CN-121770176-B - Capacitor capacity abnormity dynamic monitoring method based on edge calculation

Abstract

The invention relates to the technical field of power system state monitoring and intelligent operation and maintenance, and discloses a capacitor capacity abnormity dynamic monitoring method based on edge calculation. By connecting the edge monitoring unit and the series current limiting resistor in parallel in the target capacitor loop, a unified voltage sampling period and a time stamp are set, and safe and controllable discharging and accurate time sequence recording are realized. And continuously collecting capacitor voltage by the edge node, constructing full-sequence original data, and aligning the full-sequence original data with sampling values point by point through a theoretical discharge model to form a capacity error evaluation function. The system optimizes in the capacity searching range, judges convergence and error distribution stability, automatically outputs capacity abnormality judgment, and does not need manual intervention. All monitoring results are synchronously stored with the unique identification and the time stamp, the sliding window algorithm fuses historical data in real time, the continuous degradation trend of the capacity is captured, and the real-time performance, the precision and the early warning capability of monitoring are improved.

Inventors

• MIAO JIAOHONG

Assignees

• 江阴市君利莱电子科技有限公司

Dates

Publication Date

20260512

Application Date

20251231

Claims (9)

1. 1. The method for dynamically monitoring the capacitor capacity abnormality based on edge calculation is characterized by comprising the following steps: An edge monitoring unit and a series current limiting resistor are connected in parallel in an electric loop where a target capacitor is located, a preset voltage sampling period is set, an initial time stamp is recorded, and an initial voltage value and a nominal capacitance value of a relative time zero point moment are obtained; Collecting voltages at two ends of a target capacitor according to a preset voltage sampling period from the moment of a relative time zero point to form an original voltage sequence data set; Establishing a theoretical discharge voltage model taking initial voltage values at two ends of a target capacitor, resistance values of a current limiting resistor and candidate capacitance values as parameters, calculating theoretical voltages at all sampling moments, and generating a theoretical voltage sequence data set corresponding to an original sequence one by one; calculating residual errors of the voltage actual measurement value and the corresponding theoretical voltage under each candidate capacitance value, and squaring and summing to obtain a capacitance error evaluation function taking the capacitance value as an independent variable; calculating a capacity error evaluation function in a preset capacity searching range, selecting the smallest sum of squares of total residual errors as a capacity optimal estimated value, and comparing the smallest sum of squares of total residual errors with a preset residual error difference threshold value by the difference between the smallest sum of squares and the next smallest sum of squares, so as to judge convergence and residual error distribution stability; Calculating the capacity absolute deviation and capacity relative error ratio of the capacity optimal estimated value and the nominal capacity value when the capacity optimal estimated value is converged and stabilized, comparing the capacity absolute deviation and capacity relative error ratio with a capacity relative error ratio threshold value, and outputting abnormal capacity or normal capacity; Generating a capacity monitoring result record by using the capacity optimal estimated value, the capacity relative error proportion and the capacity state mark, and writing the capacity monitoring result record into an edge node database by combining the unique identification code of the target capacitor and the initial timestamp; setting the length of a sliding window, selecting the most recent capacity optimal estimated value in the window, calculating the error ratio of the sliding average capacity value to the sliding capacity, comparing the error ratio with the threshold value of the relative error ratio of the capacity, identifying the continuous degradation trend and outputting a capacity degradation early warning mark.
2. 2. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 1, wherein the edge monitoring unit and the series current limiting resistor are connected in parallel in an electrical loop where a target capacitor is located, a preset voltage sampling period is set, a starting time stamp is recorded, and an initial voltage value and a nominal capacitance value at a relative time zero point moment are obtained, specifically comprising: edge monitoring units are connected in parallel at two ends of the target capacitor, each edge monitoring unit has a voltage acquisition function and a local calculation processing capacity, and a unique number is allocated to each edge monitoring unit; A current limiting resistor is connected in series in an electric loop where a target capacitor is located, the resistance value of the current limiting resistor is preset and recorded, and the resistance value of the current limiting resistor is used as a fixed resistance parameter in a discharge loop; Configuring a voltage sampling period in the edge monitoring unit, and setting the voltage sampling period as a preset voltage sampling period; Before a target capacitor discharge test is executed, recording the current actual calendar time, recording the actual calendar time as a discharge test starting time stamp, and defining a discharge starting time as a relative time zero point time; At the moment of starting discharging of the target capacitor, collecting initial voltage values at two ends of the target capacitor through an edge monitoring unit, and recording the initial voltage values at the two ends of the target capacitor as initial voltage values at the moment of relative time zero point; And obtaining the nominal capacitance value of the target capacitor from the nameplate of the target capacitor or the factory technical data, and taking the nominal capacitance value as a reference capacity parameter in the capacity identification and abnormality judgment process.
3. 3. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 2, wherein the step of collecting voltages at both ends of the target capacitor according to a preset voltage sampling period from the moment of relative time zero to form an original voltage sequence data set specifically comprises the steps of: In the discharging process of the target capacitor, starting from the relative time zero point moment, performing continuous equidistant sampling according to a preset voltage sampling period to obtain a plurality of sampling moments which are arranged according to time sequence, wherein the total number of time and voltage sampling point pairs is not less than two; at each sampling moment, respectively acquiring voltage actual measurement values of two ends of a target capacitor through an edge monitoring unit, and forming a group of time and voltage sampling point pairs by each sampling moment and the corresponding voltage actual measurement value; and sequentially storing all time and voltage sampling point pairs into a data set according to a time sequence to form an original voltage sequence data set for describing the discharging process of the target capacitor.
4. 4. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 3, wherein the establishing a theoretical discharge voltage model taking initial voltage values at two ends of a target capacitor, resistance values of a current limiting resistor and candidate capacitance values as parameters, calculating theoretical voltages at each sampling moment, and generating a theoretical voltage sequence data set corresponding to an original sequence one by one specifically comprises: Based on a discharge circuit formed by connecting a target capacitor and a current-limiting resistor in series, constructing a theoretical discharge voltage model by adopting an exponential decay characteristic of serial discharge of the resistor and the capacitor, wherein the theoretical discharge voltage model takes an initial voltage value at two ends of the target capacitor, a resistance value of the current-limiting resistor and a candidate capacitance value as input parameters, and outputting theoretical voltage values at two ends of the target capacitor under a given time condition; for each sampling time recorded in an original voltage sequence data set, under the condition of the same candidate capacitance value, invoking a theoretical discharge voltage model to calculate theoretical voltage values at two ends of a target capacitor corresponding to the corresponding sampling time; And sequentially arranging theoretical voltage values corresponding to each sampling time according to the time sequence of the original voltage sequence data set to generate theoretical voltage sequence data sets which are in one-to-one correspondence with the original voltage sequence data sets at the sampling time.
5. 5. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 4, wherein the calculating the residual error between the measured voltage value and the corresponding theoretical voltage value and summing the squares of the residual errors under each candidate capacitance value to obtain a capacity error evaluation function with the capacitance value as an independent variable comprises: Under the condition of each candidate capacitance value, calculating the difference value between the actual measurement value of the voltage in the corresponding time and voltage sampling point pair and the theoretical voltage value of the corresponding time and voltage sampling point pair in the theoretical voltage sequence data set for each time and voltage sampling point pair in the original voltage sequence data set, and taking the difference value between the actual measurement value of the voltage and the theoretical voltage value as a single point residual error of the corresponding time and voltage sampling point pair; under the same candidate capacitance value, squaring single-point residuals of all time and voltage sampling point pairs respectively, and executing summation operation within the range of all time and voltage sampling point pairs to obtain total residual square sum aiming at the current candidate capacitance value; And forming a capacity error evaluation function with independent variables of candidate capacitance values and function values of corresponding total residual error square sums in all candidate capacitance value ranges, wherein the capacity error evaluation function only changes along with the candidate capacitance values, and the rest parameters remain unchanged in the capacity identification process.
6. 6. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 5, wherein calculating a capacity error evaluation function in a preset capacity search range, selecting the minimum sum of squares of total residuals as a capacity optimal estimated value, and comparing the difference between the minimum value and the second minimum value with a preset residual difference threshold value to determine convergence and residual distribution stability, comprises: Setting a capacity searching lower limit to be half of a nominal capacitance value according to the nominal capacitance value of the target capacitor, setting a capacity searching upper limit to be the nominal capacitance value, dividing a plurality of discrete capacity points between the capacity searching lower limit and the capacity searching upper limit according to a preset capacity step length, and constructing a capacity searching set consisting of a plurality of candidate capacitance values; For each candidate capacitance value in the capacity search set, calling a capacity error evaluation function to calculate the total residual square sum corresponding to each candidate capacitance value, and recording the corresponding relation between the candidate capacitance value and the corresponding total residual square sum; Searching a candidate capacitance value with the minimum total residual error square sum in the capacity searching set, taking the candidate capacitance value with the minimum total residual error square sum as a capacity optimal estimated value in the discharging process, and recording the corresponding minimum total residual error square sum as the current minimum total residual error square sum; After the candidate capacitance value with the minimum total residual error square sum is removed, searching a candidate capacitance value with the next smallest total residual error square sum in a residual candidate capacitance value set, and recording the total residual error square sum corresponding to the next smallest candidate capacitance value as the current next smallest total residual error square sum; calculating the absolute value of the difference between the current total residual square sum and the current minimum total residual square sum, comparing the absolute value of the difference with a preset residual difference threshold, wherein the physical quantity unit of the preset residual difference threshold is voltage square, the magnitude order of the physical quantity unit is one part per million, when the difference is smaller than the preset residual difference threshold, the optimal estimated value of the determined capacity meets the convergence condition and the residual distribution is stable, and when the difference is larger than or equal to the preset residual difference threshold, the optimal estimated value of the determined capacity does not meet the stability requirement.
7. 7. The method for dynamically monitoring capacitor capacity anomaly based on edge calculation according to claim 6, wherein the calculating the capacity absolute deviation of the capacity optimal estimated value from the nominal capacity value and the capacity relative error ratio at the time of convergence and stabilization, and comparing with the capacity relative error ratio threshold value, outputs the capacity anomaly or the capacity anomaly, specifically comprises: After the capacity optimal estimated value is obtained and the convergence condition is met, calculating the absolute value of the difference between the capacity optimal estimated value and the nominal capacitance value to obtain the capacity absolute deviation; taking the ratio of the absolute deviation of the capacity to the nominal capacitance value as the capacity relative error ratio, and obtaining a capacity relative error ratio value by dividing the absolute deviation of the capacity by the nominal capacitance value; presetting a capacity relative error proportion threshold value, wherein the value of the capacity relative error proportion threshold value is set to be zero point one, and the corresponding percentage form is ten percent; And comparing the capacity relative error proportion with a capacity relative error proportion threshold, judging the capacity state of the target capacitor as abnormal capacity when the capacity relative error proportion is larger than the capacity relative error proportion threshold, and judging the capacity state of the target capacitor as normal capacity when the capacity relative error proportion is smaller than or equal to the capacity relative error proportion threshold.
8. 8. The method for dynamically monitoring capacitor capacity anomaly based on edge calculation according to claim 7, wherein the generating a capacity monitoring result record by using the capacity optimal estimated value, the capacity relative error ratio and the capacity status flag, and writing the capacity monitoring result record into the edge node database by combining the unique identification code of the target capacitor with the start time stamp, specifically comprises: Constructing a capacity monitoring result record according to the capacity optimal estimated value, the capacity relative error proportion and the capacity state mark in the current discharging process, wherein the capacity state mark adopts binary coding, a first value represents that the capacity is judged to be in an abnormal state, and a zero value represents that the capacity is judged to be in a normal state; the method comprises the steps of obtaining a unique identification code of a target capacitor and a discharge test starting time stamp, taking the unique identification code of the target capacitor and the discharge test starting time stamp as index information, and combining the index information with a capacity optimal estimated value, a capacity relative error proportion and a capacity state mark to form a capacity monitoring result record; And writing the capacity monitoring result records into an edge node database, storing according to the index information, and maintaining a structured capacity monitoring result record set in the edge node database.
9. 9. The method for dynamically monitoring abnormal capacitor capacity based on edge calculation according to claim 8, wherein the step of setting a sliding window length, selecting the most recent capacity optimal estimation value in the window, calculating the ratio of a sliding average capacity value to a sliding capacity error, comparing the ratio with a capacity relative error ratio threshold, identifying a continuous degradation trend and outputting a capacity degradation early warning mark comprises the following steps: Selecting the capacity optimal estimated value corresponding to the most recent multiple discharge tests of the same target capacitor according to a time sequence in an edge node database, presetting the sliding window length to be more than one, and forming the most recent multiple capacity optimal estimated values into a sliding window capacity estimated set; Calculating the optimal estimated value of the capacity in the sliding window in an arithmetic average mode during each degradation trend analysis to obtain a sliding average capacity value; calculating the absolute value of the difference between the sliding average capacitance value and the nominal capacitance value to obtain the sliding capacity absolute deviation, and taking the ratio of the sliding capacity absolute deviation to the nominal capacitance value as the sliding capacity error ratio; Comparing the sliding capacity error proportion with a capacity relative error proportion threshold, when the sliding capacity error proportion is larger than the capacity relative error proportion threshold, judging that the capacity of the target capacitor has a continuous degradation trend, generating a capacity degradation early warning mark, and when the sliding capacity error proportion is smaller than or equal to the capacity relative error proportion threshold, judging that the capacity change of the target capacitor is in an allowable range, and not generating the capacity degradation early warning mark.

Description

Capacitor capacity abnormity dynamic monitoring method based on edge calculation Technical Field The invention relates to the technical field of power system state monitoring and intelligent operation and maintenance, in particular to a capacitor capacity abnormity dynamic monitoring method based on edge calculation. Background Along with the continuous expansion of the scale of the power system and the improvement of the intelligent level, a large number of parallel capacitors are widely applied to the scenes of electric energy quality control, reactive compensation and the like. The accuracy of the capacitor capacity is directly related to reactive power balance and equipment safety of a power distribution system, however, in the long-term operation process, the capacitor is easy to generate capacity attenuation and even faults due to medium aging, overvoltage impact, partial discharge and the like, so that the actual capacity deviates from a nominal value, and the stable operation of the power system is further affected. Therefore, it is important to timely and accurately detect and monitor the abnormal capacitor capacity. The prior art mainly relies on modes such as manual regular inspection, off-line test instruments or background big data statistical analysis to detect the capacity of the capacitor. For example, part of transformer substations adopt portable capacity testers for periodic spot check, power failure wire disconnection is needed each time, manual measurement is performed, and the transformer substations have long period and are easy to miss. Part of the system attempts to estimate the running state of the capacitor through a static model or an empirical criterion based on voltage and current data acquired by the main station, but the result is extremely susceptible to load fluctuation, power grid disturbance and the like, and the accuracy and the instantaneity are limited. In recent years, the communication network is utilized to transmit the original data back to the background server, and the capacity information is processed by means of the centralized algorithm, but the method faces the real problems of bandwidth bottleneck, network delay, mass data centralized storage and the like, and is difficult to meet the requirement of monitoring the distributed large-scale capacitor. In addition, the traditional method often ignores dynamic modeling and real-time parameter identification of physical characteristics of the equipment body, only depends on single-point data or static rules, lacks pertinence, and cannot realize early detection and trend tracking of capacity abnormality. Meanwhile, because the field working condition is complex, the environment interference is large, the existing detection means have uncertainty in links such as data acquisition, boundary criterion setting, capacity change curve analysis and the like, so that misjudgment and missed judgment phenomena are frequent, and the operation and maintenance efficiency and the safety level of the power system are seriously affected. Therefore, the scheme aims to provide a capacitor capacity anomaly dynamic monitoring method based on edge calculation, which is characterized in that a high-resolution voltage sequence is obtained through a preset sampling period, a corresponding theoretical voltage sequence is generated by combining an RC discharge theoretical model, capacity is estimated through a residual error square sum function, and accurate, real-time and low-delay capacity anomaly detection and degradation early warning are realized by utilizing two mechanisms of residual error convergence judgment and sliding window trend analysis. Disclosure of Invention The invention provides a capacitor capacity abnormity dynamic monitoring method based on edge calculation, which facilitates solving the problems in the background art. The invention provides a capacitor capacity abnormity dynamic monitoring method based on edge calculation, which comprises the following steps: An edge monitoring unit and a series current limiting resistor are connected in parallel in an electric loop where a target capacitor is located, a preset voltage sampling period is set, an initial time stamp is recorded, and an initial voltage value and a nominal capacitance value of a relative time zero point moment are obtained; Collecting voltages at two ends of a target capacitor according to a preset voltage sampling period from the moment of a relative time zero point to form an original voltage sequence data set; Establishing a theoretical discharge voltage model taking initial voltage values at two ends of a target capacitor, resistance values of a current limiting resistor and candidate capacitance values as parameters, calculating theoretical voltages at all sampling moments, and generating a theoretical voltage sequence data set corresponding to an original sequence one by one; calculating residual errors of the voltage actual measurement value and th