A digital data intelligent analysis system and method for green agricultural development

Abstract

This application provides a digital data intelligent analysis system and method for green agricultural development, relating to the field of digital agriculture technology. It identifies explicit correlations between multi-source agricultural data based on the direct changing relationships between agricultural input factors and ecological environment indicators, generating explicit correlation labels characterizing the direct ecological impact of agricultural inputs. By utilizing the distribution characteristics of these explicit correlation labels among different ecological elements, it constructs agricultural factor impact propagation paths. Based on these propagation paths, it identifies indirect correlations between multi-source agricultural data, generating implicit correlation labels characterizing the indirect ecological impact of agricultural inputs. Based on the explicit and implicit correlation labels, it constructs an agricultural ecological impact correlation diagram. Based on this diagram, it performs intelligent analysis of the relationship between agricultural resource utilization status and ecological environment changes, obtaining the analysis results. This solution enables a systematic characterization of the direct and indirect impacts of agricultural inputs on the ecological environment, thereby improving the accuracy of evaluating the state of green agricultural development.

Description

Technical Field

[0001] This application relates to the field of digital agriculture technology, and more specifically, to a digital data intelligent analysis system and method for green agricultural development. Background Technology

[0002] In the digital process of green agricultural development, relying on the Internet of Things, big data and artificial intelligence technologies to collect and analyze multi-source data on input behavior, ecological environment response and crop growth status in agricultural production has become an important means to promote precision agricultural management and collaborative protection of the ecological environment. By conducting correlation analysis between input factors such as fertilization and irrigation and environmental indicators such as soil and water, existing technologies have initially realized the quantitative assessment of agricultural resource utilization efficiency and the dynamic monitoring of ecological environment changes, providing data support for the green transformation of agriculture.

[0003] However, most existing agricultural ecological environment impact analysis techniques are limited to direct correlations between single input factors and single environmental indicators. Their core logic is based on observable explicit changes, making it difficult to systematically reveal the indirect ecological effects of agricultural inputs transmitted through multiple environmental media. For example, fertilizer application not only directly increases soil nutrient content but may also affect the quality of surrounding water bodies through runoff or influence atmospheric nitrogen deposition through volatilization and sedimentation processes. These indirect impact pathways transmitted through intermediate ecological elements are often fragmented or ignored in traditional analysis methods, leading to an underestimation or even misjudgment of the true scope and transmission mechanism of agricultural inputs' impact on the ecological environment. Therefore, how to achieve a systematic characterization of the direct and indirect impacts of agricultural inputs on the ecological environment, thereby improving the accuracy of agricultural green development status assessment capabilities, has become a challenge for the industry. Summary of the Invention

[0004] This application provides a digital data intelligent analysis system and method for green agricultural development, which can systematically characterize the direct and indirect impacts of agricultural inputs on the ecological environment, thereby improving the ability to accurately evaluate the status of green agricultural development.

[0005] Firstly, this application provides a digital data intelligent analysis method for green agricultural development, including:

[0006] Collect agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data within agricultural production areas, and construct a multi-source agricultural dataset;

[0007] Based on the direct changing relationship between agricultural input factors and ecological environment indicators, explicit correlation identification is performed on multi-source agricultural data to generate explicit correlation labels that characterize the direct ecological impact of agricultural inputs.

[0008] By constructing the agricultural factor influence propagation path through the distribution characteristics of the explicit association tags among different ecological elements, and then identifying the indirect association of multi-source agricultural data based on the agricultural factor influence propagation path, implicit association tags characterizing the indirect ecological impact relationship of agricultural inputs are generated.

[0009] An agricultural ecological impact correlation diagram is constructed based on the explicit and implicit correlation labels. Then, based on the agricultural ecological impact correlation diagram, a multi-dimensional index fusion analysis is performed on the relationship between agricultural resource utilization status and ecological environment change to generate agricultural green development status analysis results.

[0010] In this embodiment, based on the direct changing relationship between agricultural input factors and ecological environment indicators, explicit correlation identification is performed on multi-source agricultural data to generate explicit correlation labels characterizing the direct ecological impact of agricultural inputs. Specifically, this includes:

[0011] Spatiotemporal alignment processing is performed on multi-source agricultural data to establish a matching relationship between agricultural input behavior data and ecological environment monitoring data under the same spatiotemporal coordinate system;

[0012] Determine the rate of change in the application of agricultural input factors within a preset time window;

[0013] Determine the fluctuation range of ecological and environmental indicators within a preset time window;

[0014] The correlation statistical characteristic values ​​between agricultural input factors and ecological environment indicators within the preset time window are determined based on the application rate change rate and the volatility.

[0015] Based on the correlation statistical characteristic values, input-ecological element pairs that exceed the first correlation threshold are selected.

[0016] For the selected input-ecological element pairs, extract the time-series matching features of their changing trends, including consistency of change direction and change lag window;

[0017] The explicit association label is generated based on the time-series matching features. The explicit association label includes input element identifier, ecological environment indicator identifier, and association type identifier.

[0018] In this embodiment, constructing the agricultural factor influence propagation path based on the distribution characteristics of the explicit association tags among different ecological elements specifically includes:

[0019] Obtain all explicit association tags and construct an element-level association graph with ecological elements as nodes and explicit association relationships as edges;

[0020] Identify the shared ecological element nodes in the element-level association graph, where each shared ecological element node simultaneously serves as the target or starting point of at least two explicit association relationships.

[0021] Based on the shared ecological element nodes, multiple explicit relationships are linked and combined in the order of their inherent dependencies to obtain a candidate propagation path from the initial ecological element that has an explicit relationship with agricultural input elements to the final ecological element through intermediate ecological elements.

[0022] The candidate propagation paths are validated for path directionality, and paths that do not conform to the time sequence of ecological processes are eliminated to generate the propagation paths of agricultural factor influence.

[0023] In this embodiment, the indirect correlation identification of multi-source agricultural data based on the propagation path of agricultural factor influences, and the generation of implicit correlation labels characterizing the indirect ecological impact relationship of agricultural inputs, specifically includes:

[0024] Extract the initial agricultural input element nodes and the final ecological element nodes corresponding to the propagation path of the agricultural factors to form the input-final element pairs to be analyzed;

[0025] Along the propagation path of the agricultural factors, the monitoring data changes of intermediate ecological factor nodes are traced segment by segment to verify whether the data fluctuations of the initial agricultural input factors can be transmitted to the end ecological factors along the path.

[0026] If the time series of data changes of the end ecological elements and the time series of data changes of the starting agricultural input elements have a statistical correlation under path constraints, and the statistical correlation cannot be directly explained by the explicit association between the two, then it is identified that there is an implicit association between the starting agricultural input elements and the end ecological elements.

[0027] Implicit association labels are generated based on implicit association relationships to characterize the indirect ecological impact of agricultural inputs. These implicit association labels record the initial agricultural input elements, the final ecological elements, and the propagation paths they take.

[0028] In this embodiment, constructing the agricultural ecological impact correlation diagram based on the explicit association labels and the implicit association labels specifically includes:

[0029] A basic framework for a heterogeneous graph network is established, using agricultural input factors and ecological factors as two types of nodes.

[0030] Explicit association labels are used as the first type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the first type of edges are given direct association attributes.

[0031] Implicit association labels are used as the second type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the second type of edges are given indirect association attributes and association propagation path information.

[0032] By integrating the first type of edge and the second type of edge, the agricultural ecological impact correlation graph containing multiple types of nodes and multiple types of associated edges is constructed.

[0033] In this embodiment, based on the agricultural ecological impact correlation diagram, a multi-dimensional index fusion analysis is performed on the relationship between agricultural resource utilization status and ecological environment change to generate agricultural green development status analysis results, specifically including:

[0034] In the above agricultural ecological impact correlation diagram, each agricultural input element node is taken as the starting point, and all ecological element nodes affected by it through explicit and implicit correlations are traversed to calculate the ecological impact node coverage of agricultural input elements.

[0035] Extract the key propagation path connecting agricultural input element nodes and ecological element nodes in the agricultural ecological impact correlation diagram, analyze the edge weights and node state changes on the key propagation path, identify the main effect path of impact transmission, and determine the transmission efficiency of the main effect path. The transmission efficiency is used to reflect the transmission efficiency of the impact of agricultural resource input on ecological environment change.

[0036] The coverage of the ecological impact nodes and the transmission efficiency of the main effect path are integrated to construct an evaluation index for green agricultural development.

[0037] Based on the agricultural green development evaluation index and the distribution characteristics of ecological element nodes in the agricultural ecological impact correlation diagram, an agricultural green development status analysis result including bottleneck identification results is generated.

[0038] In this embodiment, the agricultural input behavior data includes fertilizer application, pesticide application, irrigation amount, and agricultural machinery operation information.

[0039] In this embodiment, the ecological environment monitoring data includes soil nutrient content, soil moisture, soil temperature, and water environment indicators.

[0040] In this embodiment, the crop growth monitoring data includes crop plant height, leaf area index, and crop growth image information.

[0041] Secondly, this application provides a digital data intelligent analysis system for green agricultural development, comprising:

[0042] The data acquisition module is used to collect agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data within the agricultural production area, and to construct a multi-source agricultural dataset.

[0043] The explicit association module is used to identify explicit associations in multi-source agricultural data based on the direct changing relationship between agricultural input factors and ecological environment indicators, and to generate explicit association labels that characterize the direct ecological impact of agricultural inputs.

[0044] The implicit association module is used to construct the agricultural factor influence propagation path through the distribution characteristics of the explicit association tags among different ecological elements, and then to identify the indirect association of multi-source agricultural data based on the agricultural factor influence propagation path, and generate implicit association tags that characterize the indirect ecological impact relationship of agricultural inputs.

[0045] The fusion analysis module is used to construct an agricultural ecological impact correlation map based on the explicit and implicit correlation tags, and then perform multi-dimensional index fusion analysis on the relationship between agricultural resource utilization status and ecological environment change based on the agricultural ecological impact correlation map to generate agricultural green development status analysis results.

[0046] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects:

[0047] This application achieves a systematic characterization and accurate evaluation of the direct and indirect impacts of agricultural inputs on the ecological environment through the construction of multi-source agricultural datasets, joint mining of explicit and implicit associations, modeling of agricultural ecological impact association graphs, and fusion analysis of multi-dimensional indicators. First, explicit associations are identified based on the direct changing relationships between agricultural input factors and ecological environment indicators. By calculating the correlation statistical characteristic values ​​between the application rate change rate and indicator volatility, input-ecological element pairs exceeding the correlation threshold are screened, and their temporal matching features are extracted to generate explicit association labels for quantifying the direction and intensity of the direct impact of agricultural inputs on the ecological environment. Second, the distribution characteristics of explicit association labels among different ecological elements are used to construct the agricultural element impact propagation path. Shared ecological element nodes are identified, and multiple explicit associations are linked and combined according to their inherent dependency order. After directional verification, a propagation path reflecting the impact transmission chain is formed. Then, data changes at intermediate nodes are traced segment by segment along this path to verify whether the data fluctuations of the initial input can be transmitted to the final ecological element, uncovering statistical factors that cannot be directly explained by explicit associations. This paper generates implicit association labels that record the indirect ecological impacts transmitted through intermediate ecological elements and their complete paths, breaking through the limitations of traditional methods that are limited to single direct association comparisons. Then, based on explicit and implicit association labels, an agricultural ecological impact association graph is constructed with agricultural input elements and ecological elements as two types of nodes and direct and indirect association edges as edges. Based on this association graph, the ecological impact node coverage is calculated by traversing all ecological element nodes affected by each agricultural input element node as the starting point. The key propagation paths connecting input nodes and ecological nodes are extracted to identify the main effect paths and determine the transmission efficiency. The coverage and transmission efficiency are integrated to construct an agricultural green development evaluation index. At the same time, bottleneck links are identified by combining the distribution characteristics of ecological element nodes, and an agricultural green development status analysis result containing the bottleneck link identification results is generated. In summary, this application realizes a systematic deconstruction of the direct and indirect impacts of agricultural inputs on the ecological environment through the synergistic effect of explicit and implicit association joint mining and graph-based multidimensional analysis, which significantly improves the interpretability of agricultural green development status evaluation. Attached Figure Description

[0048] Figure 1 This is a schematic diagram illustrating an application scenario of the digital data intelligent analysis system for green agricultural development, based on some embodiments of this application.

[0049] Figure 2 This is an exemplary flowchart of a digital data intelligent analysis method for green agricultural development, as shown in some embodiments of this application.

[0050] Figure 3 This is a schematic diagram of the structure of an intelligent digital data analysis system for green agricultural development, as shown in some embodiments of this application.

[0051] Figure 4This is a schematic diagram of the structure of a computer device for realizing a digital data intelligent analysis method for green agricultural development, according to some embodiments of this application. Detailed Implementation

[0052] To better understand the technical solution of this application, the technical solution of this application will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0053] refer to Figure 1 This figure is a schematic diagram of an application scenario for an intelligent analysis system for digital data on green agricultural development, based on some embodiments of this application. The figure includes data acquisition devices, a server, a communication network, and terminals. The data acquisition devices include acquisition devices 1 to N, where N is an integer at least greater than 6. Each acquisition device interacts with the server via a communication link to exchange data and transmit commands. The terminals are connected to the server system via the communication network. The server acquires agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data provided by the acquisition devices. After cleaning, standardizing, and initially classifying the various data types, a multi-source agricultural dataset containing multi-dimensional data items is constructed. Then, based on the direct relationship between agricultural input factors and ecological environment indicators, explicit correlation identification is performed on the multi-source agricultural data to generate explicit correlation labels characterizing the direct ecological impact of agricultural inputs. Through the distribution characteristics of these explicit correlation labels among different ecological elements, the propagation path of agricultural factor influence is constructed, and then based on the... This paper describes the indirect correlation identification of agricultural inputs and ecological environment changes by analyzing the propagation path of agricultural factors. It mines the potential correlation between agricultural inputs and ecological environment changes, generating implicit correlation tags that characterize the indirect ecological impact of agricultural inputs. Then, based on the explicit and implicit correlation tags, an agricultural ecological impact correlation map is constructed. Furthermore, based on this map, multi-dimensional indicators such as agricultural resource utilization efficiency, ecological environment carrying capacity, and crop growth adaptability are integrated to conduct a comprehensive analysis of the relationship between agricultural resource utilization status and ecological environment changes, generating an analysis result of agricultural green development status. When a request for analysis and adjustment targeting a specific agricultural production area, agricultural input type, or ecological environment monitoring dimension is received via a terminal, the server feeds back the agricultural green development status analysis result, correlation tag details, and multi-dimensional analysis indicators to the terminal for viewing by agricultural managers, growers, ecological monitoring personnel, or agricultural technical guidance personnel. This provides data support for agricultural green production adjustments, resource optimization, and ecological protection decisions.

[0054] The data collection equipment may include agricultural input behavior recording terminals, soil moisture monitoring instruments, meteorological monitoring stations, crop growth sensors, irrigation and fertilization data collection modules, and ecological environment sampling equipment; the terminals may be, but are not limited to, agricultural green development management back-ends, smart terminals for growers, ecological environment monitoring platforms, or agricultural technology guidance terminals; the servers may be local data computing nodes deployed in agricultural industrial parks, or distributed agricultural green development data intelligent analysis service platforms built on agricultural cloud platforms.

[0055] refer to Figure 1 The figure is an exemplary flowchart of a digital data intelligent analysis method for green agricultural development according to some embodiments of this application. The digital data intelligent analysis method for green agricultural development mainly includes the following steps:

[0056] In step 101, agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data are collected within the agricultural production area, and a multi-source agricultural dataset is constructed.

[0057] In this embodiment, it should be noted that the agricultural input behavior data in this application includes fertilizer application amount, pesticide application amount, irrigation amount, and agricultural machinery operation information; the ecological environment monitoring data includes soil nutrient content, soil moisture, soil temperature, and water environment indicators; the crop growth monitoring data includes crop plant height, leaf area index, and crop growth image information; during the collection process, the agricultural input behavior data in the agricultural production area can be collected in the following ways: flow sensors or power monitoring modules can be installed on agricultural machinery such as fertilizer applicators and irrigation pumps to record operation information such as fertilizer application amount and irrigation amount in real time; at the same time, the execution log can be automatically exported by the control system of the intelligent water and fertilizer integrated machine to obtain input data; the collection of ecological environment monitoring data can be achieved in the following ways: soil sensors can be buried in the rhizosphere area of ​​crops to automatically collect soil temperature, humidity, conductivity, and pH value according to a set time period, and at the same time, field sensors can be deployed... CO2 temperature and humidity sensors collect environmental indicators such as air temperature and humidity, with a set time period of once every 10 minutes. Crop growth monitoring data can be collected in the following ways: high-resolution crop growth images can be obtained by periodically taking aerial photos with a drone equipped with a multispectral or camera along a planned route, while ground personnel carrying handheld chlorophyll meters, positioning devices, and plant measurement tools measure plant height, leaf area index, and other data at selected sampling points for verification and supplementation. All sensors are connected to a local edge controller via a bus, and data is aggregated according to the Modbus-RTU protocol. The data is then transmitted to a cloud server or data platform via a message queue telemetry transmission protocol through a 4G / 5G network. Finally, the agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data collected in the agricultural production area are normalized according to a preset format, and the normalized dataset is used as a multi-source agricultural dataset.

[0058] In step 102, explicit association identification is performed on multi-source agricultural data based on the direct change relationship between agricultural input factors and ecological environment indicators, generating explicit association labels that characterize the direct ecological impact relationship of agricultural inputs.

[0059] In this embodiment, the explicit correlation identification of multi-source agricultural data based on the direct changing relationship between agricultural input factors and ecological environment indicators, and the generation of explicit correlation labels characterizing the direct ecological impact of agricultural inputs, can be achieved through the following steps:

[0060] Spatiotemporal alignment processing is performed on multi-source agricultural data to establish a matching relationship between agricultural input behavior data and ecological environment monitoring data under the same spatiotemporal coordinate system;

[0061] Determine the rate of change in the application of agricultural input factors within a preset time window;

[0062] Determine the fluctuation range of ecological and environmental indicators within a preset time window;

[0063] The correlation statistical characteristic values ​​between agricultural input factors and ecological environment indicators within the preset time window are determined based on the application rate change rate and the volatility.

[0064] Based on the correlation statistical characteristic values, input-ecological element pairs that exceed the first correlation threshold are selected.

[0065] For the selected input-ecological element pairs, extract the time-series matching features of their changing trends, including consistency of change direction and change lag window;

[0066] The explicit association label is generated based on the time-series matching features. The explicit association label includes input element identifier, ecological environment indicator identifier, and association type identifier.

[0067] It should be noted that the application rate change rate in this application is an indicator that quantifies the magnitude of change in the application amount of agricultural input factors within a preset time window; the volatility is a statistical measure of the degree of fluctuation or variation of ecological and environmental indicators within a preset time window; the correlation statistical feature value is a value that comprehensively reflects the strength of the statistical correlation between changes in the application amount of agricultural input factors and fluctuations in ecological and environmental indicators; the input-ecological element pair is a pairing combination that identifies a possible direct correlation between agricultural input factors and ecological and environmental indicators; the time series matching feature is a parameter used to describe the synchronicity and lag characteristics of the changing trends of input-ecological element pairs in a time series; and the explicit association label is a structured identification information used to mark the direct ecological impact relationship between agricultural input factors and ecological and environmental indicators.

[0068] In practical applications, firstly, the collected agricultural input behavior data and ecological environment monitoring data may have inconsistent timestamps and geographical locations due to their different sources. Therefore, it is necessary to integrate the two types of data according to a unified time and spatial coordinate system. In the time dimension, based on a preset sampling interval, linear interpolation or nearest neighbor matching methods are used to align the recording time of agricultural input behavior data with that of ecological environment monitoring data to the same time point. In the spatial dimension, based on the specific field location where agricultural input behavior occurs and the deployment location of ecological environment monitoring sensors, geographic coordinate matching or buffer analysis is used to group input data and environmental data belonging to the same field or adjacent area into the same spatial unit. Through the above processing, a unified time and spatial coordinate system is established. First, establish a matching relationship between agricultural input behavior data at each moment and spatial location and the corresponding ecological environment monitoring data, using the generated matching relationship as the data foundation for subsequent correlation analysis. Second, for the agricultural input behavior data with established matching relationships, select a fixed-length time window, such as seven consecutive days or one month. Within this time window, calculate the total application amount of a certain agricultural input factor, then compare the application amounts of two adjacent time windows. Subtract the application amount of the previous window from the application amount of the later window, and then divide by the application amount of the previous window to obtain the rate of change of the application amount of that input factor. If the application amount of the previous window is zero, then directly use the application amount of the later window as the change amount, and mark the rate of change specially. Then, for the agricultural input behavior data... For ecological and environmental monitoring data within the same time window, select the ecological and environmental indicator to be analyzed, such as soil nitrate nitrogen content, and calculate the standard deviation of the indicator within the time window to quantify its fluctuation range. Alternatively, the difference between the maximum and minimum values ​​within the window can be used as a measure of volatility. If there are many data points within the time window, the sum of squared residuals after moving average can also be used to reflect the degree of volatility. The larger the volatility value, the more drastic the change in the ecological and environmental indicator during that period. The calculated volatility is used as a characteristic value to measure the intensity of environmental response. Then, based on the application rate change rate and the volatility, determine the correlation statistical characteristic value between agricultural input factors and ecological and environmental indicators within the preset time window. This can be achieved in the following way: within the same time... Within the window, the calculated rate of change and volatility of application rate are used as a pair of observations. The above calculation is repeated for multiple consecutive time windows to form two time series. Then, statistical methods are used to calculate the correlation between these two series, such as the Pearson correlation coefficient, to measure the degree of linear correlation between the rate of change and volatility of application rate. If the data does not conform to a normal distribution, the Spearman rank correlation coefficient can also be used. The obtained correlation coefficient is then used as the correlation statistical characteristic value, which reflects the synchronous change relationship between input changes and environmental fluctuations. Furthermore, based on the correlation statistical characteristic value, input-ecological element pairs that exceed the first correlation threshold are screened out. This can be achieved in the following way: a first correlation threshold is preset, for example, the absolute value of the correlation coefficient is greater than 0.6. The first correlation threshold can be set based on the mean of the experimental data, and is not limited here. Then, the correlation statistical characteristic value of each pair of agricultural input factors and ecological environment indicators is compared with the first correlation threshold. If the correlation statistical characteristic value exceeds the threshold, it is determined that there is a statistically significant correlation between the input factor and the ecological environment indicator, and this pair is recorded to form an input-ecological element pair. Simultaneously, the correlation statistical characteristic value corresponding to this pair can be retained as a weight reference for subsequent analysis. All selected input-ecological element pairs are used as a candidate set for further extraction of time-series features. Furthermore, for the selected input-ecological element pairs, the extraction of their trend-based time-series matching features can be achieved in the following way: For each selected input-ecological element pair, its original time-series data, i.e., the input factor application rate sequence and the ecological environment indicator sequence, are extracted. Through cross-correlation analysis or dynamic time warping, the similarity of the two sequences at different time offsets is calculated, and the time offset that maximizes the similarity is found. This time offset is then used as... The time lag window is used as the time offset. Simultaneously, the consistency of the change direction is determined based on the change direction of the two sequences under this time offset. If the environmental indicator also increases when the input increases, it is considered positive consistency; if the environmental indicator decreases when the input increases, it is considered negative consistency. These parameters are then combined to form the time-series matching feature of the input-ecological element pair. Finally, the explicit association label is generated based on the time-series matching feature. This can be achieved in the following way: the obtained time-series matching feature is integrated with the corresponding agricultural input elements and ecological environment indicators to generate a structured data label. This label contains three basic components: input element identifier, indicating which agricultural input, such as nitrogen fertilizer or phosphate fertilizer; ecological environment indicator identifier, indicating which environmental indicator, such as soil nitrate nitrogen or groundwater total phosphorus; and association type identifier, describing the nature of the direct association, including positive or negative association, time lag length, and association strength level. All identified explicit association labels are stored in the database, and the generated explicit association labels are used as direct relationship edges in the subsequent construction of the agricultural ecological impact association diagram.

[0069] In step 103, the distribution characteristics of the explicit association tags among different ecological elements are used to construct the agricultural element influence propagation path. Then, based on the agricultural element influence propagation path, the indirect association of multi-source agricultural data is identified, and implicit association tags representing the indirect ecological impact relationship of agricultural inputs are generated.

[0070] In this embodiment, constructing the propagation path of agricultural factor influence based on the distribution characteristics of the explicit association tags among different ecological elements can be achieved through the following steps:

[0071] Obtain all explicit association tags and construct an element-level association graph with ecological elements as nodes and explicit association relationships as edges;

[0072] Identify the shared ecological element nodes in the element-level association graph, where each shared ecological element node simultaneously serves as the target or starting point of at least two explicit association relationships.

[0073] Based on the shared ecological element nodes, multiple explicit relationships are linked and combined in the order of their inherent dependencies to obtain a candidate propagation path from the initial ecological element that has an explicit relationship with agricultural input elements to the final ecological element through intermediate ecological elements.

[0074] The candidate propagation paths are validated for path directionality, and paths that do not conform to the time sequence of ecological processes are eliminated to generate the propagation paths of agricultural factor influence.

[0075] It should be noted that the element-level association diagram in this application is a graphical model representing the network structure formed by the interconnection of ecological elements through explicit association relationships; the shared ecological element node refers to the intersection of multiple explicit association relationships, connecting key ecological elements with different direct ecological impact paths; the agricultural element impact propagation path is an effective path used to truly reflect the ecological impact of agricultural input elements through the temporal causal chain between ecological elements.

[0076] In practical application, firstly, all generated explicit association labels are retrieved from the database. Each explicit association label contains an input element identifier, an ecological environment indicator identifier, and an association type identifier. Using the ecological element corresponding to each ecological environment indicator identifier as a node in the graph, and the association relationships recorded in the explicit association labels as edges connecting the nodes, a network graph is constructed that only contains ecological element nodes and their direct association relationships. During the construction process, for each explicit association label, if the label involves two ecological elements, an edge is established between these two ecological element nodes, and the association type is recorded as the edge's attribute. After traversing all explicit association labels, a network graph consisting of ecological element nodes and edges representing direct ecological impact relationships is obtained. The first step involves constructing an element-level association graph. Then, within this graph, all ecological element nodes are traversed, and the number of edges connected to each node and the number of times it acts as the starting or ending point of an edge are counted. If a node simultaneously acts as the ending point of at least two explicit associations, or simultaneously acts as the starting point of at least two explicit associations, or simultaneously possesses both of these conditions, then that node is marked as a shared ecological element node. These identified shared ecological element nodes represent the intersection points between ecological elements and are potential transit nodes in the influence propagation path. All identified shared ecological element nodes are used as hubs for subsequent concatenated candidate propagation paths. Finally, each starting ecological element that has an explicit association with an agricultural input element is used as the starting point... The starting ecological element refers to the ecological element directly connected to agricultural input elements through explicit association tags. In the element-level association diagram, starting from the starting ecological element, following the direction indicated by the explicit association relationships, it sequentially passes through the identified common ecological element nodes, and gradually extends according to the inherent dependency order between ecological elements, connecting multiple explicit association relationships to form a sequence from the starting ecological element through intermediate ecological elements to the final ecological element. All possible starting points and paths are traversed to obtain all possible transmission sequences, which are then used as candidate propagation paths. Finally, the directionality of each candidate propagation path is verified based on the objective temporal order of changes in various elements during the ecological process. Specifically… Obtain time-series monitoring data corresponding to each ecological element node in the candidate propagation path, and check whether the order of changes between adjacent nodes in the path conforms to ecological common sense or actual observed temporal relationships. For example, the change time of the initial ecological element should be earlier than the change time of the intermediate ecological element, and the change time of the intermediate ecological element should be earlier than the change time of the final ecological element. If there is a time order reversal or inconsistency in a candidate propagation path, the path is determined to be inconsistent with the actual ecological process and is eliminated. All paths that pass the directionality check are retained. These paths are the agricultural element impact propagation paths that can truly reflect the transmission process of agricultural input factors. The generated agricultural element impact propagation paths are used as the basis for subsequent implicit association identification.

[0077] In this embodiment, the indirect correlation identification of multi-source agricultural data based on the propagation path of agricultural factor influences, and the generation of implicit correlation labels characterizing the indirect ecological impact relationship of agricultural inputs, can be achieved through the following steps:

[0078] Extract the initial agricultural input element nodes and the final ecological element nodes corresponding to the propagation path of the agricultural factors to form the input-final element pairs to be analyzed;

[0079] Along the propagation path of the agricultural factors, the monitoring data changes of intermediate ecological factor nodes are traced segment by segment to verify whether the data fluctuations of the initial agricultural input factors can be transmitted to the end ecological factors along the path.

[0080] If the time series of data changes of the end ecological elements and the time series of data changes of the starting agricultural input elements have a statistical correlation under path constraints, and the statistical correlation cannot be directly explained by the explicit association between the two, then it is identified that there is an implicit association between the starting agricultural input elements and the end ecological elements.

[0081] Implicit association labels are generated based on implicit association relationships to characterize the indirect ecological impact of agricultural inputs. These implicit association labels record the initial agricultural input elements, the final ecological elements, and the propagation paths they take.

[0082] It should be noted that the implicit association in this application is used to characterize the inherent connection between the initial agricultural input factors and the final ecological factors, which is transmitted through intermediate ecological factors; the implicit association tag is a structured identification information used to record the indirect ecological impact relationship of agricultural inputs.

[0083] In practical application, firstly, the starting point and ending point of each agricultural factor impact propagation path are extracted. The starting point is the initial agricultural input node directly associated with the agricultural input, and the ending point is the end ecological element node pointed to by the path. For each agricultural factor impact propagation path, the corresponding initial agricultural input node and end ecological element node are combined to form a pair to be analyzed. This process is performed on all agricultural factor impact propagation paths, resulting in multiple input-end element pairs. Each pair clearly indicates that a certain agricultural input factor may indirectly affect a certain end ecological element through a specific propagation path. Secondly, for each input-end element pair, the corresponding agricultural factor impact propagation path is obtained. The seeding path starts from the initial agricultural input node, passes through several intermediate ecological element nodes, and finally reaches the final ecological element node. Along this path, starting from the initial agricultural input node, data sequences of application rate changes for that input over multiple time windows are acquired. Monitoring data change sequences for the first intermediate ecological element node within the same time period are also acquired. The similarity between the two sequences is calculated to verify whether changes in the initial input have caused changes in the intermediate ecological element. If the verification is successful, the path continues downwards, verifying whether changes in the previous intermediate ecological element have caused changes in the next, until the final ecological element node is reached. Time series analysis methods, such as cross-correlation, are used in each verification step. The function determines whether changes in preceding nodes have predictive power for changes in subsequent nodes. Then, after completing segment-by-segment tracing verification along the path, it obtains the time series of data changes for initial agricultural inputs and final ecological elements. Under the constraint of the known propagation path—that is, considering the time lags and transmission effects that may be introduced by intermediate nodes—it calculates the statistical correlation between these two time series. Specifically, based on the total transmission lags determined by segment-by-segment verification along the path, it applies a corresponding time offset to the time series of changes in initial inputs, and then calculates the correlation coefficient with the time series of changes in final ecological elements. If the calculated correlation coefficient exceeds a preset second correlation threshold, it indicates that there is a statistical correlation between the two under the path constraint. Simultaneously, it checks... The database is checked to see if there is an explicit correlation label between the initial agricultural input and the final ecological element. If no such label exists, it means that the statistical correlation cannot be explained by a direct explicit correlation between the two. Therefore, it is determined that there is an implicit correlation between the initial agricultural input and the final ecological element. It should be further noted that the second correlation threshold in this embodiment can be set according to historical experimental data, and can be set to the mean or other feature values. No limitation is made here. Finally, for each identified implicit correlation, a structured data label is generated. This label contains three core components: an initial agricultural input identifier, indicating the type of agricultural input that has an indirect impact; and a final ecological element identifier, indicating the environmental indicators that are indirectly affected.The information on the propagation path is recorded in the form of a node sequence in the labels, clearly showing the complete chain of impact transmission. In addition, attributes such as statistical correlation strength and transmission lag under path constraints can be added. All generated implicit association labels are stored in the database and used as indirect relationship edges in the subsequent construction of the agricultural ecological impact relationship diagram.

[0084] In step 104, an agricultural ecological impact correlation diagram is constructed based on the explicit and implicit correlation labels. Then, based on the agricultural ecological impact correlation diagram, a multi-dimensional index fusion analysis is performed on the relationship between agricultural resource utilization status and ecological environment change to generate agricultural green development status analysis results.

[0085] In this embodiment, constructing the agricultural ecological impact correlation diagram based on the explicit association labels and the implicit association labels can be achieved through the following steps:

[0086] A basic framework for a heterogeneous graph network is established, using agricultural input factors and ecological factors as two types of nodes.

[0087] Explicit association labels are used as the first type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the first type of edges are given direct association attributes.

[0088] Implicit association labels are used as the second type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the second type of edges are given indirect association attributes and association propagation path information.

[0089] By integrating the first type of edge and the second type of edge, the agricultural ecological impact correlation graph containing multiple types of nodes and multiple types of associated edges is constructed.

[0090] It should be noted that the first type of edge in this application is a connection line used to represent the direct ecological impact relationship between agricultural input factors and ecological factors in a heterogeneous graph network; the second type of edge is a connection line used to represent the indirect ecological impact relationship between agricultural input factors and ecological factors in a heterogeneous graph network, with attached propagation path information; the agricultural ecological impact correlation graph is a graphical model used to comprehensively display the complete network structure formed by the direct and indirect connections between agricultural input factors and ecological factors.

[0091] In practical application, firstly, from the generated explicit and implicit association labels, the identifiers of all appearing agricultural input elements and ecological elements are obtained respectively. Each agricultural input element identifier is treated as an agricultural input element node, and each ecological element identifier as an ecological element node. Then, a blank graph data structure is built. This graph structure clearly distinguishes between the two node types and is prepared for adding edges connecting these nodes later. During the construction process, a unique identifier is assigned to each node, and its type, i.e., agricultural input element type or ecological element type, is recorded. This establishes the basic framework of a heterogeneous graph network containing all agricultural input element nodes and all ecological element nodes, but without any edges yet added. Secondly, the graph is traversed... All generated explicit association labels in the database, each containing an input element identifier and an ecological environment indicator identifier, are used. Within the established heterogeneous graph network framework, the agricultural input element node corresponding to the input element identifier and the ecological environment indicator node corresponding to the ecological environment indicator identifier are identified. An edge is added between these two nodes and marked as a first-type edge. Simultaneously, association type identifiers, including positive association, negative association, time lag window, and association strength, are extracted from the explicit association labels and stored as attributes of the first-type edge, assigning it a direct association attribute. After performing the above operations on all explicit association labels, all direct associations are added to the heterogeneous graph network as first-type edges. The first type of edge is added as a graph element representing direct ecological impact relationships. Then, all generated implicit association labels in the database are traversed. Each implicit association label contains the identifier of the initial agricultural input element, the identifier of the final ecological element, and the information of the propagation path. In the existing basic framework of the heterogeneous graph network, the agricultural input element node corresponding to the identifier of the initial agricultural input element and the ecological element node corresponding to the identifier of the final ecological element are found. An edge is added between these two nodes and marked as a second type edge. At the same time, the propagation path information recorded in the implicit association label, i.e., the sequence of intermediate ecological element nodes, and the possible indirect association strength attributes, are stored as attributes of the second type edge and assigned to it. After performing the above operations on all implicit association labels, all indirect associations are added to the heterogeneous graph network as second-type edges, and the added second-type edges are used as graph elements representing indirect ecological impact relationships. Finally, the basic framework of the heterogeneous graph network, which includes all agricultural input element nodes and all ecological element nodes, is integrated with all first-type edges and all second-type edges. During the integration process, the established basic framework of the heterogeneous graph network, which already includes all agricultural input element nodes and ecological element nodes, is used as the basis. All first-type edges are traversed, and each first-type edge is added to the framework according to the input element nodes and ecological element nodes it connects, and the direct association attributes are attached to the edges.Iterate through all second-type edges, adding each edge to the framework according to the input element nodes and ecological element nodes it connects, and attaching indirect association attributes and propagation path information to the edges. For cases where both first-type and second-type edges exist between the same pair of nodes during the addition process, retain both types of edges as multi-edges, thus obtaining a complete network containing all nodes and all associated edges. This complete network is then used as the agricultural ecological impact correlation graph.

[0092] In this embodiment, the analysis of the relationship between agricultural resource utilization status and ecological environment change based on the agricultural ecological impact correlation diagram, using multi-dimensional index fusion analysis, can generate agricultural green development status analysis results through the following steps:

[0093] In the above agricultural ecological impact correlation diagram, each agricultural input element node is taken as the starting point, and all ecological element nodes affected by it through explicit and implicit correlations are traversed to calculate the ecological impact node coverage of agricultural input elements.

[0094] Extract the key propagation path connecting agricultural input element nodes and ecological element nodes in the agricultural ecological impact correlation diagram, analyze the edge weights and node state changes on the key propagation path, identify the main effect path of impact transmission, and determine the transmission efficiency of the main effect path. The transmission efficiency is used to reflect the transmission efficiency of the impact of agricultural resource input on ecological environment change.

[0095] The coverage of the ecological impact nodes and the transmission efficiency of the main effect path are integrated to construct an evaluation index for green agricultural development.

[0096] Based on the agricultural green development evaluation index and the distribution characteristics of ecological element nodes in the agricultural ecological impact correlation diagram, an agricultural green development status analysis result including bottleneck identification results is generated.

[0097] It should be noted that the ecological impact node coverage in this application is a breadth indicator that quantifies the number of ecological elements that agricultural input factors can influence through direct and indirect connections; the main effect path is used to identify the key path that plays a dominant role in the transmission of impacts in the agricultural ecological impact correlation diagram; and the agricultural green development evaluation index is a quantitative score that comprehensively reflects the degree of impact of agricultural resource utilization on the ecological environment and its transmission characteristics.

[0098] In practical application, firstly, in the constructed agricultural ecological impact correlation graph, an agricultural input element node is selected as the starting point for analysis. Starting from this point, a graph traversal algorithm is used to expand along all first-type and second-type edges connected to this node, visiting all reachable ecological element nodes. During the traversal, all visited ecological element nodes are recorded, duplicate nodes are removed, and the total number of these unique ecological element nodes is counted. The ratio between this total number and the total number of all nodes in the agricultural ecological impact correlation graph is then used as the ecological impact node coverage of the agricultural input element node. The larger this value, the wider the impact range of the agricultural input element on the ecological environment. Secondly, all paths connecting agricultural input element nodes and ecological element nodes are extracted from the agricultural ecological impact correlation graph. These paths include direct paths composed of first-type edges and indirect paths composed of second-type edges. For each extracted path, the weights of each edge on the path are obtained. The edge weights can be assigned based on the correlation strength in the explicit correlation label or the statistical correlation strength in the implicit correlation label. Simultaneously, the ecological element nodes on the path are obtained. By analyzing the monitoring data change sequence within a historical time window, and further analyzing the cumulative sum of edge weights and the transmission speed of node state changes along the path, the importance of each path in the overall impact transmission is comprehensively evaluated. The path with the highest weight, the fastest transmission speed, or the strongest explanatory power for changes in terminal nodes is identified as the main effect path. Further, for the identified main effect path, its transmission efficiency is calculated. This efficiency can be quantified by the ratio of the change amplitude of terminal ecological element nodes to the change amplitude of initial agricultural input element nodes, or the reciprocal of the time required for change transmission. Then, the calculated ecological impact node coverage and the transmission efficiency of the determined main effect path are fused. The fusion method can employ weighted summation or product normalization. For example, the ecological impact node coverage is first normalized to a range of 0 to 1, and the transmission efficiency is also normalized. Then, based on preset weight coefficients, such as a coverage weight of 0.4 and a transmission efficiency weight of 0, the process is repeated.6. A weighted sum of the two values ​​yields a comprehensive score, which serves as the agricultural green development evaluation index. A higher index value indicates a more limited scope and lower transmission efficiency of the impact of agricultural resource utilization on the ecological environment, signifying a higher level of green development. Conversely, a lower index value indicates a wider scope and faster transmission of ecological impact, indicating a lower level of green development. Finally, the constructed agricultural green development evaluation index is used as the basic numerical output of the analysis results, intuitively reflecting the current level of green development in agricultural resource utilization. This is then combined with the distribution characteristics of each ecological element node in the agricultural ecological impact correlation diagram to identify bottlenecks. These distribution characteristics include... The number of connections a node has in the association graph (i.e., degree centrality) and the number of times it appears as an intermediate bridge on multiple paths (i.e., betweenness centrality) are considered. Specifically, the focus is on ecological element nodes with high degree centrality or betweenness centrality, as these nodes are often pivotal points affecting transmission. If their state is abnormal, such as soil nutrient imbalance, they may become bottlenecks restricting overall ecological health. The identified bottlenecks are integrated with the agricultural green development evaluation index to generate an agricultural green development status analysis result that includes the bottleneck identification results. This generated agricultural green development status analysis result is used as the final output to guide the optimization and adjustment of agricultural production.

[0099] It should be noted that the existing agricultural ecological environment impact analysis is usually limited to the direct correlation comparison of a single element, making it difficult to systematically reveal the indirect ecological effects of agricultural inputs transmitted through multiple media environmental elements, and lacking the technical bottleneck of multi-dimensional data fusion and evaluation capabilities for inputs, environment, and crops. This invention proposes a digital data intelligent analysis method for green agricultural development. Compared with existing technologies, the inventiveness of this invention lies in: constructing a progressive analysis chain of explicit correlation identification, propagation path construction, and implicit correlation mining, thus decoupling and graphically representing the direct and indirect impacts of agricultural inputs on the ecological environment for the first time; particularly through explicit correlation tags in different ecological elements... By constructing the distribution characteristics of elements to establish the impact propagation path, and then mining implicit correlations based on this path, the technical challenge of traditional methods being unable to capture the indirect impact mechanism of agricultural inputs being transmitted step by step from intermediate ecological elements such as soil and water to end-point environmental indicators is solved. On this basis, by constructing a heterogeneous network of agricultural ecological impacts that integrates explicit and implicit correlations, a multi-dimensional index fusion analysis of the relationship between agricultural resource utilization status and ecological environment changes is achieved. This elevates fragmented monitoring data to a systematic green development status evaluation result that includes the breadth of impact, transmission efficiency, and bottleneck links, achieving the technical effect of accurately identifying the key transmission paths and control nodes of the complex impact of agricultural production activities on the ecological environment.

[0100] On the other hand, in some embodiments, this application provides a digital data intelligent analysis system for green agricultural development, with reference to... Figure 3The figure is a schematic diagram of the structure of a digital data intelligent analysis system for green agricultural development according to some embodiments of this application. The digital data intelligent analysis system for green agricultural development includes: a data acquisition module 301, an explicit association module 302, an implicit association module 303, and a fusion analysis module 304, which are described below:

[0101] The data acquisition module 301 is used to collect agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data within the agricultural production area, and to construct a multi-source agricultural dataset.

[0102] The explicit association module 302 is used to identify explicit associations in multi-source agricultural data based on the direct change relationship between agricultural input factors and ecological environment indicators, and to generate explicit association labels that characterize the direct ecological impact relationship of agricultural inputs.

[0103] The implicit association module 303 is used to construct the agricultural factor influence propagation path through the distribution characteristics of the explicit association labels among different ecological elements, and then to perform indirect association identification on multi-source agricultural data based on the agricultural factor influence propagation path, and generate implicit association labels that characterize the indirect ecological impact relationship of agricultural inputs.

[0104] The fusion analysis module 304 is used to construct an agricultural ecological impact correlation diagram based on the explicit and implicit correlation labels, and then perform multi-dimensional index fusion analysis on the relationship between agricultural resource utilization status and ecological environment change based on the agricultural ecological impact correlation diagram to generate agricultural green development status analysis results.

[0105] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to acquire the code and execute the above-described intelligent analysis method for digital data on green agricultural development.

[0106] In some embodiments, reference Figure 4 The figure is a schematic diagram of the structure of a computer device for implementing a digital data intelligent analysis method for green agricultural development, according to some embodiments of this application. The digital data intelligent analysis method for green agricultural development described in the above embodiments can... Figure 4 The computer device shown is used to implement this, and the computer device 400 includes at least one processor 401, a communication bus 402, a memory 403, and at least one communication interface 404.

[0107] Processor 401 can be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0108] The communication bus 402 can be used to transmit information between the aforementioned components.

[0109] The memory 403 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disks or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 403 may exist independently and be connected to the processor 401 via the communication bus 402. The memory 403 may also be integrated with the processor 401.

[0110] The memory 403 stores program code for executing the scheme of this application, and its execution is controlled by the processor 401. The processor 401 executes the program code stored in the memory 403. The program code may include one or more software modules. In the above embodiments, the digital data intelligent analysis method for green agricultural development can be implemented by the processor 401 and one or more software modules in the program code in the memory 403.

[0111] Communication interface 404 uses any transceiver-like device to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0112] In a specific implementation, as one example, a computer device may include multiple processors, each of which may be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, a processor may refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0113] The aforementioned computer device can be a general-purpose computer device or a special-purpose computer device. In specific implementations, the computer device can be a desktop computer, a portable computer, a network server, a handheld digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. This application does not limit the type of computer device.

[0114] In addition, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above-mentioned intelligent analysis method for digital data on green agricultural development.

[0115] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0116] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. An agricultural green development digital data intelligent analysis method, characterized in that, include: Collect agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data within agricultural production areas, and construct a multi-source agricultural dataset; Based on the direct changing relationship between agricultural input factors and ecological environment indicators, explicit correlation identification is performed on multi-source agricultural data to generate explicit correlation labels that characterize the direct ecological impact of agricultural inputs. By constructing the agricultural factor influence propagation path through the distribution characteristics of the explicit association tags among different ecological elements, and then identifying the indirect association of multi-source agricultural data based on the agricultural factor influence propagation path, implicit association tags characterizing the indirect ecological impact relationship of agricultural inputs are generated. An agricultural ecological impact correlation diagram is constructed based on the explicit and implicit correlation labels. Then, based on the agricultural ecological impact correlation diagram, a multi-dimensional index fusion analysis is performed on the relationship between agricultural resource utilization status and ecological environment change to generate agricultural green development status analysis results. Specifically, based on the direct changing relationship between agricultural input factors and ecological environment indicators, explicit correlation identification is performed on multi-source agricultural data to generate explicit correlation labels characterizing the direct ecological impact of agricultural inputs. These labels include: Spatiotemporal alignment processing is performed on multi-source agricultural data to establish a matching relationship between agricultural input behavior data and ecological environment monitoring data under the same spatiotemporal coordinate system; Determine the rate of change in the application of agricultural input factors within a preset time window; Determine the fluctuation range of ecological and environmental indicators within a preset time window; The correlation statistical characteristic values ​​between agricultural input factors and ecological environment indicators within the preset time window are determined based on the application rate change rate and the volatility. Based on the correlation statistical characteristic values, input-ecological element pairs that exceed the first correlation threshold are selected. For the selected input-ecological element pairs, extract the time-series matching features of their changing trends, including consistency of change direction and change lag window; The explicit association label is generated based on the time-series matching features. The explicit association label includes input element identifier, ecological environment indicator identifier, and association type identifier. Specifically, the process of identifying indirect correlations among multi-source agricultural data based on the propagation paths of agricultural factor influences, and generating implicit association labels characterizing the indirect ecological impacts of agricultural inputs, includes: Extract the initial agricultural input element nodes and the final ecological element nodes corresponding to the propagation path of the agricultural factors to form the input-final element pairs to be analyzed; Along the propagation path of the agricultural factors, the monitoring data changes of intermediate ecological factor nodes are traced segment by segment to verify whether the data fluctuations of the initial agricultural input factors can be transmitted to the end ecological factors along the path. If the time series of data changes of the end ecological elements and the time series of data changes of the starting agricultural input elements have a statistical correlation under path constraints, and the statistical correlation cannot be directly explained by the explicit association between the two, then it is identified that there is an implicit association between the starting agricultural input elements and the end ecological elements. Implicit association labels are generated based on implicit association relationships to characterize the indirect ecological impact of agricultural inputs. These implicit association labels record the initial agricultural input elements, the final ecological elements, and the propagation paths they take. Specifically, constructing the agricultural ecological impact correlation diagram based on the explicit and implicit association labels includes: A basic framework for a heterogeneous graph network is established, using agricultural input factors and ecological factors as two types of nodes. Explicit association labels are used as the first type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the first type of edges are given direct association attributes. Implicit association labels are used as the second type of edges connecting input element nodes and ecological element nodes in the heterogeneous graph network, and the second type of edges are given indirect association attributes and association propagation path information. By integrating the first type of edge and the second type of edge, the agricultural ecological impact correlation graph containing multiple types of nodes and multiple types of associated edges is constructed; Specifically, the analysis of the relationship between agricultural resource utilization status and ecological environment change based on the aforementioned agricultural ecological impact correlation diagram, through multi-dimensional index fusion analysis, generates the following agricultural green development status analysis results: In the above agricultural ecological impact correlation diagram, each agricultural input element node is taken as the starting point, and all ecological element nodes affected by it through explicit and implicit correlations are traversed to calculate the ecological impact node coverage of agricultural input elements. Extract the key propagation path connecting agricultural input element nodes and ecological element nodes in the agricultural ecological impact correlation diagram, analyze the edge weights and node state changes on the key propagation path, identify the main effect path of impact transmission, and determine the transmission efficiency of the main effect path. The transmission efficiency is used to reflect the transmission efficiency of the impact of agricultural resource input on ecological environment change. The coverage of the ecological impact nodes and the transmission efficiency of the main effect path are integrated to construct an evaluation index for green agricultural development. Based on the agricultural green development evaluation index and the distribution characteristics of ecological element nodes in the agricultural ecological impact correlation diagram, an agricultural green development status analysis result including bottleneck identification results is generated.

2. The method of claim 1, wherein, Constructing the propagation path of agricultural factor influence based on the distribution characteristics of the explicit association tags among different ecological elements specifically includes: Obtain all explicit association tags and construct an element-level association graph with ecological elements as nodes and explicit association relationships as edges; Identify the shared ecological element nodes in the element-level association graph, where each shared ecological element node simultaneously serves as the target or starting point of at least two explicit association relationships. Based on the shared ecological element nodes, multiple explicit relationships are linked and combined in the order of their inherent dependencies to obtain a candidate propagation path from the initial ecological element that has an explicit relationship with agricultural input elements to the final ecological element through intermediate ecological elements. The candidate propagation paths are validated for path directionality, and paths that do not conform to the time sequence of ecological processes are eliminated to generate the propagation paths of agricultural factor influence.

3. The method as described in claim 1, characterized in that, The agricultural input data includes fertilizer application, pesticide application, irrigation, and agricultural machinery operation information.

4. The method as described in claim 1, characterized in that, The ecological environment monitoring data includes soil nutrient content, soil moisture, soil temperature, and water environment indicators.

5. The method as described in claim 1, characterized in that, The crop growth monitoring data includes crop height, leaf area index, and crop growth image information.

6. A digital data intelligent analysis system for green agricultural development, which performs intelligent analysis using the method described in any one of claims 1 to 5, characterized in that, The system includes: The data acquisition module is used to collect agricultural input behavior data, ecological environment monitoring data, and crop growth monitoring data within the agricultural production area, and to construct a multi-source agricultural dataset. The explicit association module is used to identify explicit associations in multi-source agricultural data based on the direct changing relationship between agricultural input factors and ecological environment indicators, and to generate explicit association labels that characterize the direct ecological impact of agricultural inputs. The implicit association module is used to construct the agricultural factor influence propagation path through the distribution characteristics of the explicit association tags among different ecological elements, and then to identify the indirect association of multi-source agricultural data based on the agricultural factor influence propagation path, and generate implicit association tags that characterize the indirect ecological impact relationship of agricultural inputs. The fusion analysis module is used to construct an agricultural ecological impact correlation map based on the explicit and implicit correlation tags, and then perform multi-dimensional index fusion analysis on the relationship between agricultural resource utilization status and ecological environment change based on the agricultural ecological impact correlation map to generate agricultural green development status analysis results.

Images