A data calling risk monitoring system and method applied to facility agriculture

Abstract

The application discloses a data calling risk monitoring system and method applied to facility agriculture, relates to the technical field of facility agriculture monitoring, and determines a production state risk mark based on the analysis of an environmental signal accumulation deviation index based on environmental data, the analysis of a texture uniformity deviation based on a canopy image, and the contradiction verification of agricultural operation records and actual regulation and control; data packet attributes are mapped into discrete symbols to generate data attribute state symbol sequences, the conversion density and conversion rhythm variation coefficient of each sequence are analyzed, cross-attribute difference degrees are calculated, and state risk indexes are analyzed; delay phenomena of processing time lengths between service nodes are analyzed, abnormal nodes and delay accumulation risks are identified, abnormal calling chain proportions are counted and converted into service application risk indexes; risk lag time and change trends are monitored, and graded response treatments are triggered according to thresholds. The application improves the confidentiality of data analysis and the analysis efficiency of fault source positioning.

Description

Technical Field

[0001] This invention relates to the field of facility agriculture monitoring technology, specifically a data retrieval risk monitoring system and method applied to facility agriculture. Background Technology

[0002] The trusted data space for facility agriculture involves complex business scenarios such as multi-entity data access, cross-entity data circulation, and multi-party collaborative modeling. The security of data flow and the reliability of service calls are key to ensuring the normal operation of industry collaboration.

[0003] In existing technologies, risk monitoring in facility agriculture mainly revolves around setting fixed thresholds for alarms based on sensor data from the production process, with risk assessment logic focused on whether a single data point exceeds the limit. However, in scenarios involving multi-entity data access and cross-entity collaborative applications within a trusted data space, risks not only arise at the production site but can also breed and spread within the data flow and service call chains. Existing solutions for detecting data flow anomalies generally employ traffic statistics or behavioral baseline comparisons, requiring the maintenance of historical behavioral archives for each entity. This increases system overhead and faces the problem of baseline failure due to natural changes in entity behavior during production stages. Furthermore, it cannot identify abnormal behaviors such as sudden data accumulation and data packet forgery based solely on the temporal structure characteristics of the data flow itself without parsing the data packet content. Risk perception of the service call chain is similarly limited to monitoring the independent response time and success rate of each node. When a service node experiences a delay, existing technologies cannot track the cascading amplification effect of this delay between upstream and downstream services, nor can they locate the key abnormal nodes causing the delay accumulation.

[0004] Therefore, this invention discloses a data retrieval risk monitoring system and method for facility agriculture to solve the above problems. Summary of the Invention

[0005] The purpose of this invention is to provide a data retrieval risk monitoring system and method for facility agriculture, in order to solve the problems raised in the prior art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a data retrieval risk monitoring method applied to facility agriculture, the method comprising the following steps: S1: Based on environmental data analysis of cumulative deviation indicators of environmental signals, based on canopy image analysis of texture uniformity deviation, combined with the verification of contradictions between agricultural operation records and actual control, the risk marker of production status is determined. S2: Map data packet attributes to discrete symbols to generate a sequence of data attribute state symbols, analyze the transformation density and transformation rhythm variation coefficient of each sequence, calculate the cross-attribute difference, and analyze the state risk index; S3: Track the processing time of each service node in the service call chain, analyze the latency phenomenon between service nodes, identify abnormal nodes and latency accumulation risks, and statistically analyze the proportion of abnormal call chains and convert them into service application risk indicators. S4: Generate a comprehensive risk value by combining production status risk markers, data usage status risk indicators, and service application risk indicators; monitor risk lag time and trends; and trigger tiered response and handling based on the trend of the comprehensive risk value and the risk lag time threshold.

[0007] S1 includes the following: S101: Obtain environmental data inside the facility agriculture entity, the environmental data including at least temperature, humidity and light intensity; obtain planting attribute data registered by the facility agriculture entity; the planting attribute data including at least planting variety, planting area, planting date and expected harvest date; Based on the planted variety, retrieve the corresponding suitable upper and lower limits of temperature, humidity, and light from the preset knowledge base; For each moment within the historical time window, determine whether the measured values ​​of temperature, humidity, and light intensity fall within their respective suitable upper and lower limits; if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 0, and if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 1; the sum of the indicator function outputs of each environmental data at each moment within the historical time window is recorded as the environmental signal cumulative deviation index. S102: Obtain visible light images of crop canopies within the facility of the main agricultural facility, convert the visible light images of the canopies into grayscale images, and analyze the texture uniformity; Extract the historical texture uniformity baseline of the corresponding facility agriculture entity during normal operation. The historical texture uniformity baseline is the median value of the texture uniformity of all images on environmentally suitable days in the past several days of the facility agriculture entity. The cumulative deviation index of the environmental signal on the environmentally suitable days is less than the cumulative deviation threshold of the environmental signal. If the number of environmentally suitable days is less than the threshold of suitable days, the historical texture uniformity baseline is equal to the historical median value of the texture uniformity of the facility agriculture entity in the same period of the previous production cycle. Analyze the texture uniformity deviation of the grayscale image; the texture uniformity deviation is equal to the ratio of the absolute value of the difference between the current texture uniformity and the historical texture uniformity baseline to the historical texture uniformity baseline. S103: Analysis of production status risk indicators based on cumulative deviation index of environmental signals and relative deviation of grayscale images; If the production status risk index is greater than the production status risk threshold, and there is a contradiction between the agricultural operation records declared by the corresponding facility agriculture entity in the public data within the historical time window and the actual environmental control data, the production status risk of the corresponding facility agriculture entity will be marked as 1; otherwise, the production status risk of the facility agriculture entity will be marked as 0. The contradictions include at least the following: the declaration has been made that ventilation and cooling operations have been implemented, but the actual temperature remains above the upper limit of suitable temperature for several consecutive hours after the declaration, and the temperature difference exceeds the temperature difference threshold; or the declaration has been made that humidification operations have been implemented, but the actual humidity remains below the lower limit of suitable humidity.

[0008] Existing agricultural risk monitoring technologies rely solely on sensor data for anomaly detection. This solution cross-references actual environmental measurements with administrative records. When environmental signals deviate from the suitable range for registered varieties for an extended period, it not only reflects production anomalies but may also reveal fraudulent use of entity qualifications. A single calculation simultaneously serves the dual goals of production safety and policy compliance. When the number of suitable environmental days is insufficient, texture data from the same period of the previous production cycle is automatically retrieved as a replacement baseline. This effectively addresses the problem of lacking normal samples for reference in newly planted crops or years with frequent extreme weather events, ensuring the robustness of the solution under sparse sample conditions.

[0009] S2 includes the following: S201: Capture the continuous data packet sequence uploaded by each facility agriculture entity in a fixed-length time window, number each data packet according to the arrival order, and extract the data attributes of each data packet; the data attributes include arrival time interval, data packet size, and permission type; Map the data attributes of each data packet to preset discrete symbols; Each data packet is mapped to a preset discrete symbol, and a data attribute status symbol sequence is generated according to the data attribute type. S202: For each data attribute state symbol sequence, analyze the transition density and transition rhythm variation coefficient of the data attribute; The specific analysis method for transformation density is as follows: analyze the consistency between two adjacent discrete symbols. If two adjacent discrete symbols are the same discrete symbols, the corresponding transformation mark is recorded as 1; otherwise, it is recorded as 0. The mean of the transformation marks is recorded as the transformation density. The specific analysis method for the coefficient of variation of the transition rhythm is as follows: extract the position index corresponding to all transitions marked as 1; calculate the number of interval steps between adjacent transition events to generate a transition interval sequence, and record the coefficient of variation of the transition interval sequence as the coefficient of variation of the transition rhythm; the coefficient of variation is equal to the ratio of the standard deviation to the mean of the number of interval steps in the transition interval sequence. S203: Analyze the degree of difference in the coefficient of variation of transformation density and transformation rhythm; The difference in transition density is equal to the sum of the absolute values ​​of the differences in transition density between any two data attribute state symbol sequences; the difference in transition rhythm coefficient of variation is equal to the sum of the absolute values ​​of the differences in transition rhythm coefficient of variation between any two data attribute state symbol sequences. The data from the analysis of the differences in the coefficients of variation of conversion density and conversion rate were used as state risk indicators. This solution maps the arrival time interval, packet size, and permission type of each data packet to three independent symbol sequences. It determines data flow anomalies by analyzing whether the transition rhythms of these three sequences are synchronized. This analysis is entirely based on the internal structure of the current window data, requiring no historical baselines or external standards, thus fundamentally avoiding baseline maintenance and conceptual drift issues. The disruption of the three-channel phase consistency is a characteristic shared by two entirely different types of anomalies: data hoarding and sudden data transmission, and deliberate splicing and tampering. Data hoarding and sudden data transmission manifests as abrupt changes in the rhythm of the interval channels but stability in the large and small channels; deliberate splicing and tampering manifests as the disruption of the original correlation between the three channels and the inability to coordinate their rhythms. A single analysis method can capture all anomalies without distinguishing between attack types, significantly improving detection efficiency. The data packet arrival time interval, packet size, and permission type are attributes that the access connector can extract at the network layer, eliminating the need to parse the packet payload content and avoiding data privacy decryption, thus improving the confidentiality of data analysis.

[0010] S3 includes the following: S301: Track the complete service call chain generated by each call to data product services by various facility agriculture entities. The call chain starts from the initial service node and includes downstream service nodes in sequence. The data product services include, but are not limited to, agricultural credit loan assessment services and agricultural product price forecasting services. For each service node in the call chain, record the timestamp of receiving the request and the timestamp of returning the response, and record the difference between the timestamp of returning the response and the timestamp of receiving the request as the processing time of the corresponding service node; S302: For two adjacent service nodes in the call chain, the ratio of the processing time of the downstream service node to the processing time of the upstream service node is recorded as the single-level delay amplification factor; the single-level delay amplification factors from the starting service node to the corresponding service node are multiplied together and recorded as the cumulative delay amplification factor of the corresponding service node. S303: When any single-level latency amplification factor is greater than the amplification factor threshold, mark the corresponding adjacent service nodes as having a processing time amplification anomaly, and record the downstream service node that first appears to have an amplification anomaly as a critical abnormal node; when the cumulative latency amplification factor at the end of the call chain is greater than the amplification factor threshold, determine that the entire call chain has a latency accumulation risk; if the processing time of any service node exceeds the preset timeout threshold, determine that the corresponding call chain has a break risk, and mark the corresponding service node as an abnormal node; Within a preset time window, the total number of all call chains initiated by the facility agriculture entity is counted, as well as the number of call chains that are judged to have accumulated delay risk or abnormal nodes; the ratio of the number of call chains judged to have accumulated delay risk or abnormal nodes to the total number of all call chains is recorded as the abnormal call chain ratio; the smaller value between the product of the abnormal call chain ratio and the abnormal coefficient and 1 is recorded as the service application risk indicator. This solution constructs a latency propagation view on the service call chain by tracking the single-level latency amplification factor between adjacent service layers and the cumulative latency amplification factor across layers. When the processing speed of a certain layer slows down, the effect will accumulate and amplify level by level in the call chain. This solution can track this amplification trajectory and accurately locate the key anomaly node where the first abnormal amplification occurs, providing deterministic guidance for operations and maintenance personnel to quickly locate the source of the fault.

[0011] S4 includes the following: S401: Weighted fusion of production status risk markers, data usage status risk indicators, and service application risk indicators to generate a comprehensive risk value; Monitor changes in the comprehensive risk value, and record the moment when the comprehensive risk value first exceeds the preset warning threshold as the comprehensive risk moment; monitor the business impact indicators of facility agriculture entities, which include at least data sharing volume and service call failure rate. When the data sharing volume decreases by more than a first percentage threshold from the normal baseline, or the service call failure rate increases by more than a second percentage threshold from the normal baseline, record the current moment as the business impact moment; record the difference between the business impact moment and the comprehensive risk moment as the risk lag time. S402: Collect comprehensive risk values ​​at a fixed sampling period. If the comprehensive risk values ​​of the three most recent consecutive sampling periods increase sequentially and the difference between comprehensive risk values ​​is greater than the threshold of the difference between comprehensive risk values, then the comprehensive risk value is determined to be on an upward trend; otherwise, it is determined to be on a stable trend. If the risk delay time is less than the first threshold of delay time and the comprehensive risk value shows an upward trend, a Level 1 rapid response is triggered. The Level 1 rapid response includes at least: downgrading the data product access permissions of the facility agriculture entity from read-write to read-only, and suspending cross-entity data query permissions. If the risk delay time is greater than or equal to the first delay time threshold but less than or equal to the second delay time threshold, and the comprehensive risk value shows an upward trend, a level-two routine response is triggered. The level-two routine response includes at least: pushing a risk notification to the facility agriculture entity and synchronizing the risk information to the facility agriculture entity's service partners. If the risk delay time is greater than the second threshold of delay time or the comprehensive risk value shows a stable trend, a level 3 observation and response is triggered. The level 3 observation and response only writes the current data to the monitoring log and does not perform active intervention. This solution innovatively introduces a time-domain dimension indicator: the time difference between a risk exceeding a threshold and actual business damage. This indicator reflects the speed at which risk propagates to the business: a shorter lag time indicates faster risk propagation and greater destructive power, requiring rapid intervention; a longer lag time indicates that the risk is still in its incubation period and can be observed and addressed. This indicator provides a second dimension of decision-making basis beyond the magnitude of risk for tiered handling, making the handling strategy more refined.

[0012] A data access risk monitoring system for facility agriculture is provided. The system is implemented using the aforementioned data access risk monitoring method for facility agriculture. The system includes a production status analysis module, a data usage status analysis module, a service application status analysis module, and a risk management module. The production status analysis module is used to analyze the cumulative deviation index of environmental signals based on environmental data, analyze the texture uniformity deviation based on canopy images, and determine the production status risk marker by combining the contradiction verification between agricultural operation records and actual control. The data use state analysis module is used to map data packet attributes to discrete symbols to generate a data attribute state symbol sequence, analyze the transformation density and transformation rhythm variation coefficient of each sequence, calculate the cross-attribute difference degree, and analyze state risk indicators. The service application status analysis module is used to track the processing time of each service node in the service call chain, analyze the delay phenomenon of processing time between service nodes, identify abnormal nodes and the risk of accumulated delay, and count the proportion of abnormal call chains and convert them into service application risk indicators. The risk management module is used to generate a comprehensive risk value by integrating production status risk markers, data usage status risk indicators, and service application risk indicators, monitor risk lag time and changing trends, and trigger graded response and management according to thresholds.

[0013] The production status analysis module includes an environmental signal deviation analysis unit, a texture deviation analysis unit, and a production status integration unit. The environmental signal deviation analysis unit is used to acquire environmental data and planting attributes of facility agriculture entities, retrieve suitable upper and lower limits, and determine whether the measured values ​​at each historical moment exceed the limits. The cumulative deviation index of environmental signals is analyzed by an indicator function. The texture deviation analysis unit is used to convert the canopy visible light image into grayscale, extract the grayscale matrix to calculate the texture uniformity, obtain the historical texture uniformity baseline, and analyze the texture uniformity deviation of the grayscale image. The production status integration unit is used to analyze production status risk indicators based on the cumulative deviation index of environmental signals and the relative deviation of grayscale images; and to analyze production status risk markers by combining the contradiction between agricultural operation records and actual environmental control data. The data usage status analysis module includes a data attribute symbol analysis unit, a symbol sequence analysis unit, and a data usage comprehensive unit. The data attribute symbol analysis unit is used to capture data packets in a fixed time window, extract the data attributes of the data packets, map them into discrete symbols according to preset rules, and generate data attribute status symbol sequences according to attribute types. The symbol sequence analysis unit is used to analyze the transition density and transition rhythm variation coefficient of each data attribute state symbol sequence; The data is analyzed using a comprehensive unit to assess the difference in the coefficients of variation of conversion density and conversion rhythm; the analysis of the difference in the coefficients of variation of conversion density and conversion rhythm is performed using state risk indicators. The service application status analysis module includes a service call chain processing unit, a latency analysis unit, and a service application risk analysis unit. The service call chain processing unit is used to record the receiving and returning timestamps for each data product service call chain, from the beginning to each downstream service node, and the difference is used as the processing time of each service node. The latency analysis unit uses the ratio of downstream to upstream processing time of adjacent service nodes as the single-level latency amplification factor; the cumulative latency amplification factor is obtained by multiplying from the start to the corresponding service node. The service application risk analysis unit is used to combine the single-level latency amplification coefficient and the cumulative latency amplification coefficient to judge abnormal situations, count the proportion of abnormal call chains, and analyze service application risk indicators. The risk management module includes a comprehensive risk analysis unit and a tiered response unit; The comprehensive risk analysis unit is used to weight and fuse production status risk markers, data usage status risk indicators, and service application risk indicators to generate a comprehensive risk value; monitor changes in the comprehensive risk value, and record the moment when the comprehensive risk value first exceeds a preset warning threshold as the comprehensive risk moment; monitor the business impact indicators of facility agriculture entities, which include at least data sharing volume and service call failure rate. When the data sharing volume decreases by more than a first percentage threshold from the normal baseline, or the service call failure rate increases by more than a second percentage threshold from the normal baseline, the current moment is recorded as the business impact moment; the difference between the business impact moment and the comprehensive risk moment is recorded as the risk lag time. The graded response unit is used to collect the comprehensive risk value at a fixed sampling period. If the comprehensive risk value of the most recent three consecutive sampling periods increases sequentially and the difference between the comprehensive risk values ​​is greater than the threshold of the difference between the comprehensive risk values, then the comprehensive risk value is determined to be on an upward trend; otherwise, it is determined to be on a stable trend. If the risk delay time is less than the first threshold of delay time and the comprehensive risk value shows an upward trend, a Level 1 rapid response is triggered. The Level 1 rapid response includes at least: downgrading the data product access permissions of the facility agriculture entity from read-write to read-only, and suspending cross-entity data query permissions. If the risk delay time is greater than or equal to the first delay time threshold but less than or equal to the second delay time threshold, and the comprehensive risk value shows an upward trend, a level-two routine response is triggered. The level-two routine response includes at least: pushing a risk notification to the facility agriculture entity and synchronizing the risk information to the facility agriculture entity's service partners. If the risk delay time is greater than the second threshold of delay time or the comprehensive risk value shows a stable trend, a level 3 observation and response is triggered. The level 3 observation and response only writes the current data to the monitoring log and does not perform active intervention. Compared with existing technologies, the advantages of this invention are as follows: Existing agricultural risk monitoring technologies rely solely on sensor data for anomaly detection, while this solution cross-compares actual environmental measurements with administrative records. When environmental signals deviate from the suitable range for the registered varieties for an extended period, it not only reflects production anomalies but may also reveal fraudulent use of entity qualifications. A single calculation simultaneously serves the dual goals of production safety and policy compliance. When the number of suitable environmental days is insufficient, texture data from the same period of the previous production cycle is automatically retrieved as a replacement baseline, effectively solving the problem of no normal samples available for reference in newly planted crops or years with frequent extreme weather events, ensuring the robustness of the solution under sparse sample conditions.

[0014] This solution maps the arrival time interval, packet size, and permission type of each data packet to three independent symbol sequences. It determines data flow anomalies by analyzing whether the transition rhythms of these three sequences are synchronized. This analysis is entirely based on the internal structure of the current window data, requiring no historical baselines or external standards, thus fundamentally avoiding baseline maintenance and conceptual drift issues. The disruption of the three-channel phase consistency is a characteristic shared by two entirely different types of anomalies: data hoarding and sudden data transmission, and deliberate splicing and tampering. Data hoarding and sudden data transmission manifests as abrupt changes in the rhythm of the interval channels but stability in the large and small channels; deliberate splicing and tampering manifests as the disruption of the original correlation between the three channels and the inability to coordinate their rhythms. A single analysis method can capture all anomalies without distinguishing between attack types, significantly improving detection efficiency. The data packet arrival time interval, packet size, and permission type are attributes that the access connector can extract at the network layer, eliminating the need to parse the packet payload content and avoiding data privacy decryption, thus improving the confidentiality of data analysis.

[0015] This solution constructs a latency propagation view on the service call chain by tracking the single-level latency amplification coefficient between adjacent service layers and the cumulative latency amplification coefficient across layers. When the processing speed of a certain layer slows down, the effect will accumulate and amplify level by level in the call chain. This solution can track this amplification trajectory and accurately locate the key anomaly node where the abnormal amplification first occurs, thus improving the analysis efficiency of fault source localization.

[0016] This solution innovatively introduces a time-domain dimension indicator: the time difference between a risk exceeding a threshold and actual business damage. This indicator reflects the speed at which risk propagates to the business: a shorter lag time indicates faster risk propagation and greater destructive power, requiring rapid intervention; a longer lag time indicates that the risk is still in its incubation period and can be observed and addressed. This indicator provides a second dimension of decision-making basis beyond the magnitude of risk for tiered handling, making the handling strategy more refined. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings: Figure 1 This is a schematic flowchart of a data retrieval risk monitoring method applied to facility agriculture according to the present invention; Figure 2 This is a schematic diagram of a data retrieval risk monitoring system for facility agriculture according to the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Please see Figure 1 The present invention provides a technical solution: a data retrieval risk monitoring method applied to facility agriculture, the method comprising the following steps: S1: Based on environmental data analysis of cumulative deviation indicators of environmental signals, based on canopy image analysis of texture uniformity deviation, combined with the verification of contradictions between agricultural operation records and actual control, the risk marker of production status is determined. S1 includes the following: S101: Obtain environmental data within the facility agriculture entity, including at least temperature, humidity, and light intensity; obtain planting attribute data registered by the facility agriculture entity; planting attribute data includes at least the planting variety, planting area, planting date, and expected harvest date; Based on the planted variety, the corresponding upper and lower limits of suitable temperature, humidity, and light intensity are retrieved from the preset knowledge base. For each moment within the historical time window, determine whether the measured values ​​of temperature, humidity, and light intensity fall within their respective suitable upper and lower limits; if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 0, and if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 1; the sum of the indicator function outputs of each environmental data at each moment within the historical time window is recorded as the environmental signal cumulative deviation index. S102: Obtain visible light images of crop canopies within the facility of the main agricultural facility, convert the visible light images of the canopies into grayscale images, and analyze the texture uniformity; Example 1: In this example, the grayscale matrix of the grayscale image in four directions (0°, 45°, 90° and 135°) is extracted, and the arithmetic mean of the contrast values ​​in the four directions is used as the texture uniformity. Extract the historical texture uniformity baseline of the corresponding facility agriculture entity during normal operation. The historical texture uniformity baseline is the median value of texture uniformity of all images on environmentally suitable days in the past several days for the facility agriculture entity. The cumulative deviation index of environmental signal on environmentally suitable days is less than the cumulative deviation threshold of environmental signal. If the number of environmentally suitable days is less than the threshold of suitable days, the historical texture uniformity baseline is equal to the historical median value of texture uniformity of the facility agriculture entity in the same period of the previous production cycle. Analyze the texture uniformity deviation of a grayscale image; the texture uniformity deviation is equal to the ratio of the absolute value of the difference between the current texture uniformity and the historical texture uniformity baseline to the historical texture uniformity baseline. S103: Analysis of production status risk indicators based on cumulative deviation index of environmental signals and relative deviation of grayscale images; Example 2: In this example, the production status risk index P = A × min (τ / τ) max ,1)+(1-A)×min(△U,1;where τ represents the cumulative deviation index of the environmental signal, τ max The cumulative deviation threshold of the environmental signal is represented by ΔU; the deviation of texture uniformity is represented by ΔU; and the weighting coefficient of the preset environmental signal is represented by A. In actual deployment, the weighting coefficient of the preset environmental signal is adjusted according to the sensitivity of different crops. For leafy vegetables that are sensitive to temperature and humidity, the weighting coefficient of the preset environmental signal is increased to 0.60, and for fruit vegetables, the weighting coefficient of the preset environmental signal is decreased to 0.50. The cumulative deviation threshold of the environmental signal is dynamically adjusted according to the crop growth stage. It is shortened to 2 hours during the flowering and fruiting period and relaxed to 4 hours during the vegetative growth period. If the production status risk index is greater than the production status risk threshold, and there is a contradiction between the agricultural operation records declared by the corresponding facility agriculture entity in the public data within the historical time window and the actual environmental control data, the production status risk of the corresponding facility agriculture entity will be marked as 1; otherwise, the production status risk of the facility agriculture entity will be marked as 0. The contradictions include at least the following: the declaration has been made that ventilation and cooling operations have been implemented, but the actual temperature has remained above the upper limit of the suitable temperature for several consecutive hours after the declaration, and the temperature difference exceeds the temperature difference threshold; or the declaration has been made that humidification operations have been implemented, but the actual humidity has remained below the lower limit of the suitable humidity.

[0020] S2: Map data packet attributes to discrete symbols to generate a sequence of data attribute state symbols, analyze the transformation density and transformation rhythm variation coefficient of each sequence, calculate the cross-attribute difference, and analyze the state risk index; S2 includes the following: S201: Capture the continuous data packet sequence uploaded by each facility agriculture entity in a fixed-length time window, number each data packet according to the arrival order, and extract the data attributes of each data packet; the data attributes include arrival time interval, data packet size, and permission type; Map the data attributes of each data packet to preset discrete symbols; Example 3: In this example, the mapping of arrival time intervals is as follows: when the arrival time interval is less than 1 second, it is mapped to symbol a; when the arrival time interval is greater than or equal to 1 second and less than or equal to 5 seconds, it is mapped to symbol b; when the arrival time interval is greater than 5 seconds, it is mapped to symbol c. For the mapping of packet size: when the packet size is less than 10 KB, it is mapped to symbol x; when the packet size is greater than or equal to 10 KB and less than or equal to 100 KB, it is mapped to symbol y; when the packet size is greater than 100 KB, it is mapped to symbol z. For permission type mapping: when the permission type is read-only query, it is mapped to symbol α; when the permission type is data upload, it is mapped to symbol β; when the permission type is command issuance, it is mapped to symbol γ; when the permission type is system configuration, it is mapped to symbol δ. Each data packet is mapped to a preset discrete symbol, and a data attribute status symbol sequence is generated according to the data attribute type. S202: For each data attribute state symbol sequence, analyze the transition density and transition rhythm variation coefficient of the data attribute; The specific analysis method for transformation density is as follows: analyze the consistency between two adjacent discrete symbols. If two adjacent discrete symbols are the same discrete symbols, the corresponding transformation mark is recorded as 1; otherwise, it is recorded as 0. The mean of the transformation marks is recorded as the transformation density. The specific analysis method for the coefficient of variation of the transition rhythm is as follows: extract the position index corresponding to all transitions marked as 1; calculate the number of interval steps between adjacent transition events to generate a transition interval sequence, and record the coefficient of variation of the transition interval sequence as the coefficient of variation of the transition rhythm; the coefficient of variation is equal to the ratio of the standard deviation of the number of interval steps in the transition interval sequence to the mean. S203: Analyze the degree of difference in the coefficient of variation of transformation density and transformation rhythm; The difference in transition density is equal to the sum of the absolute values ​​of the differences in transition density between any two data attribute state symbol sequences; the difference in transition rhythm coefficient of variation is equal to the sum of the absolute values ​​of the differences in transition rhythm coefficient of variation between any two data attribute state symbol sequences. The data from the analysis of the differences in the coefficients of variation of conversion density and conversion rate were used as state risk indicators. Example 4: In this example, the data usage status risk index Q = min((B+C) / (θ)). B +θ c ),1); where B represents the degree of difference in transformation density, C represents the degree of difference in the coefficient of variation of transformation rhythm, θ B θ represents the threshold of difference in conversion density. c This represents the threshold for the degree of difference in the coefficient of variation of the conversion rhythm; S3: Track the processing time of each service node in the service call chain, analyze the latency phenomenon between service nodes, identify abnormal nodes and latency accumulation risks, and statistically analyze the proportion of abnormal call chains and convert them into service application risk indicators. S3 includes the following: S301: Track the complete service call chain generated by each call to data product services by various facility agriculture entities. The call chain starts from the initial service node and includes downstream service nodes in sequence. Data product services include, but are not limited to, agricultural credit loan assessment services and agricultural product price forecasting services. For each service node in the call chain, record the timestamp of receiving the request and the timestamp of returning the response, and record the difference between the timestamp of returning the response and the timestamp of receiving the request as the processing time of the corresponding service node; S302: For two adjacent service nodes in the call chain, the ratio of the processing time of the downstream service node to the processing time of the upstream service node is recorded as the single-level delay amplification factor; the single-level delay amplification factors from the starting service node to the corresponding service node are multiplied together and recorded as the cumulative delay amplification factor of the corresponding service node. S303: When any single-level latency amplification factor is greater than the amplification factor threshold, mark the corresponding adjacent service nodes as having a processing time amplification anomaly, and record the downstream service node that first appears to have an amplification anomaly as a critical abnormal node; when the cumulative latency amplification factor at the end of the call chain is greater than the amplification factor threshold, determine that the entire call chain has a latency accumulation risk; if the processing time of any service node exceeds the preset timeout threshold, determine that the corresponding call chain has a break risk, and mark the corresponding service node as an abnormal node; Within a preset time window, the total number of all call chains initiated by the facility agriculture entity is counted, as well as the number of call chains that are judged to have accumulated delay risk or abnormal nodes; the ratio of the number of call chains judged to have accumulated delay risk or abnormal nodes to the total number of all call chains is recorded as the abnormal call chain ratio; the smaller value between the product of the abnormal call chain ratio and the abnormal coefficient and 1 is recorded as the service application risk indicator. S4: Generate a comprehensive risk value by combining production status risk markers, data usage status risk indicators, and service application risk indicators; monitor risk lag time and trends; and trigger tiered response and handling based on the trend of the comprehensive risk value and the risk lag time threshold.

[0021] S4 includes the following: S401: Weighted fusion of production status risk markers, data usage status risk indicators, and service application risk indicators to generate a comprehensive risk value; Monitor changes in the comprehensive risk value and record the moment when the comprehensive risk value first exceeds the preset warning threshold as the comprehensive risk moment; monitor the business impact indicators of facility agriculture entities, which include at least data sharing volume and service call failure rate. When the data sharing volume decreases by more than the first percentage threshold from the normal baseline, or the service call failure rate increases by more than the second percentage threshold from the normal baseline, record the current moment as the business impact moment; record the difference between the business impact moment and the comprehensive risk moment as the risk lag time. S402: Collect comprehensive risk values ​​at a fixed sampling period. If the comprehensive risk values ​​of the three most recent consecutive sampling periods increase sequentially and the difference between comprehensive risk values ​​is greater than the threshold of the difference between comprehensive risk values, then the comprehensive risk value is determined to be on an upward trend; otherwise, it is determined to be on a stable trend. If the risk delay time is less than the first threshold of delay time and the comprehensive risk value shows an upward trend, a Level 1 rapid response is triggered. The Level 1 rapid response includes at least: downgrading the data product access permissions of the facility agriculture entity from read-write to read-only, and suspending cross-entity data query permissions. If the risk delay time is greater than or equal to the first delay time threshold but less than or equal to the second delay time threshold, and the overall risk value shows an upward trend, a level-two routine response is triggered. The level-two routine response includes at least: sending a risk notification to the facility agriculture entity and synchronizing the risk information to the facility agriculture entity's service partners. If the risk delay time is greater than the second threshold of delay time or the comprehensive risk value shows a stable trend, a level 3 observation and response is triggered. The level 3 observation and response only writes the current data to the monitoring log and does not perform active intervention. Please see Figure 2 The present invention provides a technical solution: a data access risk monitoring system for facility agriculture, the system comprising a production status analysis module, a data usage status analysis module, a service application status analysis module, and a risk management module; The production status analysis module is used to analyze the cumulative deviation index of environmental signals based on environmental data, analyze the deviation of texture uniformity based on canopy images, and determine the risk marker of production status by combining the contradiction verification between agricultural operation records and actual control. The data usage status analysis module is used to map data packet attributes to discrete symbols to generate a data attribute status symbol sequence, analyze the transition density and transition rhythm variation coefficient of each sequence, calculate cross-attribute variability, and analyze status risk indicators. The service application status analysis module is used to track the processing time of each service node in the service call chain, analyze the latency phenomenon between service nodes, identify abnormal nodes and latency accumulation risks, and statistically analyze the proportion of abnormal call chains and convert them into service application risk indicators. The risk management module is used to generate a comprehensive risk value by integrating production status risk markers, data usage status risk indicators, and service application risk indicators, monitor risk lag time and changing trends, and trigger graded response and management according to thresholds.

[0022] The production status analysis module includes an environmental signal deviation analysis unit, a texture deviation analysis unit, and a production status integration unit; The environmental signal deviation analysis unit is used to acquire environmental data and planting attributes of facility agriculture entities, retrieve suitable upper and lower limits, and determine whether the measured values ​​at each historical time point exceed the limits. The cumulative deviation index of environmental signals is analyzed by using an indicator function. The texture deviation analysis unit is used to convert the canopy visible light image to grayscale, extract the grayscale matrix to calculate the texture uniformity, obtain the historical texture uniformity baseline, and analyze the texture uniformity deviation of the grayscale image. The production status integration unit is used to analyze production status risk indicators based on the cumulative deviation index of environmental signals and the relative deviation of grayscale images; and to analyze production status risk markers by combining the contradiction between agricultural operation records and actual environmental control data. The data usage status analysis module includes a data attribute symbol analysis unit, a symbol sequence analysis unit, and a data usage comprehensive unit; The data attribute symbol analysis unit is used to capture data packets within a fixed time window, extract the data attributes of the data packets, map them into discrete symbols according to preset rules, and generate data attribute status symbol sequences according to attribute types. The symbol sequence analysis unit is used to analyze the transition density and transition rhythm variation coefficient of each data attribute state symbol sequence; The data uses a composite unit to analyze the difference in the coefficients of variation of transformation density and transformation rhythm; the data used in the analysis of the difference in the coefficients of variation of transformation density and transformation rhythm are based on state risk indicators. The service application status analysis module includes a service call chain processing unit, a latency analysis unit, and a service application risk analysis unit. The service call chain processing unit is used to record the receiving and returning timestamps for each data product service call chain, from the beginning to each downstream service node, and the difference is used as the processing time of each service node. The delay analysis unit uses the ratio of downstream to upstream processing time of adjacent service nodes as the single-level delay amplification factor; the cumulative delay amplification factor is obtained by multiplying from the start to the corresponding service node. The service application risk analysis unit is used to combine the single-level latency amplification coefficient and the cumulative latency amplification coefficient to judge abnormal situations, count the proportion of abnormal call chains, and analyze service application risk indicators. The risk management module includes a comprehensive risk analysis unit and a tiered response unit; The comprehensive risk analysis unit is used to weight and fuse production status risk markers, data usage status risk indicators, and service application risk indicators to generate a comprehensive risk value; monitor changes in the comprehensive risk value, and record the moment when the comprehensive risk value first exceeds the preset warning threshold as the comprehensive risk moment; monitor the business impact indicators of facility agriculture entities, which include at least data sharing volume and service call failure rate. When the data sharing volume decreases by more than the first percentage threshold from the normal baseline, or the service call failure rate increases by more than the second percentage threshold from the normal baseline, the current moment is recorded as the business impact moment; the difference between the business impact moment and the comprehensive risk moment is recorded as the risk lag time. The graded response unit is used to collect the comprehensive risk value at a fixed sampling period. If the comprehensive risk value of the three most recent consecutive sampling periods increases sequentially and the difference between the comprehensive risk values ​​is greater than the threshold of the comprehensive risk value difference, then the comprehensive risk value is determined to be on an upward trend; otherwise, it is determined to be on a stable trend. If the risk delay time is less than the first threshold of delay time and the comprehensive risk value shows an upward trend, a Level 1 rapid response is triggered. The Level 1 rapid response includes at least: downgrading the data product access permissions of the facility agriculture entity from read-write to read-only, and suspending cross-entity data query permissions. If the risk delay time is greater than or equal to the first delay time threshold but less than or equal to the second delay time threshold, and the overall risk value shows an upward trend, a level-two routine response is triggered. The level-two routine response includes at least: sending a risk notification to the facility agriculture entity and synchronizing the risk information to the facility agriculture entity's service partners. If the risk delay time is greater than the second threshold of delay time or the comprehensive risk value shows a stable trend, a level 3 observation and response is triggered. The level 3 observation and response only writes the current data to the monitoring log and does not perform active intervention. It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "include," "contain," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0023] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A data retrieval risk monitoring method applied to facility agriculture, characterized in that, The method includes the following steps: S1: Based on environmental data analysis of cumulative deviation indicators of environmental signals, based on canopy image analysis of texture uniformity deviation, combined with the verification of contradictions between agricultural operation records and actual control, the risk marker of production status is determined. S2: Map data packet attributes to discrete symbols to generate a sequence of data attribute state symbols, analyze the transformation density and transformation rhythm variation coefficient of each sequence, calculate the cross-attribute difference, and analyze the state risk index; S3: Track the processing time of each service node in the service call chain, analyze the latency phenomenon between service nodes, identify abnormal nodes and latency accumulation risks, and statistically analyze the proportion of abnormal call chains and convert them into service application risk indicators. S4: Generate a comprehensive risk value by combining production status risk markers, data usage status risk indicators, and service application risk indicators; monitor risk lag time and trends; and trigger tiered response and handling based on thresholds.

2. The data retrieval risk monitoring method applied to facility agriculture according to claim 1, characterized in that: S1 includes the following: S101: Obtain environmental data inside the facility agriculture entity, the environmental data including at least temperature, humidity and light intensity; obtain planting attribute data registered by the facility agriculture entity; the planting attribute data including at least planting variety, planting area, planting date and expected harvest date; Based on the planted variety, retrieve the corresponding suitable upper and lower limits of temperature, humidity, and light from the preset knowledge base; For each moment within the historical time window, determine whether the measured values ​​of temperature, humidity, and light intensity fall within their respective suitable upper and lower limits; if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 0, and if the measured values ​​fall within the corresponding suitable upper and lower limits, the indicator function output is 1; the sum of the indicator function outputs of each environmental data at each moment within the historical time window is recorded as the environmental signal cumulative deviation index. S102: Obtain visible light images of crop canopies within the facility of the main agricultural facility, convert the visible light images of the canopies into grayscale images, and analyze the texture uniformity; Extract the historical texture uniformity baseline of the corresponding facility agriculture entity during normal operation. The historical texture uniformity baseline is the median value of the texture uniformity of all images on environmentally suitable days in the past several days of the facility agriculture entity. The cumulative deviation index of the environmental signal on the environmentally suitable days is less than the cumulative deviation threshold of the environmental signal. If the number of environmentally suitable days is less than the threshold of suitable days, the historical texture uniformity baseline is equal to the historical median value of the texture uniformity of the facility agriculture entity in the same period of the previous production cycle. Analyze the texture uniformity deviation of the grayscale image; the texture uniformity deviation is equal to the ratio of the absolute value of the difference between the current texture uniformity and the historical texture uniformity baseline to the historical texture uniformity baseline. S103: Analysis of production status risk indicators based on cumulative deviation index of environmental signals and relative deviation of grayscale images; If the production status risk index is greater than the production status risk threshold, and there is a contradiction between the agricultural operation records declared by the corresponding facility agriculture entity in the public data within the historical time window and the actual environmental control data, the production status risk of the corresponding facility agriculture entity will be marked as 1; otherwise, the production status risk of the facility agriculture entity will be marked as 0.

3. The data retrieval risk monitoring method applied to facility agriculture according to claim 2, characterized in that: The contradictions include at least the following: the declaration has been made that ventilation and cooling operations have been implemented, but the actual temperature remains above the upper limit of suitable temperature for several consecutive hours after the declaration, and the temperature difference exceeds the temperature difference threshold; or the declaration has been made that humidification operations have been implemented, but the actual humidity remains below the lower limit of suitable humidity.

4. The data retrieval risk monitoring method applied to facility agriculture according to claim 3, characterized in that: S2 includes the following: S201: Capture the continuous data packet sequence uploaded by each facility agriculture entity in a fixed-length time window, number each data packet according to the arrival order, and extract the data attributes of each data packet; the data attributes include arrival time interval, data packet size, and permission type; Map the data attributes of each data packet to preset discrete symbols; Each data packet is mapped to a preset discrete symbol, and a data attribute status symbol sequence is generated according to the data attribute type. S202: For each data attribute state symbol sequence, analyze the transition density and transition rhythm variation coefficient of the data attribute; S203: Analyze the degree of difference in the coefficient of variation of transformation density and transformation rhythm; The difference in transition density is equal to the sum of the absolute values ​​of the differences in transition density between any two data attribute state symbol sequences; the difference in transition rhythm coefficient of variation is equal to the sum of the absolute values ​​of the differences in transition rhythm coefficient of variation between any two data attribute state symbol sequences. The data used in the analysis of the differences in the coefficients of variation of conversion density and conversion rhythm were analyzed using state risk indicators.

5. The data retrieval risk monitoring method applied to facility agriculture according to claim 4, characterized in that: The specific analysis method for transformation density is as follows: analyze the consistency between two adjacent discrete symbols. If two adjacent discrete symbols are the same discrete symbols, the corresponding transformation mark is recorded as 1; otherwise, it is recorded as 0. The mean of the transformation marks is recorded as the transformation density. The specific analysis method for the coefficient of variation of the transition rhythm is as follows: extract the position index corresponding to all transitions marked as 1; calculate the number of interval steps between adjacent transition events, generate a transition interval sequence, and record the coefficient of variation of the transition interval sequence as the coefficient of variation of the transition rhythm.

6. The data retrieval risk monitoring method applied to facility agriculture according to claim 5, characterized in that: S3 includes the following: S301: Track the complete service call chain generated by each call to data product services by various facility agriculture entities. The call chain starts from the initial service node and includes downstream service nodes in sequence. The data product services include, but are not limited to, agricultural credit loan assessment services and agricultural product price forecasting services. For each service node in the call chain, record the timestamp of receiving the request and the timestamp of returning the response, and record the difference between the timestamp of returning the response and the timestamp of receiving the request as the processing time of the corresponding service node; S302: For two adjacent service nodes in the call chain, the ratio of the processing time of the downstream service node to the processing time of the upstream service node is recorded as the single-level delay amplification factor. The cumulative latency amplification factor of the corresponding service node is calculated by multiplying all single-level latency amplification factors from the starting service node to the corresponding service node. S303: When any single-level latency amplification factor is greater than the amplification factor threshold, mark the corresponding adjacent service nodes as having a processing time amplification anomaly, and record the downstream service node that first appears to have an amplification anomaly as a critical abnormal node; when the cumulative latency amplification factor at the end of the call chain is greater than the amplification factor threshold, determine that the entire call chain has a latency accumulation risk; if the processing time of any service node exceeds the preset timeout threshold, determine that the corresponding call chain has a break risk, and mark the corresponding service node as an abnormal node; Within a preset time window, the total number of all call chains initiated by the facility agriculture entity is counted, as well as the number of call chains that are judged to have accumulated delay risk or abnormal nodes. The ratio of the number of call chains judged to have accumulated delay risk or abnormal nodes to the total number of all call chains is recorded as the abnormal call chain ratio. The smaller value between the product of the abnormal call chain ratio and the abnormal coefficient and 1 is recorded as the service application risk indicator.

7. The data retrieval risk monitoring method applied to facility agriculture according to claim 6, characterized in that: S4 includes the following: S401: Weighted fusion of production status risk markers, data usage status risk indicators, and service application risk indicators to generate a comprehensive risk value; Monitor changes in the comprehensive risk value, and record the moment when the comprehensive risk value first exceeds the preset warning threshold as the comprehensive risk moment; monitor the business impact indicators of facility agriculture entities, which include at least data sharing volume and service call failure rate. When the data sharing volume decreases by more than a first percentage threshold from the normal baseline, or the service call failure rate increases by more than a second percentage threshold from the normal baseline, record the current moment as the business impact moment; record the difference between the business impact moment and the comprehensive risk moment as the risk lag time. S402: Collect comprehensive risk values ​​at a fixed sampling period. If the comprehensive risk values ​​of the three most recent consecutive sampling periods increase sequentially and the difference between comprehensive risk values ​​is greater than the threshold of the difference between comprehensive risk values, then the comprehensive risk value is determined to be on an upward trend; otherwise, it is determined to be on a stable trend. Tiered response measures are triggered based on the trend of the comprehensive risk value and the risk lag time threshold.

8. The data retrieval risk monitoring method applied to facility agriculture according to claim 7, characterized in that: The specific details of the tiered response and handling triggered by the trend and risk lag time threshold based on the comprehensive risk value are as follows: If the risk delay time is less than the first threshold of delay time and the comprehensive risk value shows an upward trend, a Level 1 rapid response is triggered. The Level 1 rapid response includes at least: downgrading the data product access permissions of the facility agriculture entity from read-write to read-only, and suspending cross-entity data query permissions. If the risk delay time is greater than or equal to the first delay time threshold but less than or equal to the second delay time threshold, and the comprehensive risk value shows an upward trend, a level-two routine response is triggered. The level-two routine response includes at least: pushing a risk notification to the facility agriculture entity and synchronizing the risk information to the facility agriculture entity's service partners. If the risk delay time exceeds the second threshold of delay time or the comprehensive risk value shows a stable trend, a level 3 observation and response is triggered. The level 3 observation and response only writes the current data to the monitoring log and does not perform active intervention.

9. A data retrieval risk monitoring system for facility agriculture, wherein the system is implemented using the data retrieval risk monitoring method for facility agriculture as described in any one of claims 1-8, characterized in that, The system includes a production status analysis module, a data usage status analysis module, a service application status analysis module, and a risk management module; The production status analysis module is used to analyze the cumulative deviation index of environmental signals based on environmental data, analyze the texture uniformity deviation based on canopy images, and determine the production status risk marker by combining the contradiction verification between agricultural operation records and actual control. The data use state analysis module is used to map data packet attributes to discrete symbols to generate a data attribute state symbol sequence, analyze the transformation density and transformation rhythm variation coefficient of each sequence, calculate the cross-attribute difference degree, and analyze state risk indicators. The service application status analysis module is used to track the processing time of each service node in the service call chain, analyze the delay phenomenon of processing time between service nodes, identify abnormal nodes and the risk of accumulated delay, and count the proportion of abnormal call chains and convert them into service application risk indicators. The risk management module is used to generate a comprehensive risk value by integrating production status risk markers, data usage status risk indicators, and service application risk indicators, monitor risk lag time and changing trends, and trigger graded response and management according to thresholds.

Images