Intelligent control method and system for phase selection circuit breaker based on big data

Abstract

The invention relates to the technical field of industrial big data, and discloses a phase selection circuit breaker intelligent control method and system based on big data, and the method comprises the steps: collecting the original operation data of each node, extracting the load peak valley, voltage fluctuation and current instability characteristics, completing the state evaluation, and generating an initial operation data set; carrying out standardization, denoising and seasonal leveling treatment to obtain a cleaned load feature set; fusing the frequency deviation and the power factor change, executing trend operation, and forming a load trend prediction data set; in combination with real-time voltage fluctuation and a temperature coefficient, load response characteristics are analyzed, and a preliminary control opportunity adjustment scheme is determined; the current instability index and the harmonic interference intensity are corrected based on the scheme, and an operation parameter combination is formed; finally, according to the parameter combination, a control instruction is generated, time sequence scheduling is completed, and a control execution plan is output. The method can solve the problem of response lag of the circuit breaker under complex working conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial big data, and particularly relates to a phase selector circuit breaker intelligent control method and system based on big data. BACKGROUND

[0002] With the continuous evolution of industrial big data technology, power grid operation gradually enters the intelligent era with data driving as the core. In the distribution system, the circuit breaker as a key control device, its intelligent level is directly related to the stability and safety of system operation. With the fusion analysis capability of industrial big data on massive operation parameters, environmental factors and load behavior, the control logic with self-learning and self-adaptive characteristics is gradually becoming the core path of intelligent upgrading of distribution system.

[0003] The current circuit breaker control method mainly depends on preset parameters or rules, and through the collection of load peak and valley information, voltage and current parameters, etc., it can cope with the change of conventional load to a certain extent. However, the application effect is not ideal in complex environment. In the face of sudden load mutation, harmonic disturbance or operation abnormality caused by extreme weather in mountainous area power grid, the traditional control means often cannot timely capture the change trend of key indicators, causing response lag or dispatching failure.

[0004] In summary, the prior art has the problem of circuit breaker response lag under complex working conditions. SUMMARY

[0005] The present application provides a phase selector circuit breaker intelligent control method and system based on big data to solve the problem of circuit breaker response lag under complex working conditions.

[0006] In a first aspect, to solve the above technical problems, the present application provides a phase selector circuit breaker intelligent control method based on big data, comprising: obtaining an original operation data set of each node; performing feature extraction and state evaluation according to the original operation data set to obtain an initial operation data set; performing standardization processing on the initial operation data set, and performing noise filtering, continuity completion and seasonal leveling operations to obtain a cleaned load feature set; According to the load feature set, the frequency offset amplitude and power factor change in the power grid system operation parameters are fused, and trend operation is performed to obtain a load trend prediction data set; According to the load trend prediction data set, combining the real-time collected voltage fluctuation interval and temperature influence coefficient, load response characteristic analysis is performed to determine the preliminary control timing adjustment scheme; According to the preliminary control timing adjustment scheme, performing parameter dynamic correction to obtain an adjusted operation parameter combination; According to the operation parameter combination, control instruction generation and time series scheduling are performed to determine a final control execution plan.

[0007] Preferably, according to the original operation data set, feature extraction and state evaluation are performed to obtain an initial operation data set, including: The original operation data set includes: a load peak time series, a load valley time series, a voltage fluctuation interval, and a current instability index; Based on the load peak time series and the load valley time series, a time series analysis method is used to extract periodicity features and fluctuation amplitude features of the load; If the periodicity feature is greater than or equal to a preset periodicity threshold, and the fluctuation amplitude feature is greater than or equal to a preset fluctuation threshold, then the voltage fluctuation interval and the current instability index are used to calculate a stability index of the power grid operation state; According to the stability index, a clustering analysis method is used to classify the operation states of each node to obtain the initial operation data set.

[0008] Preferably, the initial operation data set is standardized and noise filtering, continuity completion, and seasonal leveling operations are performed to obtain a cleaned load feature set, including: Obtain a load daily average curve and seasonal load features; According to the initial operation data set, median filtering denoising processing is performed to obtain a denoised load distribution set; Based on the load daily average curve and the seasonal load features, the load distribution set is smoothed to obtain a stable interval; If the fluctuation range of the stable interval exceeds a preset fluctuation threshold, the stable interval is completed to restore the time series continuity of the load distribution data to obtain completed load data; According to the load data, combined with a preset seasonal weight coefficient, data standardization processing is performed to obtain the cleaned load feature set.

[0009] Preferably, according to the load feature set, the frequency offset amplitude and power factor change in the power grid system operation parameters are fused to perform trend operation to obtain a load trend prediction data set, including: Based on the frequency offset amplitude and the power factor change, joint analysis is performed on the load feature set to obtain a fused load comprehensive data set; The load comprehensive data set is subjected to time series extraction, and the load fluctuation frequency is analyzed to obtain a fluctuation frequency interval; judging a change direction of the fluctuation frequency interval relative to a load history, if the change direction exceeds a preset change threshold, a weighted calculation method is used for correction to obtain an adjusted load trend change sequence; In combination with historical data benchmarks, the load trend change sequence is subjected to trend operation to obtain a load trend prediction data set.

[0010] Preferably, in the load response characteristic analysis and control opportunity calibration, a preliminary control opportunity adjustment scheme is determined according to the load trend prediction data set in combination with a real-time collected voltage fluctuation interval and a temperature influence coefficient, including: According to the load trend prediction data set, a multi-dimensional fusion processing is performed in combination with a real-time collected voltage fluctuation interval and a temperature influence coefficient to obtain a load response characteristic data set; According to the load response characteristic data set, a load segmentation extraction is performed in combination with an environmental interference factor under extreme weather conditions to obtain a correlation between the load response characteristic and the grid operation state; If the load response characteristic distribution and the grid operation state history correlation are greater than or equal to a preset correlation threshold, the load response characteristic data set is corrected to obtain an adjusted response characteristic data set; Based on the response characteristic data set, a final control opportunity calibration is performed to determine a preliminary control opportunity adjustment scheme.

[0011] Preferably, in the parameter dynamic correction according to the preliminary control opportunity adjustment scheme, an adjusted operation parameter combination is obtained, including: The preliminary control opportunity adjustment scheme is executed, and a current instability index and a harmonic interference intensity after control execution are collected; The current instability index and the harmonic interference intensity are subjected to multi-dimensional matching analysis with a target operation state parameter benchmark value established in advance to obtain deviation distribution characteristics of current stability and harmonic interference intensity; If any of the deviation distribution characteristics of current stability or the deviation distribution characteristics of harmonic interference intensity is greater than or equal to a corresponding preset threshold, a temperature and humidity factor in a grid operation environment is combined, and a parameter correction tool is used to dynamically adjust the characteristics exceeding the threshold to obtain a corrected temporary parameter set; Based on the temporary parameter set, in combination with the correlation between the load response characteristic and the grid operation state, a weighted calculation method is used for optimization processing and adaptive verification to obtain an operation parameter combination meeting control requirements.

[0012] Preferably, in the control instruction generation and time sequence scheduling according to the operation parameter combination, a final control execution plan is determined, including: Based on the combination of operating parameters, real-time environmental change data and load fluctuation data of multiple nodes are collected, and differential analysis is performed on different nodes to construct a personalized environmental adaptation dataset. Based on the personalized environmental adaptation dataset, mutation identification and instruction draft generation are performed to obtain a preliminary control scheme; Based on the aforementioned preliminary control scheme, and combined with the parameter recovery cycle and execution priority requirements of the nodes, the allocation and sorting of the operation time sequence are performed to obtain the instruction execution time sequence combination of each node; Based on the combination of the instruction execution time sequence, instruction integration and adaptive calibration are performed to determine the final control execution plan.

[0013] Secondly, the present invention provides an intelligent control system for phase-selective circuit breakers based on big data, comprising: The data acquisition module is used to acquire the raw runtime data sets of each node; The initial running data group module is used to perform feature extraction and state evaluation based on the original running data group to obtain the initial running data group; The load feature set module is used to standardize the initial running data set and perform noise filtering, continuous completion and seasonal leveling operations to obtain the cleaned load feature set. The prediction data group module is used to perform trend calculations based on the load characteristic set, integrating the frequency offset amplitude and power factor changes in the power grid system operating parameters, to obtain the load trend prediction data group. The control adjustment module is used to analyze the load response characteristics based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, so as to determine the initial control timing adjustment scheme. The operating parameter module is used to dynamically correct parameters according to the initial control timing adjustment scheme to obtain the adjusted operating parameter combination; The control execution module is used to generate control instructions and schedule time sequences based on the combination of operating parameters, thereby determining the final control execution plan.

[0014] Thirdly, the present invention also provides an electronic device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor executes the computer program to implement the big data-based intelligent control method for phase-selective circuit breakers described in any one of the above.

[0015] Fourthly, the present invention also provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the intelligent control method for phase-selective circuit breakers based on big data as described above.

[0016] Compared with the prior art, the present invention has the following beneficial effects: (1) By acquiring the original operating data of the node and performing feature extraction, noise filtering and time series completion, the present invention can ensure that the load information remains complete and continuous under various environments, thereby improving the reliability of subsequent control data.

[0017] (2) When processing load characteristics, the present invention combines the changes in frequency offset and power factor to make trend judgments, which can capture the changes in the direction of load fluctuation in a timely manner, provide a more reasonable basis for phase control of the circuit breaker, and reduce the false trip rate.

[0018] (3) In the control and adjustment stage, the present invention introduces the current instability index and harmonic interference intensity as evaluation basis, and dynamically corrects the operating parameters according to the deviation size to ensure that the control strategy can still maintain consistency with the operating state after actual execution.

[0019] (4) In the instruction execution phase, the present invention allocates specific operation time according to the recovery time and priority order of different nodes, and integrates and compares all control instructions, thereby ensuring that the response sequence of the circuit breaker is more reasonable and the execution rhythm is more in line with the actual needs of the power grid. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the intelligent control method for phase selection circuit breakers based on big data provided in the first embodiment of the present invention; Figure 2 This is a schematic diagram of the intelligent control system for phase selection circuit breakers based on big data, provided in the second embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Reference Figure 1 The first embodiment of the present invention provides a smart control method for phase-selective circuit breakers based on big data, comprising the following steps: S11, obtain the raw running data group of each node; S12, Based on the original running data set, perform feature extraction and state evaluation to obtain the initial running data set; S13, standardize the initial running data set and perform noise filtering, continuous completion and seasonal leveling operations to obtain the cleaned load feature set; S14. Based on the load characteristic set, the frequency offset amplitude and power factor change in the power grid system operating parameters are integrated to perform trend calculation and obtain the load trend prediction data set. S15. Based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, perform load response characteristic analysis to determine the preliminary control timing adjustment scheme. S16, Based on the preliminary control timing adjustment scheme, perform dynamic parameter correction to obtain the adjusted combination of operating parameters; S17. Based on the combination of operating parameters, control instructions are generated and time sequence scheduling is performed to determine the final control execution plan.

[0023] In step S11, the original running data sets of each node are obtained; It is worth noting that the system acquires raw operational data sets from various key nodes in the power grid. This data mainly includes time series of load peak and valley values, voltage fluctuation ranges, and current instability index data. The acquisition of these data relies on various sensors installed on-site, which continuously record various operational information at a set frequency. For example, load data is statistically analyzed hourly, recording the maximum load value within that time period as the peak value and the minimum load value as the valley value, and arranging these values ​​in chronological order.

[0024] The acquisition of the voltage fluctuation range relies on the real-time monitoring function of the voltage sensor. The system determines the fluctuation range for that period by recording the maximum and minimum voltage values ​​in each sampling period and calculating the difference. For example, if the highest voltage recorded at a certain node within 30 minutes is 236V and the lowest voltage is 222V, then the voltage fluctuation range for that node during that period is 14V.

[0025] The current instability index is a comprehensive indicator reflecting the intensity of node current fluctuations. It is calculated based on the product of the standard deviation of the current values ​​and a fluctuation factor. In practice, a moving standard deviation is calculated for the current sampling sequence per unit time (e.g., 60 data points per minute, sampled once per second), reflecting the overall dispersion of the current. Simultaneously, a corresponding fluctuation factor β is set for correction based on the frequency characteristics of the current fluctuations. For example, if the sampled data is [9.2A, 9.8A, 9.5A, 10.1A], with a standard deviation of 0.34A, and a fluctuation factor of 1.2, the current instability index would be 0.408.

[0026] The value of the fluctuation factor is mainly set according to the sensitivity of the power grid to load fluctuations. If there are a large number of nonlinear loads or frequently started and stopped equipment in the area, the power grid is more sensitive to fluctuations, and the fluctuation factor value is adjusted accordingly. If the power grid operates stably and has a high tolerance, the value is lower. The general range is 1.0 to 1.5, and in mountainous power grids, it is often set to 1.2 or 1.3 to reflect the impact of the environment on current fluctuations.

[0027] In step S12, feature extraction and state evaluation are performed based on the original running data set to obtain an initial running data set, including: The original operating data set includes: load peak time series, load valley time series, voltage fluctuation range, and current instability index; Based on the load peak time series and load trough time series, time series analysis methods are used to extract the periodicity and fluctuation amplitude characteristics of the load. If the periodicity characteristic is greater than or equal to a preset periodic threshold and the fluctuation amplitude characteristic is greater than or equal to a preset fluctuation threshold, then for the voltage fluctuation range and the current instability index, calculate the stability index of the power grid operation state. Based on the stability index, cluster analysis is used to classify the operating status of each node to obtain the initial operating data group.

[0028] It's worth noting that time series refers to load data with sampled time points arranged in an orderly fashion. For example, for a substation node in a mountainous area, the peak load values ​​collected over seven consecutive days are 4800kW, 5200kW, 5000kW, 5300kW, 4900kW, 5100kW, and 5000kW. Time series analysis methods are then used to extract data features, primarily focusing on two types of indicators: periodicity and fluctuation amplitude.

[0029] Based on the load peak and trough sequences of each node for the same time period each day, corresponding time series data are constructed; then, autocorrelation analysis is performed on the sequence. The analysis uses a lag interval of one day (24 hours) to calculate the correlation coefficient between the original sequence and its lagged sequence. If the correlation coefficient after a 1-day lag is greater than 0.9, the node is considered to have stable daily periodicity; if the correlation coefficient after a 7-day lag remains at a high level, it can also be identified as having periodicity.

[0030] The fluctuation amplitude feature measures the degree of variability in load data over a continuous time range. Its extraction can be achieved using the "standard deviation ratio method," which calculates the standard deviation of the load peak or trough sequence and divides it by the mean to obtain the degree of variation. For example, based on the above sequence, with a mean of 5042.86 kW and a standard deviation of 151.66 kW, the fluctuation amplitude feature is calculated as 151.66 ÷ 5042.86 ≈ 3.01%. If the system sets the fluctuation amplitude threshold to 5%, the load change at that node can be considered to be within a stable range. If the feature value exceeds the threshold, the node needs to be marked as an area with abnormal operational fluctuations.

[0031] In addition, to ensure the accuracy of the results, the system also needs to jointly verify the judgment results of periodicity characteristics and fluctuation amplitude characteristics: if both indicators exceed the system threshold (e.g., periodic autocorrelation coefficient > 0.9 and fluctuation amplitude > 5%), it indicates that the node has both load periodic regularity and fluctuation disturbance characteristics. The system will include the node in the key monitoring scope and trigger voltage and current stability analysis; if only one indicator exceeds the threshold, the system will set a delayed confirmation mechanism for the node, continue to collect subsequent data, and conduct a status assessment after the trend is clear; if neither indicator exceeds the threshold, the system will identify the node as a stable load operation segment, and no additional monitoring strategy is required.

[0032] If the periodic characteristic value is greater than the system-set periodic threshold, and the fluctuation amplitude characteristic value is also greater than the fluctuation threshold, it indicates that the node exhibits typical periodic disturbance behavior. The system will further introduce voltage fluctuation range and current instability index for stability assessment. The system will extract the voltage fluctuation range and current instability index for each node from two perspectives: voltage fluctuation and current fluctuation. The voltage fluctuation range represents the difference between the maximum and minimum voltage values ​​of a node within a sampling period, in volts. The current instability index is obtained by collecting continuous current values ​​per unit time, calculating its standard deviation through a sliding window, and then multiplying it by a fluctuation factor. The final result is a value reflecting the degree of fluctuation, ranging from 0 to 1.

[0033] To unify the processing of two parameters with different dimensions, the system normalizes the voltage fluctuation range. For example, if the maximum allowable voltage fluctuation is set to 12V, then if the voltage fluctuation at a certain node is 6V, the normalized result is 0.5. This normalized result is then weighted and averaged with the current instability index to generate a stability index. For example, voltage fluctuation and current instability each account for 50%. If the normalized voltage fluctuation at a certain node is 0.5 and the current instability index is 0.6, then the stability index is calculated as: (0.5 × 0.5) + (0.6 × 0.5) = 0.55. The larger this value, the more pronounced the fluctuation at that node, and the less stable its operation.

[0034] After the stability index is calculated, the system further classifies all nodes. The classification method employs conventional clustering analysis, using the K-means algorithm to perform unsupervised clustering analysis on all nodes. This process uses the node's stability index as the sole clustering dimension, initially setting the number of clusters to two, representing "stable operation" and "fluctuating operation" respectively. In the initial stage, the system randomly selects two nodes' stability index values ​​as initial cluster centers, for example, choosing 0.48 and 0.80 as initial mean centers.

[0035] Subsequently, the system calculates the Euclidean distance between the stability index of each node and the two cluster centers, and assigns the node to the nearest cluster based on the distance. For example, node A has a stability index of 0.42, a distance of 0.48 of 0.06, and a distance of 0.80 of 0.38, therefore it is assigned to the cluster centered at 0.48. After completing the initial grouping of all nodes, the system recalculates the average stability index of all nodes within each group and uses it as the new cluster center for the next iteration.

[0036] This process repeats continuously: in each iteration, the system reallocates nodes based on the new cluster centers and updates the centers until the cluster centers converge, meaning the change in centers between two iterations is within a set precision range (e.g., ±0.001). This ultimately forms a stable cluster structure, completing the automatic partitioning of the operating state.

[0037] Taking 10 nodes of a power grid as an example, their stability indices are [0.42, 0.48, 0.55, 0.61, 0.63, 0.70, 0.75, 0.80, 0.85, 0.90]. In the initial iteration, the system sets the cluster centers to 0.48 and 0.80. After three iterations, the two cluster centers stabilize at 0.48 and 0.78, respectively. Finally, node [0.42, 0.48, 0.55, 0.61] is classified as stable, and node [0.63, 0.70, 0.75, 0.80, 0.85, 0.90] is classified as fluctuating.

[0038] Ultimately, each node will correspond to a stability index value and its category label. These data, along with the node's basic characteristics (load peak, valley, voltage and current sequence, etc.), constitute the initial operating data set.

[0039] In step S13, the initial running data set is standardized, and noise filtering, continuity completion, and seasonal leveling operations are performed to obtain a cleaned load feature set, including: Obtain the daily average load curve and seasonal load characteristics; Based on the initial running data set, median filtering and noise reduction processing is performed to obtain the denoised load distribution set; Based on the daily average load curve and the seasonal load characteristics, the load distribution set is smoothed to obtain a stable range; If the fluctuation range of the stable interval exceeds the preset fluctuation threshold, the stable interval is completed to restore the time series continuity of the load distribution data and obtain the completed load data. Based on the load data, combined with a preset seasonal weighting coefficient, data standardization processing is performed to obtain a cleaned load feature set.

[0040] It's worth noting that the daily average load curve is obtained by averaging the daily load time series data of a node over a past period (such as a month or a quarter) on an hourly basis. For example, assuming a node collects load values ​​hourly from 0:00 to 24:00 daily for 30 consecutive days, the 30 data points for each hour can be averaged to obtain a 24-hour daily average load curve. This curve reflects the basic load variation pattern of the node under normal operation and is suitable for subsequent smoothing and anomaly identification. Seasonal load characteristics, on the other hand, are obtained by statistically analyzing typical time periods in historical data on a quarterly or monthly basis. Specifically, these seasonal load characteristics are used to calculate the average load level and fluctuation range of each node in each season, thereby deriving the load value and load range representing the characteristics of that season. For example, the system divides the year into four seasons: spring (March–May), summer (June–August), autumn (September–November), and winter (December–February). Based on the historical operating data of a certain mountain node, the system summarizes its daily hourly load data in summer and calculates that the average load in summer is 4,800 kilowatts with a fluctuation range of ±600 kilowatts; while in winter, the average load of the same node is 4,200 kilowatts with a fluctuation range of only ±300 kilowatts.

[0041] Next, the load data in the initial running data set is cleaned. The first step in cleaning is noise filtering, which is accomplished using median filtering. During the operation, the load time series is divided into several fixed windows, such as every 5 sampling points as a group. The median value in each group is selected as the representative value and replaces the original data. This effectively removes maxima or minima caused by sampling anomalies or external disturbances, thus obtaining a denoised load distribution set. For example, if the data collected at a certain node during a certain time period is [3100, 3200, 5000, 3150, 3100] kilowatts, then 5000 is clearly an outlier. After median filtering, this value is replaced with 3150 kilowatts, eliminating the abnormal peak value.

[0042] Then, the hourly load data of each node over the past 7 days is processed by a moving average. The commonly used window size is 3 hours. That is, the load value at each moment is replaced by the average of several data points before and after that moment, thereby smoothing out short-term fluctuation noise.

[0043] After applying the moving average, the load value at each moment is compared with the historical daily average for the corresponding period to determine the degree of deviation. For example, if a node shows a daily load of approximately 3400 kW at 1 PM in the summer daily average curve, while the current moving average is 3300 kW, the difference is less than ±5%, indicating that the node is in a stable operating state. Similarly, deviation rate analysis is performed on the entire time series. If the deviation rate within a continuous time period is consistently less than a set threshold (e.g., ±10%), that continuous time period is defined as a stable interval. Conversely, if the load in certain time periods deviates from the typical range of the historical curve or exhibits drastic fluctuations, these segments are excluded from the stable interval.

[0044] The specific fluctuation threshold is set between ±10% and ±15%, and the specific value can be adjusted according to the load fluctuation characteristics of the power grid area. For example, in a small power grid in a mountainous area, considering that the load fluctuation is relatively drastic, the system fluctuation threshold is set to 12%. If the fluctuation range of the load data in the stable range during a certain night reaches 17%, which significantly exceeds the set standard, it is considered an unstable section and must be supplemented.

[0045] The data completion process uses linear interpolation. Taking a node with missing data from 22:00 to 2:00 the next day as an example, if the load at 21:00 is 3100 kW and the load at 3:00 is 3300 kW, then the load values ​​for the intermediate four hours can be estimated linearly in ascending order, at 3150, 3200, 3250, and 3300 kW respectively, thus restoring the continuity of the time series. This approach is both simple and effective, while maintaining the consistency of the data trend.

[0046] After data completion, significant seasonal fluctuations remain, necessitating standardization to uniformly measure load characteristics across different seasons. Based on historical seasonal load mean and variance, Z-score normalization is employed for seasonal standardization. Specifically, the current load value is subtracted from the historical mean for the corresponding season, and then divided by the standard deviation for that season to obtain the dimensionless standardized load value. For example, a mountainous node has a historical winter load mean of 4200 kW and a standard deviation of 300 kW. The current collected value is 4500 kW. Therefore, its seasonally standardized load is: (4500 - 4200) ÷ 300 ≈ 1.0.

[0047] The final cleaned load feature set includes the smoothed load sequence, the completed continuous time series data, the standardized load values, and the stability label or interpolation mark used to identify the data processing status.

[0048] In step S14, based on the load characteristic set, frequency offset amplitude and power factor changes in the power grid system operating parameters are integrated to perform trend calculations, resulting in a load trend prediction data set, including: Based on the frequency offset amplitude and the power factor change, the load feature set is jointly analyzed to obtain a fused load comprehensive dataset. The load aggregate dataset is subjected to time series extraction, and the load fluctuation frequency is analyzed to obtain the fluctuation frequency range; Determine the direction of change of the fluctuation frequency range relative to the load history. If the direction of change exceeds a preset change threshold, a weighted calculation method is used for correction to obtain an adjusted load trend change sequence. By combining historical data as a benchmark, trend calculations are performed on the load trend change sequence to obtain a load trend prediction data set.

[0049] It is worth noting that, based on the previously obtained load feature set after cleaning, two important dynamic parameters in power grid operation are further introduced: frequency offset amplitude and power factor change. The system first adopts a multi-dimensional mapping method to jointly organize these two types of system parameters with the load feature set. Specifically, a unified data framework is constructed using time as a guide, and load information, frequency data, and power factor data recorded within the same time period are mapped one-to-one according to timestamps. For example, in a certain industrial node, with a sampling interval set to 15 minutes, the system will put the daily average load, frequency offset of 0.6 Hz, and power factor of 0.91 collected at 14:00 on January 15, 2024, into the same recording unit. Through this mapping method, a comprehensive dataset integrating multiple key parameters is formed, which lays the foundation for subsequent trend identification and extrapolation.

[0050] Next, the system will perform time series extraction on this fused comprehensive dataset. Operationally, based on the set sliding window width (e.g., 1 hour or 3 hours), continuous load changes will be extracted into a series of fluctuation segments. Then, the system will analyze the fluctuation amplitude and frequency of these segments to extract the fluctuation frequency range over a period of time. For example, in a certain park node, the load fluctuation over the past 4 hours may have been between ±300 kilowatts, with a significant increase or decrease occurring on average every 30 minutes. This change characteristic is recorded as a fluctuation frequency of 0.033 times / minute. Simultaneously, the system will generate corresponding fluctuation frequency change labels based on the relative deviation between the current frequency value and historical frequencies. These labels include three categories: frequency increase, frequency decrease, and frequency stability, used to characterize the load fluctuation trend characteristics within the current prediction period.

[0051] Once the fluctuation frequency range is identified, the system compares it with historical load behavior. This comparison primarily focuses on the "direction," i.e., whether the current fluctuation is more intense or more stable than in the past. For example, if the fluctuation frequency in the same time period was generally around 0.02 times / minute in the past, but now 0.05 times / minute is detected, it indicates that the system is experiencing more frequent load fluctuations, and this change is labeled as "frequency-enhancing fluctuation." The system sets a preset change threshold (e.g., ±0.015 times / minute); if the current change exceeds this range, further adjustments are required.

[0052] During the correction phase, the system considers the impact of different time periods on the trend, calculating a more reasonable trend change sequence through weighted calculations. For example, if a certain time period falls during the summer peak, the system assigns a higher weight to that data segment (e.g., 0.7), while the nighttime low period is assigned only a weight of 0.3. This way, even if some sudden data deviations occur, they will not significantly interfere with the overall trend assessment. After correction, the system forms a smoother and more representative trend change curve.

[0053] Finally, after obtaining the adjusted trend change sequence, the system will perform predictive calculations in conjunction with historical data benchmarks. Operationally, it will select historical load data that matches the operating characteristics of the current node, such as continuous load curves over the past 30 or 60 days, as a reference benchmark. The system will align and match the current trend sequence with historical trends, analyzing their similarity in terms of rate of increase, fluctuation range, and duration.

[0054] To achieve this, the system employs a load forecasting method based on historical trend slope matching. Its core principle is to first extract the load change slope (e.g., 0.18 kW / min) of the current trend segment (e.g., the last 6 hours) on the time axis. Then, it searches for upward trend segments with similar slopes in historical data and calculates the average load increase of these segments over the following 24 hours as a basis for estimation. For example, if multiple similar trend segments are found in the historical records with an average increase of 900 kW, and the current starting load value is 4200 kW, the system predicts that the peak load in the next 24 hours will be approximately 5100 kW.

[0055] Finally, the load trend forecast data set includes the load change sequence during the forecast period, fluctuation frequency change labels, adjusted weight records, and change direction indicators.

[0056] In step S15, based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, load response characteristics analysis and control timing calibration are performed to determine a preliminary control timing adjustment scheme, including: Based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, a multi-dimensional fusion processing is performed to obtain the load response characteristic dataset. Based on the load response characteristic dataset, combined with environmental interference factors under extreme weather conditions, load segmentation is performed to obtain the correlation between load response characteristics and power grid operating status. If the correlation between the load response characteristic distribution and the historical power grid operating status is greater than or equal to a preset correlation threshold, the load response characteristic dataset is corrected to obtain an adjusted response characteristic dataset. Based on the aforementioned response characteristic dataset, final control timing calibration is performed to determine a preliminary control timing adjustment scheme.

[0057] It is worth noting that the system will extract the part with the largest load change amplitude in the change sequence from the load trend prediction data group, which is the "load change amplitude"; at the same time, it will also extract the load data corresponding to the peak time to form the "load peak record".

[0058] While acquiring these predictive features, the system also retrieves real-time data collected during the current time period, especially voltage fluctuation range and temperature influence coefficient. The voltage fluctuation range reflects the grid's stability under load variations; a larger range indicates a more unstable power supply. The temperature influence coefficient quantifies the impact of environmental conditions on electrical load. In mountainous areas and other regions with significant diurnal temperature variations, temperature fluctuations directly drive substantial changes in heating and cooling loads. For example, during the cold winter months, an increase in the temperature influence coefficient can exacerbate sudden load changes during morning and evening peak hours.

[0059] Next, the system will use a multi-dimensional data integration method to merge the predicted data and real-time collected data into a unified dataset according to time series, constructing a comprehensive load response characteristic dataset containing information such as load trends, voltage fluctuations, and temperature disturbances. Based on this, the system will extract data in segments according to predefined segmentation rules, that is, extract the dynamic change patterns of the load in time segments. Each data segment will be accompanied by current environmental condition indicators, such as temperature below 5℃, voltage fluctuation exceeding ±6V, and load change rate exceeding 300 kW / h. The system will determine the actual degree of coupling between load changes and environmental parameters by statistically analyzing the correlation between them in each data segment.

[0060] When the system detects that the coupling between load response and environmental disturbances exceeds a set threshold (e.g., a correlation coefficient exceeding 0.8) within certain time periods, it indicates that the current control strategy is insufficient to handle such high-disturbance scenarios. In this case, adjustments and corrections to the original response characteristics are necessary. The correction method is based on weighted calculations, assigning different weights to different time periods according to their importance or sensitivity. For example, during the morning peak hours, due to concentrated changes in electricity load, the weight is set to 0.7, while during the nighttime off-peak hours, the weight is 0.3. In this way, the system amplifies the data changes during high-impact periods and appropriately downplays those during low-impact periods, ultimately obtaining a set of adjusted response characteristic data.

[0061] The system inputs this adjusted set of response characteristic data into a regression analysis tool for processing. The regression analysis tool used here is a multiple linear regression method. This method comprehensively considers multiple input variables such as the current load fluctuation magnitude, voltage fluctuation intensity, and temperature influence coefficient, and performs a fitting analysis with the load response behavior under similar historical weather conditions to calculate the optimal execution period for control. The entire analysis process is conducted hourly. By comparing the changing trends between current data and historical samples, the system can accurately identify the critical point before the load begins to fluctuate drastically under specific conditions.

[0062] For example, in a mountainous power grid scenario, during a winter cold wave, the system identified that the peak load fluctuation reached a critical threshold between 7:30 and 8:00 AM. Based on the fitted curve and extrapolation, the system suggested issuing control signals half an hour earlier, around 7:00 AM, to initiate load management response, effectively mitigating subsequent peak impacts. This assessment constitutes a preliminary control timing adjustment plan.

[0063] In step S16, based on the preliminary control timing adjustment scheme, parameters are dynamically corrected to obtain the adjusted combination of operating parameters, including: The preliminary control timing adjustment scheme is executed, and the current instability index and harmonic interference intensity are collected after the control is executed. The current instability index and the harmonic interference intensity are matched with the pre-established target operating state parameter benchmark values ​​in a multi-dimensional matching analysis to obtain the deviation distribution characteristics of current stability and harmonic interference intensity. If either the deviation distribution characteristic of the current stability or the deviation distribution characteristic of the harmonic interference intensity is greater than or equal to its corresponding preset threshold, then the temperature and humidity factors in the power grid operating environment are combined, and the parameter correction tool is used to dynamically adjust the characteristics that exceed the threshold, thereby obtaining a corrected temporary parameter set. Based on the temporary parameter set, and considering the correlation between the load response characteristics and the power grid operating status, a weighted calculation method is used for optimization, and an adaptability verification is performed to obtain a combination of operating parameters that meets the control requirements.

[0064] It's worth noting that the system executes a preliminary control timing adjustment plan and, based on this, collects the current instability index and harmonic interference intensity in real time after the control execution. The current instability index reflects the frequency of current fluctuations within a short period, while the harmonic interference intensity measures the degree of interference from high-frequency components in the power grid. These parameters are acquired at a frequency of once per minute or every five minutes to ensure the system can monitor the fluctuation response after the control action is executed.

[0065] Next, the system will compare and analyze the currently collected current instability index and harmonic interference intensity with the previously established target operating state parameter benchmarks. Specifically, this involves multi-dimensional matching of real-time and predicted data, including absolute difference, fluctuation trend difference, and abnormal amplitude. Matching analysis yields a set of deviation distribution characteristics, indicating the degree of agreement between the current power grid state and the preset state. For example, if the predicted current instability index is 6.5 and the actual collected value is 8.0, the absolute difference is 1.5; if the predicted harmonic interference intensity is 5.2 and the actual value is 6.1, the absolute difference is 0.9. The fluctuation trend difference is used to measure the consistency of the two parameter sequences in terms of their temporal variation trends. The system uses slope difference and correlation coefficient as quantification methods. For example, if the slope of the predicted current instability sequence is 0.125 and the slope of the real-time sequence is 0.225, the trend difference is 0.1. If the correlation coefficient between the two is lower than a set threshold (e.g., 0.9), it indicates that the trends are inconsistent. Anomaly magnitude is used to assess the deviation in volatility between two sequences, and is usually calculated using standard deviation. For example, if the standard deviation of the predicted sequence is 0.18 and the standard deviation of the real-time sequence is 0.38, then the anomaly magnitude deviation is 0.2. If it exceeds the system's set threshold (such as 0.15), it indicates that significant abnormal volatility has occurred. The above three indicators together constitute the deviation distribution characteristics of the system.

[0066] In this situation, if the deviation exceeds the system's set threshold (e.g., ±0.8), the system will trigger a parameter correction process. At this point, external environmental interference will be considered, particularly the combined effects of humidity, temperature, and air pressure in mountainous environments. For example, in humid winter weather, increased humidity may exacerbate harmonic interference. The system will read the environmental coefficients at that time, such as a humidity weight of 0.7 and a temperature weight of 0.9, and perform dynamic correction based on the deviation data. The correction method uses the target range as a reference, employing fine-tuning to compress or relax the parameters, making the corrected current instability index and harmonic intensity closer to the actual operating conditions.

[0067] After the corrections are completed, the system will generate a set of temporary parameters. Although these parameters have incorporated the impact of environmental disturbances, their stability and operational adaptability still require further evaluation. Therefore, in the next step, the system will perform correlation analysis between this set of temporary parameters and the previously generated load response characteristic data and power grid operating status. Taking a real-world business scenario as an example, during periods of significant load fluctuations in the morning peak, the system will assign higher weights to parameters for that period, such as a current instability coefficient weight of 0.65, while the weight is only 0.35 during nighttime off-peak periods. The weighted calculation tool will perform secondary calibration of the temporary parameters based on the operating pressure at different times, making the final combination of operating parameters more adaptable to the fluctuation characteristics of different time periods.

[0068] Finally, to determine whether the current combination of operating parameters meets the control timing adjustment target, the system employs a multiple linear regression method for adaptability verification. This method is based on historical multi-scenario operating data, comprehensively considering multiple key variables such as current instability index, harmonic interference intensity, and stability indicators, and combining this with historical trends in control response time for fitting analysis. During the verification phase, the system inputs the currently corrected parameter set into the regression analysis tool to analyze its corresponding optimal control execution time. Subsequently, the system compares this suggested control time with the target trigger time set in the initial control timing adjustment scheme. If the difference is within the system's set tolerance range (e.g., no more than 10 minutes), the parameter combination is deemed to have good adaptability; if it exceeds the tolerance, it prompts that the control parameters need to be corrected or the control strategy re-evaluated.

[0069] For example, at a mountainous site, the current operating parameter combination is a current instability index of 7.9, a harmonic interference intensity of 5.6, and a stability index of 0.72. The recommended control time after analysis is 6:12, while the initial plan sets the time to 6:30, a time difference of 18 minutes, which exceeds the system's allowable range. Therefore, the system determines that this parameter combination does not meet the current control objective and needs to be re-optimized and corrected. If the recommended time is 6:24, with a difference of only 6 minutes, it is considered a reasonable fit and the verification is passed.

[0070] The final generated set of operating parameters is a multi-dimensional set of parameters used to guide control decisions, consisting of information such as the corrected current instability index, harmonic interference intensity, weighted stability index, and control execution period.

[0071] In step S17, based on the combination of operating parameters, control instructions are generated and time-series scheduling is performed to determine the final control execution plan, including: Based on the combination of operating parameters, real-time environmental change data and load fluctuation data of multiple nodes are collected, and differential analysis is performed on different nodes to construct a personalized environmental adaptation dataset. Based on the personalized environmental adaptation dataset, mutation identification and instruction draft generation are performed to obtain a preliminary control scheme; Based on the aforementioned preliminary control scheme, and combined with the parameter recovery cycle and execution priority requirements of the nodes, the allocation and sorting of the operation time sequence are performed to obtain the instruction execution time sequence combination of each node; Based on the combination of the instruction execution time sequence, instruction integration and adaptive calibration are performed to determine the final control execution plan.

[0072] It is worth noting that, based on the obtained combination of operating parameters, the system collects real-time environmental change data and load fluctuations from multiple power grid nodes, conducts differential analysis, and constructs a personalized environmental adaptation dataset for each node. This data includes temperature, humidity, wind speed, voltage and current fluctuations, and is categorized according to the node's geographical conditions and historical load response patterns, making the control logic more targeted and effective during implementation.

[0073] Based on this personalized environmental adaptation dataset, the system further identifies parameter abrupt changes at each node. For example, if a node experiences a sudden surge in load or a sharp drop in power factor under extreme weather conditions, the system will generate a preliminary draft of control instructions. This draft is not executed directly, but rather identifies which nodes require priority control and how those nodes should intervene, forming an initial control scheme to be optimized. The goal of this stage is to ensure that the generation of each control instruction is based on the actual operating state of the current node and can positively regulate the overall stability of the power grid.

[0074] Subsequently, the system enters the control time sequence scheduling phase. Specifically, the control tasks of each node in the initial control scheme are sequentially arranged. Considering that some nodes have long parameter recovery periods, the system prioritizes the execution of instructions for these nodes to avoid time conflicts or execution delays. For example, if a node's equipment response is slow under low-temperature conditions, with a recovery period of 20 minutes, the system will schedule the control execution signal for that node to be issued 25 minutes before the expected event occurs, allowing sufficient reaction window. Simultaneously, for nodes with high-frequency fluctuations, the system adopts a segmented scheduling method, gradually injecting control signals to avoid interference with the overall power grid. The combination of operation time sequences formed in this stage constitutes the time skeleton of the entire execution plan, ensuring that tasks are completed both on time and reasonably.

[0075] During the instruction integration phase, the system performs a one-to-one mapping and calibration of the control instructions for each node against the aforementioned time sequence. If a control instruction is found to be unable to meet its corresponding environmental adaptation requirements at the current time point—for example, if the instruction requires a power switching operation at a node, but the node's temperature is below -10℃ and humidity exceeds 90%—the system determines that there is a risk to the equipment's thermal stability and will adjust the instruction parameters accordingly. Adjustments may include reducing the power switching ratio from the originally planned 20% to 10% to alleviate equipment load; or postponing an instruction originally scheduled for 6:30 to 7:15 to avoid periods of significant load fluctuation. This fine-tuning or delay is based on a comprehensive consideration of the node's current state and execution risks, ensuring that the instruction execution neither compromises equipment safety nor triggers cascading grid fluctuations, thus making the final control execution plan more robust and in line with the actual needs of the complex operating environment in mountainous areas.

[0076] Reference Figure 2 The second embodiment of the present invention provides an intelligent control system for phase-selective circuit breakers based on big data, comprising: The data acquisition module is used to acquire the raw runtime data sets of each node; The initial running data group module is used to perform feature extraction and state evaluation based on the original running data group to obtain the initial running data group; The load feature set module is used to standardize the initial running data set and perform noise filtering, continuous completion and seasonal leveling operations to obtain the cleaned load feature set. The prediction data group module is used to perform trend calculations based on the load characteristic set, integrating the frequency offset amplitude and power factor changes in the power grid system operating parameters, to obtain the load trend prediction data group. The control adjustment module is used to analyze the load response characteristics based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, so as to determine the initial control timing adjustment scheme. The operating parameter module is used to dynamically correct parameters according to the initial control timing adjustment scheme to obtain the adjusted operating parameter combination; The control execution module is used to generate control instructions and schedule time sequences based on the combination of operating parameters, thereby determining the final control execution plan.

[0077] It should be noted that the intelligent control device for phase selection circuit breakers based on big data provided in this embodiment of the invention is used to execute all the process steps of the intelligent control method for phase selection circuit breakers based on big data in the above embodiment. The working principles and beneficial effects of the two are one-to-one, so they will not be described again.

[0078] This invention also provides an electronic device. The electronic device includes a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a big data-based intelligent control program for phase-selective circuit breakers. When the processor executes the computer program, it implements the steps described in the various big data-based intelligent control method embodiments for phase-selective circuit breakers, for example... Figure 1 The step S11 shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above-described device embodiments, such as the data acquisition module.

[0079] For example, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0080] The electronic device may be a desktop computer, laptop, handheld computer, or smart tablet, etc. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will understand that the above components are merely examples of electronic devices and do not constitute a limitation on the electronic device. It may include more or fewer components than described above, or combine certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, etc.

[0081] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the electronic device, connecting all parts of the electronic device via various interfaces and lines.

[0082] The memory can be used to store the computer programs and / or modules. The processor implements various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and by calling data stored in the memory. The memory may mainly include a program storage area and a data storage area. The program storage area may store the operating system, at least one application program required for a function (such as sound playback function, image playback function, etc.), etc.; the data storage area may store data created according to the use of the mobile phone (such as audio data, phonebook, etc.). In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as hard disk, memory, plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0083] Wherein, if the modules / units integrated in the electronic device are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc. It should be noted that the content included in the computer-readable medium can be appropriately added or removed according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, computer-readable media do not include electrical carrier signals and telecommunication signals.

[0084] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can be specifically implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0085] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A smart control method for phase-selective circuit breakers based on big data, characterized in that, include: Obtain the raw runtime data sets for each node; Based on the original running data set, feature extraction and state evaluation are performed to obtain the initial running data set; The initial running data set is standardized, and noise filtering, continuous completion, and seasonal leveling operations are performed to obtain a cleaned load feature set. Based on the load characteristic set, frequency offset amplitude and power factor changes in the power grid system operating parameters are integrated to perform trend calculations and obtain a load trend prediction data set. Based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, load response characteristics are analyzed to determine the initial control timing adjustment scheme. Based on the preliminary control timing adjustment scheme, the parameters are dynamically corrected to obtain the adjusted combination of operating parameters; Based on the combination of operating parameters, control commands are generated and time-series scheduling is performed to determine the final control execution plan.

2. The intelligent control method for phase-selective circuit breakers based on big data according to claim 1, characterized in that, The step of performing feature extraction and state evaluation based on the original running data set to obtain the initial running data set includes: The original operating data set includes: load peak time series, load valley time series, voltage fluctuation range, and current instability index; Based on the load peak time series and load trough time series, time series analysis methods are used to extract the periodicity and fluctuation amplitude characteristics of the load. If the periodicity characteristic is greater than or equal to a preset periodic threshold and the fluctuation amplitude characteristic is greater than or equal to a preset fluctuation threshold, then for the voltage fluctuation range and the current instability index, calculate the stability index of the power grid operation state. Based on the stability index, cluster analysis is used to classify the operating status of each node to obtain the initial operating data group.

3. The intelligent control method for phase-selective circuit breakers based on big data according to claim 1, characterized in that, The initial running data set is standardized, and noise filtering, continuous completion, and seasonal leveling operations are performed to obtain a cleaned load feature set, including: Obtain the daily average load curve and seasonal load characteristics; Based on the initial running data set, median filtering and noise reduction processing is performed to obtain the denoised load distribution set; Based on the daily average load curve and the seasonal load characteristics, the load distribution set is smoothed to obtain a stable range; If the fluctuation range of the stable interval exceeds the preset fluctuation threshold, the stable interval is completed to restore the time series continuity of the load distribution data and obtain the completed load data. Based on the load data, combined with a preset seasonal weighting coefficient, data standardization processing is performed to obtain a cleaned load feature set.

4. The intelligent control method for phase-selective circuit breakers based on big data according to claim 1, characterized in that, The process involves performing trend calculations based on the load characteristic set, integrating frequency offset amplitude and power factor changes from the power grid system operating parameters, to obtain a load trend prediction data set, including: Based on the frequency offset amplitude and the power factor change, the load feature set is jointly analyzed to obtain a fused load comprehensive dataset. The load aggregate dataset is subjected to time series extraction, and the load fluctuation frequency is analyzed to obtain the fluctuation frequency range; Determine the direction of change of the fluctuation frequency range relative to the load history. If the direction of change exceeds a preset change threshold, a weighted calculation method is used for correction to obtain an adjusted load trend change sequence. By combining historical data as a benchmark, trend calculations are performed on the load trend change sequence to obtain a load trend prediction data set.

5. The intelligent control method for phase-selective circuit breakers based on big data according to claim 1, characterized in that, The process involves analyzing load response characteristics and calibrating control timing based on the load trend prediction data set, combined with real-time collected voltage fluctuation ranges and temperature influence coefficients, to determine a preliminary control timing adjustment scheme, including: Based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, a multi-dimensional fusion processing is performed to obtain the load response characteristic dataset. Based on the load response characteristic dataset, combined with environmental interference factors under extreme weather conditions, load segmentation is performed to obtain the correlation between load response characteristics and power grid operating status. If the correlation between the load response characteristic distribution and the historical power grid operating status is greater than or equal to a preset correlation threshold, the load response characteristic dataset is corrected to obtain an adjusted response characteristic dataset. Based on the aforementioned response characteristic dataset, final control timing calibration is performed to determine a preliminary control timing adjustment scheme.

6. The intelligent control method for phase-selective circuit breakers based on big data according to claim 5, characterized in that, The step of dynamically correcting parameters according to the preliminary control timing adjustment scheme to obtain the adjusted operating parameter combination includes: The preliminary control timing adjustment scheme is executed, and the current instability index and harmonic interference intensity are collected after the control is executed. The current instability index and the harmonic interference intensity are matched with the pre-established target operating state parameter benchmark values ​​in a multi-dimensional matching analysis to obtain the deviation distribution characteristics of current stability and harmonic interference intensity. If either the deviation distribution characteristic of the current stability or the deviation distribution characteristic of the harmonic interference intensity is greater than or equal to its corresponding preset threshold, then the temperature and humidity factors in the power grid operating environment are combined, and the parameter correction tool is used to dynamically adjust the characteristics that exceed the threshold, thereby obtaining a corrected temporary parameter set. Based on the temporary parameter set, and considering the correlation between the load response characteristics and the power grid operating status, a weighted calculation method is used for optimization, and an adaptability verification is performed to obtain a combination of operating parameters that meets the control requirements.

7. The intelligent control method for phase-selective circuit breakers based on big data according to claim 1, characterized in that, The step of generating control commands and scheduling time sequences based on the combination of operating parameters to determine the final control execution plan includes: Based on the combination of operating parameters, real-time environmental change data and load fluctuation data of multiple nodes are collected, and differential analysis is performed on different nodes to construct a personalized environmental adaptation dataset. Based on the personalized environmental adaptation dataset, mutation identification and instruction draft generation are performed to obtain a preliminary control scheme; Based on the aforementioned preliminary control scheme, and combined with the parameter recovery cycle and execution priority requirements of the nodes, the allocation and sorting of the operation time sequence are performed to obtain the instruction execution time sequence combination of each node; Based on the combination of the instruction execution time sequence, instruction integration and adaptive calibration are performed to determine the final control execution plan.

8. A phase-selection circuit breaker intelligent control system based on big data, characterized in that, include: The data acquisition module is used to acquire the raw runtime data sets of each node; The initial running data group module is used to perform feature extraction and state evaluation based on the original running data group to obtain the initial running data group; The load feature set module is used to standardize the initial running data set and perform noise filtering, continuous completion and seasonal leveling operations to obtain the cleaned load feature set. The prediction data group module is used to perform trend calculations based on the load characteristic set, integrating the frequency offset amplitude and power factor changes in the power grid system operating parameters, to obtain the load trend prediction data group. The control adjustment module is used to analyze the load response characteristics based on the load trend prediction data set, combined with the real-time collected voltage fluctuation range and temperature influence coefficient, so as to determine the initial control timing adjustment scheme. The operating parameter module is used to dynamically correct parameters according to the initial control timing adjustment scheme to obtain the adjusted operating parameter combination; The control execution module is used to generate control instructions and schedule time sequences based on the combination of operating parameters, thereby determining the final control execution plan.

9. An electronic device, characterized in that, The system includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein the processor, when executing the computer program, implements the intelligent control method for phase-selective circuit breakers based on big data as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the intelligent control method for phase-selective circuit breakers based on big data as described in any one of claims 1 to 7.

Images