of building a predictive model includes:

[0116] Obtain image feature datasets and period feature datasets, and construct training sets;

[0117] Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data;

[0118] The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data;

[0119] The mapping relationship model is adjusted and optimized based on a number of image feature data and period feature data obtained after segmentation to obtain an image-period feature mapping;

[0120] The image-period feature map is trained, and a prediction model is built based on the training results.

[0121] Furthermore, the process of updating the Bayesian network association evaluation model, inputting the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtaining the prevention and control strategy of the alpine grassland, and obtaining the relationship data of the shrubbery trend of the alpine grassland includes:

[0122] The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model;

[0123] Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold;

[0124] The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor;

[0125] Determine the size of the impact factor and impact threshold;

[0126] If so, the influence of the periodic features on the image features is determined based on the image-period feature mapping and the multidimensional benefit, the weight of the influence of the periodic features on the image features under the multidimensional benefit condition is determined, and the periodic data of the alpine grassland is adjusted based on the calculated weight;

[0127] The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is:

[0128] WC(x t)=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ

[0129] Where W is the weight coefficient, C(x t ) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 is the state quantity of the image feature data at time t-1;

[0130] If not, the alpine grassland will be maintained in its original state.

[0131] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A system for monitoring and controlling shrubbery in alpine grasslands, characterized by: include The division and collection module is used to demarcate the alpine grassland area, and collect image data and periodic data within the period of the demarcated area through satellites, drones and manual work. The periodic data includes the time and time interval of periodic data collection and the environmental data of the demarcated area at that time; The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area; A data processing module, used for dividing and processing the image data and the periodic data collected by the division and collection module to obtain an image feature data set and a periodic feature data set; A model management module, which constructs a prediction model based on the image feature data set and the period feature data set, and is also used to update the Bayesian network association evaluation model; The decision-making and prevention module is used to input the real-time information of the alpine grassland into the Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

2. The system for monitoring and controlling shrubbery in alpine grasslands according to claim 1, characterized in that: The data processing module divides and processes the image data and the periodic data collected by the division and collection module to obtain an image feature data set and a periodic feature data set, including: Data segmentation: image data and periodic data are segmented into several time periods based on the collected periodic time; Data processing, deleting duplicate data and abnormal data in the segmented image data and periodic data, and renaming or defining the several image data and periodic data formed after segmentation to obtain several image feature data and periodic feature data.

3. The system for monitoring and controlling shrubbery in alpine grassland according to claim 2, characterized in that: The process of constructing a prediction model by the model management module includes: Obtain image feature datasets and period feature datasets, and construct training sets; Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data; The target image data and the target period data are processed by using a neural network to generate an image-period feature map; The image-period feature map is trained, and a prediction model is built based on the training results.

4. The system for monitoring and controlling shrubbery in alpine grassland according to claim 3, characterized in that: The process of processing the target image data and the target period data by using a neural network to generate an image-period feature map includes: The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data; Based on a number of image feature data and period feature data obtained after segmentation, the mapping relationship model is adjusted and optimized to obtain an image-period feature mapping.

5. The system for monitoring and controlling shrubbery in alpine grassland according to claim 3, characterized in that: The decision-making control module updates the Bayesian network association evaluation model and inputs the real-time information into the Bayesian network association evaluation model to obtain the control strategy of the alpine grassland, which includes: The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model; The real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategy of the alpine grassland.

6. The system for monitoring and controlling shrubbery in alpine grasslands according to claim 5, characterized in that: The real-time information of the alpine grassland is input into the target Bayesian network association evaluation model to obtain the prevention and control strategies of the alpine grassland, including: Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold; The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor; Determine the size of the impact factor and impact threshold; If so, the periodic data of alpine grasslands are adjusted based on the differences between influencing factors and impact thresholds and multidimensional benefits; If not, the alpine grassland will be maintained in its original state.

7. The system for monitoring and controlling shrubbery in alpine grasslands according to claim 6, characterized in that: The process of adjusting the periodic data of the alpine grassland based on the difference between the influencing factors and the influencing thresholds and the multi-dimensional benefits includes: Based on the image-period feature mapping and multi-dimensional benefits, the influence of the periodic features on the image features is determined, the weight of the influence of the periodic features on the image features under the multi-dimensional benefits condition is calculated, and the periodic data of the alpine grassland is adjusted based on the calculated weight; The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is: WC(x t )=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ Where W is the weight coefficient, C(x t ) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 It is the state quantity of the image feature data at time t-1.

8. A method for monitoring and controlling the thicketization of alpine grasslands, using the system for monitoring and controlling the thicketization of alpine grasslands according to any one of claims 1 to 7, characterized in that: include: Delineate the alpine grassland area, collect image data and periodic data within the period of the delineated area through satellite, UAV and manual, the periodic data includes the time of periodic data collection, time interval and environmental data of the delineated area at that time; The environmental data of the area demarcated at the time includes all parameters related to plant growth in the area; The collected image data and periodic data are divided and processed based on a preset time interval to obtain an image feature data set and a periodic feature data set; Building a prediction model based on a graphic feature dataset and a period feature dataset, wherein the period feature dataset is an input of the prediction model and the image feature dataset is an output of the prediction model; Update the Bayesian network association evaluation model, input the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtain the prevention and control strategy of the alpine grassland, and obtain the relationship data of the shrubbery trend of the alpine grassland; Based on the alpine grassland shrubbery trend relationship data, a prevention and control management plan for the alpine grassland to be detected is obtained; The process of building a predictive model includes: Obtain image feature datasets and period feature datasets, and construct training sets; Perform feature extraction on the image feature data set and the period feature data set respectively to obtain target image data and target period data; The target cycle data is classified, and the target image data and the classified target cycle data are used as training data, and a neural network model algorithm is used for training to obtain a mapping relationship model with the target image data and the target cycle data; The mapping relationship model is adjusted and optimized based on a number of image feature data and period feature data obtained after segmentation to obtain an image-period feature mapping; The image-period feature map is trained, and a prediction model is built based on the training results.

9. A method for monitoring and controlling shrubbery in alpine grasslands according to claim 8, characterized in that: The process of updating the Bayesian network association evaluation model, inputting the real-time information of the alpine grassland into the Bayesian network association evaluation model, obtaining the prevention and control strategy of the alpine grassland, and obtaining the relationship data of the shrubbery trend of the alpine grassland includes: The state quantity, control quantity, environmental variable and weight variable of the Bayesian network association evaluation model are updated by using image-period feature mapping and multi-dimensional benefits to obtain the target Bayesian network association evaluation model; Determine preset data corresponding to the instant information of the alpine grassland based on the image-period feature mapping, obtain preset impact data based on the impact of the preset data on the multi-dimensional benefits, and use the preset impact data as the impact threshold; The subsequent variable data of the alpine grassland instant information is calculated through the target Bayesian network association evaluation model, and the impact on multi-dimensional benefits is determined through the subsequent variable data to obtain the subsequent impact data, and the subsequent impact data is used as the impact factor; Determine the size of the impact factor and impact threshold; If so, the influence of the periodic features on the image features is determined based on the image-period feature mapping and the multidimensional benefit, the weight of the influence of the periodic features on the image features under the multidimensional benefit condition is determined, and the periodic data of the alpine grassland is adjusted based on the calculated weight; The weight formula of the period feature and the image feature under the multi-dimensional benefit condition is: WC(x t )=λp(x t |z t )∫p(x t |x t-1 ,y t )C(x t )dx t-1 +δ Where W is the weight coefficient, C(x t ) is the parameter after the normalization of the image feature data, λ is the influencing factor of the multi-benefit, p(x t |z t ) is the image-period feature prediction model at time t, p(x t |x t-1 ,y t ) is the parameter offset after normalization of image feature data, x t is the state quantity of the image feature data at time t, y t is the environmental variable of the periodic characteristic data at time t, Z t is the control amount of the image feature data at time t, x t-1 is the state quantity of the image feature data at time t-1; If not, the alpine grassland will be maintained in its original state.

Images