Real-time monitoring method for charging and discharging states of intelligent capacitor

Abstract

The invention relates to the technical field of power electronic equipment monitoring, and discloses an intelligent capacitor charging and discharging state real-time monitoring method, which comprises the following steps: acquiring real-time charging and discharging data of a capacitor, and screening out voltage and current parameters which change along with time, including instantaneous values and accumulated values; establishing a state evaluation model according to the parameters, and deducing monitoring state data of a plurality of time periods backwards to obtain initial monitoring state data; the method comprises the following steps: calculating state variation of adjacent time periods in initial monitoring state data, when the variation reaches a set limit, determining a key monitoring area, collecting historical state records of a plurality of monitoring nodes in a set coverage range, generating a monitoring adjustment instruction according to the historical state records, and distributing the monitoring adjustment instruction to each monitoring node to enhance the monitoring density of a specific time period. According to the method, the charging and discharging states of the intelligent capacitor are dynamically and accurately monitored, the monitoring efficiency and the early warning reliability are improved, and the monitoring resource configuration is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of power electronic equipment monitoring, and in particular to a real-time monitoring method for the charging and discharging status of an intelligent capacitor. Background Art

[0002] In power systems, capacitors are crucial reactive power compensation devices, and their operational stability directly impacts the power quality, safety, and reliability of the system. As power systems become increasingly intelligent, the need for real-time, accurate monitoring of capacitor charge and discharge status is becoming increasingly urgent. However, existing intelligent capacitor charge and discharge status monitoring technologies still have many shortcomings.

[0003] Traditional monitoring methods often use a fixed monitoring frequency, making it difficult to dynamically adjust the monitoring density based on the actual operating status of the capacitor. For example, when the capacitor's operating status changes dramatically, a fixed monitoring frequency may result in missing critical data and fail to capture sudden changes in status in a timely manner. Conversely, when the operating status is relatively stable, excessively high monitoring frequencies waste resources and increase the data processing burden.

[0004] Existing models for evaluating capacitor charge and discharge status are often based on historical data or empirical formulas, lacking dynamic updates and adaptive adjustments to real-time data. This significantly reduces prediction accuracy when faced with complex operating conditions or aging equipment, making it difficult to accurately reflect the actual state of the capacitor.

[0005] Traditional methods typically employ uniform monitoring across the entire area, failing to accurately identify key monitoring areas based on changes in capacitor operating conditions. This one-size-fits-all approach not only fails to effectively identify potential fault hazards but also leads to irrational allocation of monitoring resources, reducing monitoring efficiency.

[0006] Existing early warning mechanisms often rely on single parameter thresholds, lacking comprehensive analysis and trend prediction of multiple parameters, such as voltage and current. When capacitors experience progressive failures, single parameter threshold warnings often lag behind the actual development of the fault, making early warning difficult and impacting the safe operation of the power system.

[0007] In terms of historical data utilization, existing technologies fail to fully tap the value of historical status records and are unable to optimize monitoring strategies based on historical data. For example, it is impossible to determine the appropriate monitoring frequency for different areas based on historical data, resulting in a lack of targeted and scientific monitoring strategies. Summary of the Invention

[0008] The purpose of the present invention is to provide a method for real-time monitoring of the charging and discharging status of an intelligent capacitor to solve the problems raised in the above background technology.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for real-time monitoring of the charge and discharge status of an intelligent capacitor, the method comprising:

[0010] Acquire real-time charge and discharge data of capacitors and screen voltage and current parameters that change over time, including instantaneous and cumulative values;

[0011] A status assessment model is established based on the selected parameters, and the monitoring status data of several time periods are deduced backward to obtain the initial monitoring status data;

[0012] Calculating the state change amount of adjacent time periods in the initial monitoring state data, and when the change amount reaches a set limit, determining the key monitoring area based on the corresponding relationship between the change amount and the capacitor operation mode;

[0013] Collect historical status records of multiple monitoring nodes within the coverage area according to key monitoring areas;

[0014] Generating a monitoring adjustment instruction based on the historical status record, and distributing the adjustment instruction to multiple monitoring nodes, wherein the adjustment instruction is used to instruct to increase the monitoring density of a specific time period based on the historical status record;

[0015] Based on the status assessment model, the actual status data of the latest period is supplemented, and the updated monitoring status data of the next period is obtained. The next period is compared based on the initial monitoring status data and the updated monitoring status data to generate a comparison result. Based on the comparison result, it is determined whether to trigger the early warning mechanism.

[0016] Preferably, the obtaining of real-time charge and discharge data of the capacitor and screening of voltage and current parameters that vary with time include:

[0017] Get the real-time charge and discharge data of the capacitor within the set time range;

[0018] Based on the set time range, select all time periods in the same cycle as the current monitoring period;

[0019] Filter the voltage and current parameters of all time periods in real-time charge and discharge data.

[0020] Preferably, the state assessment model is established based on the screened parameters, and the monitoring state data of several time periods are deduced backward to obtain the initial monitoring state data, including:

[0021] Establishing a state assessment model based on voltage and current parameters of all time periods to generate a first matching curve that meets a first matching condition;

[0022] The monitoring status data of several time periods are deduced backward according to the first matching curve to obtain the initial monitoring status data.

[0023] Preferably, the calculating of the state change amount between adjacent time periods in the initial monitoring state data, and when the change amount reaches a set limit, determining the key monitoring area according to the correspondence between the change amount and the capacitor operation mode includes:

[0024] Obtain the operating mode characteristics of a single monitoring point in the state change;

[0025] According to the operation mode characteristics and the total number of monitoring points in the state change amount, a key monitoring area is determined, and the key monitoring area includes specific locations that require enhanced monitoring.

[0026] Preferably, generating a monitoring adjustment instruction based on the historical status record and distributing the adjustment instruction to multiple monitoring nodes includes:

[0027] Taking the key monitoring area as the center, the set coverage range is divided into sections in the order from near to far, and several sub-areas within the set sub-range are obtained;

[0028] Select monitoring nodes in each sub-area and determine the basic monitoring frequency within each sub-area to ensure that the deviation is within the allowable range based on the historical status records of the monitoring nodes in each sub-area;

[0029] Determine the monitoring nodes that meet the basic monitoring frequency in each sub-area, and obtain multiple monitoring nodes, wherein the sum of the basic monitoring frequencies of the multiple monitoring nodes is greater than or equal to the target monitoring frequency;

[0030] Generate and issue monitoring adjustment instructions based on multiple monitoring nodes and the basic monitoring frequency of each monitoring node;

[0031] The historical status records of multiple monitoring nodes within the set coverage area collected according to the key monitoring area include:

[0032] With the key monitoring area as the center, the set coverage area is divided into three layers of areas with increasing radius;

[0033] For the monitoring nodes in each layer area, the voltage and current historical data of the same time period every day in the past three months are collected to form a historical status record of each node.

[0034] Preferably, the actual status data of the latest period is supplemented based on the status assessment model, and updated monitoring status data of the next period is obtained, and the comparison of the next period is performed based on the initial monitoring status data and the updated monitoring status data to generate the comparison result, which includes:

[0035] The state assessment model is updated according to the state data of the latest period to generate a second matching curve that meets the second matching condition;

[0036] Identify updated monitoring status data for the next period in the second matching curve, perform a fluctuation comparison based on the initial monitoring status data and the updated monitoring status data, and generate a comparison result.

[0037] Preferably, performing fluctuation comparison based on the initial monitoring state data and the updated monitoring state data to generate a comparison result includes:

[0038] Extracting the predicted voltage value and the predicted current value of the next period in the initial monitoring state data, and updating the actual voltage value and the actual current value of the next period in the monitoring state data;

[0039] The deviation rate between the voltage prediction value and the actual value and the deviation rate between the current prediction value and the actual value are calculated respectively, and the average value of the two deviation rates is used as the fluctuation comparison result.

[0040] Preferably, determining whether to trigger an early warning mechanism based on the comparison result includes:

[0041] When the difference between the monitoring data of the next period in the updated monitoring state data and the initial monitoring state data is within the allowable difference, a first instruction for maintaining the current monitoring density is generated to fully maintain the monitoring density of the specific period;

[0042] When the updated monitoring status data is less than the monitoring data of the next period in the initial monitoring status data, and the difference between the two is greater than the allowed difference, based on the distribution of basic monitoring frequencies among the multiple monitoring nodes, an enhanced monitoring and early warning coexistence instruction is generated for some nodes, so that at least one of the multiple monitoring nodes coordinates monitoring of other nodes except itself, and maintains the monitoring density of the at least one node;

[0043] When the updated monitoring status data is greater than the monitoring data of the next period in the initial monitoring status data, and the difference between the two is greater than the allowed difference, a second instruction for comprehensive enhanced monitoring is generated to fully enhance the monitoring density of the specific period.

[0044] Preferably, establishing a state assessment model based on the voltage and current parameters of all time periods to generate a first matching curve that meets the first matching condition includes:

[0045] Arrange the voltage and current parameters of all time periods in time series to construct a training data set;

[0046] The historical data cross-validation method is used to train the model on the training data set, and the model parameters are adjusted until the fitting error of the model output is less than the first matching threshold, thereby generating a first matching curve.

[0047] Preferably, the calculation of the state change amount in adjacent time periods in the initial monitoring state data includes:

[0048] Select the voltage and current parameters of the continuous time period in the initial monitoring state data, and calculate the voltage difference and current difference between the current time period and the previous time period;

[0049] The sum of the absolute values ​​of the voltage difference and the current difference is taken as the state change amount, which is used to reflect the degree of fluctuation of the charge and discharge states in adjacent time periods.

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] The real-time monitoring method for intelligent capacitor charge and discharge status, provided by this invention, obtains real-time capacitor charge and discharge data and screens voltage and current parameters, providing accurate basic data for status assessment. By establishing a status assessment model and deducing the monitoring status data backward, the operating status of the capacitor can be predicted in advance, providing a basis for subsequent monitoring and early warning. By calculating the state changes between adjacent time periods and determining key monitoring areas accordingly, this method achieves precise allocation of monitoring resources, improves monitoring efficiency, and avoids resource waste.

[0052] By collecting historical status records from multiple monitoring nodes centered around key monitoring areas and generating monitoring adjustment instructions based on these records, the system can dynamically adjust monitoring density based on historical data, making monitoring strategies more scientific and rational. For example, the monitoring frequency can be increased near key areas to ensure effective monitoring of the operating status of critical locations, while maintaining a basic monitoring frequency in non-key areas, thus optimizing the allocation of monitoring resources.

[0053] The latest actual status data is supplemented and compared with the status assessment model, generating a comparison result to determine whether to trigger the early warning mechanism. This enables dynamic tracking of the capacitor's operating status and real-time early warning. By comprehensively analyzing the deviation rate between the predicted and actual voltage and current values, the operating status of the capacitor can be more accurately determined, avoiding the limitations of single-parameter early warning.

[0054] When the difference in monitoring data is within different ranges, different instructions are generated, such as maintaining monitoring density, partially enhancing monitoring and early warning at the same time, or comprehensively enhancing monitoring, making the early warning mechanism more flexible and accurate, and able to take corresponding measures in a timely manner according to actual conditions, thereby improving the safety and reliability of the system.

[0055] The historical data cross-validation method is used to train the model on the training data set, which ensures the accuracy and reliability of the state assessment model, enables the model to better adapt to the actual operation of the capacitor, and improves the accuracy of state prediction.

[0056] Defining the state change as the sum of the absolute values ​​of the voltage difference and the current difference can more comprehensively reflect the degree of fluctuation of the charging and discharging states in adjacent time periods, and provide a more scientific basis for the determination of key monitoring areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a working principle diagram of the real-time monitoring method for the charging and discharging status of an intelligent capacitor according to the present invention;

[0058] Figure 2 A flow chart for updating the state assessment model and generating comparative results;

[0059] Figure 3 A flow chart for generating comparison results for performing fluctuation comparison;

[0060] Figure 4 Generate a design diagram of the first matching curve for establishing a state assessment model. DETAILED DESCRIPTION

[0061] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0062] See also Figure 1-Figure 4 The present invention relates to a method for real-time monitoring of the charging and discharging status of an intelligent capacitor, and the specific implementation steps are as follows:

[0063] Acquire real-time capacitor charge and discharge data and filter out time-varying voltage and current parameters, including instantaneous and cumulative values. Specifically, acquire real-time capacitor charge and discharge data within a set time range. Based on this set time range, select all time periods that fall within the same cycle as the current monitoring period, and then filter out the voltage and current parameters for all time periods in the real-time charge and discharge data.

[0064] A state assessment model is established based on the selected parameters, and the monitoring state data for several time periods is deduced backward to obtain the initial monitoring state data. Specifically, the state assessment model is established based on the voltage and current parameters of all time periods, generating a first matching curve that meets the first matching condition. The monitoring state data for several time periods is then deduced backward based on the first matching curve to obtain the initial monitoring state data.

[0065] The initial monitoring state data is calculated for adjacent time periods. When the change reaches a set limit, a key monitoring area is identified based on the correspondence between the change and the capacitor's operating mode. Specifically, the operating mode characteristics of individual monitoring points in the state change are obtained. Then, based on the operating mode characteristics and the total number of monitoring points in the state change, a key monitoring area is identified, including specific locations requiring enhanced monitoring.

[0066] Collect historical status records for multiple monitoring nodes within the designated coverage area based on the key monitoring area. The designated coverage area is divided into three layers of increasing radius, with the key monitoring area as the center. For each monitoring node within each layer, historical voltage and current data for the same time period every day over the past three months is collected to form a historical status record for each node.

[0067] A monitoring adjustment instruction is generated based on historical status records and distributed to multiple monitoring nodes. The adjustment instruction is used to instruct to increase the monitoring density of a specific time period based on the historical status records. With the key monitoring area as the center, the set coverage range is segmented in order from near to far, obtaining a number of sub-areas within the set sub-range. Monitoring nodes within each sub-area are selected. Based on the historical status records of the monitoring nodes in each sub-area, a basic monitoring frequency within each sub-area is determined, and the monitoring nodes in each sub-area that meet the basic monitoring frequency are determined. Multiple monitoring nodes are obtained, wherein the sum of the basic monitoring frequencies of the multiple monitoring nodes is greater than or equal to the target monitoring frequency. Based on the multiple monitoring nodes and the basic monitoring frequency of each monitoring node, a monitoring adjustment instruction is generated and issued.

[0068] Based on the state assessment model, the actual state data of the latest time period is supplemented, and the updated monitoring state data of the next time period is obtained. The initial monitoring state data and the updated monitoring state data are compared for the next time period to generate a comparison result. Based on the comparison result, it is determined whether to trigger the early warning mechanism. The state assessment model is updated based on the state data of the latest time period to generate a second matching curve that meets the second matching condition. The updated monitoring state data of the next time period in the second matching curve is identified. The predicted voltage and predicted current values ​​of the next time period in the initial monitoring state data are extracted, as are the actual voltage and actual current values ​​of the next time period in the updated monitoring state data. The deviation rate between the predicted voltage value and the actual value and the deviation rate between the predicted current value and the actual value are calculated respectively. The average of the two deviation rates is used as the fluctuation comparison result. Based on this comparison result, it is determined whether to trigger the early warning mechanism. When the difference between the monitoring data of the next time period in the updated monitoring status data and the initial monitoring status data is within the allowable difference, a first instruction is generated to maintain the current monitoring density to maintain the monitoring density of the specific time period; when the updated monitoring status data is less than the monitoring data of the next time period in the initial monitoring status data, and the difference between the two is greater than the allowable difference, based on the distribution of basic monitoring frequencies in multiple monitoring nodes, an instruction for enhanced monitoring and early warning coexistence of some nodes is generated to enable at least one of the multiple monitoring nodes to coordinate monitoring of other nodes except itself and maintain the monitoring density of at least one node; when the updated monitoring status data is greater than the monitoring data of the next time period in the initial monitoring status data, and the difference between the two is greater than the allowable difference, a second instruction for comprehensive enhanced monitoring is generated to enhance the monitoring density of the specific time period.

[0069] Example 1: This example mainly focuses on obtaining real-time charge and discharge data of capacitors and screening voltage and current parameters that change over time. The specific implementation method is as follows:

[0070] Obtain real-time charge and discharge data for capacitors within a set time range. The set time range here needs to be determined based on actual monitoring needs. For example, it can be set to one day, one week, one month, or other reasonable durations. The determination of this time range requires comprehensive consideration of factors such as the capacitor's application scenario, operating rules, and monitoring accuracy requirements. For example, for capacitors used in high-frequency charging and discharging scenarios, a shorter time range may be required to capture their state changes more promptly; while for capacitors operating at low frequencies, a longer time range may be more conducive to analyzing their long-term operating trends.

[0071] When acquiring real-time charge and discharge data, it is necessary to sample the operating status of the capacitor in real time through corresponding sensors or monitoring devices. These sensors should have high accuracy and reliability to ensure that the collected data can truly reflect the charge and discharge situation of the capacitor. During the data collection process, it is necessary to proceed according to a certain sampling frequency, and the setting of the sampling frequency also needs to be determined according to the characteristics of the capacitor and the monitoring requirements, such as sampling once per second, sampling once per minute, etc. The collected data include but are not limited to parameters such as voltage, current, power, temperature, etc., and the focus of Example 1 is on voltage and current parameters.

[0072] After obtaining the real-time charge and discharge data within the set time range, it is necessary to select all time periods in the same cycle as the current monitoring period based on the time range. The "same cycle" here needs to be defined according to the operating cycle characteristics of the capacitor. For example, if the capacitor is regularly charged and discharged according to a cycle of 24 hours a day, and the current monitoring period is set to 8:00-10:00 every day, then the time period of the same cycle is 8:00-10:00 every day within the set time range. For another example, if the operating cycle of the capacitor is weekly, and the current monitoring period is 14:00-16:00 every Monday, then the time period of the same cycle is the time period of every Monday within the set time range.

[0073] When determining time periods within the same cycle, accurate parsing and matching of time data is required. This requires that data records contain accurate timestamp information so that filtering by time period can be performed. For example, by analyzing the data's timestamp, information such as the date, hour, and minute can be extracted. Then, based on the periodic characteristics of the current monitoring period, all time period data that matches that period can be filtered.

[0074] After selecting the time periods within the same cycle, the voltage and current parameters for all of these time periods must be filtered from the real-time charge and discharge data. These parameters include instantaneous and cumulative values. The instantaneous value refers to the voltage and current values ​​of the capacitor at a specific moment. It can intuitively reflect the capacitor's charge and discharge status at that moment, such as whether it is in the charging process and the charging or discharging rate. The cumulative value is the accumulated amount of voltage or current over a certain period of time. It can be used to analyze the total amount of charge and discharge of the capacitor during that period, such as the amount of charge or discharge in a certain period of time.

[0075] When screening voltage and current parameters, data needs to be categorized and organized. For each period that fits within the same cycle, the instantaneous voltage and current values ​​at all sampling points within that period must be extracted, and the cumulative voltage and current values ​​for that period must be calculated. The cumulative voltage value can be calculated by integrating or adding up all the instantaneous voltage values ​​within that period, and the same applies to the cumulative current value.

[0076] For example, if the sampling frequency is once per minute for a 2-hour monitoring period, there are 120 sampling points in that period, each of which corresponds to an instantaneous voltage and current value. By adding or integrating these 120 instantaneous voltage values, the cumulative voltage value for that period can be obtained. The same calculation can be performed on the instantaneous current values ​​to obtain the cumulative current value.

[0077] The selected voltage and current parameters must be stored and managed for subsequent analysis and processing. Data integrity and accuracy must be ensured during storage to avoid data loss or errors. Furthermore, to facilitate subsequent retrieval and use, the data should be stored in a standardized format. For example, data can be stored in chronological order, with appropriate identifiers and descriptions for each parameter.

[0078] Furthermore, during the data screening process, abnormal data may need to be processed. For example, sensor failure or external interference may cause abnormal values ​​in the collected data. In this case, specific algorithms or methods are needed to identify and eliminate abnormal data to ensure the accuracy of the filtered voltage and current parameters. Common methods for processing abnormal data include threshold judgment and sliding average.

[0079] Through the above steps, we can obtain real-time capacitor charge and discharge data and filter voltage and current parameters. This filtered and processed data provides a reliable foundation for subsequent operations such as establishing a condition assessment model and deducing monitoring status data. Accurate voltage and current parameters can help more precisely describe the capacitor's charge and discharge status, laying the foundation for the effectiveness of the entire monitoring method.

[0080] Example 2: This example mainly focuses on the process of establishing a state assessment model based on the screened voltage and current parameters and deducing the initial monitoring state data. The specific implementation method is as follows:

[0081] The voltage and current parameters filtered out from all time periods need to be arranged in time series to construct a training data set. The arrangement of the time series must strictly follow the chronological order of data collection to ensure that the voltage and current parameters corresponding to each time point can accurately reflect the operating status of the capacitor at that moment. For example, if the filtered time period is 8:00-10:00 every day and data is collected continuously for 30 days, the voltage and current parameters in each of the same time periods in these 30 days need to be arranged in order of date to form a continuous time series data set. During the arrangement process, it is necessary to ensure that the timestamp information of the data is accurate to avoid deviations in subsequent model training due to disordered time sequence.

[0082] When constructing a training dataset, data preprocessing is required. This preprocessing step includes operations such as data cleaning, missing value handling, and normalization. Data cleaning primarily involves identifying and removing outliers in the data, such as sudden changes in voltage and current values ​​caused by sensor failure or external electromagnetic interference. Missing values ​​can be filled using methods such as the mean of adjacent time points, linear interpolation, or predicted values ​​based on historical data to ensure the integrity of the dataset. Normalization converts voltage and current parameters of different dimensions into a unified numerical range to prevent differences in parameter dimensions from affecting model training accuracy. For example, voltage and current values ​​are normalized to the range [0, 1].

[0083] After the training dataset is constructed, the model is trained using historical data cross-validation. The core idea of ​​historical data cross-validation is to divide the dataset into several subsets and evaluate model performance and optimize parameters through multiple training and validation cycles. Specifically, the training dataset can be divided into k subsets, with k-1 subsets selected as the training set and the remaining subset as the validation set. The training process is repeated k times, and the average of the k validation results is taken as the evaluation metric for model performance. This method can effectively avoid model evaluation bias caused by different dataset partitioning methods and improve the model's generalization ability.

[0084] During model training, it's important to select the appropriate model type. Considering the nonlinear and time-varying nature of the charge and discharge states of smart capacitors, machine learning regression models, such as support vector machine regression (SVR), random forest regression, or neural network models, can be employed. Taking a neural network model as an example, a multilayer perceptron can be constructed, consisting of an input layer, hidden layers, and an output layer. The number of nodes in the input layer corresponds to the dimensions of the voltage and current parameters, while the number of nodes in the output layer corresponds to the number of state parameters to be predicted. The number of hidden layers and nodes can be adjusted based on data complexity.

[0085] The key to model training is adjusting parameters until the fitting error of the model output is less than the first matching threshold. Fitting error is typically measured using metrics such as mean squared error (MSE) and mean absolute error (MAE). During training, the model's weights and bias parameters are continuously updated through the backpropagation algorithm, gradually reducing the model's fitting error to the training data. For example, if the first matching threshold is set to 0.05, when the model's mean squared error on the validation set is less than 0.05, the model is considered to have achieved the desired fit and training can be stopped. If the threshold is not reached, parameter adjustments are continued until the required fit is met.

[0086] Once model training is complete and the fitting error is less than the first matching threshold, a first matching curve that meets the first matching criteria is generated. This first matching curve is the result of fitting the historical voltage and current parameters based on the trained model, and can intuitively reflect the trend of the capacitor's charge and discharge status over time. This curve must accurately capture key features of the capacitor's charge and discharge process, such as charging peaks and discharge plateaus, to ensure the reliability of the subsequently derived monitoring status data.

[0087] The initial monitoring status data is derived by extrapolating the monitoring status data for several time periods backward based on the first matching curve. The number of time periods to be extrapolated depends on the actual monitoring needs and the model's predictive capabilities. For example, if the capacitor status needs to be monitored for the next 24 hours, and each time period is one hour, 24 time periods need to be extrapolated backward. During this extrapolation process, the voltage, current, and other status data for each time period are predicted in chronological order based on the first matching curve.

[0088] During the prediction process, it is necessary to consider the uncertainties inherent in capacitor operation, such as load variations and ambient temperature fluctuations, that can impact the prediction results. To improve the accuracy of the predicted data, a dynamic adjustment mechanism can be incorporated into the model to revise the predicted results in real time based on the latest monitoring data. For example, after completing each monitoring period, the actual data for that period is fed into the model, the first matching curve is updated, and the state data for subsequent periods is re-derived to minimize the accumulation of prediction errors.

[0089] After initial monitoring status data is generated, it needs to be stored and managed. This storage must include key information such as the time stamp for each period, as well as predicted voltage and current values, to facilitate subsequent state change calculations and comparative analysis. Furthermore, to facilitate query and access, the initial monitoring status data can be stored in a database, with a corresponding indexing mechanism established.

[0090] Furthermore, during the process of establishing the condition assessment model and deducing the initial monitoring status data, the model needs to be regularly updated and optimized. As the capacitor's operating time increases, its internal characteristics may change, such as capacitance decay and increased equivalent series resistance, resulting in a decrease in the model's prediction accuracy. Therefore, it is necessary to regularly collect new monitoring data, retrain the model, and generate a new first matching curve to ensure that the model can continue to accurately reflect the actual operating status of the capacitor.

[0091] Through the above steps, the state assessment model was established and the initial monitoring state data was deduced. This process fully utilized the time series characteristics of historical voltage and current parameters. Through scientific model training and verification methods, the accuracy and reliability of the model were ensured, providing an important predictive basis for subsequent operations such as determining key monitoring areas, generating monitoring adjustment instructions, and triggering early warning mechanisms.

[0092] Example 3: This example mainly focuses on calculating the state change amount of adjacent time periods in the initial monitoring state data and determining the key monitoring area when the change amount reaches the set limit. The specific implementation method is as follows:

[0093] Calculate the state change between adjacent time periods in the initial monitoring state data. The initial monitoring state data is the predicted state data for each time period, derived by backward extrapolating the state assessment model. It contains parameters such as voltage and current for each time period. When calculating the change, select the voltage and current parameters for consecutive time periods in the initial monitoring state data. For example, select data from consecutive, adjacent time periods, such as the nth and n+1th time periods, or the n+1th and n+2th time periods. For each group of adjacent time periods, calculate the voltage and current differences between the current time period and the previous time period.

[0094] The voltage difference is calculated by subtracting the predicted voltage value for the previous period from the predicted voltage value for the current period. The current difference is calculated by subtracting the predicted current value for the previous period from the predicted current value for the current period. It is important to note that the difference calculation must strictly correspond to the parameters of the same monitoring point to ensure data consistency and comparability. For example, if the initial monitoring status data contains voltage and current parameters for multiple monitoring points per period, the voltage and current differences between adjacent periods must be calculated separately for each monitoring point.

[0095] After obtaining the voltage and current differences, the sum of their absolute values ​​is used as the state change. This absolute value sum is used to avoid the signs of the voltage and current differences canceling each other out, thereby more comprehensively reflecting the degree of fluctuation in the charge and discharge state between adjacent time periods. For example, if the voltage difference between adjacent time periods is +2V and the current difference is -1A, the sum of their absolute values ​​is 3. A larger value indicates a more significant state fluctuation between adjacent time periods; conversely, a smaller value indicates a more gradual fluctuation.

[0096] After calculating the state change for each adjacent time period, a reasonable threshold value needs to be set to determine whether specific areas require focused monitoring. This threshold value can be determined based on the operating characteristics of the smart capacitor, historical data patterns, and actual monitoring needs. For example, by analyzing a large amount of historical monitoring data, it can be determined that when the state change exceeds a certain threshold, the capacitor may be in an abnormal operating state. This threshold value can be used as the set threshold value.

[0097] When the state change between adjacent time periods reaches a set limit, further analysis of the monitoring area corresponding to the change is required to determine the key monitoring area. At this point, the operating mode characteristics of a single monitoring point in the state change are first obtained. Operating mode characteristics refer to a series of characteristic parameters that can reflect the operating rules of the monitoring point during the charging and discharging process, including but not limited to the voltage change trend (such as continuous rise, fall, or fluctuation), the current change amplitude (such as the size of the change), and the charge and discharge state transition frequency (such as the number of switches from charging to discharging).

[0098] For example, the voltage difference at a certain monitoring point between adjacent time periods is +5V, and the current difference is +3A. The sum of the absolute values ​​is 8, exceeding the set limit of 6. At this point, analyzing the operating mode characteristics of this monitoring point reveals a rapid increase in voltage and current, which may indicate that the capacitors in the area where the monitoring point is located are in an abnormal charging state and require special attention.

[0099] After obtaining the operational pattern characteristics of individual monitoring points, key monitoring areas are identified based on the total number of monitoring points in the state change data. The total number of monitoring points refers to the total number of monitoring points involved in calculating the state change data in the initial monitoring state data. By analyzing the relationship between operational pattern characteristics and the total number of monitoring points, the scope and severity of the state change can be determined, thereby identifying specific locations requiring enhanced monitoring.

[0100] Specifically, if the operating mode characteristics of a monitoring point show abnormal state fluctuations and occupy a certain proportion of the total number of monitoring points (for example, more than 10% of the monitoring points show similar abnormal characteristics), then the area where the monitoring point is located can be determined as a key monitoring area. The determination of key monitoring areas requires comprehensive consideration of factors such as the spatial distribution of the monitoring points, the similarity of the operating mode characteristics, and the magnitude of the state change. For example, if the state changes of multiple adjacent monitoring points exceed the set limit and their operating mode characteristics are similar (for example, they all show a rapid drop in voltage and an abnormal increase in current), then the continuous area where these monitoring points are located can be designated as a key monitoring area.

[0101] The scope of the key monitoring area can be adjusted based on actual conditions. For example, a specific area requiring enhanced monitoring can be expanded outward from the abnormal monitoring point. This area must have clear geographic boundaries or equipment location identifiers so that subsequent monitoring nodes can accurately adjust the monitoring density in this area.

[0102] When determining key monitoring areas, the physical structure and electrical connection methods of the capacitors must also be considered. For example, if capacitors are installed in groups and connected in parallel, and the capacitor bank at a particular monitoring point experiences an abnormal change in state, the physical location of that capacitor bank can be designated as the key monitoring area, allowing for more targeted monitoring.

[0103] Furthermore, to improve the accuracy of identifying key monitoring areas, a historical data comparison and analysis mechanism can be introduced. The currently calculated state changes and corresponding operating mode characteristics are compared with abnormal cases in historical data. If similar characteristic patterns are found, the rationality of the key monitoring areas can be further confirmed, and the monitoring strategy can be adjusted based on historical experience.

[0104] Through the above steps, the calculation of the state change amount in adjacent time periods in the initial monitoring state data and the determination of the key monitoring area when the change amount reaches the set limit are realized. This process can timely detect abnormal fluctuations in the charge and discharge state of the capacitor and accurately locate the areas that need to be focused on, providing a clear goal for the subsequent collection of historical state records and generation of monitoring adjustment instructions, thereby realizing the refined monitoring and management of the operating status of the intelligent capacitor. In actual application, it is necessary to reasonably set the limit value of the state change amount according to the specific installation layout of the capacitor, the distribution density of the monitoring nodes and the operational safety requirements, and optimize the analysis method of the operating mode characteristics to ensure that the operation of Example 3 can accurately and efficiently determine the key monitoring area and ensure the safe and stable operation of the capacitor.

[0105] Example 4: This example mainly involves collecting historical status records of multiple monitoring nodes within a set coverage area according to a key monitoring area, and generating monitoring adjustment instructions based on these records and distributing them to multiple monitoring nodes. The following describes the implementation method in detail with reference to specific examples:

[0106] Assume that the key monitoring area is determined to be Area A, where a certain smart capacitor bank is located, located within a substation in the power system. When collecting historical status records, the set coverage area is divided into three layers of increasing radius, with Area A as the center. For example, the first layer is within a radius of 5 meters, including Area A itself and nearby directly associated monitoring nodes; the second layer is within a radius of 5 to 15 meters, covering adjacent equipment areas that are electrically connected to Area A; and the third layer is within a radius of 15 to 30 meters, including other equipment areas within the substation that belong to the same power supply network as Area A.

[0107] For each monitoring node within each layer, historical voltage and current data for the same time period each day over the past three months must be collected. For example, if the current monitoring period is 2:00 PM to 4:00 PM daily, voltage and current data for each monitoring node from 2:00 PM to 4:00 PM daily for the three months preceding the current date must be collected. For example, if the current date is June 26, 2025, data from 2:00 PM to 4:00 PM daily from March 26, 2025, to June 25, 2025, must be collected. Historical data for each monitoring node must include a specific timestamp, instantaneous voltage and current values, and the corresponding cumulative values ​​to ensure data integrity and temporal continuity.

[0108] Taking the monitoring nodes N1, N2, and N3 within the first-level area as an example, node N1 is located in the center of area A and directly monitors the input voltage and output current of the target capacitor bank. Node N2 is located at the edge of area A and monitors the bus voltage of the capacitor bank. Node N3 is located in an adjacent distribution cabinet and monitors the current of the line connected in series with the capacitor bank. Voltage and current data for these three nodes is collected daily from 2:00 PM to 4:00 PM over the past three months to form a historical status record for each node. For example, the instantaneous voltage and current values ​​of node N1 at 2:00 PM on April 10th were 405 V and 20 A, respectively; and the instantaneous voltage and current values ​​at 2:30 PM on May 20th were 398 V and 22 A, respectively. This data must be recorded completely and in chronological order.

[0109] After collecting historical status records, the generation of monitoring adjustment instructions begins. First, with key monitoring area A as the center, the set coverage range is segmented from near to far, resulting in several sub-areas within the set sub-range. For example, the coverage range is divided into sub-areas S1 (0-10 meters), S2 (10-20 meters), and S3 (20-30 meters), where S1 corresponds to the first-level area, S2 corresponds to the second-level area, and S3 corresponds to the third-level area.

[0110] For each sub-area, select the monitoring nodes. For example, sub-area S1 includes monitoring nodes N1, N2, and N3; sub-area S2 includes monitoring nodes N4, N5, and N6; and sub-area S3 includes monitoring nodes N7, N8, and N9. Then, based on the historical status records of the monitoring nodes within each sub-area, determine the base monitoring frequency within each sub-area to ensure that the deviation is within the allowable range. The allowable range can be set based on the capacitor's rated parameters and industry standards. For example, the allowable voltage deviation range is ±5% of the rated value, and the allowable current deviation range is ±10% of the rated value.

[0111] Taking subregion S1 as an example, we analyze the historical data of nodes N1, N2, and N3, and calculate the frequency of data deviations within the acceptable range for each node during the daily 2:00 PM to 4:00 PM period over the past three months. Assuming that node N1 has data deviations within the acceptable range for 80 of its 90 days of records, its basic monitoring frequency can be set to twice per hour. Node N2 has 75 days of acceptable data, and its basic monitoring frequency can be set to once per hour. Node N3 has 85 days of acceptable data, and its basic monitoring frequency can be set to three times per hour.

[0112] Within each sub-area, we identify monitoring nodes that meet the basic monitoring frequency. Multiple monitoring nodes are obtained, and the sum of their basic monitoring frequencies must be greater than or equal to the target monitoring frequency. The target monitoring frequency is set based on the importance of the key monitoring area and the monitoring requirements. For example, the target monitoring frequency for sub-area S1 is 5 times per hour. Nodes N1, N2, and N3 have basic monitoring frequencies of 2, 1, and 3 times per hour, respectively. The sum of these frequencies is 6 times per hour, meeting the target frequency of 5 times per hour. Therefore, all three nodes are selected as meeting the criteria.

[0113] Based on multiple monitoring nodes and each node's basic monitoring frequency, monitoring adjustment instructions are generated and issued. For example, a monitoring adjustment instruction for sub-area S1 could be expressed as: "Monitoring node N1 collects voltage and current parameters twice per hour, node N2 collects once per hour, and node N3 collects three times per hour. The collection time must be evenly distributed throughout each hour to enhance real-time monitoring of key monitoring area A."

[0114] When dividing areas, the physical location and communication links of the monitoring nodes must also be considered. For example, nodes N4, N5, and N6 in sub-area S2 may be located in different power distribution cabinets. It is necessary to ensure that instructions can be accurately delivered to each node via the communication network. Furthermore, the adjustment instructions must clearly specify the transmission path and storage method for collected data. For example, node N4 may be required to transmit collected data in real time to the monitoring center's server, while node N5 may be required to upload data every 10 minutes after local storage.

[0115] Furthermore, if the sum of the basic monitoring frequencies of monitoring nodes within a sub-area fails to meet the target frequency, the node selection for that sub-area needs to be re-evaluated or the target frequency needs to be adjusted. For example, if the target monitoring frequency for sub-area S3 is 4 times per hour, and the sum of the basic monitoring frequencies of nodes N7, N8, and N9 is 3 times per hour, consider increasing the number of monitoring nodes in that sub-area or appropriately lowering the target frequency to 3 times per hour to ensure the execution of the instruction.

[0116] In practice, the basic monitoring frequency for different sub-areas can be dynamically adjusted based on the degree of fluctuation in historical data. For example, if historical data from sub-area S1 indicates significant recent voltage fluctuations, the basic monitoring frequency for node N1 can be increased from twice per hour to three times per hour to more closely capture voltage changes. This dynamic adjustment mechanism requires real-time analysis of historical status records, such as a weekly review of historical data for each node to re-determine the basic monitoring frequency.

[0117] Through the above steps, the process of collecting historical status records and generating monitoring adjustment instructions based on key monitoring areas is realized. This process is based on specific regional divisions and combines historical data analysis to accurately determine the monitoring frequency of each monitoring node, thereby achieving differentiated and refined monitoring of key areas. For example, in the above example, by adjusting the frequency of monitoring nodes in area A and its surrounding sub-areas, it can ensure that when the capacitor's charging and discharging status fluctuates abnormally, key data can be obtained in a timely and comprehensive manner, providing strong support for subsequent status assessment and early warning.

[0118] Example 5: This example mainly focuses on the process of supplementing the latest period actual status data based on the status assessment model, obtaining updated monitoring status data, performing comparative analysis, and determining whether to trigger the early warning mechanism. The specific implementation method is as follows:

[0119] The actual state data for the latest period is supplemented based on the state assessment model. This data represents the actual operating parameters of the capacitor during the current period, obtained through real-time monitoring. These parameters include instantaneous and cumulative values ​​of voltage and current. For example, assuming the current period is time t (e.g., 3:00 PM to 3:30 PM on June 26, 2025), real-time voltage and current data for that period are collected through monitoring nodes located at key locations across the capacitor. This data must include the timestamp and corresponding parameter value for each sampling point to ensure data accuracy and timeliness.

[0120] After obtaining the latest actual state data, the state assessment model needs to be updated. The state assessment model was previously trained using historical data and is used to fit and predict the charge and discharge state of the capacitor. When the latest actual data is added, the model will incorporate this new data into the training set and re-optimize the parameters and adjust the model. For example, if the original state assessment model was trained based on the previous 30 days of data, after obtaining the actual data for the 31st day, the 31st day of data will be added to the training set and the model will be re-trained. The model parameters such as weights and bias will be adjusted to ensure that the model can better reflect the current operating characteristics of the capacitor.

[0121] After the model update is complete, a second matching curve is generated that meets the second matching criteria. Similar to the first matching curve generated earlier, the second matching curve fits the changing patterns of the capacitor's charge and discharge states. However, the second matching curve incorporates the latest actual data, thus more accurately reflecting the capacitor's current state trends. For example, if the latest actual data indicates an abnormal downward trend in the capacitor's voltage, the second matching curve will adjust accordingly to reflect this change.

[0122] Identify the updated monitoring status data for the next time period in the second matching curve. The next time period is the monitoring cycle following the most recent time period. For example, if the current time period is 3:00 PM to 3:30 PM, the next time period is 3:30 PM to 4:00 PM. The voltage, current, and other parameters for this time period are predicted using the second matching curve to obtain updated monitoring status data. This data includes the predicted instantaneous voltage and current values, as well as the cumulative values, for this time period.

[0123] Extract the predicted voltage and current values ​​for the next time period from the initial monitoring status data. The initial monitoring status data is derived by backward extrapolating the first matching curve and contains predicted data for multiple future time periods. The data for the next time period corresponding to the updated monitoring status data must be selected for comparison. For example, the predicted voltage and current values ​​for the 3:30 PM to 4:00 PM period in the initial monitoring status data are 400 V and 25 A, respectively. However, the actual voltage and current values ​​for the same time period, derived from the second matching curve in the updated monitoring status data, are (predicted) 395 V and 28 A, respectively.

[0124] Calculate the deviation rate between the voltage prediction value and the actual value, and the deviation rate between the current prediction value and the actual value. The calculation method of the deviation rate is to divide the absolute value of the difference between the prediction value and the actual value by the prediction value, and then multiply it by 100%. For example, if the voltage prediction value is 400V and the actual value is 395V, the deviation rate is ; The predicted current value is 25A, the actual value is 28A, and the deviation rate is The average of the two deviation rates is taken as the fluctuation comparison result, that is, .

[0125] The triggering of the early warning mechanism is determined based on the fluctuation comparison results, which is divided into the following three situations:

[0126] The first case: when the difference between the monitoring data of the next period in the updated monitoring status data and the initial monitoring status data is within the allowable difference, the first instruction to maintain the current monitoring density is generated. The allowable difference is a threshold value set according to the operating standard and actual needs of the capacitor. For example, the allowable voltage difference is ±2%, and the allowable current difference is ±5%. If the voltage deviation rate of 1.25% in the above example is within the range of ±2%, and the current deviation rate of 12% exceeds the range of ±5%, but because the average value of the fluctuation comparison result is 6.625%, it is necessary to judge in combination with the specific allowable difference standard. Assuming that the average value threshold of the allowable difference is 10%, 6.625% is less than 10% and is within the allowable range. At this time, the first instruction is generated to instruct each monitoring node to maintain the current monitoring frequency, for example, collect data every 30 minutes.

[0127] The second scenario: When the updated monitoring status data is less than the monitoring data for the next period in the initial monitoring status data, and the difference between the two is greater than the allowable difference, instructions for both enhanced monitoring and early warning are generated for some nodes. For example, the initial monitoring status data predicts a voltage of 400V for the next period, while the updated monitoring status data shows the actual voltage to be 380V, with a difference of 20V. If the allowable difference is 10V, then 20V is greater than 10V. In this case, it is necessary to determine which nodes require enhanced monitoring based on the distribution of basic monitoring frequencies among multiple monitoring nodes. For example, if the basic monitoring frequency of node A is twice per hour, node B is once per hour, and node C is three times per hour, nodes A and C can be instructed to increase their monitoring frequency to four times per hour. This will also trigger the early warning mechanism, sending a warning message about an abnormal voltage drop to the monitoring center so that staff can promptly troubleshoot the problem.

[0128] The third scenario: When the updated monitoring status data is greater than the monitoring data for the next period in the initial monitoring status data, and the difference between the two is greater than the allowable difference, a second instruction for comprehensive enhanced monitoring is generated. For example, if the initial predicted current is 25A and the actual current after the update is 35A, the difference is 10A, and the allowable difference is 5A. 10A is greater than 5A. In this case, it is considered that the capacitor may be in an overloaded state, and the monitoring density of all monitoring nodes needs to be comprehensively enhanced. For example, the monitoring frequency of all nodes can be increased from once every 30 minutes to once every 15 minutes to collect data more frequently, monitor the state changes of the capacitor in real time, and prevent equipment damage due to overload.

[0129] When triggering the early warning mechanism, the content and recipients of the warning message must be clearly defined. This information should include key information such as the time and location of the anomaly, as well as parameter deviations. For example, "At 3:30 PM on June 26, 2025, the current in the capacitor bank in area A increased abnormally. The current predicted value is 35A, exceeding the initial predicted value by 25A, with a deviation rate of 40%. Please check immediately." This information should be sent to mobile devices of maintenance personnel and servers at the monitoring center, ensuring timely delivery of the warning message.

[0130] During the comparative analysis process, the time synchronization of data and the consistency of monitoring nodes must also be considered. For example, initial and updated monitoring status data must correspond to the same set of monitoring nodes and have the same time range to avoid biased comparison results due to data asynchrony. Furthermore, comprehensive analysis is required for comparative data from multiple monitoring nodes to avoid misjudgments due to data anomalies at a single node.

[0131] Through the above steps, the entire process of supplementing the latest period data, updating the model, comparing status, and triggering warnings is completed. This process can promptly detect abnormal fluctuations in the capacitor's charge and discharge status. Through different levels of warning instructions and monitoring adjustments, it ensures real-time monitoring and safety protection of the capacitor's operating status.

[0132] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0133] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for real-time monitoring of the charge and discharge status of an intelligent capacitor, characterized in that: The method comprises: Acquire real-time charge and discharge data of capacitors and screen voltage and current parameters that change over time, including instantaneous and cumulative values; A status assessment model is established based on the selected parameters, and the monitoring status data of several time periods are deduced backward to obtain the initial monitoring status data; Calculating the state change amount of adjacent time periods in the initial monitoring state data, and when the change amount reaches a set limit, determining the key monitoring area based on the corresponding relationship between the change amount and the capacitor operation mode; Collect historical status records of multiple monitoring nodes within the coverage area according to key monitoring areas; Generating a monitoring adjustment instruction based on the historical status record, and distributing the adjustment instruction to multiple monitoring nodes, wherein the adjustment instruction is used to instruct to increase the monitoring density of a specific time period based on the historical status record; Based on the status assessment model, the actual status data of the latest period is supplemented, and the updated monitoring status data of the next period is obtained. The next period is compared based on the initial monitoring status data and the updated monitoring status data to generate a comparison result. Based on the comparison result, it is determined whether to trigger the early warning mechanism.

2. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 1, characterized in that: The method of obtaining real-time charge and discharge data of the capacitor and screening voltage and current parameters that change over time includes: Get the real-time charge and discharge data of the capacitor within the set time range; Based on the set time range, select all time periods in the same cycle as the current monitoring period; Filter the voltage and current parameters of all time periods in real-time charge and discharge data.

3. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 2, characterized in that: The state assessment model is established based on the screened parameters, and the monitoring state data of several time periods are deduced backward to obtain the initial monitoring state data, including: Establishing a state assessment model based on voltage and current parameters of all time periods to generate a first matching curve that meets a first matching condition; The monitoring status data of several time periods are deduced backward according to the first matching curve to obtain the initial monitoring status data.

4. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 3, characterized in that: The calculating of the state change amount in adjacent time periods in the initial monitoring state data, and when the change amount reaches a set limit, determining the key monitoring area according to the corresponding relationship between the change amount and the capacitor operation mode includes: Obtain the operating mode characteristics of a single monitoring point in the state change; According to the operation mode characteristics and the total number of monitoring points in the state change amount, a key monitoring area is determined, and the key monitoring area includes specific locations that require enhanced monitoring.

5. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 1, characterized in that: Generating a monitoring adjustment instruction based on the historical status record and distributing the adjustment instruction to multiple monitoring nodes includes: Taking the key monitoring area as the center, the set coverage range is divided into sections in the order from near to far, and several sub-areas within the set sub-range are obtained; Select monitoring nodes in each sub-area and determine the basic monitoring frequency within each sub-area to ensure that the deviation is within the allowable range based on the historical status records of the monitoring nodes in each sub-area; Determine the monitoring nodes that meet the basic monitoring frequency in each sub-area, and obtain multiple monitoring nodes, wherein the sum of the basic monitoring frequencies of the multiple monitoring nodes is greater than or equal to the target monitoring frequency; Generate and issue monitoring adjustment instructions based on multiple monitoring nodes and the basic monitoring frequency of each monitoring node; The historical status records of multiple monitoring nodes within the set coverage area collected according to the key monitoring area include: With the key monitoring area as the center, the set coverage area is divided into three layers of areas with increasing radius; For the monitoring nodes in each layer area, the voltage and current historical data of the same time period every day in the past three months are collected to form a historical status record of each node.

6. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 3, characterized in that: The state assessment model is used to supplement the actual state data of the latest period, and to obtain the updated monitoring state data of the next period. The comparison of the next period is performed based on the initial monitoring state data and the updated monitoring state data to generate the comparison result, including: The state assessment model is updated according to the state data of the latest period to generate a second matching curve that meets the second matching condition; Identify updated monitoring status data for the next period in the second matching curve, perform a fluctuation comparison based on the initial monitoring status data and the updated monitoring status data, and generate a comparison result.

7. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 6, characterized in that: The generating of the comparison result by performing fluctuation comparison based on the initial monitoring state data and the updated monitoring state data includes: Extracting the predicted voltage value and the predicted current value of the next period in the initial monitoring state data, and updating the actual voltage value and the actual current value of the next period in the monitoring state data; The deviation rate between the voltage prediction value and the actual value and the deviation rate between the current prediction value and the actual value are calculated respectively, and the average value of the two deviation rates is used as the fluctuation comparison result.

8. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 1 or 7, characterized in that: Determining whether to trigger the early warning mechanism according to the comparison result includes: When the difference between the monitoring data of the next period in the updated monitoring state data and the initial monitoring state data is within the allowable difference, a first instruction for maintaining the current monitoring density is generated to fully maintain the monitoring density of the specific period; When the updated monitoring status data is less than the monitoring data of the next period in the initial monitoring status data, and the difference between the two is greater than the allowed difference, based on the distribution of basic monitoring frequencies among the multiple monitoring nodes, an enhanced monitoring and early warning coexistence instruction is generated for some nodes, so that at least one of the multiple monitoring nodes coordinates monitoring of other nodes except itself, and maintains the monitoring density of the at least one node; When the updated monitoring status data is greater than the monitoring data of the next period in the initial monitoring status data, and the difference between the two is greater than the allowed difference, a second instruction for comprehensive enhanced monitoring is generated to fully enhance the monitoring density of the specific period.

9. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 3, characterized in that: The step of establishing a state assessment model based on the voltage and current parameters of all time periods to generate a first matching curve that meets the first matching condition includes: Arrange the voltage and current parameters of all time periods in time series to construct a training data set; The historical data cross-validation method is used to train the model on the training data set, and the model parameters are adjusted until the fitting error of the model output is less than the first matching threshold, thereby generating a first matching curve.

10. The method for real-time monitoring of the charging and discharging status of an intelligent capacitor according to claim 4, characterized in that: The calculation of the state change amount in adjacent time periods in the initial monitoring state data includes: Select the voltage and current parameters of the continuous time period in the initial monitoring state data, and calculate the voltage difference and current difference between the current time period and the previous time period; The sum of the absolute values ​​of the voltage difference and the current difference is taken as the state change amount, which is used to reflect the degree of fluctuation of the charge and discharge states in adjacent time periods.

Images