Power distribution network edge analysis method and device for passive tag communication multi-sensor synchronization

Abstract

The application discloses a power distribution network edge analysis method and device for passive tag communication multi-sensor synchronization, and relates to the technical field of intelligent monitoring and data processing of power distribution networks. The method comprises the following steps: collecting tag environment synchronous sensing data in real time, performing data preprocessing, and fusing and behavior mapping of the tag environment synchronous sensing data on a unified time axis; early identifying the weak disturbance characteristics of the tag, constructing a sensitive feature matrix, and feeding back and adjusting the communication and operation state abnormalities; judging the probability of occurrence of a weak fault of the tag, identifying the weak fault of the power equipment, and performing alarm; judging the decision threshold of the weak fault of the tag, and judging whether the equipment exceeds the tolerance fluctuation range. The method solves the problem that when part of the equipment of the power distribution network has an early fault, the passive tag response mode changes slightly but does not constitute a significant anomaly, so that the abnormal state is difficult to identify.

Description

Technical Field

[0001] This invention relates to the field of intelligent monitoring and data processing technology for power distribution networks, specifically to a method and apparatus for edge analysis of power distribution networks using passive tag communication and multi-sensor synchronization. Background Technology

[0002] In smart distribution networks, passive tags are widely used for equipment status sensing and identification due to their advantages of requiring no external power supply, low cost, and ease of deployment. However, with the expansion of system scale and the increasing complexity of operating environments, traditional tag response data often faces challenges such as difficulty in identifying weak disturbances and difficulty in quantifying minor fault states. To improve the accuracy and response efficiency of equipment status monitoring, constructing a multi-source fusion analysis mechanism that integrates tag communication characteristics and operating status data has become a key path to achieving edge intelligent diagnosis, predictive maintenance, and refined operation and maintenance management.

[0003] For example, invention patent CN118260233A discloses a passive tag I2C bus driver circuit, a passive chip, and a communication system. This invention discloses a passive tag I2C bus driver circuit integrated within a chip, including an I2C interface module configured with a data input terminal and a data output terminal. The data input terminal receives data from a sensor; the data output terminal transmits data from the chip to the sensor; an inverting tri-state gate circuit controls the data transmission direction under the action of the enable signal EN; when the enable signal EN is enabled, it controls the transmission of data from the chip to the sensor; when the enable signal EN is disabled, it controls the transmission of data from the sensor to the chip or maintains a high impedance state; a pull-up module pulls the I2C bus level to the power supply voltage; and an amplification and detection circuit processes the data transmitted from the I2C bus before transmitting it to the I2C interface module. This circuit design solves the limitation of the speed-power consumption contradiction in I2C bus communication, enabling the chip to be applied to passive IoT self-powered nodes.

[0004] However, in smart distribution network scenarios, when some devices experience early faults such as slight loosening, unstable connections, or cable aging, the communication response parameters of the tags will exhibit subtle but continuous fluctuations. These fluctuations are often on the edge of the normal fluctuation range, making them difficult to accurately identify using traditional fixed threshold models or one-time detection mechanisms. Because passive tags are highly sensitive to environmental changes, these micro-perturbations may manifest as intermittent weak signals, decreased signal-to-noise ratio, or slight jitter in response delay, easily confused with normal noise and read / write interference, leading to misjudgments or missed detections.

[0005] Therefore, in order to address the above problems, there is an urgent need for a power distribution network edge analysis method and device for passive tag communication and multi-sensor synchronization. Summary of the Invention

[0006] Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a method and apparatus for edge analysis of power distribution networks using passive tag communication and multi-sensor synchronization. This solves the problem that when some equipment in the power distribution network experiences early-stage faults, the response modes of passive tags change slightly but do not constitute significant anomalies, making it difficult to identify abnormal states.

[0008] Technical solution

[0009] To achieve the above objectives, this invention provides the following technical solution: a power distribution network edge analysis method for passive tag communication multi-sensor synchronization, comprising the following steps: real-time acquisition of tag environment synchronization sensing data; data preprocessing of the tag environment synchronization sensing data and establishment of a unified time axis; fusion and behavior mapping of the tag environment synchronization sensing data on the unified time axis; early identification of weak disturbance characteristics of tags based on the tag environment synchronization sensing data on the unified time axis; construction of a sensitive feature matrix based on the early identification results of weak disturbance characteristics; feedback adjustment for communication and operation status anomalies; determination of the probability of tag weak faults occurring using the tag environment synchronization sensing data in conjunction with the sensitive feature matrix; identification of weak faults in power equipment and alarm based on the determination results of the tag weak fault occurrence probability; fusion of the determination results of the tag weak fault occurrence probability and the tag environment synchronization sensing data to determine the decision threshold of tag weak faults; and determination of whether the equipment exceeds the tolerable fluctuation range based on the tag weak fault decision threshold.

[0010] Furthermore, the real-time acquisition of tag environment synchronous sensing data, the preprocessing of this data, and the establishment of a unified timeline for the fusion and behavior mapping of the tag environment synchronous sensing data on this unified timeline are as follows: Real-time acquisition of tag environment synchronous sensing data is performed using passive tag readers and edge sensor nodes. This data includes the tag's RSSI, phase, response delay, and the voltage, current, and temperature of the device's operating status. All passive tags are uniformly numbered, and GPS timing is used to precisely synchronize the acquisition time of each sensor, adding a unified and precise timestamp to each tag environment synchronous sensing data point. The system sorts all tag environment synchronization sensing data according to timestamps to establish a unified timeline; it performs outlier removal and missing value imputation operations, and uses moving average filtering for data noise reduction and smoothing; it also standardizes and normalizes the tag environment synchronization sensing data; a tag environment synchronization sensing database is established to store the tag environment synchronization sensing data; on the unified timeline, each moment corresponds to a tag environment synchronization sensing data record; mutual information correlation analysis is used to identify the correlation between tag communication behavior characteristics and device operating status; and regression analysis is used to establish the mapping relationship between tag communication behavior characteristics and device operating status.

[0011] Furthermore, based on the synchronous sensing data of the tag environment on a unified time axis, the specific process for early identification of weak perturbation features of the tags is as follows: Divide the time axis into time windows, acquire RSSI, phase, and response delay data for each time window to obtain RSSI, phase, and response delay sequences; calculate the standard deviation of the phase sequence, and simultaneously calculate the sample entropy value of the RSSI sequence using an entropy algorithm after acquiring the RSSI sequence; acquire the response delay at the current time and the previous time, divide the difference between the response delay at the current time and the response delay at the previous time by the response delay at the previous time to obtain the response delay change rate, and calculate the standard deviation of the response delay sequence; multiply the standard deviation of the phase sequence, the sample entropy value of the RSSI sequence, and the response delay change rate by 1 to obtain the perturbation change intensity value; add the standard deviation of the response delay sequence to 1 to obtain the delay stability value; divide the perturbation change intensity value by the delay stability value to obtain the weak feature sensitivity value.

[0012] Furthermore, based on the early identification results of weak disturbance characteristics, a sensitive feature matrix is ​​constructed. The specific process for feedback adjustment of communication and operational status anomalies is as follows: The weak feature sensitivity value for each time window is calculated, and the timestamp, standard deviation of the phase sequence, sample entropy value of the RSSI sequence, response delay change rate, standard deviation of the response delay sequence, and weak feature sensitivity value within each time window are extracted to construct a sensitive feature matrix, which is then written into the edge device. The real-time calculated weak feature sensitivity value is compared with a dynamic sensitivity threshold. If the weak feature sensitivity value is greater than the dynamic sensitivity threshold, it is marked as an early abnormal state. The weak feature sensitivity value is resampled and recalculated every 20 seconds. If the weak feature sensitivity value drops, the current anomaly mark is automatically removed. If weak signal and path instability are identified, the antenna power is increased and the angle is adjusted. If the node current and voltage fluctuations are synchronized with the increase in the weak feature sensitivity value, the edge device actively triggers voltage stabilization and load limiting commands. Simultaneously, the daily sensitive feature matrix and early abnormal state marks are written into the tag environment synchronous perception database.

[0013] Furthermore, combining the sensitive feature matrix and utilizing the tag environment synchronous sensing data, the specific process for determining the probability of tag weak faults is as follows: The sensitive feature matrix is ​​received, and the RSSI, phase, and response delay sequences of the current time window are obtained. The maximum and minimum values ​​of the RSSI in the current time window are filtered, and the difference is calculated to obtain the RSSI fluctuation amplitude. Simultaneously, the average value of the RSSI in the current time window is calculated. The maximum and minimum values ​​of the phase in the current time window are filtered, and the difference is calculated to obtain the phase fluctuation amplitude. Simultaneously, the average value of the phase in the current time window is calculated. The maximum and minimum values ​​of the response delay in the current time window are filtered again, and the difference is calculated to obtain the response delay fluctuation amplitude. Simultaneously, the response delay in the current time window is calculated. The average delay is calculated as follows: the signal strength fluctuation value is obtained by dividing the RSSI fluctuation amplitude by the average RSSI and then multiplying it by the signal strength fluctuation weighting factor; the phase stability value is obtained by dividing the phase fluctuation amplitude by the average phase and then multiplying it by the phase stability weighting factor; the response delay fluctuation value is obtained by dividing the response delay fluctuation amplitude by the average response delay and then multiplying it by the communication delay fluctuation weighting factor; the signal strength fluctuation value, phase stability value, and response delay fluctuation value are added together to obtain the total disturbance metric value; the negative number of the total disturbance metric value is taken as the exponent, and natural exponentiation is performed to obtain the exponential decay term; the exponential decay term is added by 1 and the reciprocal is taken to obtain the tag weak fault behavior probability value.

[0014] Furthermore, the specific process of identifying and alarming weak faults in power equipment based on the probability of occurrence of tag-based weak faults is as follows: the tag-based weak fault behavior probability value is compared with the fault threshold in real time on the local edge device. If the tag-based weak fault behavior probability value is greater than the fault threshold, it is automatically marked as a weak fault state. If the weak fault behavior probability value shows an upward trend in multiple consecutive time windows, a real-time alarm mechanism will be triggered and pushed to the operation and maintenance terminal. All tag-based weak fault behavior probability values, tag numbers, timestamps, and trigger result information are synchronously written into the tag environment synchronous perception database. The database is analyzed and identified in real time to identify long-term evolution trends, frequent fluctuation points, and potential risk areas, enabling predictive maintenance, automatic task scheduling, and maintenance priority ranking.

[0015] Furthermore, the specific process of integrating the judgment result of the tag weak fault occurrence probability with the tag environment synchronous perception data to determine the tag weak fault decision threshold is as follows: real-time acquisition of tag weak fault behavior probability values ​​within past time windows, and calculation of the mean and standard deviation of tag weak fault behavior probability values ​​respectively; acquisition of voltage, current, and temperature data, calculation of the variance of voltage, current, and temperature data respectively, and summing of the calculated variances of voltage, current, and temperature data to obtain the state fluctuation variance; addition of the mean and standard deviation of tag weak fault behavior probability values ​​to obtain the tag weak fault tolerance upper limit, and division of the tag weak fault tolerance upper limit by the sum of the state fluctuation variance and 1 to obtain the tag weak fault decision value.

[0016] Furthermore, the specific process for determining whether the equipment exceeds the tolerable fluctuation range based on the decision threshold of tag weak faults is as follows: In each real-time monitoring window period, the current tag weak fault behavior probability value is compared with the tag weak fault decision value calculated at the corresponding time point. If the tag weak fault behavior probability value is less than or equal to the tag weak fault decision value, it indicates that the equipment is within the acceptable fluctuation range, and monitoring continues without response. If the tag weak fault behavior probability value is greater than the tag weak fault decision value, it indicates that the tag behavior has exceeded the current tolerable fluctuation limit, i.e., the weak fault state is upgraded to an alarm state, immediately triggering the local early warning mechanism and initiating alarm output. If the tag weak fault behavior probability value is greater than the tag weak fault decision value for three consecutive periods, the task is prioritized for inspection. If tag angle offset or environmental reflection interference is detected, the antenna transmission power, frequency, and time slot are automatically adjusted. If minor loosening or ringing is identified... If environmental fluctuations cause anomalies, the tag reading interval is extended to avoid excessive polling. Simultaneously, the current tag number, timestamp, tag weak fault behavior probability value, and tag weak fault decision value are recorded to the edge device. The obtained weak feature sensitivity values, tag weak fault behavior probability values, and tag weak fault decision values ​​are structured, integrated, and graphically presented for transparent monitoring of device status and operational support. Tag status and tag weak fault behavior probability value trends are displayed through multiple layers, supporting tag search, status filtering, historical trend tracking, and real-time alarm response via an interactive interface, and also supporting manual annotation and false alarm correction. After each warning is triggered, an automatic pop-up notification and linkage to the work order system are generated, and operation feedback is recorded. Furthermore, equipment operation stability reports and operational priority suggestions are generated regularly, comprehensively considering high-incidence weak fault points and regional operational health status to construct a closed-loop mechanism for alarm response, trend analysis, and predictive maintenance.

[0017] The second aspect of this invention provides a power distribution network edge analysis device for passive tag communication multi-sensor synchronization, comprising: a data synchronization acquisition and fusion module, used to acquire tag environment synchronization sensing data in real time, preprocess the tag environment synchronization sensing data, establish a unified time axis, and perform tag environment synchronization sensing data fusion and behavior mapping on the unified time axis; a weak feature early extraction module, used to identify weak disturbance features of tags early based on tag environment synchronization sensing data on the unified time axis, construct a sensitive feature matrix based on the early identification results of weak disturbance features, and provide feedback adjustment for communication and operation status anomalies; a tag weak fault intelligent behavior recognition module, used to combine the sensitive feature matrix and tag environment synchronization sensing data to determine the probability of tag weak fault occurrence, identify weak faults in power equipment and issue alarms based on the determination results of tag weak fault occurrence probability; and an edge dynamic decision and early warning module, used to fuse the determination results of tag weak fault occurrence probability with tag environment synchronization sensing data, determine the decision threshold of tag weak fault, and determine whether the equipment exceeds the tolerable fluctuation range based on the tag weak fault decision threshold.

[0018] A third aspect of the present invention provides a storage medium for power distribution network edge analysis with passive tag communication and multi-sensor synchronization, comprising: the storage medium having one or more programs, the one or more programs being executed by one or more processors to implement the power distribution network edge analysis method with passive tag communication and multi-sensor synchronization as described in claims 1-8.

[0019] Beneficial effects

[0020] The present invention has the following beneficial effects:

[0021] (1) This invention integrates passive tag communication features such as RSSI, phase, response delay and equipment operating status such as voltage, current and temperature data, and combines a unified time axis and a multi-dimensional feature matrix to accurately identify minor disturbance-level anomalies such as slight loosening, unstable connection and cable aging, so as to achieve early capture and marking of weak fault features and provide a first-hand signal for predictive maintenance.

[0022] (2) This invention avoids misjudgment and missed reporting caused by fixed thresholds by combining the upper limit of the label's weak fault tolerance with the dynamic label weak fault decision value of the state fluctuation; and can automatically adjust the tolerance boundary according to different environmental conditions such as temperature rise and current fluctuation, so as to realize flexible fault-tolerant judgment and dynamic early warning triggering of equipment status, with strong adaptability and low false alarm rate.

[0023] (3) This invention improves stability by integrating edge self-rescue and intelligent response mechanisms: when a weak fault is detected but does not reach the strong fault threshold, self-rescue measures such as channel adaptation, tag resampling, and reading interval adjustment can be implemented to alleviate mild disturbances without human intervention, effectively prevent fault spread and communication interruption, and make the equipment operation more stable.

[0024] (4) This invention displays the label status, disturbance trend, early warning record and operation and maintenance feedback through a visual interface. It supports status filtering, historical tracking, manual annotation and work order linkage. Combined with stability assessment report and priority ranking suggestions, it forms a closed loop of intelligent operation and maintenance process from identification, judgment, response and backtracking, which greatly improves inspection efficiency and the scientific nature of resource scheduling.

[0025] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0026] Figure 1 Flowchart of a power distribution network edge analysis method for multi-sensor synchronization of passive tag communication;

[0027] Figure 2 A schematic diagram of a power distribution network edge analysis device for passive tag communication and multi-sensor synchronization;

[0028] Figure 3 This is a trend chart of indicator changes based on label-based weak fault decision values;

[0029] Figure 4 A visualization of intelligent monitoring of weak faults labeled with tags. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. As those skilled in the art will understand, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figures 1-4 This invention provides a technical solution: a method and apparatus for edge analysis of a power distribution network using passive tag communication and multi-sensor synchronization, comprising the following steps: real-time acquisition of tag environment synchronization sensing data; data preprocessing of the tag environment synchronization sensing data and establishment of a unified time axis; fusion and behavior mapping of the tag environment synchronization sensing data on the unified time axis; early identification of weak disturbance characteristics of tags based on the tag environment synchronization sensing data on the unified time axis; construction of a sensitive feature matrix based on the early identification results of weak disturbance characteristics; feedback adjustment for communication and operation status anomalies; determination of the probability of tag weak faults occurring using the tag environment synchronization sensing data in conjunction with the sensitive feature matrix; identification of weak faults in power equipment and alarm based on the determination results of the tag weak fault occurrence probability; fusion of the determination results of the tag weak fault occurrence probability and the tag environment synchronization sensing data to determine the decision threshold of tag weak faults; and determination of whether the equipment exceeds the tolerable fluctuation range based on the decision threshold of tag weak faults.

[0032] Specifically, the process of real-time acquisition of tag environment synchronous sensing data, data preprocessing of the tag environment synchronous sensing data, and establishment of a unified time axis, followed by the fusion and behavior mapping of the tag environment synchronous sensing data on the unified time axis, is as follows: Tag environment synchronous sensing data is acquired in real time through passive tag readers and edge sensor nodes. This data includes the tag's RSSI, phase, response delay, and the voltage, current, and temperature of the device's operating status. All passive tags are uniformly numbered, and the acquisition time of each sensor is precisely synchronized using GPS timing. A unified and precise timestamp is added to each tag environment synchronous sensing data point, and all tag environment synchronous sensing data are sorted according to the timestamps to establish a unified time axis. Outlier removal and missing value imputation are performed, and data noise reduction and smoothing are achieved through moving average filtering. Simultaneously, the tag environment synchronous sensing data is standardized and normalized. A tag environment synchronous sensing data fusion and behavior mapping process is then established. The database stores tag-based environmental synchronous sensing data. On a unified timeline, each moment corresponds to a tag-based environmental synchronous sensing data record. Mutual information correlation analysis is used to identify the correlation between tag communication behavior characteristics and equipment operating status. Mutual information correlation analysis is a statistical method that measures the nonlinear dependency between two variables. By calculating the information gain between their joint distribution and their respective marginal distributions, it reveals the potential coupling patterns between communication characteristics such as RSSI fluctuations and state parameters such as voltage jitter, thereby identifying implicit behavioral connections under nonlinear disturbances. Regression analysis is used to establish a mapping relationship between tag communication behavior characteristics and equipment operating status. The regression analysis uses linear regression, with tag communication behavior characteristics as input variables and equipment operating status as the target output, to train a mapping model that reflects the functional relationship between the two. This model is used to predict the trend of equipment status changes with tag response, improving behavioral explanatory power and state estimation accuracy.

[0033] This implementation scheme, by fusing tag communication characteristics and equipment operating status data on a unified time axis and introducing mutual information correlation analysis and regression analysis methods, can not only accurately identify hidden weak fault behaviors under nonlinear coupling, but also construct a stable and effective feature mapping model to achieve dynamic response prediction of equipment status to communication changes. This multi-source fusion and precise modeling mechanism significantly improves the sensitivity to perturbation changes, the reliability of fault identification, and the foresight of status assessment, laying a solid foundation for building a high-precision, highly adaptable intelligent sensing and operation and maintenance system for the distribution network edge.

[0034] Specifically, the process of early identification of weak perturbation characteristics of tags based on tag-environment synchronous sensing data on a unified time axis is as follows: The unified time axis is divided into time windows, and RSSI, phase, and response delay data are acquired for each time window to obtain RSSI, phase, and response delay sequences. The standard deviation of the phase sequence is calculated, reflecting the dispersion of phase changes during tag communication within that time window. A larger value indicates more severe phase fluctuations, potentially indicating unstable connections or slight changes in tag attitude. Simultaneously, after acquiring the RSSI sequence, the sample entropy value of the RSSI sequence is calculated using an entropy algorithm. Sample entropy is an indicator of the complexity and uncertainty of a time series. By comparing the similarity of adjacent subsequences and the repeatability of statistical patterns, a negative logarithmic value is calculated, reflecting the stability of the sequence. A larger value indicates a more irregular and complex signal; a smaller value indicates a more stable and orderly signal. In passive tag communication scenarios, sample entropy can be used to extract perturbation features of RSSI signals, which is an important indicator for identifying early micro-faults. The response delay at the current and previous timestamps is obtained, and the difference between the current and previous response delays is divided by the previous response delay to obtain the response delay change rate. The response delay change rate is used to capture the degree of abrupt change in tag response delay between adjacent sampling points, suitable for identifying potential communication anomalies caused by link jitter. The standard deviation of the response delay sequence is calculated. Multiplying the standard deviation of the phase sequence, the sample entropy value of the RSSI sequence, and the response delay change rate by 1 yields the perturbation change intensity value. This perturbation change intensity value integrates information from multiple weak anomaly dimensions and is a core indicator characterizing the overall perturbation level of the tag within the current time window. Adding the standard deviation of the response delay sequence to 1 yields the delay stability value. Dividing the perturbation change intensity value by the delay stability value yields the weak feature sensitivity value.

[0035] The specific formula for the weak feature sensitivity value is as follows:

[0036] ;

[0037] In the formula, The weak feature sensitivity value is a comprehensive quantitative indicator that measures the sensitivity to weak perturbation features in the tag response and is used to identify minor early fault signals. The standard deviation of the phase sequence represents the degree of phase jitter and indicates whether the electromagnetic coupling state is stable. The sample entropy value represents the RSSI sequence, indicating RSSI complexity. A higher value indicates a more chaotic signal with more disturbances. It represents the rate of change of response delay, measures the drastic change in tag communication time, and reflects the fluctuation of the tag reader path; It represents the standard deviation of the response delay sequence, which measures the stability of the delayed signal. The smaller the value, the more stable the signal; the larger the value, the less stable the signal.

[0038] In this implementation scheme, a weak perturbation feature recognition mechanism based on a unified time axis is constructed. This method integrates multi-dimensional indicators such as phase fluctuation, RSSI complexity, and response delay variation to quantify the intensity of perturbation changes and delay stability, and finally calculates the weak feature sensitivity value. This process enables accurate capture of non-mutational and minor latent faults, effectively improving the recognition sensitivity and discrimination resolution of tag perturbation behavior, laying a high-quality data foundation for subsequent weak fault probability calculation and dynamic decision-making, and significantly enhancing the edge nodes' ability to perceive early abnormal states and their intelligent response efficiency.

[0039] Specifically, the process of constructing a sensitive feature matrix based on the early identification results of weak disturbance characteristics and providing feedback adjustment for communication and operational anomalies is as follows: The weak feature sensitivity value for each time window is calculated, and the timestamp, standard deviation of the phase sequence, sample entropy value of the RSSI sequence, response delay change rate, standard deviation of the response delay sequence, and weak feature sensitivity value within each time window are extracted to construct a sensitive feature matrix. This matrix is ​​organized in units of time windows, with each row representing the quantitative characteristics of multidimensional disturbance behavior within a time period, facilitating subsequent clustering identification, trend analysis, and intelligent response judgment. The sensitive feature matrix is ​​then written to the edge device. The real-time calculated weak feature sensitivity value is compared with a dynamic sensitivity threshold. If the weak feature sensitivity value is greater than the dynamic sensitivity threshold, it is marked as an early abnormal state. The weak feature sensitivity value is resampled and recalculated every 20 seconds to effectively filter out instantaneous abnormal signal interference and improve early warning judgment. The system maintains stability and fault tolerance. If the sensitivity value of weak features drops, the current abnormal label is automatically removed. If weak signal and unstable path are identified, antenna power is increased and the angle is adjusted to improve tag reading success rate and link quality, ensuring communication integrity. If node current and voltage fluctuations are synchronized with the increase in the sensitivity value of weak features, the edge device actively triggers voltage stabilization and load limiting commands. The voltage stabilization command can control the output of the local transformer module and voltage regulation chip to set the voltage value, while the load limiting command dynamically adjusts the upper limit of load power based on load-side sampling feedback, triggering relays and smart circuit breakers for current limiting control to prevent voltage drops, cable overheating, or joint failures caused by increased fluctuations, thus achieving a rapid local fault risk suppression mechanism. At the same time, the daily sensitivity feature matrix and early abnormal state labels are written into the tag environment synchronous perception database, providing a structured data foundation for subsequent long-term evolution trend analysis and predictive maintenance strategies.

[0040] In this implementation plan, a sensitive feature matrix based on time windows is constructed to achieve high-precision quantitative expression of tag communication behavior and environmental disturbance characteristics. Combined with dynamic threshold comparison and resampling mechanisms, the accuracy and robustness of early anomaly identification are effectively improved. At the same time, local response strategies such as tag path optimization and voltage stabilization and load limiting are used to enhance adaptive control capabilities and avoid equipment degradation caused by the accumulation of micro-disturbances. The supporting structured data writing and trend archiving mechanisms provide strong support for subsequent anomaly clustering, trend prediction and operation and maintenance scheduling, and comprehensively improve the intelligence level of distribution network edge state perception.

[0041] Specifically, the process of determining the probability of tag weak faults by combining the sensitive feature matrix and using tag environment synchronous sensing data is as follows: The sensitive feature matrix is ​​received, and the RSSI, phase, and response delay sequences of the current time window are obtained. The maximum and minimum values ​​of the RSSI in the current time window are filtered, and the difference is calculated to obtain the RSSI fluctuation amplitude. Simultaneously, the average value of the RSSI in the current time window is calculated. The maximum and minimum values ​​of the phase in the current time window are filtered, and the difference is calculated to obtain the phase fluctuation amplitude. Simultaneously, the average value of the phase in the current time window is calculated. The maximum and minimum values ​​of the response delay in the current time window are filtered again, and the difference is calculated to obtain the response delay fluctuation amplitude. Simultaneously, the average value of the response delay in the current time window is calculated. These calculation processes reflect the intensity and fluctuation trend of each communication feature within the current time window, providing direct support for quantifying signal stability. The RSSI fluctuation amplitude is divided by the average value of the RSSI, and then... The signal strength fluctuation value is obtained by multiplying the signal strength fluctuation by a signal strength fluctuation weighting factor; the phase stability value is obtained by dividing the phase fluctuation amplitude by the average phase value and then multiplying by the phase stability weighting factor; the response delay fluctuation value is obtained by dividing the response delay fluctuation amplitude by the average response delay value and then multiplying by the communication delay fluctuation weighting factor; the signal strength fluctuation value, phase stability value, and response delay fluctuation value are added together to obtain the total disturbance metric value. The total disturbance metric value comprehensively describes the fluctuation aggregation effect of the current tag communication behavior in multiple dimensions and serves as an important reference indicator for potential fault symptoms; the total disturbance metric value is negatively raised to the exponent and natural exponential operation is performed to obtain the exponential decay term. The exponential decay term is added by 1 and the reciprocal is taken to obtain the tag weak fault behavior probability value. This exponential mapping method can compress the disturbance value to the (0,1) interval and has higher discrimination for medium and high values, thereby enhancing the response sensitivity to high-risk disturbance states.

[0042] The specific formula for the probability value of weak fault behavior of the tag is as follows:

[0043] ;

[0044] In the formula, This represents the probability value of a tag's weak fault behavior, which is the confidence level in determining whether a tag is currently in a weak fault state. The closer the value is to 1, the greater the probability of a weak fault; the closer the value is to 0, the more normal the communication state. Indicates the RSSI fluctuation amplitude, representing the degree of fluctuation in signal strength; This represents the average value of RSSI, and the standardized fluctuation range of RSSI. It indicates the amplitude of phase fluctuation, reflecting the fluctuation of electromagnetic coupling state; This represents the average value of the phase, and regularizes the magnitude of the phase change. This indicates the magnitude of response delay fluctuations and the degree of response delay jitter. This represents the average response delay, and the standardized response delay fluctuation range. The signal strength fluctuation weighting factor is dynamically calculated by dividing the standard deviation of RSSI by the sum of the standard deviations of RSSI, phase, and response delay, and its value ranges from 0 to 1. The phase stability weighting factor is dynamically calculated by dividing the phase standard deviation by the sum of the standard deviations of RSSI, phase standard deviation, and response delay standard deviation, and its value ranges from 0 to 1. The communication delay fluctuation weighting factor is dynamically calculated by dividing the standard deviation of response delay by the sum of the standard deviation of RSSI, the standard deviation of phase, and the standard deviation of response delay, and its value ranges from 0 to 1.

[0045] In this implementation scheme, a tag communication behavior disturbance modeling mechanism based on a sensitive feature matrix is ​​constructed. This mechanism integrates the volatility assessment of key features such as RSSI, phase, and response delay in their respective dimensions to quantify the comprehensive disturbance measure that constitutes signal stability. By combining the introduction of dynamic weighting factors and an exponential mapping compression strategy, accurate estimation of the probability of weak fault behavior is achieved. This not only enhances the sensitivity of edge devices to early latent abnormal states but also provides highly reliable and interpretable criteria support for subsequent intelligent response and maintenance priority determination. As a result, the accuracy of passive tag communication anomaly identification and the real-time performance of operation and maintenance response in the distribution network system are significantly improved.

[0046] Specifically, the process of identifying and alarming weak faults in power equipment based on the probability of occurrence of tag-based weak faults is as follows: The local edge device compares the probability value of tag-based weak fault behavior with the fault threshold in real time. If the probability value is greater than the threshold, it is automatically marked as a weak fault. This marking process is autonomously executed by the edge intelligent processing module without manual intervention, supporting state transition and data reporting within milliseconds. If the probability value of weak fault behavior shows an increasing trend within multiple consecutive time windows, a real-time alarm mechanism is triggered and pushed to the operation and maintenance terminal. The trend detection is based on fitting the increasing trend of tag-based weak fault behavior probability values ​​within a sliding window, ensuring robust identification of continuous abnormal changes and avoiding false alarms due to short-term fluctuations. All tag-based weak fault behavior probability values, tag numbers, timestamps, and trigger result information are synchronously written into the tag environment synchronization perception database in real time. The system analyzes and identifies long-term evolution trends, frequent fluctuation points, and potential risk areas. Clustering algorithms are used to identify high-incidence areas of abnormal behavior. Combined with equipment operation cycle models, a heat map of potential hazards in spatial and temporal dimensions is constructed for predictive maintenance, automatic task scheduling, and maintenance priority ranking. Automatic task scheduling and maintenance priority ranking employ a multi-factor decision-making method, using standardized input features such as the probability value of weak fault behavior, fault duration, fluctuation trend slope, equipment operating status, and importance indicators. An analytic hierarchy process (AHP) is applied to score fault task priorities. High-priority tasks automatically generate work orders and push them to the maintenance terminal or scheduling platform, while low-priority tasks enter a periodic inspection queue. Simultaneously, based on task spatial distribution and personnel path planning, a nearest neighbor priority and workload balancing strategy are used for dynamic scheduling, forming a maintenance execution plan that minimizes response time and maximizes handling efficiency.

[0047] In this implementation plan, by monitoring and evaluating the probability values ​​of weak fault behaviors of tags in real time on edge devices, an automatic status identification and early warning triggering mechanism without human intervention is realized, which significantly improves the ability to perceive and respond to early anomalies. With the help of structured databases and time series indexes, efficient management and accurate retrieval of data records are ensured. At the same time, cluster analysis and spatiotemporal model construction are integrated to form a potential risk map, providing reliable support for predictive maintenance. Furthermore, by combining multi-factor priority scoring and dynamic task scheduling strategies, a fully automatic closed-loop processing mechanism from data collection and intelligent judgment to task push is realized, which greatly improves the configuration efficiency of operation and maintenance resources and the initiative of equipment health management.

[0048] Specifically, the process of determining the decision threshold for tag weak faults by integrating the judgment results of tag weak fault occurrence probability with tag environment synchronous perception data is as follows: Real-time acquisition of tag weak fault behavior probability values ​​within past time windows, and calculation of the mean and standard deviation of these probability values. The mean reflects the overall offset trend of tag communication status within historical time windows, while the standard deviation characterizes its short-term fluctuation amplitude, helping to determine the fluctuation tolerance boundary of fault behavior; acquisition of voltage, current, and temperature data, and calculation of the variance of these data to reflect the stability of the equipment in its operating state. More severe fluctuations indicate a more unstable physical state. The variances of the calculated voltage, current, and temperature data are then summed to obtain the state fluctuation variance. The difference comprehensively reflects the degree of influence of the external operating environment on the stability of tag communication and is one of the key factors for decision rationality. The upper limit of tag weak fault tolerance is obtained by adding the mean of tag weak fault behavior probability values ​​and the standard deviation of tag weak fault behavior probability values. This upper limit serves as a dynamic boundary for allowing communication disturbances. It is dynamically adjusted in combination with historical behavior distribution to enhance the algorithm's adaptability and discrimination ability to weak fault fluctuations in variable scenarios. The tag weak fault decision value is obtained by dividing the tag weak fault tolerance upper limit by the sum of the state fluctuation variance and 1. This decision value is essentially the ratio between the tag weak fault state tolerance and the current operating fluctuation intensity. After normalization, it is convenient to uniformly judge the trigger threshold in various operating conditions and serve as the heuristic signal basis for subsequent edge response mechanisms.

[0049] The specific formula for the tag-based weak fault decision value is as follows:

[0050] ;

[0051] In the formula, This represents the tag's weak fault decision value, used to determine whether the current tag's weak fault behavior probability value has a significant impact and whether an alert should be triggered. Indicates a time window; The mean of the probability values ​​of weak fault behavior of the tag represents the overall level of the probability of weak faults in the recent period and provides a reference baseline; The standard deviation of the probability value of weak fault behavior of the tag indicates the degree of recent probability fluctuation and reflects instability or abnormal upward trend. State fluctuation variance, which is the sum of the variances of voltage, current, and temperature data, is a quantitative indicator of the stability of the current operating state of distribution network equipment. It integrates the fluctuations of these three key operating parameters and is used to dynamically adjust the early warning sensitivity to minor faults.

[0052] Within five different time windows, the mean and standard deviation of the tag weak fault behavior probability values, as well as the different state fluctuation variances, were used to calculate the tag weak fault decision value. Table 1 shows the tag weak fault decision value data.

[0053] Table 1. Data Table of Tag-Based Weak Fault Decision Values

[0054] like Figure 3 As shown, this is a trend chart of the index changes of the tag-based weak fault decision value provided in the embodiments of this application. It shows the changing trend of each key index in the tag-based weak fault decision value formula with time window, reveals the superimposed influence of the mean and fluctuation of the disturbance on the dynamic threshold, and emphasizes the suppressive effect of the state fluctuation variance on the overall decision value; according to Table 1 and Figure 3 It can be seen that, over time, the mean of the tag weak fault behavior probability value and the tag weak fault decision value tend to be consistent, indicating that the tag weak fault decision value can dynamically respond to changes in the fault behavior probability. The standard deviation of the tag weak fault behavior probability value is small, but fluctuations exist, which also affect the sensitivity of the final tag weak fault decision value.

[0055] In this implementation plan, the dynamic decision-making process integrates the statistical distribution characteristics of the probability values ​​of tag weak fault behaviors with the fluctuation information of equipment operating status to construct tag weak fault decision values ​​that are both adaptive and scenario-adaptable. This not only enables dynamic adjustment of the upper limit of disturbance tolerance, but also allows for flexible judgment of risk levels based on differences in the intensity of environmental fluctuations. This effectively improves the robustness of anomaly detection and the accuracy of edge response, providing a scientific and reliable quantitative basis for subsequent fault response, early warning output, and operation and maintenance scheduling.

[0056] Specifically, the process of determining whether a device exceeds its tolerable fluctuation range based on the decision threshold for tag weak faults is as follows: In each real-time monitoring window period, the current tag weak fault behavior probability value is compared with the tag weak fault decision value calculated at the corresponding time point. If the tag weak fault behavior probability value is less than or equal to the tag weak fault decision value, it indicates that the device is within the acceptable fluctuation range, and monitoring continues without response. If the tag weak fault behavior probability value is greater than the tag weak fault decision value, it indicates that the tag behavior has exceeded the current tolerable fluctuation limit, i.e., the weak fault state is upgraded to an alarm state, immediately triggering the local early warning mechanism and initiating alarm output. This alarm can be presented through various methods such as a local buzzer, LED flashing, and system interface highlighting, and can be linked to the local HMI and remote maintenance terminal for synchronous notification. If the tag weak fault behavior probability value is greater than the tag weak fault decision value for three consecutive periods, the task is given priority inspection. This mechanism uses a sliding window to determine the persistence of the anomaly, ensuring that non-transient jitter will not falsely trigger high-priority handling, effectively improving stability and alarm accuracy. If tag angle offset or environmental reflection interference is detected, the antenna transmission power, frequency, and... The time slot adjustment is executed by the dynamic control module of the radio frequency parameters in the edge node. First, based on channel state feedback, it automatically calls the device's built-in RF control interface and dynamically adjusts the transmit power by modifying the output level of the power amplifier. Second, based on the current channel interference map, it executes frequency hopping algorithms, such as the frequency hopping spread spectrum (FHSS) strategy, to avoid interference frequencies and dynamically allocate available working frequency bands. The time slot adjustment is based on the fluctuation analysis results of tag communication response time. By adjusting the TDMA time slot configuration file, it extends and shortens the tag response interval time window, thereby avoiding tag response overlap or timing misalignment, improving anti-interference capability and tag identification success rate. If the anomaly is identified as slight looseness or environmental fluctuation, the tag reading interval is extended to avoid excessive polling, thereby reducing energy consumption, reducing false judgments, and extending the device's lifespan, reflecting a differentiated response strategy for anomaly types. At the same time, the current tag number, timestamp, tag weak fault behavior probability value, and tag weak fault decision value are recorded to the edge device. These records will serve as key data support for subsequent backtracking optimization, maintenance strategy evaluation, and anomaly reproduction, and will be periodically uploaded to the cloud for historical record archiving and global analysis.The obtained weak feature sensitivity values, tag weak fault behavior probability values, and tag weak fault decision values ​​are structured, integrated, and graphically presented for transparent monitoring of equipment status and operation and maintenance assistance. The structured integration uses a unified data model, storing multi-dimensional indicators normalized along a time axis and dynamically loading them on the front end in the form of line charts, heatmaps, scatter plots, etc., forming a multi-perspective linked expression of trend, local, and sudden anomalies. Tag status and tag weak fault behavior probability value trends are displayed through multiple layers. The multi-layer display function is supported by the front-end GIS, constructing multiple overlayable layers based on the spatial location information and operational status data of the tags. It supports tag search, status filtering, historical trend tracking, and real-time alarm response through an interactive interface. It supports manual annotation and false alarm correction, implemented through built-in interactive components such as annotation dialog boxes and false alarm correction forms. Users can click on graphical nodes and abnormal curve points to trigger an annotation window, where they can select labels, define annotation types, and add notes. After each alert is triggered, a pop-up notification automatically appears and links to the work order system, recording operation feedback. In implementing automatic pop-up notifications, a message push function is implemented based on a front-end listening mechanism and an event triggering model. Specifically, when an edge node or central service receives an event packet indicating a weak fault behavior probability exceeding a threshold, the event message is pushed in real-time using the WebSocket protocol. The linkage with the work order system can exchange data with the local work order platform via an asynchronous interface based on message queues such as Kafka. Furthermore, it regularly generates equipment operation stability reports and maintenance priority suggestions, comprehensively considering high-incidence points of weak faults and regional operational health status to construct a closed-loop mechanism for alarm response, trend analysis, and predictive maintenance.

[0057] At five time points, the weak feature sensitivity value, weak fault behavior probability value, and weak fault decision value of different labels were calculated, and the triggering of an early warning was determined accordingly. Table 2 shows the intelligent monitoring data for weak faults in labels.

[0058] Table 2. Intelligent Monitoring Data for Labeled Weak Faults

[0059] like Figure 4 As shown, this is a visualization diagram of intelligent monitoring of tag-based weak faults provided in an embodiment of this application. The diagram visually displays the changing trends of the weak feature sensitivity value, tag-based weak fault behavior probability value, and tag-based weak fault decision value of the distribution network tags over five time points through a combination of area, broken lines, and dashed lines. High-risk warning points are highlighted in red with crosses. According to Table 2 and... Figure 4 It can be seen that the probability value of the tag's weak fault behavior exceeds the tag's weak fault decision value at the third time point and is marked as an over-threshold warning point. At other time points, the equipment status is normal. Through dynamic comparison of multiple indicators, timely identification and visual early warning of equipment abnormal risks are achieved.

[0060] This implementation plan effectively constructs an intelligent operation and maintenance system that integrates anomaly detection, intelligent judgment, adaptive handling, and full-cycle closed-loop feedback. By comparing the probability of weak tag fault behaviors with decision values ​​in real time, it achieves refined fluctuation management and anomaly escalation judgment. Furthermore, leveraging multi-source sensing data, dynamic radio frequency control, and classification response strategies, it significantly improves the stability of tag identification and the accuracy of fault identification in complex interference environments. Combined with sliding window judgment and task priority adjustment mechanisms, it enhances sensitivity to persistent risks and the rationality of resource scheduling. Simultaneously, it records and uploads key data in real time, providing a solid data foundation and guarantee for subsequent trend analysis, fault tracking, and predictive maintenance. By integrating and graphically presenting multi-dimensional data from weak disturbance identification, weak fault probability judgment, and dynamic decision-making results, a device status monitoring and maintenance platform with high transparency, strong interactivity, and intelligent response capabilities has been constructed. It supports multi-layered linked display, multi-dimensional trend analysis, real-time alarms, and manual correction feedback mechanisms. Through WebSocket message push and Kafka work order linkage interfaces, it achieves closed-loop control between the edge, cloud, and human, effectively improving the timeliness of anomaly identification and fault response efficiency. Furthermore, based on spatial thermal analysis and structured indicator evaluation, it regularly generates maintenance priority suggestions, achieving closed-loop technical support throughout the entire process from anomaly discovery and response triggering to maintenance decision-making, significantly enhancing the intelligence level of equipment maintenance and predictive maintenance capabilities.

[0061] Reference Figure 2 As shown, the second aspect of the present invention provides a power grid edge analysis device for passive tag communication multi-sensor synchronization, applied to the aforementioned power grid edge analysis method for passive tag communication multi-sensor synchronization, comprising: a data synchronization acquisition and fusion module, used to acquire tag environment synchronization sensing data in real time, preprocess the tag environment synchronization sensing data, establish a unified time axis, and perform tag environment synchronization sensing data fusion and behavior mapping on the unified time axis; a weak feature early extraction module, used to identify the weak disturbance features of tags early based on the tag environment synchronization sensing data on the unified time axis, construct a sensitive feature matrix based on the early identification results of the weak disturbance features, and provide feedback adjustment for communication and operation status anomalies; a tag weak fault intelligent behavior recognition module, used to combine the sensitive feature matrix and tag environment synchronization sensing data to determine the probability of tag weak fault occurrence, identify power equipment weak faults and issue alarms based on the determination result of the tag weak fault occurrence probability; and an edge dynamic decision and early warning module, used to fuse the determination result of the tag weak fault occurrence probability with the tag environment synchronization sensing data, determine the decision threshold of tag weak fault, and determine whether the equipment exceeds the tolerance fluctuation range based on the tag weak fault decision threshold.

[0062] This implementation plan achieves refined identification and intelligent operation and maintenance of abnormal states in passive tag communication by constructing an integrated full-link functional module encompassing acquisition, identification, judgment, decision-making, response, and feedback. It not only performs high-precision synchronous fusion and behavior mapping of tag environment synchronous sensing data, but also extracts weak disturbance features in the early stages, forming a sensitive feature matrix and triggering abnormal feedback. Combined with intelligent behavior recognition and probabilistic computing mechanisms, it supports dynamic judgment and adaptive response to weak fault states. Utilizing the dynamic decision-making and early warning module on the edge side, it can quickly match response strategies under different fluctuation conditions. Simultaneously, with multi-layered, interactive visualization and work order linkage mechanisms, it realizes an integrated closed loop of state identification, alarm response, trend analysis, and predictive maintenance, significantly improving the technical level and application value in communication disturbance identification, fault judgment timeliness, and intelligent operation and maintenance efficiency.

[0063] A third aspect of the present invention provides a storage medium for power distribution network edge analysis with passive tag communication and multi-sensor synchronization, comprising: the storage medium having one or more programs, the one or more programs being executed by one or more processors to implement the power distribution network edge analysis method with passive tag communication and multi-sensor synchronization as described in claims 1-8.

[0064] This implementation scheme enables efficient synchronous acquisition and fusion processing of data from multiple sensor types at the edge, improving the accuracy and consistency of power distribution network status perception. By integrating intelligent analysis and decision-making programs, it effectively enhances real-time response and local anomaly detection capabilities, reduces reliance on central servers, and minimizes data transmission energy consumption and communication load. Simultaneously, the storage medium possesses excellent compatibility and scalability, adapting to various passive tags and sensor devices, supporting flexible deployment across multiple scenarios, and improving the stability and reliability of the power distribution network monitoring system, providing solid support for the intelligent and efficient operation and maintenance of the power distribution network.

[0065] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0066] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. As those skilled in the art will understand, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A method for distribution network edge analysis based on passive tag communication and multi-sensor synchronization, characterized in that, Includes the following steps: Real-time collection of tag environment synchronous perception data, data preprocessing of tag environment synchronous perception data, establishment of a unified time axis, and fusion and behavior mapping of tag environment synchronous perception data on the unified time axis; Based on the tag environment synchronous perception data on a unified time axis, the weak disturbance characteristics of the tags are identified in the early stage. A sensitive feature matrix is ​​constructed based on the early identification results of the weak disturbance characteristics, and feedback adjustment is performed on communication and operation status anomalies. By combining the sensitive feature matrix and using the synchronous perception data of the tag environment, the probability of tag-related weak faults is determined, and weak faults in power equipment are identified and alarms are triggered based on the determination results of the tag-related weak fault occurrence probability. The system integrates the probability of weak tag faults with synchronous sensing data of the tag environment to determine the decision threshold for weak tag faults, and then uses this decision threshold to determine whether the equipment exceeds the tolerance fluctuation range.

2. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process of real-time acquisition of tag environment synchronous perception data, data preprocessing of tag environment synchronous perception data, establishment of a unified time axis, and fusion and behavior mapping of tag environment synchronous perception data on the unified time axis is as follows: The passive tag reader and edge sensor nodes collect real-time tag environment synchronous sensing data, which includes: tag RSSI, phase, response delay, and voltage, current, and temperature of the device operating status. All passive tags are uniformly numbered, and the acquisition time of each sensor is accurately synchronized using GPS time synchronization. A uniform and accurate timestamp is added to each tag's environmental synchronization sensing data point, and all tag environmental synchronization sensing data are sorted according to the timestamp to establish a unified timeline. Outlier removal and missing value filling operations are performed, and data noise reduction and smoothing are carried out through moving average filtering. At the same time, the tag environmental synchronization sensing data is standardized and normalized. A tag environmental synchronization sensing database is established to store the tag environmental synchronization sensing data. On a unified timeline, each moment corresponds to a tag environment synchronous perception data record. Mutual information correlation analysis is used to identify the correlation between tag communication behavior characteristics and device operating status, and regression analysis is used to establish the mapping relationship between tag communication behavior characteristics and device operating status.

3. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process for early identification of weak perturbation features of tags based on tag-based environmental synchronous sensing data on a unified time axis is as follows: Divide the time window according to a unified time axis, and acquire RSSI, phase, and response delay data for each time window to obtain RSSI, phase, and response delay sequences; calculate the standard deviation of the phase sequence, and calculate the sample entropy value of the RSSI sequence using an entropy algorithm after acquiring the RSSI sequence; acquire the response delay at the current time and the previous time, divide the difference between the response delay at the current time and the response delay at the previous time by the response delay at the previous time to obtain the rate of change of response delay, and calculate the standard deviation of the response delay sequence; The perturbation intensity value is obtained by multiplying the standard deviation of the phase sequence, the sample entropy value of the RSSI sequence, and the sum of the response delay rate and 1. The delay stability value is obtained by adding the standard deviation of the response delay sequence to 1. The weak feature sensitivity value is obtained by dividing the perturbation intensity value by the delay stability value.

4. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process of constructing a sensitive feature matrix based on the early identification results of weak disturbance characteristics and adjusting for communication and operational anomalies is as follows: Calculate the weak feature sensitivity value for each time window, and extract the timestamp, standard deviation of the phase sequence, sample entropy value of the RSSI sequence, response delay change rate, standard deviation of the response delay sequence, and weak feature sensitivity value within each time window to construct a sensitive feature matrix, and write the sensitive feature matrix into the edge device; The real-time calculated weak feature sensitivity value is compared with the dynamic sensitivity threshold. If the weak feature sensitivity value is greater than the dynamic sensitivity threshold, it is marked as an early abnormal state. The weak feature sensitivity value is resampled and recalculated every 20 seconds. If the weak feature sensitivity value drops, the abnormal state is automatically removed. If the signal is identified as weak and the path is unstable, the antenna power is increased and the angle is adjusted. If the node current and voltage fluctuations are synchronized with the increase in the weak feature sensitivity value, the edge device will actively trigger voltage stabilization and load limiting commands. At the same time, the daily sensitivity feature matrix and the early abnormal state markings are written into the tag environment synchronous perception database.

5. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process of combining the sensitive feature matrix and using synchronously perceived data of the tag environment to determine the probability of tag weak faults is as follows: Receive the sensitive feature matrix and obtain the RSSI, phase, and response delay sequence of the current time window. Filter the maximum and minimum values ​​of RSSI in the current time window and calculate the difference to obtain the RSSI fluctuation amplitude. At the same time, calculate the average value of RSSI in the current time window. Filter the maximum and minimum values ​​of phase in the current time window and calculate the difference to obtain the phase fluctuation amplitude. At the same time, calculate the average value of phase in the current time window. The maximum and minimum values ​​of the response delay in the current time window are filtered again and the difference is calculated to obtain the response delay fluctuation range. At the same time, the average value of the response delay in the current time window is calculated. The signal strength fluctuation value is obtained by dividing the RSSI fluctuation amplitude by the average RSSI value and then multiplying it by the signal strength fluctuation weighting factor; the phase stability value is obtained by dividing the phase fluctuation amplitude by the average phase value and then multiplying it by the phase stability weighting factor; the response delay fluctuation value is obtained by dividing the response delay fluctuation amplitude by the average response delay value and then multiplying it by the communication delay fluctuation weighting factor; the signal strength fluctuation value, phase stability value, and response delay fluctuation value are added together to obtain the total disturbance metric value. The total disturbance metric value is then negatively raised to the power of the exponent and subjected to natural exponential operation to obtain the exponential decay term. The exponential decay term is then incremented by 1 and the reciprocal is taken to obtain the tag weak fault behavior probability value.

6. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process of identifying weak faults in power equipment and issuing alarms based on the determination result of the probability of occurrence of weak faults in tags is as follows: The tag's weak fault behavior probability value is compared with the fault threshold in real time on the local edge device. If the tag's weak fault behavior probability value is greater than the fault threshold, it is automatically marked as a weak fault state. If the weak fault behavior probability value shows an upward trend in multiple consecutive time windows, a real-time alarm mechanism will be triggered and pushed to the operation and maintenance terminal. All tag weak fault behavior probability values, tag numbers, timestamps, and trigger result information are synchronously written into the tag environment synchronous perception database. Real-time analysis is used to identify long-term evolution trends, frequent fluctuation points, and potential risk areas, enabling predictive maintenance, automatic task scheduling, and maintenance priority ranking.

7. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process for determining the decision threshold for weak tag faults by combining the judgment result of the fusion tag weak fault occurrence probability with the tag environment synchronous perception data is as follows: The system acquires the probability values ​​of weak fault behaviors of tags within past time windows in real time, calculates the mean and standard deviation of the probability values ​​of weak fault behaviors of tags, acquires voltage, current and temperature data, calculates the variance of voltage, current and temperature data, and sums the calculated variances of voltage, current and temperature data to obtain the state fluctuation variance. The upper limit of the label weak fault tolerance is obtained by adding the mean of the label weak fault behavior probability value to the standard deviation of the label weak fault behavior probability value. The label weak fault decision value is obtained by dividing the upper limit of the label weak fault tolerance by the sum of the state fluctuation variance and 1.

8. The method for passive tag communication multi-sensor synchronization in power distribution network edge analysis according to claim 1, characterized in that, The specific process for determining whether a device exceeds the tolerable fluctuation range based on the decision threshold for weak faults in the tag is as follows: In each real-time monitoring window period, the current tag weak fault behavior probability value is compared with the tag weak fault decision value calculated at the corresponding time point. If the tag weak fault behavior probability value is less than or equal to the tag weak fault decision value, it means that the device is within the acceptable fluctuation range and will continue to monitor without responding. If the tag's weak fault behavior probability value is greater than the tag's weak fault decision value, it means that the tag's behavior has exceeded the current tolerable fluctuation limit. That is, the weak fault state is upgraded to an alarm state, and the local early warning mechanism is immediately triggered to start alarm output. If the tag's weak fault behavior probability value is greater than the tag's weak fault decision value for three consecutive cycles, the task will be given priority inspection. If tag angle offset or environmental reflection interference is detected, the antenna transmission power, frequency, and time slot are automatically adjusted; if the anomaly is identified as slight looseness or environmental fluctuation, the tag reading interval is extended to avoid excessive polling; at the same time, the current tag number, timestamp, tag weak fault behavior probability value, and tag weak fault decision value are recorded to the edge device. The obtained weak feature sensitivity values, label weak fault behavior probability values, and label weak fault decision values ​​are structured, integrated, and graphically presented to enable transparent monitoring of equipment status and operation and maintenance assistance. The system displays tag status and tag weak fault behavior probability trends through multiple layers, supports tag search, status filtering, historical trend tracking and real-time alarm response through an interactive interface, and supports manual annotation and false alarm correction. After each warning is triggered, a prompt will automatically pop up and the work order system will be linked, and the operation feedback will be recorded. In addition, equipment operation stability reports and maintenance priority suggestions will be generated regularly. By comprehensively considering the high incidence of weak faults and the regional operation health status, a closed-loop mechanism for alarm response, trend analysis and predictive maintenance will be constructed.

9. A power distribution network edge analysis device for passive tag communication and multi-sensor synchronization, employing any one of the passive tag communication and multi-sensor synchronization methods for power distribution networks as described in claims 1-8, characterized in that... include: The data synchronization acquisition and fusion module is used to collect tag environment synchronous perception data in real time, preprocess the tag environment synchronous perception data, establish a unified time axis, and perform tag environment synchronous perception data fusion and behavior mapping on the unified time axis. The weak feature early extraction module is used to identify weak perturbation features of tags early based on tag environment synchronous perception data on a unified time axis, construct a sensitive feature matrix based on the early identification results of weak perturbation features, and provide feedback adjustment for communication and operation status anomalies. The tag-based weak fault intelligent behavior recognition module is used to combine a sensitive feature matrix and synchronously sense data of the tag environment to determine the probability of tag-based weak faults occurring, and to identify weak faults in power equipment and issue alarms based on the determination result of the tag-based weak fault occurrence probability. The edge dynamic decision-making and early warning module is used to integrate the judgment results of the probability of tag weak faults with the synchronous perception data of tag environment, determine the decision threshold of tag weak faults, and determine whether the device exceeds the tolerance fluctuation range based on the decision threshold of tag weak faults.

10. A storage medium for distribution network edge analysis using passive tag communication and multi-sensor synchronization, employing any one of the passive tag communication and multi-sensor synchronization methods for distribution network edge analysis as described in claims 1-8, characterized in that... include: The storage medium has one or more programs, which are executed by one or more processors to implement the passive tag communication multi-sensor synchronization distribution network edge analysis method as described in claims 1-8.

Images