Forecasting method of total output value of agriculture, forestry, animal husbandry and fishery based on night light index

Abstract

The present invention relates to a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index, and belongs to the technical field of output value prediction. The present invention constructs a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, inputs a training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training, optimizes the model parameters of the total output value prediction model for agriculture, forestry, animal husbandry and fishery by a particle swarm algorithm, outputs the total output value prediction model for agriculture, forestry, animal husbandry and fishery, and finally predicts the total output value of agriculture, forestry, animal husbandry and fishery in a target area by the total output value prediction model for agriculture, forestry, animal husbandry and fishery, obtains a prediction result, and generates relevant early warning information according to the prediction result. The present invention optimizes the training set by fusing a Dice coefficient measurement algorithm and a graph neural network, so that the data of the training set can be more in line with the training requirements, thereby reducing the training time of the model and improving the prediction accuracy of the model, and can timely regulate the production plan according to actual production needs.

Description

Technical Field

[0001] The present invention relates to the technical field of output value prediction, and in particular to a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on nighttime light index. Background Art

[0002] The night light index is significantly different from the daytime remote sensing index. It has unique advantages in reflecting human society, human economic activities and other human survival activities, so the night light index has gradually been used by many scholars in many fields. The night light index has a certain relevance to agriculture, forestry, animal husbandry and fishery. For example, by providing a certain amount of light at night for agricultural plants to photosynthesize, it promotes the growth of crops, thereby affecting the total output value of agriculture; in fisheries, long-term fishing at night can affect the total output value of fisheries; however, the existing technology cannot estimate the total output value of agriculture, forestry, animal husbandry and fishery, so it is impossible to adjust the production plan in time according to actual production needs, which affects the supply of agricultural, forestry, animal husbandry and fishery products. Summary of the invention

[0003] The invention overcomes the shortcomings of the prior art and provides a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index.

[0004] To achieve the above object, the technical solution adopted by the present invention is:

[0005] The first aspect of the present invention provides a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index, which specifically comprises the following steps:

[0006] Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in each region, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis;

[0007] A training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery. A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training.

[0008] The particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output;

[0009] The total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated based on the prediction results.

[0010] Furthermore, in this method, the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery are obtained, and the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery are analyzed by grey correlation analysis, specifically including:

[0011] Obtain the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, introduce the grey correlation analysis method, and input the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery into the grey correlation analysis method for calculation;

[0012] Calculate the correlation degree information between each night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, and preset the correlation degree threshold information, and determine whether the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information;

[0013] When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery;

[0014] When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is not greater than the correlation degree threshold information, the corresponding night light index data will be regarded as night light index data that is unrelated to the total output value of agriculture, forestry, animal husbandry and fishery.

[0015] Furthermore, in this method, a training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery, specifically including:

[0016] Introduce the Dice coefficient measurement algorithm and graph neural network, obtain relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery, and input the relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery into the graph neural network;

[0017] The relevant night light index data and the corresponding historical data of the total output value of agriculture, forestry, animal husbandry and fishery are used as graph nodes of the graph neural network to construct a topological structure graph. By classifying the topological structure graph, the classified topological structure graph is obtained.

[0018] According to the classified topological structure diagram, several training subsets are constructed and initialized, and the Dice coefficient between the training subsets is calculated by the Dice coefficient measurement algorithm, and it is determined whether the Dice coefficient is greater than the preset Dice coefficient threshold;

[0019] When the Dice coefficient is greater than the preset Dice coefficient threshold, the training items in the training subset are readjusted until the Dice coefficient is no greater than the preset Dice coefficient threshold, the training subset is output, and the training set is constructed.

[0020] Furthermore, in this method, a total output value prediction model of agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network, and the training set is input into the total output value prediction model of agriculture, forestry, animal husbandry and fishery for training, which specifically includes:

[0021] A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery, and the topological structure graph in the training set is processed through a cyclic spatial attention mechanism to obtain a feature graph.

[0022] By performing a SoftMax operation on each graph node in the feature graph, a normalized feature graph with spatial attention is generated, and an inner product operation is performed on the normalized feature graph with spatial attention and the topological structure graph in the training set;

[0023] The attention is focused on the graph nodes, an attention feature map is generated, the attention feature map is input into the hidden layer, and it works together with the hidden state in the hidden layer to update the hidden layer state.

[0024] Furthermore, in this method, a particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by the particle swarm algorithm, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output, which specifically includes:

[0025] The particle swarm algorithm is introduced, and the number of iterations is set according to the particle swarm algorithm, the accuracy threshold of the model is set, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are initialized, the accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is evaluated by the average relative error, and the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is obtained;

[0026] Determine whether the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model; when the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, output the total output value prediction model of agriculture, forestry, animal husbandry and fishery;

[0027] When the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is not greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are adjusted according to the number of iterations;

[0028] Until the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are saved, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output.

[0029] Furthermore, in this method, the total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted by the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated according to the prediction results, specifically including:

[0030] Obtaining night light index data within a preset time in the target area, and inputting the night light index data within the preset time in the target area into a total output value prediction model for agriculture, forestry, animal husbandry and fishery for prediction;

[0031] Obtain the total output value of agriculture, forestry, animal husbandry and fishery in the target area within a preset time by prediction, obtain the production demand information in the target area, and determine whether the total output value of agriculture, forestry, animal husbandry and fishery of each type is within the production demand information in the preset target area;

[0032] When the total output value of agriculture, forestry, animal husbandry and fishery is not within the production demand information of the target area, an abnormal warning will be issued for the total output value of agriculture, forestry, animal husbandry and fishery of the corresponding type, and relevant warning information will be generated;

[0033] Based on the relevant early warning information, environmental early warning is carried out for the target area, and the total output value data of agriculture, forestry, animal husbandry and fishery is compiled, and industrial planning and regulation of agriculture, forestry, animal husbandry and fishery are carried out.

[0034] A second aspect of the present invention provides a system for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index. The system includes a memory and a processor. The memory includes a method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index. When the method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index is executed by the processor, the system includes the following steps:

[0035] Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in each region, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis;

[0036] A training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery. A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training.

[0037] The particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output;

[0038] The total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated based on the prediction results.

[0039] The third aspect of the present invention provides a computer-readable storage medium, which includes a program for a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index. When the program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index is executed by a processor, any step of the method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index is implemented.

[0040] The present invention solves the defects existing in the background technology and has the following beneficial effects:

[0041] The present invention obtains the historical night light index data and the total output value historical data of agriculture, forestry, animal husbandry and fishery in various regions, and analyzes the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis, and then constructs a training set according to the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and the related total output value historical data of agriculture, forestry, animal husbandry and fishery, and constructs a total output value prediction model of agriculture, forestry, animal husbandry and fishery based on a deep neural network, and inputs the training set into the total output value prediction model of agriculture, forestry, animal husbandry and fishery for training, thereby introducing a particle swarm algorithm, optimizing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery through the particle swarm algorithm, outputting the total output value prediction model of agriculture, forestry, animal husbandry and fishery, and finally predicting the total output value of agriculture, forestry, animal husbandry and fishery in the target area through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, obtaining the prediction result, and generating relevant early warning information according to the prediction result. The present invention optimizes the training set by fusing the Dice coefficient measurement algorithm and the graph neural network, so that the data of the training set can be more in line with the training requirements, thereby reducing the training time of the model and improving the prediction accuracy of the model, and can timely regulate the production plan according to the actual production needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, drawings of other embodiments can be obtained based on these drawings without paying creative work.

[0043] Figure 1 The overall method flow chart of the method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index is shown;

[0044] Figure 2 A first method flow chart of a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on nighttime light index is shown;

[0045] Figure 3 A second method flow chart of the method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index is shown;

[0046] Figure 4 The system block diagram of the total output value prediction system of agriculture, forestry, animal husbandry and fishery based on night light index is shown. DETAILED DESCRIPTION

[0047] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0049] like Figure 1 As shown, the first aspect of the present invention provides a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index, which specifically includes the following steps:

[0050] S102: Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis;

[0051] S104: constructing a training set based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery, constructing a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, and inputting the training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training;

[0052] S106: Introducing a particle swarm algorithm, optimizing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery through the particle swarm algorithm, and outputting the total output value prediction model of agriculture, forestry, animal husbandry and fishery;

[0053] S108: Predict the total output value of agriculture, forestry, animal husbandry and fishery in the target area through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, obtain the prediction results, and generate relevant early warning information based on the prediction results.

[0054] It should be noted that the present invention optimizes the training set by integrating the Dice coefficient measurement algorithm and the graph neural network, which can make the data of the training set more in line with the training requirements, thereby reducing the training time of the model and improving the prediction accuracy of the model, and can timely adjust the production plan according to actual production needs.

[0055] like Figure 2 As shown, further, in step S102 of the method, it specifically includes:

[0056] S202: Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, introduce grey correlation analysis method, and input the historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions into the grey correlation analysis method for calculation;

[0057] S204: Calculate the correlation degree information between each night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, and preset the correlation degree threshold information, and determine whether the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information;

[0058] S206: When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery;

[0059] S208: When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is not greater than the correlation degree threshold information, the corresponding night light index data is used as night light index data that is irrelevant to the total output value of agriculture, forestry, animal husbandry and fishery.

[0060] It should be noted that for the factors between two systems, the measure of the size of the correlation that changes over time or with different objects is called the correlation degree. In the process of system development, if the trends of the changes of the two factors are consistent, that is, the degree of synchronous change is high, it can be said that the correlation between the two factors is high; otherwise, it is low. Therefore, the gray correlation analysis method is a method to measure the degree of correlation between factors based on the similarity or difference in the development trends between the factors, that is, the "gray correlation degree". When the correlation degree information is greater than the preset correlation degree threshold information, it means that there is a certain correlation between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery. Among them, the night light index is a variety of night light indices (such as total night light index, average night light index, night light change index) constructed based on the night light intensity data extracted from the night light remote sensing image, and using statistics such as the sum, mean, standard deviation, and sum of squares of deviations.

[0061] like Figure 3 As shown, further, in step S104 of the method, a training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data of the total output value of agriculture, forestry, animal husbandry and fishery, specifically including:

[0062] S302: Introduce the Dice coefficient measurement algorithm and the graph neural network, obtain relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery, and input the relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery into the graph neural network;

[0063] S304: Use the relevant nighttime light index data and the corresponding historical data of the total output value of agriculture, forestry, animal husbandry and fishery as graph nodes of the graph neural network, construct a topological structure graph, and obtain a classified topological structure graph by classifying the topological structure graph;

[0064] S306: construct and initialize several training subsets according to the classified topological structure graph, calculate the Dice coefficient between the training subsets by using the Dice coefficient measurement algorithm, and determine whether the Dice coefficient is greater than a preset Dice coefficient threshold;

[0065] S308: When the Dice coefficient is greater than the preset Dice coefficient threshold, the training items in the training subset are readjusted until the Dice coefficient is not greater than the preset Dice coefficient threshold, the training subset is output, and a training set is constructed.

[0066] It should be noted that the Dice coefficient is used to measure the similarity between two sets. In fact, when selecting a training set, if there are other types of sample data in the training subset, classification can be performed through a decision tree model or a clustering algorithm. Since classification may be abnormal and the data does not belong to the same type, this will reduce the recognition accuracy and training time of the model. By introducing the Dice coefficient measurement algorithm and graph neural network, it is possible to calculate and process the training sets, determine whether there are identical and overlapping sample data between the training subsets, and make the sample data in each training subset meet the requirements, thereby improving the training speed of the model and the prediction accuracy of the model.

[0067] Furthermore, in step S104 of the method, a total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network, and the training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training, which specifically includes:

[0068] A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery, and the topological structure graph in the training set is processed through a cyclic spatial attention mechanism to obtain a feature graph.

[0069] By performing a SoftMax operation on each graph node in the feature graph, a normalized feature graph with spatial attention is generated, and an inner product operation is performed on the normalized feature graph with spatial attention and the topological structure graph in the training set;

[0070] The attention is focused on the graph nodes, an attention feature map is generated, the attention feature map is input into the hidden layer, and it works together with the hidden state in the hidden layer to update the hidden layer state.

[0071] It should be noted that deep neural networks include convolutional neural networks, long short-term memory recurrent neural networks, recursive neural networks, graph neural networks, etc. By integrating the recurrent spatial attention mechanism, they can suppress the multi-scale interference information brought by the training set, thereby improving the prediction accuracy of the model.

[0072] Furthermore, in step S106 of the method, a particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by the particle swarm algorithm, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output, which specifically includes:

[0073] The particle swarm algorithm is introduced, and the number of iterations is set according to the particle swarm algorithm, the accuracy threshold of the model is set, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are initialized, the accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is evaluated by the average relative error, and the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is obtained;

[0074] Determine whether the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model; when the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, output the total output value prediction model of agriculture, forestry, animal husbandry and fishery;

[0075] When the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is not greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are adjusted according to the number of iterations;

[0076] Until the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are saved, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output.

[0077] It should be noted that the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery include data such as the number of neurons and the number of convolutional layers. Through this method, a better total output value prediction model of agriculture, forestry, animal husbandry and fishery can be obtained.

[0078] Furthermore, in step S108 of the method, the total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted by the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction result is obtained, and relevant early warning information is generated according to the prediction result, specifically including:

[0079] Obtaining night light index data within a preset time in the target area, and inputting the night light index data within the preset time in the target area into a total output value prediction model for agriculture, forestry, animal husbandry and fishery for prediction;

[0080] Obtain the total output value of agriculture, forestry, animal husbandry and fishery in the target area within a preset time by prediction, obtain the production demand information in the target area, and determine whether the total output value of agriculture, forestry, animal husbandry and fishery of each type is within the production demand information in the preset target area;

[0081] When the total output value of agriculture, forestry, animal husbandry and fishery is not within the production demand information of the target area, an abnormal warning will be issued for the total output value of agriculture, forestry, animal husbandry and fishery of the corresponding type, and relevant warning information will be generated;

[0082] Based on the relevant early warning information, environmental early warning is carried out for the target area, and the total output value data of agriculture, forestry, animal husbandry and fishery is compiled, and industrial planning and regulation of agriculture, forestry, animal husbandry and fishery are carried out.

[0083] It should be noted that the production demand information in the target area mainly includes the production demand threshold of each type, such as the production demand threshold of agriculture, the production demand threshold of forestry and other data.

[0084] In addition, the historical night light index data of each region is obtained, which specifically includes the following steps:

[0085] Obtaining light feature data information under each environmental factor data, and inputting the environmental factor data affecting the historical night light index data into the graph neural network, using the environmental factor data as the first node and the light feature data information as the second node;

[0086] Constructing a directed edge description relationship, constructing a second topological graph by combining the first node and the second graph node according to the directed edge description relationship, acquiring a related adjacency matrix according to the second topological structure graph, and constructing a light feature prediction model based on a deep neural network;

[0087] Acquire environmental factor data for collecting the historical nighttime light index data, and input the environmental factor data for collecting the historical nighttime light index data into the light feature prediction model for prediction;

[0088] Through prediction, the light characteristic data information corresponding to the environmental factor data for collecting the historical night light index data is obtained, and the historical night light index data of each region is corrected according to the light characteristic data information corresponding to the environmental factor data for collecting the historical night light index data.

[0089] It should be noted that environmental factor data include temperature, humidity, visibility and other data. Different environmental factor data will affect the lighting feature data information (such as light intensity). Data anomalies will reduce the prediction accuracy of the model. However, this method can correct the historical night light index data of various regions according to the lighting feature data information corresponding to the environmental factor data of the historical night light index data collected, thereby improving the accuracy of the training data, avoiding data anomalies, and ensuring the prediction accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery.

[0090] In addition, obtaining the night light index data in the target area within a preset time may specifically include the following steps: obtaining environmental factor data information under each meteorological feature data through big data, and constructing a knowledge graph, and inputting the environmental factor data information under each meteorological feature data into the knowledge graph for storage; obtaining meteorological feature data information within a preset time, and inputting the meteorological feature data information within the preset time into the knowledge graph for prediction, and obtaining the predicted environmental factor data corresponding to the meteorological feature data information within the preset time; obtaining clarity data of the night light index data obtained in each time period according to the predicted environmental factor data corresponding to the meteorological feature data information within the preset time, and sorting the clarity data of the night light index data obtained in each time period to obtain the sorting result; obtaining the time period with the maximum clarity data according to the sorting result, and using the time period with the maximum clarity data as the optimal collection time period for the night light index data, and collecting the night light index data according to the optimal collection time period for the night light index data, and obtaining the night light index data in the target area within the preset time.

[0091] It should be noted that this method can select the optimal collection period to collect night light index data, thereby ensuring the accuracy of data collection, ensuring the authenticity and effectiveness of data collection, and thus improving the prediction accuracy of the total output value of agriculture, forestry, animal husbandry and fishery in the target area within the preset time.

[0092] like Figure 4 As shown, the second aspect of the present invention provides a system 4 for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index. The system 4 includes a memory 41 and a processor 42. The memory 41 includes a method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index. When the method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index is executed by the processor, the following steps are included:

[0093] Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in each region, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis;

[0094] A training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery. A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training.

[0095] The particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output;

[0096] The total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated based on the prediction results.

[0097] Furthermore, in this system, the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery are obtained, and the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery are analyzed through grey correlation analysis, specifically including:

[0098] Obtain the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, introduce the grey correlation analysis method, and input the historical night light index data of each region and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery into the grey correlation analysis method for calculation;

[0099] Calculate the correlation degree information between each night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, and preset the correlation degree threshold information, and determine whether the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information;

[0100] When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery;

[0101] When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is not greater than the correlation degree threshold information, the corresponding night light index data will be regarded as night light index data that is unrelated to the total output value of agriculture, forestry, animal husbandry and fishery.

[0102] Furthermore, in this system, a training set is constructed based on the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and the related historical data of the total output value of agriculture, forestry, animal husbandry and fishery, specifically including:

[0103] Introduce the Dice coefficient measurement algorithm and graph neural network, obtain relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery, and input the relevant night light index data and corresponding historical data on the total output value of agriculture, forestry, animal husbandry and fishery into the graph neural network;

[0104] The relevant night light index data and the corresponding historical data of the total output value of agriculture, forestry, animal husbandry and fishery are used as graph nodes of the graph neural network to construct a topological structure graph. By classifying the topological structure graph, the classified topological structure graph is obtained.

[0105] According to the classified topological structure diagram, several training subsets are constructed and initialized, and the Dice coefficient between the training subsets is calculated by the Dice coefficient measurement algorithm, and it is determined whether the Dice coefficient is greater than the preset Dice coefficient threshold;

[0106] When the Dice coefficient is greater than the preset Dice coefficient threshold, the training items in the training subset are readjusted until the Dice coefficient is no greater than the preset Dice coefficient threshold, the training subset is output, and the training set is constructed.

[0107] Furthermore, in this system, a total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network, and the training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training, specifically including:

[0108] A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery, and the topological structure graph in the training set is processed through a cyclic spatial attention mechanism to obtain a feature graph.

[0109] By performing a SoftMax operation on each graph node in the feature graph, a normalized feature graph with spatial attention is generated, and an inner product operation is performed on the normalized feature graph with spatial attention and the topological structure graph in the training set;

[0110] The attention is focused on the graph nodes, an attention feature map is generated, the attention feature map is input into the hidden layer, and it works together with the hidden state in the hidden layer to update the hidden layer state.

[0111] Furthermore, in this system, a particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery through the particle swarm algorithm, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output, which specifically includes:

[0112] The particle swarm algorithm is introduced, and the number of iterations is set according to the particle swarm algorithm, the accuracy threshold of the model is set, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are initialized, the accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is evaluated by the average relative error, and the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is obtained;

[0113] Determine whether the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model; when the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, output the total output value prediction model of agriculture, forestry, animal husbandry and fishery;

[0114] When the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is not greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are adjusted according to the number of iterations;

[0115] Until the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are saved, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output.

[0116] Furthermore, in this system, the total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated according to the prediction results, including:

[0117] Obtaining night light index data within a preset time in the target area, and inputting the night light index data within the preset time in the target area into a total output value prediction model for agriculture, forestry, animal husbandry and fishery for prediction;

[0118] Obtain the total output value of agriculture, forestry, animal husbandry and fishery in the target area within a preset time by prediction, obtain the production demand information in the target area, and determine whether the total output value of agriculture, forestry, animal husbandry and fishery of each type is within the production demand information in the preset target area;

[0119] When the total output value of agriculture, forestry, animal husbandry and fishery is not within the production demand information of the target area, an abnormal warning will be issued for the total output value of agriculture, forestry, animal husbandry and fishery of the corresponding type, and relevant warning information will be generated;

[0120] Based on the relevant early warning information, environmental early warning is carried out for the target area, and the total output value data of agriculture, forestry, animal husbandry and fishery is compiled, and industrial planning and regulation of agriculture, forestry, animal husbandry and fishery are carried out.

[0121] A second aspect of the present invention provides a system for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index. The system includes a memory and a processor. The memory includes a method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index. When the method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index is executed by the processor, the system includes the following steps:

[0122] Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in each region, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis;

[0123] A training set is constructed based on night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery. A total output value prediction model for agriculture, forestry, animal husbandry and fishery is constructed based on a deep neural network. The training set is input into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training.

[0124] The particle swarm algorithm is introduced to optimize the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery, and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output;

[0125] The total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction results are obtained, and relevant early warning information is generated based on the prediction results.

[0126] The third aspect of the present invention provides a computer-readable storage medium, which includes a program for a method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index. When the program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index is executed by a processor, any step of the method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a night light index is implemented.

[0127] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0128] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0129] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0130] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0131] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0132] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index, characterized in that: The specific steps include: Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis; Constructing a training set based on nighttime light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery, constructing a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, and inputting the training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training; Introducing a particle swarm algorithm, optimizing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by the particle swarm algorithm, and outputting the total output value prediction model of agriculture, forestry, animal husbandry and fishery; Predicting the total output value of agriculture, forestry, animal husbandry and fishery in the target area through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, obtaining the prediction result, and generating relevant early warning information according to the prediction result; The historical night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery in each region are obtained, and the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery are analyzed through grey correlation analysis, specifically including: Obtain historical nighttime light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, introduce a grey correlation analysis method, and input the historical nighttime light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions into the grey correlation analysis method for calculation; Calculate the correlation degree information between each night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, and preset the correlation degree threshold information, and determine whether the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information; When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery; When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is not greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data that is not related to the total output value of agriculture, forestry, animal husbandry and fishery; The training set is constructed based on the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and the related historical data of the total output value of agriculture, forestry, animal husbandry and fishery, specifically including: Introduce the Dice coefficient measurement algorithm and the graph neural network, obtain relevant night light index data and corresponding historical data of total output value of agriculture, forestry, animal husbandry and fishery, and input the relevant night light index data and corresponding historical data of total output value of agriculture, forestry, animal husbandry and fishery into the graph neural network; The relevant nighttime light index data and the corresponding historical data of the total output value of agriculture, forestry, animal husbandry and fishery are used as graph nodes of the graph neural network to construct a topological structure graph, and the classified topological structure graph is obtained by classifying the topological structure graph; Constructing and initializing a plurality of training subsets according to the classified topological structure graph, calculating the Dice coefficients between the training subsets by the Dice coefficient measurement algorithm, and determining whether the Dice coefficients are greater than a preset Dice coefficient threshold; When the Dice coefficient is greater than a preset Dice coefficient threshold, the training items in the training subset are readjusted until the Dice coefficient is not greater than the preset Dice coefficient threshold, the training subset is output, and a training set is constructed.

2. The method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index according to claim 1 is characterized in that: Constructing a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, and inputting the training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training, specifically including: Constructing a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, inputting the training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery, and processing the topological structure graph in the training set through a cyclic spatial attention mechanism to obtain a feature graph; Generate a normalized feature map with spatial attention by performing a SoftMax operation on each graph node in the feature map, and perform an inner product operation on the normalized feature map with spatial attention and a topological structure map in a training set; The attention is focused on the graph nodes, an attention feature map is generated, the attention feature map is input into the hidden layer, and the hidden layer state is updated by working in conjunction with the hidden state in the hidden layer.

3. The method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index according to claim 1 is characterized in that: Introducing a particle swarm algorithm, optimizing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by the particle swarm algorithm, and outputting the total output value prediction model of agriculture, forestry, animal husbandry and fishery, specifically including: Introducing a particle swarm algorithm, setting the number of iterations according to the particle swarm algorithm, setting the accuracy threshold of the model, initializing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery, evaluating the accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by average relative error, and obtaining the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery; Determine whether the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, and when the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, output the total output value prediction model of agriculture, forestry, animal husbandry and fishery; When the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is not greater than the accuracy threshold of the model, adjusting the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery according to the number of iterations; Until the model accuracy of the total output value prediction model of agriculture, forestry, animal husbandry and fishery is greater than the accuracy threshold of the model, the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery are saved and the total output value prediction model of agriculture, forestry, animal husbandry and fishery is output.

4. The method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on night light index according to claim 1 is characterized in that: The total output value of agriculture, forestry, animal husbandry and fishery in the target area is predicted by the total output value prediction model of agriculture, forestry, animal husbandry and fishery, the prediction result is obtained, and relevant early warning information is generated according to the prediction result, specifically including: Acquire the night light index data in the target area within a preset time, and input the night light index data in the target area within the preset time into the total output value prediction model of agriculture, forestry, animal husbandry and fishery for prediction; Obtain the total output value of agriculture, forestry, animal husbandry and fishery in the target area within a preset time by prediction, obtain the production demand information in the target area, and determine whether the total output value of agriculture, forestry, animal husbandry and fishery of each type is within the production demand information in the preset target area; When the total output value of agriculture, forestry, animal husbandry and fishery is not within the production demand information of the target area, an abnormal warning is issued for the total output value of agriculture, forestry, animal husbandry and fishery of the corresponding type, and relevant warning information is generated; Based on the relevant early warning information, environmental early warning is carried out for the target area, and the total output value data of agriculture, forestry, animal husbandry and fishery is counted, and industrial planning and regulation of agriculture, forestry, animal husbandry and fishery is carried out.

5. The total output value prediction system of agriculture, forestry, animal husbandry and fishery based on night light index is characterized by: The system includes a memory and a processor, wherein the memory includes a method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index, and when the method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on a nighttime light index is executed by the processor, the method program includes the following steps: Obtain historical night light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, and analyze the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery through grey correlation analysis; Constructing a training set based on nighttime light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and related historical data on the total output value of agriculture, forestry, animal husbandry and fishery, constructing a total output value prediction model for agriculture, forestry, animal husbandry and fishery based on a deep neural network, and inputting the training set into the total output value prediction model for agriculture, forestry, animal husbandry and fishery for training; Introducing a particle swarm algorithm, optimizing the model parameters of the total output value prediction model of agriculture, forestry, animal husbandry and fishery by the particle swarm algorithm, and outputting the total output value prediction model of agriculture, forestry, animal husbandry and fishery; Predicting the total output value of agriculture, forestry, animal husbandry and fishery in the target area through the total output value prediction model of agriculture, forestry, animal husbandry and fishery, obtaining the prediction result, and generating relevant early warning information according to the prediction result; The historical night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery in each region are obtained, and the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery are analyzed through grey correlation analysis, specifically including: Obtain historical nighttime light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions, introduce a grey correlation analysis method, and input the historical nighttime light index data and historical data on the total output value of agriculture, forestry, animal husbandry and fishery in various regions into the grey correlation analysis method for calculation; Calculate the correlation degree information between each night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery, and preset the correlation degree threshold information, and determine whether the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information; When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery; When the correlation degree information between the night light index data and the historical data of the total output value of agriculture, forestry, animal husbandry and fishery is not greater than the correlation degree threshold information, the corresponding night light index data is used as the night light index data that is not related to the total output value of agriculture, forestry, animal husbandry and fishery; The training set is constructed based on the night light index data related to the total output value of agriculture, forestry, animal husbandry and fishery and the related historical data of the total output value of agriculture, forestry, animal husbandry and fishery, specifically including: Introduce the Dice coefficient measurement algorithm and the graph neural network, obtain relevant night light index data and corresponding historical data of total output value of agriculture, forestry, animal husbandry and fishery, and input the relevant night light index data and corresponding historical data of total output value of agriculture, forestry, animal husbandry and fishery into the graph neural network; The relevant nighttime light index data and the corresponding historical data of the total output value of agriculture, forestry, animal husbandry and fishery are used as graph nodes of the graph neural network to construct a topological structure graph, and the classified topological structure graph is obtained by classifying the topological structure graph; Constructing and initializing a plurality of training subsets according to the classified topological structure graph, calculating the Dice coefficients between the training subsets by the Dice coefficient measurement algorithm, and determining whether the Dice coefficients are greater than a preset Dice coefficient threshold; When the Dice coefficient is greater than a preset Dice coefficient threshold, the training items in the training subset are readjusted until the Dice coefficient is not greater than the preset Dice coefficient threshold, the training subset is output, and a training set is constructed.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index. When the method program for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index is executed by a processor, the steps of the method for predicting the total output value of agriculture, forestry, animal husbandry and fishery based on the night light index as described in any one of claims 1 to 4 are implemented.

Images