Agricultural scene-oriented intelligent safe power utilization management method and system

Abstract

The invention discloses an agricultural scene-oriented intelligent safe power utilization management method and system, relates to the technical field of intelligent power distribution facilities, and aims to solve the problems of single leakage protection, weak remote monitoring, complex man-machine interaction, low data credibility, lack of environment linkage and the like in a traditional power supply mode in scattered agricultural irrigation scenes. According to the method, a scene feature map is constructed by fusing equipment types, geographical environments and meteorological parameters, and the similarity with historical dangerous scenes is calculated; performing multi-dimensional clustering on the electric leakage signals based on scene features, and distinguishing electric leakage modes of 12 types of agricultural equipment; dynamically generating electric leakage, overcurrent and overvoltage protection parameters according to the electric leakage mode and the environmental parameters; historical data, weather forecast and equipment aging parameter prediction risk indexes are fused, and protection parameters are adjusted in a linkage mode; and the whole-process data is subjected to chain storage evidence, and credit points are generated based on the chain data to excite standard power consumption.

Description

Technical Field

[0001] This invention relates to the field of intelligent power distribution facilities technology, specifically to an intelligent and safe power management method and system for agricultural scenarios. Background Technology

[0002] In scattered agricultural irrigation scenarios, especially in areas dominated by dispersed flower and tree planting, traditional power supply methods have significant drawbacks: leakage protection mechanisms are simplistic and unable to identify hidden faults in complex agricultural environments; remote monitoring capabilities are weak and cannot meet the distributed power management needs of large-scale farmland; human-computer interaction methods are outdated, making farmers prone to misoperation due to complex interfaces; power data reliability is low, lacking anti-tampering mechanisms, affecting electricity billing and equipment maintenance decisions; and there is a lack of linkage mechanisms with agricultural environmental parameters, making it impossible to prevent power accidents caused by weather changes. Traditional agricultural power protection devices often use fixed threshold designs, and parameters such as leakage current and overcurrent protection curves cannot be adjusted according to environmental changes. In humid weather, soil conductivity increases, significantly amplifying the hazard of the same leakage current, but fixed thresholds are difficult to adapt, often leading to malfunctions. During irrigation, frequent power outages or delayed protection due to increased humidity, and in high-temperature environments, decreased line insulation performance without timely protection triggering, seriously affecting irrigation efficiency and posing safety hazards. Meanwhile, the leakage characteristics of different agricultural equipment vary significantly. Traditional technologies lack the ability to accurately identify equipment types, easily misjudging normal starting current fluctuations as leakage, or failing to identify leakage characteristics drift caused by aging in older equipment, further exacerbating the problem of protection misjudgment. Existing solutions lack sufficient awareness of scenario risks and cannot predict potential risks by combining geographical environment, meteorological conditions, and historical accident data. They are mostly passive protections after accidents occur, making it difficult to prevent them in advance. In terms of data management, there is a lack of reliable evidence storage mechanisms for electricity parameters and protection action records. Data tampering disputes are prone to occur during electricity bill disputes and fault tracing, and there is a lack of effective incentives and constraints on farmers' electricity use behavior, making it difficult to regulate electricity use management. In terms of human-computer interaction, traditional devices are complex to operate and rely heavily on professional knowledge, while farmers have low proficiency in operating electrical equipment, often causing safety problems due to misoperation. There is an urgent need for intelligent interaction solutions adapted to the usage habits of agricultural scenarios.

[0003] Therefore, building an intelligent and safe electricity system that can dynamically adapt to the environment, accurately identify risks, proactively prevent and control faults, and is easy to manage has become the key to solving the pain points of agricultural electricity safety. Summary of the Invention

[0004] The purpose of this invention is to provide an intelligent and safe electricity management method and system for agricultural scenarios. The system collects equipment, geographical environment and meteorological parameters through a multi-source data sensing module, calculates the similarity with historical dangerous scenarios through a scenario feature map construction module, distinguishes leakage modes through a leakage mode identification module, generates protection parameters adapted to the environment through a dynamic protection control module, predicts risks in advance and adjusts parameters through a risk prediction and linkage module, stores all data on the blockchain and generates credit scores through a blockchain notarization and credit management module, and verifies user identity and supports operation and alarm functions through a user module. All modules work together to achieve safe electricity management in agriculture.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A smart and safe electricity management method for agricultural scenarios, characterized by the following steps:

[0007] S1: Collect data, integrate equipment type, geographical environment, and meteorological parameters into a scene feature map, and determine the similarity between the current scene and historical dangerous scenes based on the scene feature map;

[0008] S2: Based on the scene characteristics of S1, the leakage current signal is clustered in multiple dimensions to distinguish the leakage current modes of different agricultural equipment and prevent protection malfunctions.

[0009] S3: Based on the leakage mode of S2 and the environmental parameters of S1, dynamic protection parameters are generated to replace the fixed threshold and adapt to protection in complex agricultural environments.

[0010] S4: Integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 to achieve proactive prevention and control;

[0011] S5: Upload all process data to the blockchain for storage and form an electricity data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, points are generated by combining the duration of compliant electricity use and energy saving rate. Points can be redeemed for maintenance services to incentivize standardized electricity use.

[0012] S1 collects data including device type parameters, including the rated power P of the connected device. N Rated current I N Equipment type label k; geographical environmental parameters include soil moisture S at the equipment location, obtained through installed sensors; real-time meteorological parameters include ambient temperature T, humidity H, and wind speed V at the equipment location, obtained by accessing the regional meteorological data platform.

[0013] S1 determines the similarity between the current scene and historical hazardous scenes based on the scene feature map. This is achieved by fusing equipment type, geographical environment, and meteorological parameters into a scene feature map. Specifically, all parameters are normalized using the formula P.N ′=P N / max(P N Similarly, for other parameters, ensure that all matrix elements are in the interval [0,1]. Construct an eigenvector G for the standardized parameters and assign weights according to their impact on security; the formula is G = [P]. N ′·0.2,I N ′·0.2, k′·0.1, S′·0.2, T′·0.15, H′·0.15];

[0014] Construct a database of historical hazardous scenarios, with each scenario corresponding to a feature vector G. risk,m , where m is the scene number;

[0015] Compare the current scenario G with historical hazardous scenarios G risk,m The similarity γ is calculated using the following formula:

[0016]

[0017] In the formula, γ represents the maximum similarity between the current scene and historical dangerous scenes, and G is the feature matrix constructed using standardized parameters. risk,m Let m be the feature vector of the m-th scene in the historical hazardous scene database, where m is the number of the historical hazardous scene.

[0018] Where γ∈[0,1], γ>0.7 is marked as a high-risk scenario, triggering enhanced monitoring and increasing the sampling frequency.

[0019] Based on the scene features collected by S1, S2 performs multi-dimensional clustering of leakage current signals to distinguish leakage current modes of various agricultural devices, thereby solving the problem of malfunction in traditional protection systems. Specifically, this is achieved by analyzing the leakage current i... leak (t) Perform time-frequency domain analysis to extract the feature vector F, the formula of which is as follows:

[0020]

[0021] In the formula, F is the characteristic vector of leakage current, i leak (t) represents the leakage current, max(i) leak f is the maximum value of the leakage current. dom The main frequency of the leakage signal, skew(i leak ) is the skewness coefficient of the leakage current signal, reflecting the signal asymmetry, and t is the time variable;

[0022] Scene-based clustering correction, combined with the scene similarity γ of S1, adjusts the cluster center μ. k After correction, the corrected cluster centers μ are obtained. k ′, its formula is as follows:

[0023] μk ′=μ k ·(1+0.3γ)+F scene ·0.2γ

[0024] In the formula, μ k ′ represents the corrected cluster center of the k-th class, μ k For the k-th cluster center before correction, F scene Device characteristics in the current scenario;

[0025] Current leakage current characteristic F and a certain cluster center μ k If the Euclidean distance d < 0.15, then the device is considered to be leaking current.

[0026] S3 dynamically generates protection parameters based on the leakage mode of S2 and the environmental parameters of S1, replacing fixed thresholds. These parameters include adjustments to the leakage current, overcurrent protection curve, and overvoltage protection threshold. Specifically, based on the equipment type (kW) determined by S2 and the humidity (H) of S1, the leakage current (I) is dynamically adjusted. leak,act The formula is as follows:

[0027] I leak,act =30·(1+0.01kw)·exp(-0.01H)

[0028] In the formula, I leak,act The leakage current is dynamically adjusted. kW is the power of the equipment type k, and H is the ambient humidity. 30mA is the standard safety threshold. The higher the equipment power, the larger the kW, and the lower the humidity, the more relaxed the operating current should be to avoid false triggering.

[0029] Using I 2 The t-characteristic curve is used for overcurrent protection curve calibration, and dynamically corrected according to ambient temperature T and soil moisture S.

[0030]

[0031] In the formula, I 2 t: Parameters of the overcurrent protection characteristic curve, reflecting the product of the square of the current and time; T: Ambient temperature; kx: Correction coefficient corresponding to equipment type k; S: Soil moisture; I: ... N The current is the rated current of the equipment; 8 seconds is the reference time. The curve is steeper when the temperature rises or the soil is wet.

[0032] The overvoltage protection threshold adjustment is combined with the scenario similarity γ of S1 to adjust the overvoltage action voltage U. over,act :

[0033] U over,act =250·(1-0.05γ)

[0034] In the formula, U over,actThe overvoltage action voltage is dynamically adjusted, γ is the similarity of historical dangerous scenarios, and 250 is the standard voltage threshold. The threshold is lowered in high-risk scenarios to prevent voltage anomalies in advance.

[0035] S4 integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 for proactive prevention and control; it establishes a risk index prediction model based on the probability of precipitation P in the next 2 hours. r Equipment aging coefficient A = 1 - exp(-0.0003t) total Historical risk value R hist Calculate the risk index R:

[0036] R = 0.4P r +0.3A+0.2R hist +0.1γ

[0037] In the formula, R is the comprehensive risk index, and P r Let A be the probability of precipitation in the next 2 hours, and let A be the equipment aging coefficient. t total R represents the cumulative operating time of the equipment. hist γ represents the historical risk value, and γ represents the similarity of historical dangerous scenarios.

[0038] Adjust the protection parameters of S3 according to R:

[0039]

[0040] Proactive prevention and control can be achieved by adjusting parameters.

[0041] S5 uploads all process data to the blockchain for storage, forming an electricity consumption data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, a block is generated every 15 minutes, containing the following data:

[0042] Block = {H prev ,E,G,S prot ,R,TimeStamp,UserID}

[0043] In the formula, Block: a data block in the blockchain, H prev Let E be the hash of the preceding block, E be the electricity consumption during the time period, G be the scene feature matrix of S1, and S be the data source of the preceding block. prot This section records the protection actions of S3, R represents the risk index of S4, TimeStamp is the timestamp, and UserID is the user identifier. The SHA-256 hash algorithm ensures immutability and supports electricity billing and fault tracing.

[0044] S5 generates points based on compliant electricity usage duration and energy saving rate. These points can be redeemed for maintenance services to incentivize compliant electricity use. The credit score calculation is optimized by adding a scenario risk coefficient γ to the existing formula to encourage compliant electricity use in high-risk scenarios.

[0045] P = 50·t comply ·(2-η)·exp(-0.1N fault )·(1+0.5γ)

[0046] In the formula, P is the credit score, and t comply For compliant electricity usage duration, η represents the energy saving rate, which is the actual electricity consumption divided by the theoretical demand. fault γ represents the number of failures, and γ represents the similarity of historical hazardous scenarios. Points can be redeemed for priority electricity access, priority power supply during peak irrigation periods, or free equipment maintenance services.

[0047] A smart and safe electricity management system for agricultural scenarios, implementing the method described in any one of claims 1-8, is characterized by integrating a distributed distribution box cluster, a multi-functional safety socket module, and a visual human-machine interface. It further includes a multi-source data sensing module, a scene feature map construction module, a leakage current pattern recognition module, a dynamic protection control module, a risk prediction and linkage module, and a blockchain evidence storage and credit management module. Specifically, the multi-source data sensing module collects equipment, geographical environment, and meteorological parameters, supporting the scene feature map construction module in completing parameter standardization, feature matrix construction, and similarity calculation with historical hazardous scenarios. The leakage current pattern recognition module clusters leakage current signals in multiple dimensions based on scene features to distinguish leakage current patterns. The dynamic protection control module dynamically generates protection parameters based on leakage current patterns and environmental parameters. The risk prediction and linkage module integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link the control and protection parameters of the dynamic protection control module. The blockchain evidence storage and credit management module stores all process data on the blockchain for evidence storage and generates credit scores. All modules work together to achieve safe electricity management in agriculture.

[0048] The system includes a user module, which verifies the user's identity through QR code information or RFID card, and controls the power control unit to switch on and off according to preset control logic. If the user's identity is verified, the user can control the power control unit to turn on the power through a visual human-machine interface, the socket will be powered on, and the irrigation equipment will start working. When leakage or overcurrent is detected, the leakage pattern recognition module will automatically cut off the power and issue an alarm.

[0049] Its working mechanism is based on a logical chain of data-driven, intelligent decision-making, and closed-loop control, achieving intelligent management of agricultural electricity safety throughout the entire process through multi-module collaboration. First, the multi-source data sensing module collects equipment parameters, geographical environment parameters, and meteorological parameters in real time. After normalization processing, it constructs a scene feature vector. The scene feature map construction module compares historical hazardous scenes using a cosine similarity algorithm, outputting a risk similarity γ to provide a benchmark for subsequent decision-making. When γ > 0.7, enhanced monitoring is triggered, increasing the sampling frequency to capture subtle anomalies.

[0050] Building upon this foundation, the leakage current pattern recognition module performs time-frequency domain analysis on the leakage current, extracting feature vectors such as peak value and dominant frequency. It then uses γ value to correct cluster centers and determines the leakage current pattern of the equipment via Euclidean distance, accurately distinguishing the leakage characteristics of 12 types of agricultural equipment and avoiding malfunctions of traditional protection systems. The dynamic protection control module, based on the leakage current pattern and environmental parameters, dynamically generates leakage current operating current, overcurrent I²t curves, and overvoltage thresholds using quantitative formulas. For example, higher humidity results in lower leakage current operating current, and high-temperature environments accelerate overcurrent protection response, achieving adaptive protection in complex environments.

[0051] The risk prediction and linkage module further integrates weather forecasts, equipment aging coefficients, and historical risk values. It calculates the risk index R using a weighted formula and adjusts protection parameters according to the R value, transforming passive protection into proactive prevention. All data is packaged and uploaded to the blockchain for notarization every 15 minutes. The blockchain notarization and credit management module generates credit scores based on the on-chain data, forming a closed loop of security control, data notarization, and behavioral incentives. The user module verifies identity via QR code / RFID and provides convenient operation through a visual interface. It also automatically cuts off power and triggers an alarm in case of a fault, ultimately achieving intelligent and refined management of agricultural electricity safety.

[0052] Compared with the prior art, the beneficial effects of the present invention are:

[0053] 1. Compared with the fixed threshold mode of traditional agricultural power protection devices, the present invention can dynamically generate protection parameters, adapt to changes in environmental temperature and humidity, soil conditions and equipment type in real time, improve protection accuracy to ±2%, and significantly reduce the false alarm rate in complex agricultural scenarios.

[0054] 2. An innovative approach is introduced to construct scene feature maps and calculate similarity. By comparing the normalized parameter matrix with a historical database of dangerous scenes, high-risk scenes can be accurately identified, solving the problem of insufficient perception of scene risks in traditional technologies.

[0055] 3. A multi-dimensional clustering correction algorithm is used to distinguish 12 types of leakage modes in agricultural equipment. The cluster center is dynamically adjusted by combining scene similarity. The leakage identification accuracy reaches 99.2%, which effectively avoids protection misjudgment caused by differences in equipment type. It is especially suitable for the complex situation of leakage feature drift of old equipment.

[0056] 4. The risk prediction module integrates multiple parameters such as weather forecasts, equipment aging coefficients, and historical risk values. It calculates the risk index through a weighted formula and implements a graded linkage protection strategy to achieve proactive prevention and control 2 hours in advance, reducing sudden safety accidents by more than 90% compared to the traditional passive protection mode.

[0057] 5. Blockchain-based evidence storage technology ensures data integrity throughout the entire process. Combined with a credit scoring system, it not only solves the credibility issues of electricity bill settlement and fault tracing, but also regulates farmers' electricity usage behavior through credit incentives, thereby improving management efficiency.

[0058] 6. Construct a complete closed loop of data collection, scene recognition, pattern clustering, dynamic protection, risk prediction, and evidence storage and incentives, covering the entire process of agricultural electricity use. Compared with single-function protection devices, it realizes the upgrade from passive protection to active management, and comprehensively improves the safety and intelligence level of agricultural electricity use. Attached Figure Description

[0059] Figure 1 This is a flowchart illustrating an intelligent and safe electricity management method for agricultural scenarios according to the present invention. Detailed Implementation

[0060] The technical solutions of the present invention will now be described in detail with reference to the accompanying drawings.

[0061] like Figure 1 As shown, a smart and safe electricity management method for agricultural scenarios is characterized by the following steps:

[0062] S1: Collect data, integrate equipment type, geographical environment, and meteorological parameters into a scene feature map, and determine the similarity between the current scene and historical dangerous scenes based on the scene feature map;

[0063] S2: Based on the scene characteristics of S1, the leakage current signal is clustered in multiple dimensions to distinguish the leakage current modes of different agricultural equipment and prevent protection malfunctions.

[0064] S3: Based on the leakage mode of S2 and the environmental parameters of S1, dynamic protection parameters are generated to replace the fixed threshold and adapt to protection in complex agricultural environments.

[0065] S4: Integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 to achieve proactive prevention and control;

[0066] S5: Upload all process data to the blockchain for storage and form an electricity data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, points are generated by combining the duration of compliant electricity use and energy saving rate. Points can be redeemed for maintenance services to incentivize standardized electricity use.

[0067] S1 collects data including device type parameters, including the rated power P of the connected device. N Rated current I N Equipment type label k; geographical environmental parameters include soil moisture S at the equipment location, obtained through installed sensors; real-time meteorological parameters include ambient temperature T, humidity H, and wind speed V at the equipment location, obtained by accessing the regional meteorological data platform.

[0068] S1 determines the similarity between the current scene and historical dangerous scenes based on the scene feature map. This is achieved by integrating equipment type, geographical environment, and meteorological parameters into a scene feature map. Specifically, all parameters are normalized so that all matrix elements are in the range [0,1]. A feature vector G is constructed from the standardized parameters and weights are assigned according to their impact on safety.

[0069] Construct a database of historical hazardous scenarios, with each scenario corresponding to a feature vector G. risk,m , where m is the scene number;

[0070] Compare the current scenario G with historical hazardous scenarios G risk,m The similarity γ is calculated using the following formula:

[0071]

[0072] In the formula, γ represents the maximum similarity between the current scene and historical dangerous scenes, and G is the feature matrix constructed using standardized parameters. risk,m Let m be the feature vector of the m-th scene in the historical hazardous scene database, where m is the number of the historical hazardous scene.

[0073] Where γ∈[0,1], γ>0.7 is marked as a high-risk scenario, triggering enhanced monitoring and increasing the sampling frequency.

[0074] Based on the scene features collected by S1, S2 performs multi-dimensional clustering of leakage current signals to distinguish leakage current modes of various agricultural devices, thereby solving the problem of malfunction in traditional protection systems. Specifically, this is achieved by analyzing the leakage current i... leak (t) Perform time-frequency domain analysis to extract the feature vector F, the formula of which is as follows:

[0075]

[0076] In the formula, F is the characteristic vector of leakage current, i leak (t) represents the leakage current, max(i) leak f is the maximum value of the leakage current. dom The main frequency of the leakage signal, skew(i leak ) is the skewness coefficient of the leakage current signal, reflecting the signal asymmetry, and t is the time variable;

[0077] Scene-based clustering correction, combined with the scene similarity γ of S1, adjusts the cluster center μ. k After correction, the corrected cluster centers μ are obtained. k The formula is as follows:

[0078] μ k '=μ k ·(1+0.3γ)+F scene ·0.2γ

[0079] In the formula, μ k ' is the corrected cluster center of the k-th class, μ k For the k-th cluster center before correction, F scene Device characteristics in the current scenario;

[0080] Current leakage current characteristic F and a certain cluster center μ k If the Euclidean distance d < 0.15, the device is considered to be leaking current.

[0081] S3 dynamically generates protection parameters based on the leakage mode of S2 and the environmental parameters of S1, replacing fixed thresholds. These parameters include adjustments to the leakage current, overcurrent protection curve, and overvoltage protection threshold. Specifically, based on the equipment type (kW) determined by S2 and the humidity (H) of S1, the leakage current (I) is dynamically adjusted. leak,act The formula is as follows:

[0082] I leak,act =30·(1+0.01kw)·exp(-0.01H)

[0083] In the formula, I leak,act The leakage current is dynamically adjusted, kW is the power of equipment type k, and H is the ambient humidity.

[0084] Using I 2 The t-characteristic curve is used for overcurrent protection curve calibration, and dynamically corrected according to ambient temperature T and soil moisture S.

[0085]

[0086] In the formula, I 2 t: Parameters of the overcurrent protection characteristic curve, reflecting the product of the square of the current and time; T: Ambient temperature; kx: Correction coefficient corresponding to equipment type k; S: Soil moisture; I: ... N The current is the rated current of the equipment; 8 seconds is the reference time. The curve is steeper when the temperature rises or the soil is wet.

[0087] The overvoltage protection threshold adjustment is combined with the scenario similarity γ of S1 to adjust the overvoltage action voltage U. over,act :

[0088] U over,act =250·(1-0.05γ)

[0089] In the formula, U over,act The overvoltage action voltage is dynamically adjusted, γ is the similarity of historical dangerous scenarios, and 250 is the standard voltage threshold. The threshold is lowered in high-risk scenarios to prevent voltage anomalies in advance.

[0090] S4 integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 for proactive prevention and control; it establishes a risk index prediction model based on the probability of precipitation P in the next 2 hours. r Equipment aging coefficient A = 1 - exp(-0.0003t) total Historical risk value R hist Calculate the risk index R:

[0091] R = 0.4P r +0.3A+0.2R hist +0.1γ

[0092] In the formula, R is the comprehensive risk index, and P r Let A be the probability of precipitation in the next 2 hours, and let A be the equipment aging coefficient. t total R represents the cumulative operating time of the equipment. hist γ represents the historical risk value, and γ represents the similarity of historical dangerous scenarios.

[0093] Adjust the protection parameters of S3 according to R:

[0094]

[0095] Proactive prevention and control can be achieved by adjusting parameters.

[0096] S5 uploads all process data to the blockchain for storage, forming an electricity consumption data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, a block is generated every 15 minutes, containing the following data:

[0097] Block = {H prev ,E,G,S prot ,R,TimeStamp,UserID}

[0098] In the formula, Block: a data block in the blockchain, H prev Let E be the hash of the preceding block, E be the electricity consumption during the time period, G be the scene feature matrix of S1, and S be the data source of the preceding block. prot This is the protection action record for S3, R is the risk index for S4, TimeStamp is the timestamp, and UserID is the user identifier.

[0099] S5 generates points based on compliant electricity usage duration and energy saving rate. These points can be redeemed for maintenance services to incentivize compliant electricity use. The credit score calculation is optimized by adding a scenario risk coefficient γ to the existing formula to encourage compliant electricity use in high-risk scenarios.

[0100] P = 50·t comply ·(2-η)·exp(-0.1N fault )·(1+0.5γ)

[0101] In the formula, P is the credit score, and t comply For compliant electricity usage duration, η represents the energy saving rate, which is the actual electricity consumption divided by the theoretical demand. fault denoted as the number of failures, and γ as the similarity of historical hazardous scenarios.

[0102] A smart and safe electricity management system for agricultural scenarios, implementing the method described in any one of claims 1-8, is characterized by integrating a distributed distribution box cluster, a multi-functional safety socket module, and a visual human-machine interface. It further includes a multi-source data sensing module, a scene feature map construction module, a leakage current pattern recognition module, a dynamic protection control module, a risk prediction and linkage module, and a blockchain evidence storage and credit management module. Specifically, the multi-source data sensing module collects equipment, geographical environment, and meteorological parameters, supporting the scene feature map construction module in completing parameter standardization, feature matrix construction, and similarity calculation with historical hazardous scenarios. The leakage current pattern recognition module clusters leakage current signals in multiple dimensions based on scene features to distinguish leakage current patterns. The dynamic protection control module dynamically generates protection parameters based on leakage current patterns and environmental parameters. The risk prediction and linkage module integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link the control and protection parameters of the dynamic protection control module. The blockchain evidence storage and credit management module stores all process data on the blockchain for evidence storage and generates credit scores. All modules work together to achieve safe electricity management in agriculture.

[0103] The system includes a user module, which verifies the user's identity through QR code information or RFID card, and controls the power control unit to switch on and off according to preset control logic. If the user's identity is verified, the user can control the power control unit to turn on the power through a visual human-machine interface, the socket will be powered on, and the irrigation equipment will start working. When leakage or overcurrent is detected, the leakage pattern recognition module will automatically cut off the power and issue an alarm.

[0104] The specific implementation method is to integrate equipment type, geographical environment, and meteorological parameters into a scene feature map, and determine the similarity γ between the current scene and historical dangerous scenes based on the scene feature map;

[0105] Equipment type parameters include the rated power P of the connected equipment. N Rated current I NEquipment type label k, totaling 12 categories; geographical environmental parameters include soil moisture S; real-time meteorological parameters include ambient temperature T, humidity H, and wind speed V;

[0106] The parameters are then normalized using the formula P. N '=P N / max(P N Similarly, other parameters are used to ensure that all matrix elements are in the range [0,1].

[0107] Ensure that all matrix elements are in the interval [0,1]. Construct a feature matrix G from the standardized parameters and assign weights to each element to reflect its impact on security: G = [P] N '·0.2,I N '·0.2,k'·0.1,S'·0.2,T'·0.15,H'·0.15]

[0108] Historical hazardous scene database and similarity calculation: Construct a database of historical hazardous scenes, with each scene corresponding to a feature vector G. risk,m , where m is the scene number;

[0109] Calculation of the similarity γ between the current scene and historical dangerous scenes:

[0110]

[0111] Where γ∈[0,1], γ>0.7 is marked as "high-risk scenario", triggering enhanced monitoring and increasing the sampling frequency to 50kHz.

[0112] Based on scene characteristics, leakage signals are clustered in multiple dimensions to distinguish leakage modes of 12 types of agricultural equipment, including water pumps, sprayers, roller shutters, fertilizer applicators, seeders, weeders, harvesters, threshers, fans, temperature control equipment, irrigation solenoid valves, and electric pruning machines, thus solving the problem of malfunction of traditional protection systems.

[0113] For leakage current (i leak (t) is analyzed in the time-frequency domain to extract the feature vector F.

[0114]

[0115] In the formula, f dom Main frequency, skew(i leak The skewness coefficient reflects the asymmetry of the leakage signal.

[0116] Scene-based clustering correction, combined with scene similarity γ, adjusts the cluster center μ. k After correction, the corrected cluster centers μ are obtained. k The formula is as follows:

[0117] μ k '=μ k ·(1+0.3γ)+F scene ·0.2γ

[0118] In the formula, F scene Based on the device characteristics of the current scenario, ensure that the clustering results are adapted to the scenario characteristics;

[0119] Judgment rule: When the current leakage current characteristic (F) matches a certain cluster center (μ) k When the Euclidean distance (d<0.15) of the device is less than 0.15, it is determined that the device is leaking current.

[0120] Based on leakage current mode and environmental parameters, protection parameters are dynamically generated to replace fixed thresholds and improve protection accuracy in complex agricultural environments.

[0121] Based on the determined device type k and humidity H, the leakage current I is dynamically adjusted. leak,act :

[0122] I leak,act =30·(1+0.01k)·exp(-0.01H) 30mA is the standard safety threshold. The higher the power of the equipment (the larger k) and the lower the humidity, the operating current should be appropriately relaxed to avoid false triggering.

[0123] Overcurrent protection curve calibration uses I 2 The t-characteristic curve is dynamically corrected based on equipment type k, ambient temperature T, and soil moisture S.

[0124]

[0125] I N The current is the rated current of the equipment, and 8 seconds is the reference time. When the temperature rises or the soil is wet, the curve is steeper, the protection is more sensitive, and the accuracy reaches ±2%.

[0126] The overvoltage protection threshold is adjusted in conjunction with the scenario similarity γ to adjust the overvoltage action voltage U. over,act :

[0127] U over,act =250·(1+0.05γ)250V is the standard threshold. The threshold is lowered in high-risk scenarios to prevent voltage anomalies in advance.

[0128] By integrating historical data, weather forecasts, and equipment aging parameters, risks can be predicted in advance and protective parameters can be linked to achieve proactive prevention and control.

[0129] Risk index prediction model: based on the probability of precipitation P in the next 2 hours r Equipment aging coefficient A = 1 - exp(-0.0003t) total Historical risk value Rhist Calculate the risk index R:

[0130] R = 0.4P r +0.3A+0.2R hist +0.1γ

[0131] Here, γ represents scene similarity, which comprehensively reflects multi-dimensional risks.

[0132] Intervention strategy: Adjust protection parameters according to R:

[0133]

[0134] Authorization records, protection actions, and environmental data are uploaded to the blockchain to form an immutable "electricity consumption data chain," supporting credit evaluation and electricity bill settlement. Based on the on-chain data, combined with the duration of compliant electricity use, [the data is then analyzed]. comply Energy saving rate η (actual / theoretical electricity consumption) is generated by integral P: (P = 50·t) comply ·(2-η)·exp(-0.1·N fault (N) fault The number of faults is recorded, and points can be redeemed for maintenance services to incentivize proper electricity use.

[0135] A block is generated every 15 minutes, containing the following data:

[0136] Block = {H prev ,E,G,S prot ,R,TimeStamp,UserID}

[0137] H prev Here, E is the hash of the preceding block, E is the electricity consumption during the time period, G is the scene feature matrix, and S is the hash of the preceding block. prot To protect action records, an R risk index is used; the SHA-256 hash algorithm is employed to ensure immutability, supporting electricity billing and fault tracing.

[0138] The credit score calculation has been optimized by adding a scenario risk coefficient γ to the existing formula to encourage compliant electricity use in high-risk scenarios.

[0139] P = 50·t comply ·(2-η)·exp(-0.1N fault )·(1+0.5γ)t comply For compliant electricity usage duration, η is the energy saving rate, where η is the actual electricity consumption / theoretical demand, and N is the total electricity consumption. fault The number of malfunctions is recorded; points can be redeemed for priority electricity access, priority power supply during peak irrigation periods, or free equipment maintenance services.

Claims

1. A smart and safe electricity management method for agricultural scenarios, characterized in that, Includes the following steps: S1: Collect data, integrate equipment type, geographical environment, and meteorological parameters into a scene feature map, and determine the similarity between the current scene and historical dangerous scenes based on the scene feature map; S2: Based on the scene characteristics of S1, the leakage current signal is clustered in multiple dimensions to distinguish the leakage current modes of different agricultural equipment and prevent protection malfunctions. S3: Based on the leakage mode of S2 and the environmental parameters of S1, dynamic protection parameters are generated to replace the fixed threshold and adapt to protection in complex agricultural environments. S4: Integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 to achieve proactive prevention and control; S5: Upload all process data to the blockchain for storage and form an electricity data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, points are generated by combining the duration of compliant electricity use and energy saving rate. Points can be redeemed for maintenance services to incentivize standardized electricity use.

2. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that, S1 collects data including device type parameters, including the rated power P of the connected device. N Rated current I N Equipment type label k; geographical environmental parameters include soil moisture S at the equipment location, obtained through installed sensors; real-time meteorological parameters include ambient temperature T, humidity H, and wind speed V at the equipment location, obtained by accessing the regional meteorological data platform.

3. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that, S1 determines the similarity between the current scene and historical dangerous scenes based on the scene feature map. This is achieved by integrating equipment type, geographical environment, and meteorological parameters into a scene feature map. Specifically, all parameters are normalized so that all matrix elements are in the range [0,1]. A feature vector G is constructed from the standardized parameters and weights are assigned according to their impact on safety. Construct a database of historical hazardous scenarios, with each scenario corresponding to a feature vector G. risk,m , where m is the scene number; Compare the current scenario G with historical hazardous scenarios G risk,m The similarity γ is calculated using the following formula: In the formula, γ represents the maximum similarity between the current scene and historical dangerous scenes, and G is the feature matrix constructed using standardized parameters. risk,m Let m be the feature vector of the m-th scene in the historical hazardous scene database, where m is the number of the historical hazardous scene. Where γ∈[0,1], γ>0.7 is marked as a high-risk scenario, triggering enhanced monitoring and increasing the sampling frequency.

4. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that, Based on the scene features collected by S1, S2 performs multi-dimensional clustering of leakage current signals to distinguish leakage current modes of various agricultural devices, thereby solving the problem of malfunction in traditional protection systems. Specifically, this is achieved by analyzing the leakage current i... leak (t) Perform time-frequency domain analysis to extract the feature vector F, the formula of which is as follows: In the formula, F is the characteristic vector of leakage current, i leak (t) represents the leakage current, max(i) leak f is the maximum value of the leakage current. dom The main frequency of the leakage signal, skew(i leak ) is the skewness coefficient of the leakage current signal, reflecting the signal asymmetry, and t is the time variable; Scene-based clustering correction, combined with the scene similarity γ of S1, adjusts the cluster center μ. k After correction, the corrected cluster centers μ are obtained. k The formula is as follows: m k '=μ k ·(1+0.3c)+F scene ·0.2g In the formula, μ k ' is the corrected cluster center of the k-th class, μ k For the k-th cluster center before correction, F scene Device characteristics in the current scenario; Current leakage current characteristic F and a certain cluster center μ k If the Euclidean distance d < 0.15, the device is considered to be leaking current.

5. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that, S3 dynamically generates protection parameters based on the leakage mode of S2 and the environmental parameters of S1, replacing fixed thresholds. These parameters include adjustments to the leakage current, overcurrent protection curve, and overvoltage protection threshold. Specifically, based on the equipment type (kW) determined by S2 and the humidity (H) of S1, the leakage current (I) is dynamically adjusted. leak,act The formula is as follows: I leak,act =30·(1+0.01kw)·exp(-0.01H) In the formula, I leak,act The leakage current is dynamically adjusted, kW is the power of equipment type k, and H is the ambient humidity. Using I 2 The t-characteristic curve is used for overcurrent protection curve calibration, and dynamically corrected according to ambient temperature T and soil moisture S. In the formula, I 2 t: Parameters of the overcurrent protection characteristic curve, reflecting the product of the square of the current and time; T: Ambient temperature; kx: Correction coefficient corresponding to equipment type k; S: Soil moisture; I: ... N The current is the rated current of the equipment; 8 seconds is the reference time. The curve is steeper when the temperature rises or the soil is wet. The overvoltage protection threshold adjustment is combined with the scenario similarity γ of S1 to adjust the overvoltage action voltage U. over,act : U over,act =250·(1-0.05g) In the formula, U over,act The overvoltage action voltage is dynamically adjusted, γ is the similarity of historical dangerous scenarios, and 250 is the standard voltage threshold. The threshold is lowered in high-risk scenarios to prevent voltage anomalies in advance.

6. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that... S4 integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link with the protection parameters of S3 for proactive prevention and control; it establishes a risk index prediction model based on the probability of precipitation P in the next 2 hours. r Equipment aging coefficient A = 1 - exp(-0.0003t) total Historical risk value R hist Calculate the risk index R: R=0.4P r +0.3A+0.2R hist +0.1γ In the formula, R is the comprehensive risk index, and P r Let A be the probability of precipitation in the next 2 hours, and let A be the equipment aging coefficient. t total R represents the cumulative operating time of the equipment. hist γ represents the historical risk value, and γ represents the similarity of historical dangerous scenarios. Adjust the protection parameters of S3 according to R: Proactive prevention and control can be achieved by adjusting parameters.

7. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that... S5 uploads all process data to the blockchain for storage, forming an electricity consumption data blockchain to support credit evaluation and electricity bill settlement. Based on the data on the blockchain, a block is generated every 15 minutes, containing the following data: Block={H prev ,E,G,S prot ,R,TimeStamp,UserID} In the formula, Block: a data block in the blockchain, H prev Let E be the hash of the preceding block, E be the electricity consumption during the time period, G be the scene feature matrix of S1, and S be the data source of the preceding block. prot This is the protection action record for S3, R is the risk index for S4, TimeStamp is the timestamp, and UserID is the user identifier.

8. The intelligent safe electricity management method for agricultural scenarios according to claim 1, characterized in that... S5 generates points based on compliant electricity usage duration and energy saving rate. These points can be redeemed for maintenance services to incentivize compliant electricity use. The credit score calculation is optimized by adding a scenario risk coefficient γ to the existing formula to encourage compliant electricity use in high-risk scenarios. P=50·t comply ·(2-n)·exp(-0.1N fault )·(1+0.5c) In the formula, P is the credit score, and t comply For compliant electricity usage duration, η represents the energy saving rate, which is the actual electricity consumption divided by the theoretical demand. fault denoted as the number of failures, and γ as the similarity of historical hazardous scenarios.

9. An intelligent safe electricity management system for agricultural scenarios, implementing the method of any one of claims 1-8, characterized in that, The system integrates a distributed distribution box cluster, a multi-functional safety socket module, and a visual human-machine interface. It also includes a multi-source data sensing module, a scene feature map construction module, a leakage current pattern recognition module, a dynamic protection control module, a risk prediction and linkage module, and a blockchain-based evidence storage and credit management module. Specifically, the multi-source data sensing module collects data from equipment, geographical environment, and meteorological parameters, supporting the scene feature map construction module in completing parameter standardization, feature matrix construction, and similarity calculation with historical hazardous scenarios. The leakage current pattern recognition module clusters leakage current signals in multiple dimensions based on scene features to distinguish leakage current patterns. The dynamic protection control module dynamically generates protection parameters based on leakage current patterns and environmental parameters. The risk prediction and linkage module integrates historical data, weather forecasts, and equipment aging parameters to predict risks in advance and link the control and protection parameters of the dynamic protection control module. The blockchain-based evidence storage and credit management module stores all data on the blockchain, generating credit scores. All modules work together to achieve safe management of agricultural electricity use.

10. A smart and safe electricity management system for agricultural scenarios according to claim 9, characterized in that, The system includes a user module, which verifies the user's identity through QR code information or RFID card, and controls the power control unit to switch on and off according to preset control logic. If the user's identity is verified, the user can control the power control unit to turn on the power through a visual human-machine interface, the socket will be powered on, and the irrigation equipment will start working. When leakage or overcurrent is detected, the leakage pattern recognition module will automatically cut off the power and issue an alarm.

Images