A remote monitoring method and system based on a transformer area intelligent fusion terminal

Abstract

The application discloses a remote monitoring method and system based on a transformer area intelligent fusion terminal, and the method comprises the following steps: S1, collecting telemetry data of each terminal and generating original alarm events; S2, collecting original alarm events of each terminal and analyzing whether each terminal meets a fuse state triggering condition; S3, generating an alarm fuse summary event after the fuse state is triggered; S4, generating a processing operation report and executing processing; and S5, timing the fuse state and continuously judging whether a release terminal meets the fuse state triggering condition. Based on the composite fuse triggering condition of frequency, mode and dynamic baseline, the application can accurately identify and intercept a large number of original alarm events exceeding normal frequency in a unit time period in advance, then suppress and aggregate the large number of original alarm events with abnormal frequency into the alarm fuse summary event, and focus on the core fault event more directly, so that the disposal target is clear and the decision speed is improved.

Description

Technical Field

[0001] This invention relates to the field of power equipment monitoring technology, and in particular to a remote monitoring method and system based on a smart fusion terminal for power distribution areas. Background Technology

[0002] With the deepening of the intelligent transformation of distribution networks, the intelligent integrated terminal of the transformer substation connects the power grid and the user end, realizing comprehensive data collection and status awareness of low-voltage power distribution equipment. When a fault or anomaly occurs in a transformer substation, the associated three-phase voltage telemetry data will experience a chain reaction of sudden changes, triggering hundreds or even thousands of original alarm events in a short period of time. These numerous original alarm events are reported instantly, making it difficult for maintenance personnel to quickly locate the core fault. Moreover, among the numerous original alarm events, there are also cases where key information is obscured by a large number of duplicate and derivative alarms. Existing methods mostly follow the process of collection, reporting, and manual judgment, requiring personnel to spend a lot of time on manual screening, merging, and preliminary analysis, resulting in long response delays and failing to meet the requirements of modern distribution networks for rapid fault isolation and power restoration. In view of this, this solution proposes a remote monitoring method and system based on the intelligent integrated terminal of the transformer substation to solve the above problems. Summary of the Invention

[0003] The purpose of this invention is to provide a remote monitoring method and system based on a smart converged terminal for distribution areas, so as to solve the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a remote monitoring method based on a smart converged terminal for distribution areas, comprising the following operational steps:

[0005] S1 collects telemetry data from each terminal and performs real-time analysis based on mutation detection. After detecting an anomaly, it generates an original alarm event and sets the trigger conditions for the circuit breaker state.

[0006] S2 collects raw alarm events from each terminal, performs statistical analysis, pattern matching, and dynamic baseline comparison of the number of raw alarm events from each terminal within a unit time window, and analyzes whether each terminal meets the triggering conditions for the circuit breaker state. If the conditions are met, the alarm processing circuit breaker state of the corresponding terminal is triggered; if the conditions are not met, the raw alarm events are reported.

[0007] S3, after the circuit breaker state is triggered, firstly, the repeated or derived alarm events of the triggering terminal are suppressed and converted into background logs. At the same time, alarm circuit breaker summary events are generated and analyzed through root cause diagnosis to generate analysis reports.

[0008] S4: Combine the event details and analysis report of the circuit breaker summary event to generate a processing operation report. If the problem can be resolved remotely, it can be handled directly through remote control commands; if on-site handling is required, the processing is carried out according to the processing operation report.

[0009] S5, start timing the circuit breaker state. After the circuit breaker state has lasted for a preset time, release the alarm event of the corresponding terminal according to the limited release conditions, and continue to determine whether the released terminal meets the circuit breaker state trigger conditions. If it meets the conditions, continue to execute the circuit breaker state and extend the circuit breaker state time. If it does not meet the conditions, restore the initial reporting state.

[0010] Preferably, the generation of the original alarm event includes the following steps:

[0011] S101 continuously collects telemetry data from each terminal at a fixed frequency, and defines a fixed time analysis window. Within the window, the collected data is preprocessed, including removing invalid values ​​and performing low-pass filtering to smooth random noise.

[0012] S102, within each analysis window, calculates the instantaneous gradient of the three-phase voltage, the sudden increase in standard deviation within the window, and the residual abnormality of the data sequence relative to the fitting trend in real time;

[0013] S103 compares the instantaneous gradient, standard deviation burst increment, and fitting residual abnormality calculated in S102 with a preset threshold based on historical data. When any feature index continuously exceeds the threshold, it is determined that a data mutation has occurred, and a structured original alarm event containing event type, timestamp, terminal identifier, mutation parameters, and feature values ​​is generated.

[0014] Preferably, the instantaneous gradient is calculated by performing median filtering on the voltage sequence within a defined fixed-time analysis window, and then calculating the weighted difference value of the filtered voltage sequence.

[0015] Preferably, the calculation of the standard deviation spike increment within the window is to first calculate the standard deviation σ1 of the voltage data within the fixed-time analysis window, and simultaneously obtain the baseline standard deviation σ0 of the normal fluctuation level. Then, the difference between σ1 and σ0 is divided by σ0 to obtain the standard deviation spike increment.

[0016] Preferably, the calculation of the residual abnormality first uses the LOESS algorithm to perform nonlinear fitting on the data points within a fixed time analysis window, and then calculates the residual e between each data point and the LOESS fitting curve. i Then, the standardized values ​​of these residuals are calculated using the following formula:

[0017]

[0018] in, This represents the mean value obtained based on the statistical analysis of historical normal window residuals. This represents the standard deviation obtained based on the statistical analysis of historical normal window residuals.

[0019] Finally, the maximum value among all the absolute values ​​of the standardized residuals is taken as the residual constant.

[0020] Preferably, the triggering conditions for the circuit breaker state include frequency conditions, mode conditions, and baseline deviation conditions;

[0021] The measure of whether the number of alarm events triggered by the frequency condition exceeds the set frequency threshold;

[0022] The mode conditions measure the inherent correlation and repetition between alarm events and distinguish between random discrete alarms and regular fault alarms.

[0023] The baseline deviation condition measures the degree of deviation of the current alarm activity from the historical normal level of the terminal.

[0024] Preferably, the quantity statistics comparison first counts the total number of original alarm events generated by each terminal within a unit time window in real time, and then compares it with a preset frequency threshold. If the frequency threshold is exceeded, the frequency is marked as abnormal.

[0025] The pattern matching comparison first analyzes the sequence of alarm events within the same time window, and then uses a rule engine to match the real-time alarm sequence with a predefined abnormal pattern library. If any preset abnormal pattern is matched, the pattern is marked as abnormal and the matched pattern type is recorded.

[0026] The dynamic baseline comparison first calculates the comprehensive index of alarm activity intensity in the current window, and then calculates the deviation rate of the real-time index relative to the dynamic baseline. If the deviation rate exceeds the set threshold, the baseline deviation is marked as abnormal.

[0027] Preferably, after the circuit breaker state is triggered, the original alarm events of the terminal that triggered the circuit breaker state are filtered and intercepted. After filtering out duplicate alarms and derivative alarms, they are organized into a structured event log table. Then, starting from the time the circuit breaker state is triggered, an analysis window is traced back, and all original alarm events within this analysis window are counted, including counting the total number of alarms, classifying and counting by alarm type and severity level, and calculating the scope of affected next-level equipment or users based on the power grid topology. A summary event record is generated. Finally, root cause diagnosis analysis is matched and scheduled according to alarm type characteristics to generate a preliminary handling suggestion report.

[0028] Preferably, during the circuit breaker state timing and release operation, timing is first started after the terminal initiates the circuit breaker state. When the timing reaches the preset single circuit breaker time, a limited release assessment is performed to assess whether the frequency of suppressed alarms counted in the most recent assessment window meets or falls below the normal frequency. If it does, the circuit breaker state of the corresponding terminal is released and the initial reporting process is restored. If it does not meet the requirements, the circuit breaker state time is extended in a phased increment.

[0029] A remote monitoring system based on a smart converged terminal for transformer substations, used to implement the aforementioned remote monitoring method based on a smart converged terminal for transformer substations, includes a data acquisition and alarm generation unit, a statistics unit, a circuit breaker adjudication unit, a circuit breaker processing unit, a root cause diagnosis unit, a handling decision unit, and a circuit breaker management unit.

[0030] The data acquisition and alarm generation unit is used to connect to the intelligent fusion terminal of the distribution area, collect three-phase voltage telemetry data at a fixed frequency, and generate original alarm events through calculation.

[0031] The statistical unit is used to count all generated original alarm events;

[0032] The circuit breaker decision unit determines whether to trigger the circuit breaker processing state by analyzing the statistical original alarm events.

[0033] The circuit breaker processing unit is used to filter the original alarm events and transfer the suppressed alarms, and aggregate them to generate alarm circuit breaker summary events;

[0034] The root cause diagnosis unit is used to generate a diagnostic report from the root cause diagnosis of the filtered original alarm events.

[0035] The decision-making unit is used to receive the summarized events and diagnostic reports, and generate a decision-making report containing specific operational suggestions.

[0036] The circuit breaker management unit is used to maintain the circuit breaker status timer for all terminals, and after the timer expires, it initiates a limited release condition assessment.

[0037] The technical effects and advantages of this invention are as follows:

[0038] This invention, based on a composite circuit breaker triggering condition of frequency, pattern, and dynamic baseline, can accurately identify and intercept a large number of original alarm events exceeding the normal frequency within a unit time period. Then, it suppresses and aggregates the large number of original alarm events with abnormal frequency into alarm circuit breaker summary events. This completely liberates operation and maintenance personnel from information overload, enabling them to focus more directly on core fault events, making the handling objectives clear and improving decision-making speed. Attached Figure Description

[0039] Figure 1 This is a flowchart illustrating the monitoring operation implementation of the present invention.

[0040] Figure 2 This is a block diagram of the monitoring system of the present invention. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Example 1: The present invention provides as follows Figure 1 The remote monitoring method based on a smart converged terminal for distribution areas, as shown, includes the following steps:

[0043] S1 collects telemetry data from each terminal and performs real-time analysis based on mutation detection. After detecting an anomaly, it generates an original alarm event and sets the trigger conditions for the circuit breaker state.

[0044] Specifically, the generation of the original alarm event includes the following steps:

[0045] S101 continuously collects telemetry data from each terminal at a fixed frequency and defines a fixed time analysis window. Within the window, the collected data is preprocessed, including removing invalid values ​​and performing low-pass filtering to smooth random noise, providing high-quality data for subsequent analysis. When collecting three-phase voltage telemetry data from each terminal, the acquisition equipment is first deployed at the intelligent fusion terminal of each transformer area, and communication networking and data access verification are performed to ensure continuous and stable data acquisition.

[0046] The data preprocessing first verifies the validity of each data point uploaded by each terminal. When a null value or a communication anomaly is found, it is directly marked as invalid. When an out-of-range value is found, it is determined whether the data exceeds the physical possible range. When a sudden outlier value is found, a simple rate-based judgment is used. If the difference between a point and the previous point changes by more than 20% per second, it is marked as a suspicious outlier value. Data points marked as invalid are removed and do not participate in subsequent calculations.

[0047] Then, after removing invalid values, the data sequence is subjected to first-order hysteresis filtering in chronological order for each data point to obtain a smoothed data sequence, which suppresses measurement white noise and transient interference, while retaining the true trend of voltage gradual or sudden changes.

[0048] Finally, for invalid data points that are removed, no interpolation is performed to avoid introducing false information. Subsequent analysis windows will be calculated based on valid data points, and the processed data will be strictly aligned with the precise timestamp to ensure the accuracy of time series analysis.

[0049] S102, within each analysis window, calculates the instantaneous gradient of the three-phase voltage, the sudden increase in standard deviation within the window, and the residual abnormality of the data sequence relative to the fitting trend in real time;

[0050] The instantaneous gradient is calculated by performing median filtering on the voltage sequence within a defined fixed-time analysis window, and then calculating the weighted difference of the filtered voltage sequence. The three most recent sampling points are selected and denoted as (t, t-1, t-2). The calculation formula is as follows:

[0051] (1);

[0052] Where a, b, and c represent weighting coefficients, V t V represents the three-phase voltage value at time t. t-1 The three-phase voltage value at time t-1 is represented by V. t-2 This represents the three-phase voltage value at time t-2. It indicates a fixed sampling time interval; it assigns higher weight to the latest data, enabling it to capture changing acceleration trends more sensitively and resiliently to noise.

[0053] The calculation of the standard deviation spike increment within the window is as follows: First, the standard deviation σ1 of the voltage data within the fixed-time analysis window is calculated, and the baseline standard deviation σ0 of the normal fluctuation level is obtained. This baseline standard deviation is updated by the exponentially weighted moving average (EWMA) algorithm for the standard deviation of all windows over a past period. Then, the difference between σ1 and σ0 is divided by σ0 to obtain the standard deviation spike increment. Only when the current fluctuation deviates significantly from the recent adaptive baseline is it judged as an abnormal spike.

[0054] The calculation of residual abnormality first uses the LOESS algorithm to perform nonlinear fitting on the data points within a fixed-time analysis window to better adapt to local trend changes. Then, the residual e between each data point and the LOESS fitted curve is calculated. i Then, the standardized values ​​of these residuals are calculated using the following formula:

[0055] (1);

[0056] in, This represents the mean value obtained based on the statistical analysis of historical normal window residuals. This represents the standard deviation obtained based on the statistical analysis of historical normal window residuals.

[0057] Finally, the maximum value among all the absolute values ​​of the standardized residuals is taken as the residual abnormality. Through nonlinear fitting and standardization based on historical statistics, abnormal patterns that deviate from the existing trend patterns can be identified more accurately.

[0058] S103 compares the instantaneous gradient, standard deviation burst increment, and fitting residual abnormality calculated in S102 with a preset threshold based on historical data. When any feature index continuously exceeds the threshold, it is determined that a data mutation has occurred, and a structured original alarm event containing event type, timestamp, terminal identifier, mutation parameters, and feature values ​​is generated.

[0059] S2, collect the original alarm events of each terminal, count the number of original alarm events of each terminal within a unit time window, perform pattern matching and dynamic baseline comparison, and analyze whether each terminal meets the circuit breaker triggering conditions. The circuit breaker triggering conditions include frequency conditions, pattern conditions and baseline deviation conditions.

[0060] The frequency of alarm events is measured to determine whether the number of alarm events exceeds the set frequency threshold.

[0061] The pattern conditions are used to measure the inherent correlation and repetition between alarm events, and to distinguish between random discrete alarms and regular fault alarms.

[0062] Baseline deviation condition measures the degree of deviation of current alarm activity from the historical normal level of the terminal;

[0063] When the conditions are met, the alarm processing circuit breaker state of the corresponding terminal is triggered; when the conditions are not met, the original alarm event is reported.

[0064] Specifically, the quantity statistics comparison first counts the total number of original alarm events generated by each terminal within a unit time window in real time, and then compares it with the preset frequency threshold. If it exceeds the frequency threshold, it is marked as an abnormal frequency. This threshold can be dynamically adjusted according to the terminal type to determine in real time whether the total number of alarm events of a single terminal within a unit time window is abnormally high.

[0065] The pattern matching comparison first analyzes the sequence of alarm events within the same time window, and then uses a rule engine to match the real-time alarm sequence with a predefined abnormal pattern library. If any preset abnormal pattern is matched, the pattern is marked as abnormal, and the matched pattern type is recorded. The abnormal pattern library includes recurring burst patterns, causal derivation patterns, and spatial propagation patterns. Among them, the recurring burst pattern is an event with more than N alarm codes and identical parameters occurring within a very short sub-window; the causal derivation pattern is an event that will inevitably be accompanied by event B within T seconds after event A occurs; the spatial propagation pattern is multiple monitoring points on the same power supply branch reporting similar alarms in sequence, and the time difference matches the fault propagation speed.

[0066] The dynamic baseline comparison first calculates the comprehensive index of alarm activity intensity in the current window, and then calculates the deviation rate of the real-time index relative to the dynamic baseline. If the deviation rate exceeds the set threshold, the baseline deviation is marked as abnormal.

[0067] S3, after the circuit breaker state is triggered, firstly, duplicate or derivative alarm events of the triggering terminal are suppressed. Duplicate alarms are new events that are exactly the same as events already recorded in the current circuit breaker window and are directly intercepted. Derivative alarms are secondary alarms that are determined to be directly caused by an existing primary alarm according to predefined causal relationship rules and will also be intercepted. After interception, they are converted into background logs, recording their complete content, the reason for suppression, and timestamps for subsequent auditing and analysis. At the same time, an alarm circuit breaker summary event is generated, including: terminal identifier, circuit breaker trigger time, alarm summary in the backtracking window, classification statistics results, and preliminary impact scope assessment information. Root cause diagnosis is then used for analysis to generate an analysis report. After the circuit breaker state is triggered, the original alarm events of the terminal that triggered the circuit breaker state are filtered and intercepted, filtering out duplicate alarms and derivative alarms. After an alarm is triggered, it is compiled into a structured event log table. Then, starting from the moment the circuit breaker was triggered, an analysis window is traced backward, and all original value alarm events within this analysis window are statistically analyzed. This includes counting the total number of alarms, classifying and counting alarms by type and severity, and calculating the affected next-level equipment or user range based on the power grid topology, generating a summary event record. Finally, root cause diagnosis analysis is matched and scheduled based on alarm type characteristics to generate a preliminary handling suggestion report. The report includes: the most likely root cause, the inferred fault type and possible location; confidence level, the reliability of the diagnosis results; key alarm parameters and sequences supporting the diagnosis; and a handling operation report. The system automatically generates a handling operation report for maintenance personnel by combining the details of the circuit breaker summary event and the root cause analysis report. The report clearly provides specific and actionable operation steps, lists key alarm statistics and diagnostic conclusions, and reiterates the affected equipment and users to assess the handling priority.

[0068] S4 generates a processing operation report by combining the event details and analysis report of the circuit breaker summary event. If the problem can be resolved remotely, it will be handled directly via remote control commands. Before automatic execution, the remote control commands will undergo built-in anti-misoperation logic interlocking and current status verification. After passing the verification, the commands will be directly sent to the corresponding intelligent fusion terminal or distribution automation substation through the security gateway. If on-site handling is required, the processing will be carried out according to the processing operation report. The report includes an overview of the storm source, time, main alarm types and impact range, and clarifies the most likely root cause, confidence level and key evidence.

[0069] S5, start timing the circuit breaker state. After the circuit breaker state has lasted for a preset time, release the alarm event of the corresponding terminal according to the limited release conditions, and continue to determine whether the released terminal meets the circuit breaker state trigger conditions. If it meets the conditions, continue to execute the circuit breaker state and extend the circuit breaker state time. If it does not meet the conditions, restore the initial reporting state. When timing the circuit breaker state and performing the release operation, the timing is started after the terminal starts the circuit breaker state. When the timing reaches the preset single circuit breaker time, perform limited release evaluation. Evaluate whether the frequency of suppressed alarms counted in the most recent evaluation window meets or is below the normal frequency. If it meets the conditions, release the circuit breaker state of the corresponding terminal and restore the initial reporting process. If it does not meet the conditions, extend the circuit breaker state time in a step-by-step manner.

[0070] Example 2: The present invention provides as follows Figure 2 The remote monitoring system based on the intelligent converged terminal of the transformer substation is shown to implement the remote monitoring method based on the intelligent converged terminal of the transformer substation in Embodiment 1. It includes a data acquisition and alarm generation unit, a statistics unit, a circuit breaker adjudication unit, a circuit breaker processing unit, a root cause diagnosis unit, a handling decision unit, and a circuit breaker management unit.

[0071] The data acquisition and alarm generation unit is used to connect to the intelligent fusion terminal of the distribution area, collect three-phase voltage telemetry data at a fixed frequency, remove invalid values ​​and smooth noise through data cleaning, provide high-quality data stream for subsequent analysis, and generate original alarm events through calculation. Within a unit time window, it calculates three characteristic indicators in real time: instantaneous gradient, standard deviation burst increment and residual abnormality, and compares them with historical thresholds. Once a sudden change is detected, a structured original alarm event is immediately generated.

[0072] The statistics unit is used to count all generated raw alarm events, maintains a sliding time window counter for each terminal, and counts the alarm frequency in real time; at the same time, it performs simple sequence analysis to provide input for pattern matching.

[0073] The circuit breaker decision unit analyzes the statistical raw alarm events to determine whether to trigger the circuit breaker processing status. It comprehensively judges the frequency conditions, mode conditions, and baseline deviation conditions. If any terminal meets the circuit breaker triggering conditions, it issues a circuit breaker triggering command; otherwise, it marks the alarm event as normal and uploads it normally.

[0074] The circuit breaker processing unit is used to filter original alarm events and transfer suppressed alarms, and aggregate alarm circuit breaker summary events. When a terminal enters the circuit breaker state, it is immediately started to perform alarm filtering, intercept duplicate and derived alarms and transfer them to logs, and perform event aggregation to generate alarm circuit breaker summary events. Finally, the corresponding root cause diagnosis is called according to the alarm type.

[0075] The root cause diagnosis unit is used to generate a diagnostic report for the root cause diagnosis of the filtered original alarm events. After receiving the diagnostic task, it retrieves relevant data, runs the analysis, and outputs a preliminary diagnostic report containing possible root causes, evidence, and confidence levels.

[0076] The decision-making unit is used to receive and summarize events and diagnostic reports, generate a decision-making report containing specific operational recommendations, and transform the analysis conclusions into actionable plans.

[0077] The circuit breaker management unit is used to maintain the circuit breaker status timer for all terminals. After the timer expires, it initiates a limited release condition assessment and issues a circuit breaker restoration or extension command based on the assessment results.

[0078] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A remote monitoring method based on a smart converged terminal for distribution transformer areas, characterized in that, The following steps are included: S1 collects telemetry data from each terminal and performs real-time analysis based on mutation detection. After detecting an anomaly, it generates an original alarm event and sets the trigger conditions for the circuit breaker state. S2 collects raw alarm events from each terminal, performs statistical analysis, pattern matching, and dynamic baseline comparison of the number of raw alarm events from each terminal within a unit time window, and analyzes whether each terminal meets the triggering conditions for the circuit breaker state. If the conditions are met, the alarm processing circuit breaker state of the corresponding terminal is triggered; if the conditions are not met, the raw alarm events are reported. S3, after the circuit breaker state is triggered, firstly, the repeated or derived alarm events of the triggering terminal are suppressed and converted into background logs. At the same time, alarm circuit breaker summary events are generated and analyzed through root cause diagnosis to generate analysis reports. S4, combine the event details and analysis report of the circuit breaker summary event to generate a processing operation report. If it can be resolved remotely, it can be processed directly through remote control commands. If on-site handling is required, the handling shall be carried out in accordance with the handling operation report; S5, start timing the circuit breaker state. After the circuit breaker state has lasted for a preset time, release the alarm event of the corresponding terminal according to the limited release conditions, and continue to determine whether the released terminal meets the circuit breaker state trigger conditions. If it meets the conditions, continue to execute the circuit breaker state and extend the circuit breaker state time. If it does not meet the conditions, restore the initial reporting state.

2. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 1, characterized in that, The generation of the original alarm event includes the following steps: S101 continuously collects telemetry data from each terminal at a fixed frequency, and defines a fixed time analysis window. Within the window, the collected data is preprocessed, including removing invalid values ​​and performing low-pass filtering to smooth random noise. S102, within each analysis window, calculates the instantaneous gradient of the three-phase voltage, the sudden increase in standard deviation within the window, and the residual abnormality of the data sequence relative to the fitting trend in real time; S103 compares the instantaneous gradient, standard deviation burst increment, and fitting residual abnormality calculated in S102 with a preset threshold based on historical data. When any feature index continuously exceeds the threshold, it is determined that a data mutation has occurred, and a structured original alarm event containing event type, timestamp, terminal identifier, mutation parameters, and feature values ​​is generated.

3. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 2, characterized in that, The instantaneous gradient is calculated by performing median filtering on the voltage sequence within a defined fixed-time analysis window, and then calculating the weighted difference value of the filtered voltage sequence.

4. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 2, characterized in that, The calculation of the standard deviation spike increment within the window is as follows: first, the standard deviation σ1 of the voltage data within the fixed-time analysis window is calculated, and the baseline standard deviation σ0 of the normal fluctuation level is obtained. Then, the difference between σ1 and σ0 is divided by σ0 to obtain the standard deviation spike increment.

5. A remote monitoring method based on a smart converged terminal for distribution areas according to claim 2, characterized in that, The calculation of residual abnormality first uses the LOESS algorithm to perform nonlinear fitting on the data points within a fixed-time analysis window, and then calculates the residual e between each data point and the LOESS fitted curve. i Then, the standardized values ​​of these residuals are calculated using the following formula: （1）； in, This represents the mean value obtained based on the statistical analysis of historical normal window residuals. This represents the standard deviation obtained based on the statistical analysis of historical normal window residuals. Finally, the maximum value among all the absolute values ​​of the standardized residuals is taken as the residual constant.

6. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 1, characterized in that, The triggering conditions for the circuit breaker state include frequency conditions, mode conditions, and baseline deviation conditions. The measure of whether the number of alarm events triggered by the frequency condition exceeds the set frequency threshold; The mode conditions measure the inherent correlation and repetition between alarm events and distinguish between random discrete alarms and regular fault alarms. The baseline deviation condition measures the degree of deviation of the current alarm activity from the historical normal level of the terminal.

7. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 1, characterized in that, The quantity statistics comparison first counts the total number of original alarm events generated by each terminal within a unit time window in real time, and then compares it with a preset frequency threshold. If the frequency threshold is exceeded, the frequency is marked as abnormal. The pattern matching comparison first analyzes the sequence of alarm events within the same time window, and then uses a rule engine to match the real-time alarm sequence with a predefined abnormal pattern library. If any preset abnormal pattern is matched, the pattern is marked as abnormal and the matched pattern type is recorded. The dynamic baseline comparison first calculates the comprehensive index of alarm activity intensity in the current window, and then calculates the deviation rate of the real-time index relative to the dynamic baseline. If the deviation rate exceeds the set threshold, the baseline deviation is marked as abnormal.

8. The remote monitoring method based on a smart converged terminal for distribution areas according to claim 1, characterized in that, After the circuit breaker is triggered, the original alarm events of the terminal that triggered the circuit breaker are filtered and intercepted. After filtering out duplicate alarms and derivative alarms, they are organized into a structured event log table. Then, starting from the time the circuit breaker is triggered, an analysis window is traced back and all original alarm events within this analysis window are counted. This includes counting the total number of alarms, classifying and counting alarms by type and severity level, and calculating the scope of affected next-level equipment or users based on the power grid topology. A summary event record is generated. Finally, root cause diagnosis analysis is matched and scheduled according to alarm type characteristics to generate a preliminary handling suggestion report.

9. A remote monitoring method based on a smart converged terminal for distribution areas according to claim 1, characterized in that, During the circuit breaker state timing and release operation, timing is first started after the terminal starts the circuit breaker state. When the timing reaches the preset single circuit breaker time, a limited release assessment is performed. The assessment evaluates whether the frequency of suppressed alarms counted in the most recent assessment window meets or is below the normal frequency. If it meets the requirements, the circuit breaker state of the corresponding terminal is released and the initial reporting process is restored. If it does not meet the requirements, the circuit breaker state time is extended in a phased increment.

10. A remote monitoring system based on a smart converged terminal for transformer substations, used to implement the remote monitoring method based on a smart converged terminal for transformer substations as described in any one of claims 1-9, characterized in that, It includes a data acquisition and alarm generation unit, a statistics unit, a circuit breaker adjudication unit, a circuit breaker processing unit, a root cause diagnosis unit, a handling decision unit, and a circuit breaker management unit; The data acquisition and alarm generation unit is used to connect to the intelligent fusion terminal of the distribution area, collect three-phase voltage telemetry data at a fixed frequency, and generate original alarm events through calculation. The statistical unit is used to count all generated original alarm events; The circuit breaker decision unit determines whether to trigger the circuit breaker processing state by analyzing the statistical original alarm events. The circuit breaker processing unit is used to filter the original alarm events and transfer the suppressed alarms, and aggregate them to generate alarm circuit breaker summary events; The root cause diagnosis unit is used to generate a diagnostic report from the root cause diagnosis of the filtered original alarm events. The decision-making unit is used to receive the summarized events and diagnostic reports, and generate a decision-making report containing specific operational suggestions. The circuit breaker management unit is used to maintain the circuit breaker status timer for all terminals, and after the timer expires, it initiates a limited release condition assessment.

Images