Distributed coordination-based street lamp leakage monitoring and protection method and system and medium

Abstract

This invention discloses a method, system, and medium for street light leakage current monitoring and protection based on distributed collaboration. The method includes: synchronously collecting raw leakage current data through a multi-circuit detection device and a distributed detection terminal, and preprocessing it to obtain dual-end preprocessed data; an edge computing unit performing local fusion, anti-interference filtering, and key feature extraction on the dual-end preprocessed data; when the collaborative verification trigger condition is met, dual-end collaborative verification is executed: a cloud platform performs a risk pre-assessment based on the dual-end preprocessed data and environmental factors, generating a pre-assessment result; when the edge computing unit determines that a local high-risk threshold has been reached, it calls the pre-assessment result and cross-verifies it with locally cached historical data from the light pole side to generate a collaborative verification result; the edge computing unit executes circuit breaker protection based on the collaborative verification result, or performs graded early warning based on key features. This invention significantly improves the safety and operation and maintenance efficiency of street light power supply systems.

Description

Technical Field

[0001] This invention relates to the field of intelligent lighting control and electrical safety monitoring technology, specifically to a method, system, and medium for monitoring and protecting street light leakage current based on distributed collaboration. Background Technology

[0002] As a critical public infrastructure, urban street lighting systems operate in complex outdoor environments for extended periods, facing challenges such as humidity, high temperatures, material aging, and lightning strikes, resulting in a significant risk of electrical leakage. Streetlight leakage not only leads to unplanned energy loss but can also trigger serious public safety incidents, such as electric shock and electrical fires, posing a direct threat to urban safety. Therefore, developing technologies capable of accurately, promptly, and reliably monitoring and protecting against streetlight leakage has become a crucial issue for improving the safety and intelligent operation and maintenance of urban public lighting systems.

[0003] Currently, the technical solutions proposed by the industry for street light leakage current monitoring and protection mainly follow the following three technical routes, but each has significant limitations: Centralized monitoring approach: This solution centralizes detection within the regional distribution cabinet, using total circuit current monitoring to determine leakage current. Its advantage lies in its relatively simple deployment. However, this approach has fundamental drawbacks: First, it can only detect leakage across the entire circuit, failing to pinpoint the fault location. Subsequent troubleshooting relies on manual point-by-point testing, resulting in low efficiency and high maintenance costs. Second, achieving wide-range monitoring often sacrifices accuracy in detecting weak leakage currents, making it difficult to detect early-stage insulation degradation. Third, the device operates in a strong electromagnetic interference environment within the distribution cabinet, easily generating false alarms. Furthermore, all data must be uploaded to a remote center for processing, resulting in a long decision-making chain, high response delays, and difficulty meeting the need for rapid fault isolation in emergencies. Finally, its protection mechanism is typically based on simple fixed thresholds; malfunctions may lead to unnecessary power outages in non-faulty areas, affecting lighting functionality.

[0004] Distributed sensing approach: This solution deploys detection units at each light pole to attempt to monitor fault locations. While it solves the location problem, it has significant drawbacks: First, the detection units typically rely on existing lighting control system communication networks, resulting in poor independent networking and operational capabilities, limiting deployment flexibility and system reliability. Second, outdoor units face harsh environments; insufficient protection levels lead to high failure rates and significant long-term operation and maintenance pressures. Third, the lack of effective data interaction and collaborative analysis mechanisms between each detection unit and the higher-level monitoring device creates information silos, hindering the improvement of judgment reliability through cross-verification of multi-source information, and sometimes even increasing the overall system misjudgment rate due to local false alarms.

[0005] Single-parameter alarm route: This is the currently widely used simplified mode, which only alarms and provides protection based on whether the instantaneous value of leakage current exceeds a certain preset threshold. This mode has inherent drawbacks of being "one-size-fits-all": on the one hand, it may misjudge slowly developing, controllable insulation degradation as a fault requiring emergency handling, leading to a waste of maintenance resources; on the other hand, it may fail to identify rapidly deteriorating, high-risk leakage trends in a timely manner, creating safety hazards. Furthermore, this mode lacks the ability to mine and analyze data trends, cannot support preventative maintenance, and limits the intelligent transformation of maintenance from "passive response" to "proactive prevention."

[0006] In summary, existing technologies face a series of prominent contradictions in practical applications: it is difficult to balance monitoring range and accuracy; fault detection and precise location cannot be unified; there is a conflict between the need for rapid response and the latency of centralized cloud processing; the reliability of simple threshold alarms is insufficient and their adaptability to complex field environments is inadequate; and the coordination capability between various system components is lacking. These contradictions restrict further improvements in the safety, reliability, economy, and intelligence of street light leakage current monitoring and protection technology.

[0007] Therefore, there is an urgent need in this field for an innovative technical solution that can effectively integrate the advantages of distributed sensing and centralized management, and achieve rapid location of leakage faults, intelligent risk assessment, multi-level collaborative verification and reliable protection while ensuring high-precision and wide-range monitoring. Summary of the Invention

[0008] To address the problems in existing technologies, such as the contradiction between measurement range and accuracy, difficulty in locating leakage faults, high local decision-making response delay, lack of coordination capabilities between system components, and high false alarm rate, this invention provides a street light leakage monitoring and protection method, system, and medium based on distributed coordination to solve the aforementioned technical defects.

[0009] This invention proposes a method for monitoring and protecting street light leakage current based on distributed collaboration, comprising the following steps: S1. By deploying an edge computing multi-circuit detection device on the side of the power distribution cabinet and an intelligent single-circuit distributed detection terminal on the side of the street light pole, the raw leakage current data of each power supply circuit and the corresponding light pole branch are collected synchronously, and the raw leakage current data is preprocessed to obtain dual-end preprocessed data. S2. The edge computing unit in the edge computing type multi-loop detection device performs local fusion and anti-interference filtering on the dual-end preprocessed data to obtain purified multi-source leakage current data, and extracts key features from the purified multi-source leakage current data, including the current leakage current value and leakage current growth rate. S3. When the edge computing unit determines that the collaborative verification triggering condition is met based on key features, it performs dual-end collaborative verification on the distribution cabinet side and the light pole side. The dual-end collaborative verification includes: S31. The cloud platform performs a pre-assessment of leakage risk based on the pre-processed data from both ends and in combination with environmental factors, and generates a pre-assessment result. S32. When the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on key features, it calls the pre-evaluation result and cross-verifies it with the local cached historical data of the corresponding loop on the lamp post side to generate a collaborative verification result. S4. The edge computing unit performs local risk assessment and handling, including: if the collaborative verification result shows that the verification is passed, then the circuit breaker protection action is executed; otherwise, the leakage hazard level is determined based on key characteristics and the corresponding early warning action is executed.

[0010] Preferably, in step S1, the raw leakage current data is preprocessed to obtain two-terminal preprocessed data, including the following sub-steps: S11. Add a collection timestamp to the raw leakage current data collected from the distribution cabinet side and the light pole side respectively; S12. Based on the collected timestamps, synchronize the original leakage current data of the distribution cabinet side and the light pole side to obtain time synchronization data. S13. Convert the time synchronization data into a unified preset data format to obtain dual-end preprocessed data.

[0011] Preferably, in step S2, the edge computing unit performs local fusion and anti-interference filtering on the dual-end preprocessed data to obtain purified multi-source leakage current data, and extracts key features from the purified multi-source leakage current data, including the following sub-steps: S21. Real-time fusion of the dual-end preprocessed data to obtain a multi-source fused data stream. The dual-end preprocessed data includes circuit leakage current preprocessed data from the distribution cabinet side and branch leakage current preprocessed data from the light pole side. S22. The wavelet transform algorithm is used to perform preliminary filtering on the multi-source fused data stream to obtain the first intermediate data; S23. The first intermediate data is smoothed using a sliding window filtering algorithm to obtain the second intermediate data; S24. Based on the preset outlier removal rules, the second intermediate data is filtered to obtain the purified multi-source leakage current data. S25. Calculate and extract the current leakage current value and leakage current growth rate from the purified multi-source leakage current data.

[0012] Preferably, in step S31, the cloud platform performs a preliminary assessment of leakage risk based on the preprocessed data from both ends and in conjunction with environmental factors, generating a preliminary assessment result, including the following sub-steps: S311 The cloud platform receives pre-processed data from both ends and calculates environmental factors based on real-time ambient humidity, line aging years, and operating parameters. S312, the cloud platform obtains leakage current growth trend parameters through linear regression fitting based on the received dual-end preprocessed data; S313. The cloud platform uses environmental factors to correct the preset standard risk thresholds and generate environmentally adaptive risk thresholds. S314. The cloud platform combines leakage current growth trend parameters with environmental adaptive risk thresholds to calculate a comprehensive risk assessment value. S315. The cloud platform determines the risk level based on the comprehensive risk assessment value and, in conjunction with the environmental adaptive risk threshold, generates a pre-assessment result that includes the risk level and the environmental adaptive risk threshold.

[0013] More preferably, in step S32, when the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on key features, it calls the pre-evaluation result and performs cross-validation with the locally cached historical data of the corresponding loop on the lamppost side to generate a collaborative verification result, including the following sub-steps: S321. When the current leakage current value in the key features is determined to be greater than or equal to the local high-risk threshold, the edge computing unit calls the pre-evaluation result and calculates the deviation rate between the local current leakage current value and the environmental adaptive risk threshold in the pre-evaluation result. S322. If the deviation rate is less than or equal to the first preset deviation threshold, retrieve the local cached historical data of the corresponding circuit on the lamp post side; calculate the deviation rate between the local leakage current sum and the lamp post side historical data sum, and determine whether the growth trend of the local leakage current is consistent with the growth trend of the lamp post side historical data sum. S323. If the deviation rate between the total local leakage current and the total historical data on the lamp post side is less than or equal to the second preset deviation threshold, and the trend is consistent, then a collaborative verification result indicating that the verification has passed is generated; otherwise, a collaborative verification result indicating that the verification has failed is generated.

[0014] Preferably, during the execution of step S3, a routine data synchronization step performed before the dual-end collaborative verification is also included: The edge computing unit and the intelligent single-loop distributed detection terminal upload pre-processed data from both ends to the cloud platform according to a dynamic cycle; the length of the dynamic cycle is determined based on the current leakage current value and leakage current growth rate in the key features. When the current leakage current value is less than the first threshold and the leakage current growth rate is less than the third threshold, the first upload cycle is used. When the current leakage current value is greater than or equal to the first threshold or the leakage current growth rate is greater than or equal to the third threshold, a second upload cycle shorter than the first upload cycle is adopted.

[0015] Preferably, in step S4, the edge computing unit performs local risk assessment and handling, including the following sub-steps: S41. If the collaborative verification result shows that the verification is successful, then the circuit breaker protection action is executed; otherwise, the leakage hazard level is determined based on the current leakage current value, leakage current growth rate and leakage duration in the key features. S42. Based on the determined leakage hazard level, implement the corresponding early warning measures, including generating and uploading early warning information corresponding to the hazard level.

[0016] This invention also proposes a street light leakage current monitoring and protection system based on distributed collaboration, for implementing the method described in any of the above, the system comprising: An edge computing-based multi-loop detection device is deployed on the power distribution cabinet side to synchronously collect raw leakage current data of the power supply circuit. The raw leakage current data is preprocessed to obtain a portion of the dual-end preprocessed data. The dual-end preprocessed data is then fused locally and subjected to anti-interference filtering to obtain purified multi-source leakage current data. Key features are extracted from the purified multi-source leakage current data. Based on the key features, it is determined whether the collaborative verification triggering condition is met or whether the value is greater than or equal to the local high-risk threshold. The pre-evaluation results are called and cross-validated with locally cached historical data to generate collaborative verification results. Finally, risk judgment and handling are performed based on the collaborative verification results or key features. The intelligent single-loop distributed detection terminal is deployed on the side of the street light pole to synchronously collect the raw leakage current data of the corresponding light pole branch and preprocess the raw leakage current data to obtain another part of the double-end preprocessed data. The cloud platform communicates with edge computing multi-loop detection devices and intelligent single-loop distributed detection terminals to perform leakage risk pre-assessment based on dual-end pre-processed data and combined with environmental factors, and generates and distributes the pre-assessment results.

[0017] The present invention also proposes a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described distributed collaborative street light leakage current monitoring and protection methods.

[0018] The present invention also proposes a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of any of the above-described distributed collaborative street light leakage current monitoring and protection methods.

[0019] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) By constructing a collaborative acquisition architecture of an edge computing-based multi-loop detection device and an intelligent single-loop distributed detection terminal, and executing a dual-end collaborative verification process, the technical solution of this invention can correlate and compare the leakage current anomalies monitored on the distribution cabinet side with the data on the lamp post side of a specific circuit. This collaborative analysis mechanism enables the system to accurately locate the fault range from the entire circuit to a specific branch or lamp post group, thereby transforming fault diagnosis from the traditional inefficient manual pole-by-pole testing to system-assisted precise location, significantly shortening the fault diagnosis time and reducing operation and maintenance costs.

[0020] (2) This invention deploys a detection device equipped with an edge computing unit on the distribution cabinet side, enabling it to perform local data processing and rapid decision-making. This technical feature allows key decision-making logic such as data fusion, feature extraction, preliminary high-risk judgment, and final collaborative verification to be completed locally in milliseconds, achieving second-level rapid circuit breaker protection in emergency situations. This reduces the real-time dependence on cloud networks and computing resources, overcomes the slow response problem caused by data transmission and processing delays in traditional centralized solutions, and improves the timeliness and reliability of protection actions.

[0021] (3) This invention provides double insurance for core protection decisions by designing a multi-level, multi-dimensional collaborative verification mechanism that combines routine data synchronization, cloud-based risk pre-assessment, and edge-side historical data cross-verification. This technical solution can effectively identify and eliminate alarms triggered by non-real leakage factors such as electromagnetic interference in the distribution cabinet, instantaneous fluctuations in sensors, and temporary failures of single-point equipment. Through this collaborative verification process, the final decision to execute the protection action is ensured to have a very high degree of confidence, thereby significantly reducing false alarms and the resulting false power outages that are common in traditional solutions.

[0022] (4) Abandoning the rudimentary alarm mode of the existing technology of "single threshold, one-size-fits-all", the technical solution of this invention constructs a multi-dimensional dynamic risk level judgment model based on extracted key features and combined with leakage duration and collaborative verification results. According to this model, the system can intelligently distinguish different levels such as no risk, low risk, medium risk, and high risk, and execute differentiated handling strategies. This makes the protection action more accurate and reasonable, ensuring a rapid response to real high-risk faults while avoiding blind power outages caused by low-risk, continuous leakage, thus ensuring the continuous and reliable operation of the lighting system.

[0023] (5) This invention effectively overcomes the shortcomings of traditional schemes that sacrifice small current detection accuracy in pursuit of a wide range by introducing a piecewise linear correction and adaptive gain control collaborative algorithm with a wide range. This technical solution enables the system to maintain high-precision detection while covering the entire range from weak leakage current to large current short circuit, thereby being able to keenly capture the tiny leakage current signal generated in the early stage of insulation aging that is easily missed by existing technologies, and achieving true prevention. Attached Figure Description

[0024] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments, taken with reference to the accompanying drawings: Figure 1 This is a flowchart of a street light leakage current monitoring and protection method based on distributed collaboration; Figure 2 This is a schematic diagram of a street light leakage current monitoring and protection system based on distributed collaboration; Figure 3 This is a schematic diagram of the structure of a computer system suitable for implementing the embodiments of the present invention. Detailed Implementation

[0025] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0026] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0027] Figure 1 A flowchart of a street light leakage current monitoring and protection method based on distributed collaboration is shown, such as... Figure 1 As shown in the figure, this embodiment of the invention provides a street light leakage current monitoring and protection method based on distributed collaboration, including the following steps: S1. By deploying an edge computing multi-circuit detection device on the side of the power distribution cabinet and an intelligent single-circuit distributed detection terminal on the side of the street light pole, the raw leakage current data of each power supply circuit and the corresponding light pole branch are collected synchronously, and the raw leakage current data is preprocessed to obtain dual-end preprocessed data.

[0028] In one specific embodiment, step S1 is the foundation for subsequent collaborative analysis, and its core lies in synchronously acquiring comparable leakage current information from the main circuit of the distribution cabinet and each light pole branch. The specific implementation process is as follows: Firstly, in terms of hardware deployment, the edge computing-based multi-loop detection device is installed inside the regional street light distribution cabinet. This device integrates multiple high-precision leakage current sensors (e.g., open-type or through-type current transformers), signal conditioning circuits, analog-to-digital converters (ADCs), and a core edge computing unit (such as an ARM Cortex-based microprocessor). This device can simultaneously monitor multiple independent power supply loops from the distribution cabinet. Simultaneously, the intelligent single-loop distributed detection terminal is deployed inside each street light pole requiring monitoring or at each power supply access point. This terminal is an independent device, containing a single-channel leakage current detection module, an independent power management module (supporting wide voltage input, such as 85VAC-305VAC), an eSIM communication module, and a housing designed to meet harsh outdoor environmental requirements, such as aluminum alloy, with an IP67 protection rating.

[0029] During the data acquisition phase, the edge computing multi-loop detection device on the distribution cabinet side continuously acquires the analog leakage current signals of each power supply loop at a high sampling rate (e.g., 1kHz) and converts them into digital signals to form the raw leakage current data of the loop. Simultaneously, the intelligent single-loop distributed detection terminal on each light pole side acquires the leakage current signal of its respective light pole branch at the same or coordinated sampling rate, forming the raw leakage current data of the branch. To ensure the accuracy of subsequent collaborative analysis, the acquisition clocks of both ends are initially synchronized using a synchronization mechanism initiated through Network Time Protocol (NTP) or a high-precision crystal oscillator to achieve synchronized acquisition.

[0030] The raw leakage current data collected needs to be preprocessed to eliminate system differences and form directly comparable two-end preprocessed data. Preprocessing specifically includes the following sub-steps: S11, the edge computing-based multi-loop detection device and the intelligent single-loop distributed detection terminal immediately append a high-precision timestamp to the current raw leakage current data sample after each data sampling. This timestamp information provides a reference for subsequent strict time alignment.

[0031] S12, the preprocessing unit (which can be located within the edge computing unit or a separate preprocessing module) processes the raw leakage current data streams from the distribution cabinet side and the light pole side based on the timestamps added in S11. It aligns the sampling points of a circuit on the distribution cabinet side with the sampling points of all light poles belonging to that circuit that are closest in time, forming time-synchronized data pairs or segments. This process aims to eliminate time offsets caused by minor communication delays or differences in sampling start times, ensuring that the comparison is of the electrical state at the same moment or within the same time window. The aligned data constitutes time-synchronized data.

[0032] S13. Convert the obtained time synchronization data into a unified preset data format within the system. This format specifies the structure of the data packets, including fields such as device ID, timestamp, leakage current value (this value is the precise value after processing by the wide-range high-precision collaborative algorithm described below, with the unit uniformly in milliamperes (mA)), and data quality identifier. Format conversion ensures that data from both the distribution cabinet side and the light pole side can be consistently parsed and used in subsequent processing. The standardized data output after this step is the dual-end preprocessed data. This data is the direct input for all subsequent analyses, including local fusion, feature extraction, and cloud pre-evaluation.

[0033] It should be noted that in the acquisition and initial processing stage of step S1, in order to fundamentally ensure high-precision data acquisition across the entire range from weak leakage current to large short-circuit current, both the edge computing multi-loop detection device and the intelligent single-loop distributed detection terminal in this embodiment have built-in and run a wide-range high-precision collaborative algorithm. This algorithm is achieved through the coordinated operation of piecewise linear correction and adaptive gain control: 1. Piecewise Linear Calibration: The detection range (e.g., 10mA to 50A) is divided into multiple sub-intervals for precise calibration. In a preferred embodiment, it is divided into three intervals: interval 1 (10mA~1A), interval 2 (1A~30A), and interval 3 (30A~50A). Each interval is determined using a low-squares fit after high-precision calibration before leaving the factory.

[0034] in, The leakage current value (unit: mA or A, depending on the range) is the output after piecewise linear correction and is the precise value used for risk assessment. The leakage current value (unit: mA or A) is the original value collected by the sensor for edge computing multi-loop detection device or intelligent single-loop distributed detection terminal. It has not been calibrated and may contain systematic errors. The slope coefficient for the corresponding interval is used to correct the proportional error of the original collected values. Specific k-values ​​are obtained for different intervals through fitting calculations (e.g., ...). ); The intercept coefficient (unit: mA or A) for the corresponding interval is used to correct the fixed error of the original collected values. Specific b-values ​​are obtained for different intervals through fitting calculations (e.g., ...). The method for determining the coefficients is as follows: During the high-precision calibration process before the device leaves the factory, multiple standard current points are selected for each interval (for example, 5 points are selected for interval 1: 10 mA, 50 mA, 100 mA, 500 mA, 1 A; 3 points are selected for interval 2: 1 A, 10 A, 30 A; 2 points are selected for interval 3: 30 A, 50 A). For each interval, these theoretical values are input using a standard current source , and the corresponding original acquisition values of the device are recorded .

[0035] To obtain the optimal correction coefficients and , the least squares method is used for fitting, and its goal is to minimize the fitting error E. The calculation formula for the fitting error E is:

[0036] where the summation traverses all standard points within this interval. The value of E reflects the overall deviation between the predicted value of the correction equation and the true standard value. The smaller E is, the higher the fitting accuracy of the correction equation in this interval. By solving for the and that minimize E, the optimal correction coefficients for this interval can be obtained.

[0037] For example: Interval 1 uses coefficients k1 (preferred range 0.99 - 1.01) and b1 (preferred range -0.1 - 0.1 mA), obtained by fitting through multiple standard points (such as 10 mA, 50 mA, 100 mA, 500 mA, 1 A), to correct the zero point and proportional error in the small current segment.

[0038] Interval 2 uses coefficients k2 (preferred range 0.995 - 1.005) and b2 (preferred range -1 - 1 mA).

[0039] Interval 3 uses coefficients k3 (preferred range 0.98 - 1.02) and b3 (preferred range -10 - 10 mA).

[0040] The device monitors the value in real time, and automatically calls the correction coefficients of its所属区间 and calculates to output a high-precision value.

[0041] 2. Adaptive gain control: In the analog front end of signal acquisition, the signal amplification factor is dynamically adjusted through a programmable gain amplifier to ensure that leakage signals of different intensities can be sampled by the analog-to-digital converter with the best resolution. The gain switching logic is based on the original signal voltage U 原始 : When U 原始 < U1 (such as 5 mV), a high gain G1 (such as 1000 times) is adopted.

[0042] When U1 ≤ U 原始 < U2 (e.g., 50 mV), medium gain G2 (e.g., 100 times) is adopted.

[0043] When U 原始 ≥ U2, low gain G3 (e.g., 10 times) is adopted.

[0044] The gain switching is quickly responsive by hardware (e.g., ≤ 10 μs), ensuring immediate adaptation even when the signal mutates.

[0045] Through the above collaborative algorithm processing, the accuracy and reliability of the data are improved from the source, laying a solid data foundation for subsequent collaborative verification and risk determination.

[0046] Continue to refer to Figure 1 , a method for monitoring and protecting street lamp leakage based on distributed collaboration proposed by the present invention further includes the following steps: S2. The edge computing unit in the edge computing type multi-loop detection device performs local fusion and anti-interference filtering processing on the double-end preprocessed data, obtains the purified multi-source leakage current data, and extracts key features from the purified multi-source leakage current data. The key features include the current leakage current value and the leakage current growth rate.

[0047] In a specific embodiment, step S2 is a key link for the edge computing unit to exert its local intelligent processing ability. Its purpose is to convert the preliminarily aligned raw data into high-quality and information-rich features that can be used for intelligent decision-making. The specific implementation process preferably includes the following sub-steps: S21. Perform real-time fusion on the double-end preprocessed data to obtain a multi-source fusion data stream. The edge computing unit receives the double-end preprocessed data from step S1, which includes the loop leakage current preprocessed data from the power distribution cabinet side and the branch leakage current preprocessed data from each lamp post side. The edge computing unit converges and associates these multi-source data with the device's own operating state data (such as working voltage, temperature) under a unified time reference, forming a multi-source fusion data stream containing the complete space (loop and branch) and state context. This step lays a foundation for analyzing the correlation between the overall loop and the details of each branch in the same time frame.

[0048] S22. Adopt the wavelet transform algorithm to perform preliminary filtering on the multi-source fusion data stream to obtain the first intermediate data. Due to the complex on-site electromagnetic environment, the data stream contains high-frequency noise and instantaneous pulse interference. To effectively strip these interferences and retain the true leakage trend, this step uses the wavelet transform algorithm for processing. Its mathematical expression is:

[0049] Where x(t) represents the leakage current signal in the original multi-source fused data stream, and ψ(t) is the selected wavelet basis function. 'b' is the scaling factor (controlling the frequency analysis range), and 'a' is the translation factor (controlling the time position). By adjusting the scaling factor... The algorithm can analyze the characteristics of a signal in different frequency bands, thereby effectively separating and filtering out high-frequency noise. The signal output after this step is called the first intermediate data, which has filtered out most of the high-frequency interference and retained the main trend of leakage current variation.

[0050] S23. A sliding window filtering algorithm is used to smooth the first intermediate data to obtain the second intermediate data. To further suppress random fluctuations in the data and make the trend smoother and clearer, a sliding window filter is applied to the first intermediate data. The calculation formula is as follows:

[0051] in, It is the output value at the k-th time, and N is the preset sliding window length (e.g., it can be 5~10). This is the input value at time i within the window (i.e., the first intermediate data). The algorithm effectively smooths out random fluctuations by calculating the average value of data within a time window near the current time, and the resulting output is called the second intermediate data, whose curve is smoother.

[0052] S24. Based on preset outlier removal rules, the second intermediate data is filtered to obtain purified multi-source leakage current data. To further eliminate outlier data points that deviate significantly from the normal range due to sensor transient failures or extreme interference, this step uses statistical principles for filtering. The 3σ criterion is preferably used: for the second intermediate data within a sliding time window, its mean μ and standard deviation σ are calculated. If a certain data point... Satisfy | If -μ|>3σ, it is considered an outlier and removed. The data obtained after this rigorous screening process is the high-quality, purified multi-source leakage current data ultimately used for decision-making.

[0053] S25. Calculate and extract key features from the purified multi-source leakage current data. These key features include the current leakage current value, leakage current growth rate, and leakage current fluctuation amplitude. Based on the purified multi-source leakage current data, the edge computing unit calculates and extracts key features used to quantitatively characterize the leakage current state, evolution trend, and stability. These features are the core data inputs for subsequent collaborative verification triggering and risk level determination, specifically including: Current leakage current value (I): The leakage current data after purification is taken directly from the latest moment for a specific circuit or branch.

[0054] Leakage current growth rate (ΔI / Δt): Calculates the average rate of change of the leakage current value over a specified recent time period (e.g., the last 5 minutes, the duration of which is configurable). This feature is used to determine whether the leakage current is in a steady state, accumulating slowly, or accelerating.

[0055] Leakage current fluctuation amplitude (ΔI) max This feature calculates the absolute value of the maximum positive or negative deviation of the leakage current from its average value within a preset observation time window (e.g., 10 consecutive sampling periods), or the difference between the maximum and minimum values ​​within that window. This feature helps identify unstable leakage current modes caused by poor line contact, intermittent arcing, etc., where the current value may fluctuate drastically rather than change steadily.

[0056] These extracted key features are updated in real time and cached in the local edge computing unit. Together, they constitute the most direct and core data basis for triggering the collaborative verification process (step S3), performing local high-risk determination (step S32), and conducting comprehensive risk level assessment (step S4).

[0057] Continue to refer to Figure 1 The present invention proposes a street light leakage current monitoring and protection method based on distributed collaboration, which further includes the following steps: S3. When the edge computing unit determines that the collaborative verification triggering condition is met based on key features, it performs dual-end collaborative verification on the distribution cabinet side and the light pole side. The dual-end collaborative verification includes: S31. The cloud platform performs a pre-assessment of leakage risk based on the pre-processed data from both ends and in combination with environmental factors, and generates a pre-assessment result. S32. When the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on key features, it calls the pre-evaluation result and cross-verifies it with the local cached historical data of the corresponding circuit on the lamp post side to generate a collaborative verification result.

[0058] In one specific embodiment, step S3 is the core of the collaborative intelligence of the present invention, which aims to significantly improve the accuracy of high-risk leakage current determination and avoid blind power outages through a dual-insurance mechanism combining cloud pre-assessment and edge real-time verification. The execution of this step is triggered by the edge computing unit based on the key features extracted in step S2, and is divided into three stages: normal data synchronization, cloud pre-assessment, and edge verification.

[0059] First, there is the normal data synchronization phase (providing a continuous data stream to S31). This phase is not executed only when the triggering conditions are met, but is a continuous background process. The edge computing unit and each intelligent single-loop distributed detection terminal collect leakage current data at fixed intervals (e.g., 400 milliseconds), obtain dual-end preprocessed data, and cache it locally. The caching time is usually no less than 30 minutes.

[0060] Under normal operation, the system dynamically adjusts the cycle of uploading preprocessed data from both ends to the cloud platform based on the key features extracted in step S2, namely the current leakage current value (I) and the leakage current growth rate (ΔI / Δt), thereby achieving adaptive matching between communication resources and risk perception. The specific dynamic upload strategy is as follows:

[0061] The first, second, and third upload cycles are preset values ​​with progressively decreasing durations (e.g., 30 minutes, 15 minutes, and 3 minutes respectively). The strategy works as follows: when key features indicate the system is in a low-risk steady state, a longer first upload cycle is used to optimize communication resource consumption; when key features indicate an increased risk level, a moderate second upload cycle is used; and once the risk level indicated by key features reaches or exceeds a higher threshold, the system immediately switches to a shorter third upload cycle. This mechanism ensures that the cloud platform, when performing risk pre-assessment (S31), can analyze data based on a data update frequency that matches the risk level, thereby significantly improving the accuracy and timeliness of the pre-assessment results.

[0062] Next, S31 is executed. The cloud platform performs a pre-assessment of leakage risk based on the pre-processed data from both ends and in combination with environmental factors, and generates a pre-assessment result.

[0063] After receiving the above data, the cloud platform activates its intelligent evaluation model. The specific implementation includes the following sub-steps: Environmental factors Weighted Calculation: Introducing Environmental Factors into Cloud Platforms The assessment benchmark is dynamically adjusted to take into account the impact of different external conditions on leakage risk. The calculation formula is shown in the following example:

[0064] in, Value 1 when the ambient humidity is >80%, otherwise value 0. If the line has been built for more than 10 years, use 1; otherwise, use 0. Take 1 during off-peak electricity usage periods at night (e.g., 22:00-6:00), otherwise take 0. The smaller the value, the higher the risk of leakage due to environmental factors.

[0065] Trend prediction model: Based on recently uploaded (e.g., within the last hour) preprocessed data from both ends, the cloud uses a linear regression method to fit the leakage current growth curve. .in, Let be the current leakage current value, and k be the fitted growth rate (ΔI / Δt). Using this model, the leakage current value can be predicted after a certain period of time (e.g., 10 minutes). And calculate the risk probability. , This represents the Level IV protection threshold for the corresponding environment.

[0066] Comprehensive Risk Assessment and Result Generation: The cloud platform integrates environmental factors and trend prediction results to calculate a comprehensive risk value S. Based on the value of S, a preliminary assessment result is generated in the cloud, which contains at least two key pieces of information: Risk level prediction values: For example, S<30 is Level I (no risk), 30≤S<50 is Level II (low risk), 50≤S<70 is Level III (medium risk), and S≥70 is Level IV (high risk).

[0067] Environmental Adaptive Risk Threshold The protection threshold is obtained based on the environmental factor correction standard. This threshold reflects a more reasonable protection action trigger line under the current environment.

[0068] The pre-assessment results will be sent to the edge computing units in the corresponding areas in real time.

[0069] Finally, in step S32, when the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on key features, the pre-evaluation result is called and cross-validated with the local cached historical data of the corresponding loop on the lamp post side to generate a collaborative verification result.

[0070] This step is the final reliability check of the edge computing unit before it executes the final power-off protection. The edge computing unit performs this check based on the current leakage current value in the key features. When it is determined that the risk level has reached or exceeded the locally preset fixed high-risk threshold (e.g., 1A), the system does not act immediately, but instead initiates the following verification process: Step 1: Retrieve and compare the cloud-based pre-assessment results: The edge computing unit immediately retrieves the most recently issued pre-assessment results and compares them with the local detection values. Adaptive risk thresholds for the environment distributed from the cloud Compare and calculate the deviation rate If the risk level predicted by the cloud has reached high risk (Level IV) and the deviation rate is... If the deviation is within an acceptable range (e.g., ≤10%), proceed to the next verification step; if the deviation is too large (e.g., >20%), pause the process and request an update assessment from the cloud.

[0071] Step 2: Cross-validate with the historical data from the light poles stored locally. If the first step's comparison passes, the edge computing unit then retrieves the leakage current values ​​uploaded by each intelligent single-loop distributed detection terminal corresponding to the same loop within a recent period (e.g., 5 minutes) from the locally cached data. The sum of these historical leakage current values ​​from the light poles is then calculated. And compared with the total leakage current value of the circuit currently being detected locally. Compare the two and calculate the consistency deviation rate. The calculation formula is:

[0072] Meanwhile, the edge computing unit determines the increasing trend of the total leakage current of the light pole. The growth trend of local loop data Whether they are consistent, that is, whether the deviation between the two is within a preset range (e.g., ≤20%). If the percentage is ≤ 15% and the growth trends of both are consistent, the verification is considered passed; otherwise, the verification is considered failed.

[0073] Through the rigorous S3 steps described above, the system achieves dual verification of macro-trend analysis in the cloud and real-time micro-data at the edge, providing a highly confident basis for the final protection decision.

[0074] Continue to refer to Figure 1 The present invention proposes a street light leakage current monitoring and protection method based on distributed collaboration, which further includes the following steps: S4. The edge computing unit performs local risk assessment and handling, including: if the collaborative verification result shows that the verification is passed, then the circuit breaker protection action is executed; otherwise, the leakage hazard level is determined based on key characteristics and the corresponding early warning action is executed.

[0075] In one specific embodiment, step S4 is the final decision-making step for this method to achieve differentiated and intelligent responses. Based on the collaborative verification results generated in step S3 and the key features extracted in step S2, the edge computing unit performs the final risk assessment and triggers precisely corresponding handling actions. This step is implemented as follows: The edge computing unit first checks the collaborative verification result generated in step S32, which clearly indicates whether the verification has passed.

[0076] If the collaborative verification result shows that the verification is successful, it means that the excessively high leakage current (e.g., ≥1A) detected on the distribution cabinet side has been confirmed by both the cloud-based pre-assessment result comparison and the cross-verification of historical data on the light pole side, and is determined to be a real, urgent, high-risk leakage fault. In this case, the edge computing unit no longer performs complex level determination, but directly executes the circuit breaker protection action. Specifically, the edge computing unit, through its integrated control interface (such as a relay or solid-state switch), issues a trip command to the corresponding circuit breaker within a very short time (e.g., ≤3 seconds), cutting off the power supply to the faulty circuit, thereby immediately eliminating the risk of electric shock and fire. After executing the protection action, the edge computing unit generates a "protection action executed" confirmation message and uploads it to the cloud-based intelligent monitoring platform along with a detailed decision log (including trigger thresholds, verification process, action time, etc.), completing the closed-loop recording of the high-risk event.

[0077] Otherwise (i.e., collaborative verification failed, or the collaborative verification process was not triggered): This covers two situations: (a) the high-risk threshold was reached but the verification failed; (b) the high-risk threshold was not reached, and only the general collaborative verification triggering conditions were met. In these two situations, the system determines that it does not belong to the emergency high-risk category that requires immediate power outage, and instead enters the refined risk level determination process to output a non-high-risk leakage hazard level.

[0078] The risk level determination process is executed locally by the edge computing unit, and its core is a pre-stored risk level determination matrix. This matrix defines multiple hazard levels (e.g., Level I no risk, Level II low risk, Level III medium risk, Level IV high risk, Level V false alarm) and their corresponding multi-dimensional determination conditions. The determination process mainly relies on the following inputs: Key features: the current leakage current value and leakage current growth rate extracted from step S2.

[0079] Duration: Calculates the duration of the current leakage state (current value exceeding a certain reference threshold).

[0080] Derivative information from collaborative validation results, namely collaborative confidence: When collaborative validation is involved, a quantified confidence score C is calculated. The calculation of the C value integrates data consistency, trend consistency, and cloud assessment consistency. An example of its calculation formula is shown below: C = 0.4 × C1+ 0.3 × C2+ 0.3 × C3 Where C1 is the data consistency confidence level, according to (The deviation rate between the sum of light pole data and the local detection value) is calculated when... When ≤12%, C1=100%. When the value is greater than 20%, C1 = 0%, and linear interpolation is used in between.

[0081] C2 is the confidence level of trend consistency, which is calculated based on the deviation between the local and pole data growth trends. When the deviation ≤ 20%, C2 = 100%; when the deviation > 50%, C2 = 0%; and linear interpolation is used in between.

[0082] C3 is the pre - evaluation confidence level in the cloud, which is calculated based on the comprehensive risk value S issued by the cloud. When S ≥ 70, C3 = 100%; when S < 30, C3 = 0%; and linear interpolation is used in between.

[0083] The edge computing unit compares the above input parameters with the conditions in the decision matrix. For example, the matrix may stipulate: Level II (low risk): It is satisfied when 100 mA ≤ current value < 300 mA, the duration < 24 hours, the growth rate < 0.1 A / h, and at the same time 60% < C < 80%.

[0084] Level III (medium risk): It is satisfied when 300 mA ≤ current value < 1 A; or 100 mA ≤ current value < 300 mA but the duration ≥ 24 hours and the growth rate = 0.1 - 0.5 A / h, and at the same time C ≥ 80%.

[0085] Level V (false alarm): It is satisfied when the detected value of the power distribution cabinet ≥ 300 mA but C ≤ 30%.

[0086] Based on the comparison result, the edge computing unit outputs a definite leakage hazard level (such as Level II, Level III or Level V).

[0087] Finally, the corresponding early warning and disposal are executed. The edge computing unit executes the preset differential disposal strategy according to the determined leakage hazard level. The core of the early warning and disposal is to generate and upload the early warning information corresponding to the hazard level. For example: For Level II (low risk), generate a "low - risk early warning" notice and upload it to the cloud platform through the communication module (such as 4G) to prompt the operation and maintenance personnel to pay attention and conduct planned inspections.

[0088] For Level III (medium risk), generate a "medium - risk alarm". In addition to uploading it to the cloud, it can also trigger the local audible and visual alarm (if equipped) and require immediate arrangement for investigation.

[0089] For Level V (false alarm), generate a "suspected interference / false alarm" log and upload it to the cloud for system self - learning and algorithm optimization. At the same time, do not trigger any on - site alarm to avoid resource waste.

[0090] All early warning information includes detailed information such as the specific device location (circuit / pole ID), determination level, key feature values, determination time, etc., thus forming a complete non - high - risk event disposal closed - loop from monitoring discovery to intelligent determination and then to accurate early warning.

[0091] For further reference Figure 2 As an implementation of the above method, this invention also proposes an embodiment of a distributed collaborative street light leakage current monitoring and protection system 200, which can be applied to various electronic devices. The distributed collaborative street light leakage current monitoring and protection system 200 includes the following modules: The edge computing multi-loop detection device 210 is deployed on the power distribution cabinet side to synchronously collect raw leakage current data of the power supply loop, preprocess the raw leakage current data to obtain a part of the double-end preprocessed data, perform local fusion and anti-interference filtering on the double-end preprocessed data to obtain purified multi-source leakage current data, extract key features from the purified multi-source leakage current data, determine whether the collaborative verification triggering condition is met or is greater than or equal to the local high-risk threshold based on the key features, call the pre-evaluation results and cross-validate with the locally cached historical data to generate collaborative verification results, and perform risk judgment and handling based on the collaborative verification results or key features. The intelligent single-loop distributed detection terminal 220 is deployed on the side of the street light pole to synchronously collect the raw leakage current data of the corresponding light pole branch and preprocess the raw leakage current data to obtain another part of the double-end preprocessed data. The cloud platform 230 is connected to the edge computing multi-loop detection device 210 and the intelligent single-loop distributed detection terminal 220. It is used to perform leakage risk pre-assessment based on dual-end pre-processed data and combined with environmental factors, and generate and distribute the pre-assessment results.

[0092] In addition to risk pre-assessment, the cloud platform also integrates advanced functions such as intelligent operation and maintenance management and data analysis: Based on the risk level reported by the edge, the cloud platform automatically generates and dispatches tiered work orders (such as emergency, routine, and early warning), pushes them to maintenance personnel via APP or SMS, and tracks the entire handling process, forming a closed-loop management of "alarm-dispatch-handling-archiving".

[0093] Based on long-term stored historical leakage current data, environmental data, and handling records, the cloud platform uses big data analysis models to predict line aging trends and potential fault risks, and sends preventative maintenance suggestions to the operation and maintenance department in advance.

[0094] The cloud platform supports remote configuration and OTA upgrades of algorithm parameters (such as risk thresholds at all levels and collaborative verification rules) and device firmware for edge computing units and detection terminals, enabling centralized and efficient operation and maintenance of large-scale deployed devices.

[0095] The present invention also proposes a terminal device, including a memory, a processor, and a computer program stored in the memory and capable of running on the processor. When the processor executes the computer program, it implements the steps of any of the above-described distributed collaborative street light leakage current monitoring and protection methods.

[0096] The following is for reference. Figure 3 It shows a schematic diagram of the structure of a computer system 300 suitable for implementing terminal devices or servers of the present invention. Figure 3 The terminal device or server shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of the present invention.

[0097] like Figure 3 As shown, the computer system 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes based on programs stored in read-only memory (ROM) 302 or programs loaded from storage section 308 into random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the computer system 300. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0098] The following components are connected to I / O interface 305: an input section 306 including a keyboard, mouse, etc.; an output section 307 including a liquid crystal display (LCD) and speakers, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card and a modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to I / O interface 305 as needed. A removable medium 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 310 as needed so that computer programs read from it can be installed into storage section 308 as needed.

[0099] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by central processing unit (CPU) 301, it performs the functions defined in the methods of this invention. It should be noted that the computer-readable medium described in this invention can be a computer-readable signal medium or a computer-readable medium or any combination thereof. The computer-readable medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this invention, a computer-readable medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable medium that can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0100] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0101] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0102] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention is not limited to the specific combination of the above-described technical features, but also includes other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in this invention.

Claims

1. A method for monitoring and protecting street light leakage current based on distributed collaboration, characterized in that, Includes the following steps: S1. By deploying an edge computing multi-circuit detection device on the side of the power distribution cabinet and an intelligent single-circuit distributed detection terminal on the side of the street light pole, the raw leakage current data of each power supply circuit and the corresponding light pole branch are collected synchronously, and the raw leakage current data is preprocessed to obtain dual-end preprocessed data. S2. The edge computing unit in the edge computing type multi-loop detection device performs local fusion and anti-interference filtering on the dual-end preprocessed data to obtain purified multi-source leakage current data, and extracts key features from the purified multi-source leakage current data, the key features including the current leakage current value and the leakage current growth rate. S3. When the edge computing unit determines that the collaborative verification triggering condition is met based on the key features, it performs dual-end collaborative verification on the distribution cabinet side and the light pole side. The dual-end collaborative verification includes: S31. The cloud platform performs a pre-assessment of leakage risk based on the pre-processed data from both ends and in combination with environmental factors, and generates a pre-assessment result. S32. When the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on the key features, it calls the pre-evaluation result and performs cross-verification with the local cached historical data of the corresponding circuit on the lamp post side to generate a collaborative verification result. S4. The edge computing unit performs local risk assessment and handling, including: if the collaborative verification result shows that the verification is successful, then the circuit breaker protection action is executed; otherwise, based on the key features, the leakage hazard level is determined to be non-high risk, and the corresponding early warning handling is executed.

2. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 1, characterized in that, In step S1, the preprocessing of the raw leakage current data to obtain double-ended preprocessed data includes the following sub-steps: S11. Add a collection timestamp to the raw leakage current data collected from the distribution cabinet side and the light pole side respectively; S12. Based on the collected timestamp, the original leakage current data of the distribution cabinet side and the light pole side are synchronized in time to obtain time synchronization data; S13. Convert the time synchronization data into a unified preset data format to obtain the dual-end preprocessed data.

3. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 1, characterized in that, In step S2, the edge computing unit performs local fusion and anti-interference filtering on the dual-end preprocessed data to obtain purified multi-source leakage current data, and extracts key features from the purified multi-source leakage current data, including the following sub-steps: S21. The dual-end preprocessed data is fused in real time to obtain a multi-source fused data stream, wherein the dual-end preprocessed data includes loop leakage current preprocessed data from the distribution cabinet side and branch leakage current preprocessed data from the light pole side. S22. The multi-source fused data stream is initially filtered using a wavelet transform algorithm to obtain the first intermediate data; S23. The first intermediate data is smoothed using a sliding window filtering algorithm to obtain the second intermediate data; S24. Based on the preset outlier removal rules, the second intermediate data is filtered to obtain purified multi-source leakage current data. S25. Calculate and extract the current leakage current value and the leakage current growth rate from the purified multi-source leakage current data.

4. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 1, characterized in that, In step S31, the cloud platform performs a preliminary assessment of leakage risk based on the pre-processed data from both ends and in conjunction with environmental factors, generating a preliminary assessment result, including the following sub-steps: S311. The cloud platform receives the preprocessed data from both ends and calculates environmental factors based on real-time ambient humidity, line aging years, and operating parameters. S312. The cloud platform obtains leakage current growth trend parameters through linear regression fitting based on the received dual-end preprocessed data. S313. The cloud platform uses the environmental factors to correct the preset standard risk threshold and generate an environmentally adaptive risk threshold. S314. The cloud platform combines the leakage current growth trend parameter with the environmental adaptive risk threshold to calculate a comprehensive risk assessment value; S315. The cloud platform determines the risk level based on the comprehensive risk assessment value and, in conjunction with the environmental adaptive risk threshold, generates the pre-assessment result containing the risk level and the environmental adaptive risk threshold.

5. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 4, characterized in that, In step S32, when the edge computing unit determines that the value is greater than or equal to the local high-risk threshold based on the key features, it calls the pre-evaluation result and performs cross-validation with the locally cached historical data of the corresponding loop on the light pole side to generate a collaborative verification result, including the following sub-steps: S321. When the current leakage current value in the key feature is determined to be greater than or equal to the local high-risk threshold, the edge computing unit calls the pre-evaluation result and calculates the deviation rate between the local current leakage current value and the environmental adaptive risk threshold in the pre-evaluation result. S322. If the deviation rate is less than or equal to the first preset deviation threshold, retrieve the local cached historical data of the corresponding circuit on the lamp post side; calculate the deviation rate between the local leakage current sum and the sum of the historical data on the lamp post side, and determine whether the growth trend of the local leakage current is consistent with the growth trend of the sum of the historical data on the lamp post side. S323. If the deviation rate between the sum of local leakage currents and the sum of historical data on the lamp post side is less than or equal to the second preset deviation threshold, and is determined to be consistent with the growth trend, then a collaborative verification result indicating that the verification has passed is generated; otherwise, a collaborative verification result indicating that the verification has failed is generated.

6. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 1, characterized in that, The execution of step S3 also includes a routine data synchronization step performed before the dual-end collaborative verification: The edge computing unit and the intelligent single-loop distributed detection terminal upload the dual-end preprocessed data to the cloud platform according to a dynamic cycle; wherein, the length of the dynamic cycle is determined based on the current leakage current value and leakage current growth rate in the key features: When the current leakage current value is less than the first threshold and the leakage current growth rate is less than the third threshold, the first upload cycle is adopted. When the current leakage current value is greater than or equal to the first threshold or the leakage current growth rate is greater than or equal to the third threshold, a second upload cycle shorter than the first upload cycle is adopted.

7. The street light leakage current monitoring and protection method based on distributed collaboration according to claim 1, characterized in that, In step S4, the edge computing unit performs local risk assessment and handling, including the following sub-steps: S41. If the collaborative verification result shows that the verification is successful, then the circuit breaker protection action is executed; otherwise, the leakage hazard level is determined based on the current leakage current value, leakage current growth rate and leakage duration in the key features. S42. Based on the determined leakage hazard level, execute the corresponding early warning measures, wherein the early warning measures include generating and uploading early warning information corresponding to the hazard level.

8. A street light leakage current monitoring and protection system based on distributed collaboration, used to implement the method as described in any one of claims 1 to 7, characterized in that, The system includes: An edge computing-based multi-loop detection device is deployed on the power distribution cabinet side to synchronously collect raw leakage current data of the power supply loop. The raw leakage current data is preprocessed to obtain a portion of the dual-end preprocessed data. The dual-end preprocessed data is then fused locally and subjected to anti-interference filtering to obtain purified multi-source leakage current data. Key features are extracted from the purified multi-source leakage current data. Based on the key features, it is determined whether the collaborative verification trigger condition is met or whether the value is greater than or equal to the local high-risk threshold. The pre-evaluation result is called and cross-validated with locally cached historical data to generate a collaborative verification result. Risk judgment and handling are performed based on the collaborative verification result or the key features. An intelligent single-loop distributed detection terminal is deployed on the side of the street light pole to synchronously collect the raw leakage current data of the corresponding light pole branch and preprocess the raw leakage current data to obtain another part of the double-ended preprocessed data. The cloud platform is communicatively connected to the edge computing multi-loop detection device and the intelligent single-loop distributed detection terminal, and is used to perform a leakage risk pre-assessment based on the dual-end preprocessed data and combined with environmental factors, and generate and distribute the pre-assessment results.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the street light leakage current monitoring and protection method based on any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the street light leakage current monitoring and protection method based on any one of claims 1 to 7.

Images