High-voltage capacitor power harvesting system and method based on safety monitoring

By constructing adaptive state feature vectors and coupled feature parameters, the operating mode of the high-voltage capacitor power supply system is identified and hierarchical control is implemented, which solves the problem of unstable power supply to the high-voltage circuit breaker mechanism box, improves the stability and safety of the system, and reduces the risk of thermal breakdown.

CN122600491BActive Publication Date: 2026-11-10JIANGSU DANTONG ELECTRIC
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202610647532.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-05-12
Publication Date
2026-11-10
Estimated Expiration
2046-05-12

AI Technical Summary

Technical Problem

The control and monitoring units in the existing high-voltage circuit breaker mechanism box have insufficient power supply stability, resulting in problems such as large leakage current and insufficient effective output power. Furthermore, they are difficult to maintain power supply during load fluctuations and are prone to thermal breakdown, which can lead to a decline in insulation performance and affect the normal operation of the equipment.

Method used

By constructing an adaptive state feature vector, calculating coupled feature parameters and comprehensive safety, identifying operating modes, and implementing hierarchical control strategies, including gradual adjustment, output limiting, and cut-off control, and combining control lead time for correction, the power supply stability and anti-interference capability are improved.

Benefits of technology

This improved the stability of the high-voltage capacitor power supply system, reduced the risk of thermal breakdown, enhanced the system's anti-interference capability, improved the accuracy and response speed of fault diagnosis, and reduced the equipment failure rate.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122600491B_ABST
    Figure CN122600491B_ABST
Patent Text Reader

Abstract

This invention discloses a high-voltage capacitor power extraction system and method based on safety monitoring, relating to the field of power equipment power supply technology. The invention extracts the variation characteristics of operating parameters within a time window to construct an adaptive state feature vector; based on the state feature vector, it calculates the coupling strength of the time window and, based on the historical coupling strength, calculates coupling feature parameters; based on the coupling feature parameters of multiple consecutive time windows, it identifies the system operating mode; based on the operating mode, operating parameters, and coupling feature parameters, it calculates the comprehensive safety level, generates a hierarchical control strategy based on the comprehensive safety level, calculates the control lead time, and performs advance correction based on the control lead time, significantly improving system safety, enabling timely identification of potential risks, early intervention, and avoidance of sudden failures.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of power supply technology for electrical equipment, specifically a high-voltage capacitor power extraction system and method based on safety monitoring. Background Technology

[0002] The control unit, monitoring unit and actuator in the high-voltage circuit breaker mechanism box all require a stable power supply. Usually, the power is directly drawn from the three-phase power. In actual application, there are common problems such as large leakage current and insufficient effective output power. Especially when the load fluctuates greatly, it is difficult to continuously meet the power demand of various electrical components in the mechanism box, which affects the normal operation of the equipment. To alleviate this problem, some solutions add battery modules for auxiliary power supply, but they mainly play a compensatory role and still fail to fundamentally improve the stability of power supply.

[0003] During long-term operation, local temperature rise may be too high due to hot load, no-load and other conditions, which may lead to a decrease in insulation performance or even thermal breakdown, causing damage to the power module. Overvoltage protection measures are added to protect against this, but there is still a lack of effective control over the risk of continuous thermal failure under complex conditions. The power supply circuit is prone to output voltage instability under the influence of input voltage fluctuations and load disturbances, which may cause sine wave clipping and waveform distortion, affecting the normal operation of subsequent circuits.

[0004] Therefore, improving the power supply stability of high-voltage capacitor power extraction systems, reducing the risk of thermal breakdown of high-voltage transformers, and enhancing the system's anti-interference capability have become technical problems that need to be solved in this field. Summary of the Invention

[0005] The purpose of this invention is to provide a high-voltage capacitor power extraction system and method based on safety monitoring, so as to solve the problems raised in the prior art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a high-voltage capacitor power extraction method based on safety monitoring, the method comprising:

[0007] Step S100: At a preset sampling time in the high-voltage capacitor power supply system, the operating parameters of the high-voltage capacitor power supply system are collected, and continuous time series data is constructed by combining the timestamps. A preset time window is set, and based on the continuous time series data, the change features of the operating parameters within the time window are extracted to construct an adaptive state feature vector.

[0008] Step S200: Based on the state feature vector, calculate the coupling strength of the time window, and based on the coupling strength of the historical records, calculate the coupling feature parameters;

[0009] Step S300: Based on the coupling characteristic parameters of multiple consecutive time windows, the system operation mode is analyzed and the operation mode is identified;

[0010] Step S400: Calculate the comprehensive safety factor based on the operating mode, operating parameters and coupling characteristic parameters, generate a hierarchical control strategy based on the comprehensive safety factor, calculate the control lead based on the difference change, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety factor, and make advance corrections based on the control lead.

[0011] Furthermore, the construction of the adaptive state feature vector in step S100 includes:

[0012] The operating parameters include voltage parameters, current parameters, and load parameters. The variation characteristics include voltage variation characteristics, load variation characteristics, fluctuation characteristics, and time characteristics. The voltage variation characteristic calculation process is as follows: based on the voltage parameters in the continuous time series data, the voltage difference between adjacent sampling times is calculated in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and the average value is calculated. The average value is used as the voltage variation characteristic.

[0013] The load change characteristic calculation process is based on the load parameters in the continuous time series data. The load difference between adjacent sampling times is calculated in chronological order to obtain the load change amount. Within a preset time window, the load change amount is statistically analyzed, and the average value and standard deviation are calculated. The average value and standard deviation are used together as the load change characteristic.

[0014] The fluctuation characteristic calculation process is as follows: within a preset time window, the maximum and minimum values ​​of the voltage parameter within the preset time window are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation amount. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the time length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation amount, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic.

[0015] Normalize the different types of change features respectively. According to the preset weights of each change feature, each normalized change feature is weighted and calculated with the corresponding weight to obtain the adaptive state feature of each change feature. The adaptive state features of each change feature are summarized to construct an adaptive state feature vector.

[0016] By calculating the voltage difference, load change, fluctuation, and other characteristics one by one, the system's changes in different time periods can be fully captured, providing an accurate dynamic description. This can effectively distinguish between the normal and abnormal states of the system, providing strong data support for subsequent fault diagnosis or early warning.

[0017] By calculating fluctuation characteristics such as fluctuation frequency and zero-crossing frequency, abnormal behavior of the system under load fluctuation and voltage fluctuation can be effectively identified, potential instability factors can be detected in time, and the accuracy of system stability analysis can be improved.

[0018] Furthermore, the calculation of coupling characteristic parameters in step S200 includes:

[0019] In the adaptive state feature vector, voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features are extracted according to time windows. The changes of voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features within the time window are calculated respectively. The change is the difference between the maximum and minimum values ​​of the adaptive state features. The mean and standard deviation of the changes in the historical time windows are calculated. The standardized change is calculated as A = (BC) / D, where B represents the change in the current time window, C represents the mean of the historical time windows, and D represents the standard deviation of the historical time windows.

[0020] Summarize the standardized changes in voltage, load, and fluctuation within the time window, and calculate the coupling strength K = a × A1 × A2 + b × A3 for each time window, where A1, A2, and A3 represent the standardized changes in voltage, load, and fluctuation, respectively, a represents the weights of voltage and load, and b represents the weights of fluctuation.

[0021] Within a continuous time window, the coupling strength of each time window is obtained, the difference in coupling strength between adjacent time windows is calculated, and then divided by the time length of the time window to calculate the coupling change rate.

[0022] Historical records under stable operating conditions are obtained. A stable operating condition is defined as one that simultaneously meets the following conditions: the average value of voltage variation characteristics within each time window does not exceed a preset threshold; the standard deviation of load variation characteristics does not exceed a preset threshold; the amount of change in fluctuation characteristics does not exceed a preset threshold, and the rate of change of peak frequency and zero-crossing frequency does not exceed a preset threshold; the absolute difference in the coupling rate of change between adjacent time windows does not exceed a preset threshold. When all of the above conditions are met within N consecutive time windows, the operating state corresponding to that time period is determined as a stable operating condition.

[0023] The coupling strength of the historical records is collected, and the average value of the coupling strength is calculated. The average value, the coupling strength of the current time window, and the coupling change rate are weighted and summed to calculate the coupling characteristic parameters.

[0024] By extracting the adaptive state features of voltage, load, and fluctuation separately, and calculating their respective changes, the dynamic changes of each parameter in different time windows can be accurately reflected. Using the difference between the maximum and minimum values ​​as the change value can effectively highlight the amplitude of state fluctuations, improve the sensitivity to abnormal changes, and thus improve the accuracy of operating state identification.

[0025] By introducing standardized change quantities, the changes in the current time window can be compared with historical stable states, which can eliminate the impact of parameter differences under different dimensions and operating conditions, make abnormal changes more prominent, and promptly identify problems such as sudden fluctuations, abnormal load changes and voltage instability, thereby improving the system's anomaly early warning capability.

[0026] By constructing coupling strength parameters, voltage, load and fluctuation characteristics are comprehensively correlated and analyzed. It is no longer limited to single parameter judgment, but reflects the overall operating status of the system from multiple dimensions. The coupling correlation is reflected by the product relationship between voltage and load, and the system disturbance information is supplemented by fluctuation characteristics, making the state judgment more comprehensive and scientific.

[0027] By setting stable operating condition judgment conditions, multiple conditions are jointly constrained for voltage change amplitude, load standard deviation, fluctuation characteristics and coupling change rate, and the stability conditions are required to be met simultaneously for N consecutive time windows. This can effectively avoid misjudgment caused by short-term disturbances or instantaneous anomalies, and improve the reliability and accuracy of stable operating condition identification.

[0028] By calculating the rate of change of coupling between continuous time windows, the evolution trend of the system state can be reflected. This not only identifies whether the current state is abnormal, but also detects the trend of the system moving from stability to instability in advance, realizing the transformation from ex-post diagnosis to ex-ante warning and improving the safety of system operation.

[0029] By combining the historical average coupling strength under stable operating conditions, the current time window coupling strength, and the rate of change of coupling through weighted summation, coupling characteristic parameters are constructed. This allows the final result to reflect the current real-time state of the system and to be dynamically corrected based on historical stability benchmarks, thus exhibiting strong adaptability and practical representativeness.

[0030] Furthermore, identifying the operating mode in step S300 includes:

[0031] The coupling feature parameters of multiple consecutive time windows are arranged in chronological order to obtain a set of coupling feature parameters. Based on the set of coupling feature parameters, the slope and magnitude of the changes in the coupling feature parameters are calculated. The slope calculation process involves extracting the coupling feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupling feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupling feature parameters per unit time. This rate of change is used as the slope. The magnitude calculation process involves calculating the absolute value of the difference between the coupling feature parameters between adjacent time windows. This absolute value is used as the magnitude.

[0032] In a high-voltage capacitor power extraction system, an operating mode is preset based on the time evolution sequence of coupling characteristic parameters and the adaptive state characteristics of the current time window. The operating modes include an accelerated change mode, an oscillation evolution mode, and a critical instability mode.

[0033] The accelerated change mode is defined as follows: when the slope of the change of the coupled feature parameters is greater than a preset threshold, and at least two operating parameters in the adaptive state features exceed the judgment range of the stable operating condition, the system is determined to enter the accelerated change mode.

[0034] The oscillation evolution mode is defined as follows: when the change amplitude of the coupling characteristic parameter increases compared with the coupling characteristic parameter under the previous stable condition and the change amount of the fluctuation characteristic is greater than a preset threshold, the system is determined to enter the oscillation evolution mode.

[0035] The critical instability mode is defined as the system being in a critical instability mode when the difference between the coupling characteristic parameters of the system and the coupling characteristic parameters of the stable condition exceeds a preset threshold and the rate of change exceeds the preset threshold.

[0036] By constructing a set of coupled feature parameters and performing time series analysis on them, and introducing two key indicators, the slope and magnitude of change, the system's operating state can be characterized from two dimensions: the rate of change and the intensity of change, making the description of the system's dynamic behavior more comprehensive and detailed.

[0037] By classifying the system's operating state into accelerated change mode, oscillating evolution mode, and critical instability mode, and combining this with specific judgment conditions for identification, it is possible to accurately distinguish different types of abnormal evolution processes, avoid misjudgment problems caused by single indicator judgment, and improve the accuracy and reliability of operating mode identification.

[0038] Accelerated change mode reflects that system parameters deviate rapidly from the normal range, oscillation evolution mode reflects that the system has a continuous fluctuation risk, and critical instability mode indicates that the system is approaching the instability boundary. Through the hierarchical pattern recognition mechanism, it is possible to achieve layered early warning from minor anomalies to severe instability, thereby improving the pertinence and effectiveness of operation and maintenance response.

[0039] By analyzing the slope of the changes in coupling characteristic parameters, we can capture the accelerated change trend of the system state in advance; by analyzing the magnitude of the changes, we can identify whether the system has entered the oscillation or fluctuation enhancement stage, thereby enabling us to predict the evolution of the system from stability to abnormality or even instability in advance and improve the system's forward-looking monitoring capabilities.

[0040] Furthermore, step S400 includes:

[0041] Within the current time window, the difference between the coupling characteristic parameter and the critical instability threshold is obtained. The current coupling characteristic parameter is subtracted from the critical instability threshold to calculate the coupling margin. The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting the transition stage data during the transition from stable to unstable mode, performing statistical analysis based on the transition stage data, calculating the mean g and standard deviation h of the coupling characteristic parameter, and calculating the critical instability threshold L = g + h × j, where j represents the preset safety factor.

[0042] Obtain the adaptive state feature vector of the current time window, calculate the average adaptive state feature vector under stable operating conditions, calculate the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, normalize the difference, and sum the normalized difference by weight to obtain the state deviation.

[0043] Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the values ​​of coupling margin, state deviation, and operating mode, the values ​​are weighted and summed according to preset weights to calculate the comprehensive safety level.

[0044] Based on the changing trend of voltage parameters within a continuous time window, the difference change of voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are calculated and normalized. The normalized difference change, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are then weighted and summed to calculate the control lead.

[0045] The system obtains the overall safety level from historical records, calculates the mean n and standard deviation m of the overall safety level, and divides the system operating state into three safety intervals based on the range of the overall safety level. The first safety interval is [nm×u1, n+m×u1], the second safety interval is [nm×u2, nm×u1)∪(n+m×u1, n+m×u2], and the third safety interval is [-∞, nm×u2)∪(n+m×u2, +∞]. u1 and u2 are adjustable coefficients, and are matched with corresponding hierarchical control strategies. The hierarchical control strategy includes performing gradual adjustment control when the overall safety level is in the first safety interval.

[0046] When the overall safety level is in the second safety range, output limit control is executed;

[0047] When the overall safety level S is in the third safety range, the cutoff control is executed;

[0048] When executing the hierarchical control strategy, the triggering time of the control command is advanced according to the control advance amount;

[0049] By calculating coupling margin, state deviation, and overall safety in real time, the safety status of the system can be dynamically assessed, and corresponding control strategies can be implemented according to different safety ranges. This can identify potential risks in the early stages, reduce the possibility of sudden system instability, and effectively ensure the safety of system operation.

[0050] The calculation of comprehensive safety combines coupling characteristic parameters and adaptive state characteristics, making the judgment of system operating status more accurate and comprehensive. By monitoring the voltage change trend and the rate of change of coupling characteristics in real time, system anomalies can be detected in advance, triggering control commands and ensuring that the system is adjusted in a timely manner during the transition phase.

[0051] Specific control strategies are assigned according to different safety ranges, such as gradual adjustment control, output limiting control and cut-off control, which can flexibly respond to different changes in system state. When the overall safety level changes, the control mode is adjusted in time to avoid system instability or large fluctuations and improve the dynamic response capability of the system.

[0052] By calculating the control lead time, the triggering time of control commands can be corrected in advance, ensuring that corresponding preventive measures are taken before the system enters an unstable or abnormal state, reducing the lag in system response, and improving the timeliness and accuracy of control strategies.

[0053] By using precise safety assessment and graded control strategies, potential risks can be identified and corrected in the early stages of system operation, reducing the failure rate and avoiding maintenance costs caused by sudden instability, which helps to extend the lifespan of the equipment.

[0054] To better implement the above method, a high-voltage capacitor power extraction system based on safety monitoring is also proposed. The system includes an adaptive state feature vector module, a coupling feature parameter module, an operation mode module, and a correction module.

[0055] Adaptive State Feature Vector Module: At a preset sampling time in the high-voltage capacitor power extraction system, the operating parameters of the high-voltage capacitor power extraction system are collected, and continuous time series data is constructed by combining the timestamps. A preset time window is set, and based on the continuous time series data, the change features of the operating parameters within the time window are extracted to construct an adaptive state feature vector.

[0056] Coupling feature parameter module: Based on the state feature vector, calculate the coupling strength of the time window, and based on the coupling strength of the historical records, calculate the coupling feature parameters;

[0057] Operating mode module: Based on the coupled characteristic parameters of multiple consecutive time windows, the operating mode of the system is identified.

[0058] Correction module: Calculates the comprehensive safety level based on the operating mode, operating parameters and coupling characteristic parameters, generates a hierarchical control strategy based on the comprehensive safety level, calculates the control lead based on the difference change, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety level, and performs advance correction based on the control lead.

[0059] Furthermore, the adaptive state feature vector module includes a changing feature unit and a unit for constructing the adaptive state feature vector:

[0060] The variation feature unit includes voltage variation features, load variation features, fluctuation features, and time features. The voltage variation feature is calculated based on voltage parameters in continuous time series data, calculating the voltage difference between adjacent sampling times in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and an average value is calculated, which is then used as the voltage variation feature. The load variation feature is calculated based on load parameters in continuous time series data, calculating the load difference between adjacent sampling times in chronological order. The load change is obtained by taking the difference. Within a preset time window, the load change is statistically analyzed, and the average and standard deviation are calculated. The average and standard deviation are used together as the load change characteristic. The fluctuation characteristic calculation process is as follows: within the preset time window, the maximum and minimum values ​​of the voltage parameter are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic.

[0061] Constructing adaptive state feature vector units: Normalize different types of change features respectively. According to the preset weights of each change feature, calculate the weighted sum of each normalized change feature and its corresponding weight to obtain the adaptive state feature of each change feature. Summarize the adaptive state features of each change feature to construct an adaptive state feature vector.

[0062] Furthermore, the coupling characteristic parameter module includes a unit for calculating coupling strength and a unit for calculating coupling characteristic parameters:

[0063] Calculate the coupling strength unit: In the adaptive state feature vector, extract the voltage adaptive state feature, load adaptive state feature, and fluctuation adaptive state feature according to the time window. Calculate the change of the voltage adaptive state feature, load adaptive state feature, and fluctuation adaptive state feature within the time window. The change is the difference between the maximum and minimum values ​​in the adaptive state feature. Calculate the mean and standard deviation of the historical time window changes, calculate the standardized change, summarize the standardized changes of voltage, load, and fluctuation within the time window, and calculate the coupling strength of each time window.

[0064] Calculate coupling characteristic parameters: Within a continuous time window, acquire the coupling strength of each time window, calculate the difference in coupling strength between adjacent time windows, and divide it by the time length of the time window to calculate the coupling change rate; acquire historical records under stable operating conditions. A stable operating condition is determined when the following conditions are met simultaneously: the average change of voltage change characteristics within each time window does not exceed a preset threshold; the standard deviation of load change characteristics does not exceed a preset threshold; the change in fluctuation characteristics does not exceed a preset threshold, and the rate of change of peak frequency and zero-crossing frequency does not exceed a preset threshold; the absolute difference in coupling change rate between adjacent time windows does not exceed a preset threshold; when the above conditions are met within N consecutive time windows, the operating state corresponding to the time period is determined as a stable operating condition. Collect the coupling strength of the historical records and calculate the average value of the coupling strength. Calculate the coupling characteristic parameters by weighted summing of the average value, the coupling strength of the current time window, and the coupling change rate.

[0065] Furthermore, the operation mode module includes a coupled feature parameter change unit and a preset operation mode unit:

[0066] Coupled feature parameter change unit: Coupled feature parameters from multiple consecutive time windows are arranged in chronological order to obtain a set of coupled feature parameters. Based on this set, the slope and magnitude of the changes in the coupled feature parameters are calculated. The slope calculation process involves extracting the coupled feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupled feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupled feature parameters per unit time. This rate of change is used as the slope. The magnitude calculation process involves calculating the absolute value of the difference between the coupled feature parameters between adjacent time windows, and this absolute value is used as the magnitude.

[0067] Preset operating mode unit: In the high-voltage capacitor power supply system, an operating mode is preset based on the time evolution sequence of the coupling characteristic parameters and the adaptive state characteristics of the current time window. The operating modes include accelerated change mode, oscillation evolution mode and critical instability mode.

[0068] Furthermore, the correction module includes a unit for calculating the overall safety factor and a unit for calculating the control lead time.

[0069] The comprehensive safety unit calculates the following: Within the current time window, it obtains the difference between the coupling characteristic parameter and the critical instability threshold, subtracts the current coupling characteristic parameter from the critical instability threshold, and calculates the coupling margin. The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting transition stage data during the transition from stable to unstable modes, performing statistical analysis based on the transition stage data, calculating the mean and standard deviation of the coupling characteristic parameter, calculating the critical instability threshold, obtaining the adaptive state feature vector of the current time window, calculating the average adaptive state feature vector corresponding to the stable operating condition, calculating the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, and normalizing the difference. The normalized difference is then weighted and summed to obtain the state deviation. Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the coupling margin, state deviation, and operating mode assignments, they are weighted and summed according to preset weights to calculate the comprehensive safety.

[0070] The control lead calculation unit calculates the difference in voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety factor based on the changing trend of voltage parameters within a continuous time window. After normalization, the normalized difference in voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety factor are weighted and summed to obtain the control lead.

[0071] Compared with the prior art, the beneficial effects of the present invention are: by constructing an adaptive state feature vector, coupling feature parameters and comprehensive safety degree, the operating status of the high voltage capacitor power supply system can be finely evaluated. The introduction of stable operating condition judgment and critical instability threshold can identify potential instability risks in a timely manner and prevent equipment accidents due to overload or abnormal fluctuations. The hierarchical control strategy can be automatically adjusted according to the safety degree, from gradual adjustment to output limitation and then to cut-off control, to achieve dynamic safety protection.

[0072] The concept of control lead time is introduced, and the control command is corrected in advance based on voltage changes, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety. Compared with the traditional control method that only relies on a single measurement or a fixed threshold, it can intervene in potential risks in advance and improve system stability and safety margin.

[0073] It can analyze the coupling characteristic parameters of continuous time windows and automatically identify accelerated change mode, oscillation evolution mode and critical instability mode. Compared with traditional experience judgment or simple threshold alarm, this method realizes intelligent and dynamic operation mode identification, improving the accuracy of fault diagnosis and response speed.

[0074] By utilizing adaptive state feature vectors, coupled feature parameters, and comprehensive safety, a multi-dimensional analysis of the system state can be performed. Multiple indicators such as voltage, load, and fluctuation can be integrated and calculated to achieve a quantitative evaluation of the overall system risk. Compared with single-parameter monitoring methods, this method can better reflect the true operating state of the system and avoid misjudgments and omissions. Attached Figure Description

[0075] Figure 1 This is a schematic diagram of the high-voltage capacitor power extraction system based on safety monitoring according to the present invention. Detailed Implementation

[0076] 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.

[0077] Please see Figure 1 This invention provides a technical solution: a high-voltage capacitor power extraction method based on safety monitoring, the method comprising:

[0078] Step S100: At a preset sampling time in the high-voltage capacitor power supply system, the operating parameters of the high-voltage capacitor power supply system are collected, and continuous time series data is constructed by combining the timestamps. A preset time window is set, and based on the continuous time series data, the change features of the operating parameters within the time window are extracted to construct an adaptive state feature vector.

[0079] The construction of the adaptive state feature vector in step S100 includes:

[0080] The operating parameters include voltage parameters, current parameters, and load parameters. The variation characteristics include voltage variation characteristics, load variation characteristics, fluctuation characteristics, and time characteristics. The voltage variation characteristic calculation process is as follows: based on the voltage parameters in the continuous time series data, the voltage difference between adjacent sampling times is calculated in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and the average value is calculated. The average value is used as the voltage variation characteristic.

[0081] The load change characteristic calculation process is based on the load parameters in the continuous time series data. The load difference between adjacent sampling times is calculated in chronological order to obtain the load change amount. Within a preset time window, the load change amount is statistically analyzed, and the average value and standard deviation are calculated. The average value and standard deviation are used together as the load change characteristic.

[0082] The fluctuation characteristic calculation process is as follows: within a preset time window, the maximum and minimum values ​​of the voltage parameter within the preset time window are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation amount. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the time length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation amount, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic.

[0083] Normalize the different types of change features respectively. According to the preset weights of each change feature, each normalized change feature is weighted and calculated with the corresponding weight to obtain the adaptive state feature of each change feature. The adaptive state features of each change feature are summarized to construct an adaptive state feature vector.

[0084] For example, suppose in a high-voltage capacitor power supply system, the preset sampling time is once every 0.1 seconds. The operating parameters of the high-voltage capacitor power supply system include voltage parameters, current parameters, and load parameters. The collected continuous time series data is as follows: Time 0.0: Voltage 1000, Current 50, Load 200; Time 0.1: Voltage 1002, Current 52, Load 202; Time 0.2: Voltage 1001, Current 51, Load 205; Time 0.3: Voltage 1005, Current 53, Load 203; Time 0.4: Voltage 1003, Current 50, Load 204; Time 0.5: Voltage 1004, Current 52, Load 206.

[0085] The preset time window is 0.5 seconds. Feature extraction is performed on continuous time series data, and voltage change features are calculated. The voltage difference between adjacent sampling times is calculated in chronological order as (0.0-0.1)=1002-1000=2V.

[0086] (0.1-0.2)=1001-1002=-1V;

[0087] (0.2-0.3)=1005-1001=4V;

[0088] (0.3-0.4)=1003-1005=-2V;

[0089] (0.4-0.5)=1004-1003=1V;

[0090] Calculate the change in adjacent differences (0.0-0.2) = -1-2 = -3V;

[0091] (0.1-0.3)=4-(-1)=5V;

[0092] (0.2-0.4)=-2-4=-6V;

[0093] (0.3-0.5)=1-(-2)=3V;

[0094] The average change in difference within the preset time window is (-3+5-6+3) / 4=-0.25V;

[0095] Calculation of load variation characteristics, load difference between adjacent sampling times.

[0096] (0.0-0.1)=202-200=2kW;

[0097] (0.1-0.2)=205-202=3kW;

[0098] (0.2-0.3)=203-205=-2kW;

[0099] (0.3-0.4)=204-203=1kW;

[0100] (0.4-0.5)=206-204=2kW;

[0101] Average value = (2 + 3 - 2 + 1 + 2) / 5 = 1.2 kW;

[0102] The standard deviation is 2.07 kW;

[0103] The load variation characteristics are (mean 1.2kW, standard deviation 2.07kW);

[0104] Maximum voltage = 1005V, minimum voltage = 1000V, fluctuation = 1005 - 1000 = 5V;

[0105] The number of peaks occurs once within a 0.5-second time window;

[0106] If the voltage signal does not cross zero, the number of zero-crossings = 0.

[0107] Crest frequency = 1 / 0.5 = 2Hz;

[0108] Zero-crossing frequency = 0 / 0.5 = 0Hz;

[0109] The fluctuation characteristics are (fluctuation amount 5V, peak frequency 2Hz, zero crossing frequency 0Hz);

[0110] Each change characteristic is normalized and weighted according to its weight. For example, the voltage change characteristic has a weight of 0.4, the load change characteristic has a weight of 0.3, and the fluctuation characteristic has a weight of 0.3.

[0111] The voltage adaptive state characteristic is 0.45 × 0.4 = 0.18;

[0112] The average value of the load change characteristic adaptive state characteristic is 0.35 × 0.3 = 0.105;

[0113] The standard deviation adaptive state characteristic is 0.60 × 0.3 = 0.18;

[0114] The adaptive state characteristic of the fluctuation amount is 0.5 × 0.3 = 0.15;

[0115] The adaptive state characteristic of the peak frequency is 0.4 × 0.3 = 0.12;

[0116] The zero-crossing frequency adaptive state characteristic is 0 × 0.3 = 0;

[0117] Finally, the weighted adaptive state features are summarized to form an adaptive state feature vector of [0.18, 0.105, 0.18, 0.15, 0.12, 0].

[0118] Step S200: Based on the state feature vector, calculate the coupling strength of the time window, and based on the coupling strength of the historical records, calculate the coupling feature parameters;

[0119] The calculation of coupling characteristic parameters in step S200 includes:

[0120] In the adaptive state feature vector, voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features are extracted according to time windows. The changes of voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features within the time window are calculated respectively. The change is the difference between the maximum and minimum values ​​of the adaptive state features. The mean and standard deviation of the changes in the historical time windows are calculated. The standardized change is calculated as A = (BC) / D, where B represents the change in the current time window, C represents the mean of the historical time windows, and D represents the standard deviation of the historical time windows.

[0121] Summarize the standardized changes in voltage, load, and fluctuation within the time window, and calculate the coupling strength K = a × A1 × A2 + b × A3 for each time window, where A1, A2, and A3 represent the standardized changes in voltage, load, and fluctuation, respectively, a represents the weights of voltage and load, and b represents the weights of fluctuation.

[0122] Within a continuous time window, the coupling strength of each time window is obtained, the difference in coupling strength between adjacent time windows is calculated, and then divided by the time length of the time window to calculate the coupling change rate.

[0123] Historical records under stable operating conditions are obtained. A stable operating condition is defined as one that simultaneously meets the following conditions: the average value of voltage variation characteristics within each time window does not exceed a preset threshold; the standard deviation of load variation characteristics does not exceed a preset threshold; the amount of change in fluctuation characteristics does not exceed a preset threshold, and the rate of change of peak frequency and zero-crossing frequency does not exceed a preset threshold; the absolute difference in the coupling rate of change between adjacent time windows does not exceed a preset threshold. When all of the above conditions are met within N consecutive time windows, the operating state corresponding to that time period is determined as a stable operating condition.

[0124] The coupling strength of the historical records is collected, and the average value of the coupling strength is calculated. The average value, the coupling strength of the current time window, and the coupling change rate are weighted and summed to calculate the coupling characteristic parameters.

[0125] For example, to calculate the voltage change within a time window, the voltage change in each time window is calculated as the difference between adjacent sampling times, resulting in the following result: (0.0 - 0.1) = 100² - 1000 = 2V

[0126] (0.1-0.2)=1001-1002=-1V;

[0127] (0.2-0.3)=1005-1001=4V;

[0128] (0.3-0.4)=1003-1005=-2V;

[0129] (0.4-0.5)=1004-1003=1V;

[0130] Within each time window, the maximum voltage change is 4V, the minimum voltage change is -2V, and the total change is 4 - (-2) = 6V.

[0131] For example, the mean value within the historical window is (2+(-1)+4+(-2)+1) / 5=0.8V;

[0132] The standard deviation is calculated to be 2.15V;

[0133] To calculate the standardized change, assuming the change in the current window is 6V, the mean is 0.8V, and the standard deviation is 2.15V, then the standardized change is (6-0.8) / 2.15≈2.47;

[0134] For the load characteristics, the load change within each time window is calculated as follows: (0.0-0.1)=202-200=2kW;

[0135] (0.1-0.2)=205-202=3kW;

[0136] (0.2-0.3)=203-205=-2kW;

[0137] (0.3-0.4)=204-203=1kW;

[0138] (0.4-0.5)=206-204=2kW;

[0139] The maximum load change is 3kW, the minimum load change is -2kW, and the total change is 3 - (-2) = 5kW.

[0140] The average load during the historical time window is (2+3+(-2)+1+2) / 5=1.2kW;

[0141] The standard deviation is 2.07 kW, and the standardized variation is (5-1.2) / 2.07≈1.84;

[0142] The maximum and minimum voltage values ​​within the current window are 1005V and 1000V respectively, with a fluctuation of 1005-1000=5V. The standardized change of the fluctuation is A=(5-0) / 2.15≈2.33.

[0143] The weights for voltage and load are set to a=0.4, and the weight for fluctuation is b=0.3. The coupling strength is 0.4×2.47×1.84+0.3×2.33≈1.82+0.70=2.52.

[0144] Step S300: Based on the coupling characteristic parameters of multiple consecutive time windows, the system operation mode is analyzed and the operation mode is identified;

[0145] The step S300, which identifies the operating mode, includes:

[0146] The coupling feature parameters of multiple consecutive time windows are arranged in chronological order to obtain a set of coupling feature parameters. Based on the set of coupling feature parameters, the slope and magnitude of the changes in the coupling feature parameters are calculated. The slope calculation process involves extracting the coupling feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupling feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupling feature parameters per unit time. This rate of change is used as the slope. The magnitude calculation process involves calculating the absolute value of the difference between the coupling feature parameters between adjacent time windows. This absolute value is used as the magnitude.

[0147] In a high-voltage capacitor power extraction system, an operating mode is preset based on the time evolution sequence of coupling characteristic parameters and the adaptive state characteristics of the current time window. The operating modes include an accelerated change mode, an oscillation evolution mode, and a critical instability mode.

[0148] The accelerated change mode is defined as follows: when the slope of the change of the coupled feature parameters is greater than a preset threshold, and at least two operating parameters in the adaptive state features exceed the judgment range of the stable operating condition, the system is determined to enter the accelerated change mode.

[0149] The oscillation evolution mode is defined as follows: when the change amplitude of the coupling characteristic parameter increases compared with the coupling characteristic parameter under the previous stable condition and the change amount of the fluctuation characteristic is greater than a preset threshold, the system is determined to enter the oscillation evolution mode.

[0150] The critical instability mode is defined as the system being in a critical instability mode when the difference between the coupling characteristic parameters of the system and the coupling characteristic parameters of the stable condition exceeds a preset threshold and the rate of change exceeds the preset threshold.

[0151] For example, in this embodiment, it is assumed that the coupling characteristic parameters obtained from eight consecutive time windows are arranged in chronological order as follows: Time window 1 (0s-5s) is 2.15; Time window 2 (5s-10s) is 2.32; Time window 3 (10s-15s) is 2.58; Time window 4 (15s-20s) is 3.05; Time window 5 (20s-25s) is 3.74; Time window 6 (25s-30s) is 3.18; Time window 7 (30s-35s) is 4.62; Time window 8 (35s-40s) is 5.96.

[0152] After arranging the coupling feature parameters of the above consecutive time windows in chronological order, the set of coupling feature parameters is obtained: {2.15, 2.32, 2.58, 3.05, 3.74, 3.18, 4.62, 5.96};

[0153] The slope and magnitude of the change are calculated based on the set of coupled feature parameters. Since the interval between adjacent time windows is 5 seconds, the slope is the difference between the coupled feature parameters of two consecutive windows divided by 5 seconds. The magnitude of the change is the absolute value of the difference between the coupled feature parameters of adjacent time windows.

[0154] The slope of the change in the second time window relative to the first time window is (2.32 - 2.15) / 5 = 0.034, and the magnitude of the change is |2.32 - 2.15| = 0.17;

[0155] The slope of the change in the third time window relative to the second time window is (2.58 - 2.32 / 5 = 0.052, with a change range of 0.26;

[0156] The slope of the change in the fourth time window relative to the third time window is (3.05 - 2.58) / 5 = 0.094, and the change range is 0.47.

[0157] The slope of the change in the 5th time window relative to the 4th time window is (3.74 - 3.05) / 5 = 0.138, and the change range is 0.69.

[0158] The slope of the change in the 6th time window relative to the 5th time window is (3.18 - 3.74 / 5 = -0.112, with a change range of 0.56);

[0159] The slope of the change in the 7th time window relative to the 6th time window is (4.62 - 3.18) / 5 = 0.288, and the change range is 1.44.

[0160] The slope of the change in the 8th time window relative to the 7th time window is (5.96 - 4.62) / 5 = 0.268, and the change range is 1.34.

[0161] The preset thresholds for slope change are 0.12, fluctuation characteristic change, critical offset, and critical rate are 1.20, respectively. The average value of coupling characteristic parameters under stable conditions is set to 2.30.

[0162] When the system reaches the 5th time window, the current coupling characteristic parameter is 3.74, and the slope of change is 0.138, which is greater than the preset threshold of 0.12. At the same time, voltage deviation and load fluctuation are detected to exceed the limit within this time window, and the fluctuation characteristic is also close to the upper limit. That is, at least two of the three adaptive state characteristics are in the range of stable operating conditions. Therefore, it is determined that the system has entered the accelerated change mode. This mode indicates that the energy coupling relationship inside the system is rapidly enhanced and the power-taking load is in a state of continuous increase.

[0163] When the system reached the 6th time window, the coupling characteristic parameter decreased from 3.74 to 3.18, a change of 0.56. Subsequently, in the 7th time window, it rapidly increased to 4.62, a change of 1.44. Simultaneously, the fluctuation characteristic change during this phase was 1.36, exceeding the preset threshold of 1.20. Furthermore, the current coupling characteristic parameter was significantly higher than the average value of 2.30 under the previous stable operating condition. Therefore, it was determined that the system had entered an oscillation evolution mode. This mode indicates that the system has exhibited enhanced periodic fluctuations, with voltage and load in an alternating state.

[0164] When the system reaches the 8th time window, the current coupling characteristic parameter is 5.96, and the difference between it and the average value of 2.30 under stable operating conditions is 5.96-2.30=3.66, which exceeds the preset critical offset threshold of 2.50. At the same time, the slope of change is 0.268, which is greater than the critical rate threshold of 0.25. Therefore, it is determined that the system has entered the critical instability mode. This mode indicates that the high-voltage capacitor power supply system is close to the instability boundary and there is a risk of abnormal output, device overload or voltage collapse. It is necessary to immediately execute load limiting, alarm or protection actions.

[0165] Step S400: Calculate the comprehensive safety level based on the operating mode, operating parameters and coupling characteristic parameters, generate a hierarchical control strategy based on the comprehensive safety level, calculate the control lead based on the difference change, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety level, and make advance corrections based on the control lead.

[0166] Step S400 includes:

[0167] Within the current time window, the difference between the coupling characteristic parameter and the critical instability threshold is obtained. The current coupling characteristic parameter is subtracted from the critical instability threshold to calculate the coupling margin. The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting the transition stage data during the transition from stable to unstable mode, performing statistical analysis based on the transition stage data, calculating the mean g and standard deviation h of the coupling characteristic parameter, and calculating the critical instability threshold L = g + h × j, where j represents the preset safety factor.

[0168] Obtain the adaptive state feature vector of the current time window, calculate the average adaptive state feature vector under stable operating conditions, calculate the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, normalize the difference, and sum the normalized difference by weight to obtain the state deviation.

[0169] Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the values ​​of coupling margin, state deviation, and operating mode, the values ​​are weighted and summed according to preset weights to calculate the comprehensive safety level.

[0170] Based on the changing trend of voltage parameters within a continuous time window, the difference change of voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are calculated and normalized. The normalized difference change, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are then weighted and summed to calculate the control lead.

[0171] The system obtains the overall safety level from historical records, calculates the mean n and standard deviation m of the overall safety level, and divides the system operating state into three safety intervals based on the range of the overall safety level. The first safety interval is [nm×u1, n+m×u1], the second safety interval is [nm×u2, nm×u1)∪(n+m×u1, n+m×u2], and the third safety interval is [-∞, nm×u2)∪(n+m×u2, +∞]. u1 and u2 are adjustable coefficients, and are matched with corresponding hierarchical control strategies. The hierarchical control strategy includes performing gradual adjustment control when the overall safety level is in the first safety interval.

[0172] When the overall safety level is in the second safety range, output limit control is executed;

[0173] When the overall safety level S is in the third safety range, the cutoff control is executed;

[0174] When executing the hierarchical control strategy, the triggering time of the control command is advanced according to the control advance amount;

[0175] For example, by extracting sample data from the transition phase of the system from stable mode to unstable mode, the mean of the coupling characteristic parameters in this transition phase is 3.20, the standard deviation is 0.85, and the safety factor is set to 1.2. Then the critical instability threshold is 3.20 + 0.85 × 1.2 = 4.22.

[0176] In the current time window, such as the 8th time window (35s-40s), the detected coupling characteristic parameter is 5.96, and the coupling margin is 4.22-5.96=-1.74. Since the coupling margin is negative, it indicates that the current system has exceeded the critical instability threshold and is in a high-risk state.

[0177] Obtain the adaptive state feature vector for the current time window. Assume that the vector includes three parameters: voltage deviation, load fluctuation, and frequency disturbance. Calculate the differences as 0.35, 0.72, and 0.42, respectively. Normalize the above differences to obtain 0.70, 0.72, and 0.70. Set the weights of each parameter to 0.3, 0.4, and 0.3, respectively. Then the state deviation is 0.70×0.3+0.72×0.4+0.70×0.3=0.706.

[0178] Assuming the stable mode is assigned a value of 0.2, the oscillation evolution mode is assigned a value of 0.6, and the critical instability mode is assigned a value of 1.0, and the current time window is determined to be the critical instability mode, therefore the mode is assigned a value of 1.0;

[0179] After normalizing the values ​​of coupling margin, state deviation, and operating mode, the coupling margin is 0.95, the state deviation is 0.71, and the mode value is 1.00. With weights of 0.4, 0.3, and 0.3 respectively, the overall safety factor is 0.95×0.4+0.71×0.3+1.00×0.3=0.893.

[0180] Subsequently, an analysis was conducted based on the changing trend of voltage parameters within a continuous time window. The change in voltage difference was 0.65, the change rate of coupling characteristic parameters was 0.89, and the change rate of comprehensive safety was 0.60. Assuming the weights of the three factors are 0.4, 0.4, and 0.2 respectively, the control lead time is 0.65×0.4+0.89×0.4+0.60×0.2=0.736.

[0181] Based on the comprehensive safety data of historical records, the mean was 0.52 and the standard deviation was 0.18. With coefficients u1=1.0 and u2=2.0, the safety intervals are divided as follows: the first safety interval is [0.34, 0.70]; the second safety interval is [0.16, 0.34)∪(0.70, 0.88]; and the third safety interval is (-∞, 0.16)∪(0.88, +∞).

[0182] The current overall safety level is 0.893, falling into the third safety range. Therefore, the system executes the cut-off control strategy, which immediately disconnects the power output and triggers the protection mechanism. At the same time, when executing the cut-off control strategy, the triggering time of the control command is advanced according to the control advance amount of 0.736. For example, if the original trigger delay is 1 second, the corrected triggering time is 1-0.736=0.264 seconds.

[0183] To better implement the above method, a high-voltage capacitor power extraction system based on safety monitoring is also proposed. The system includes an adaptive state feature vector module, a coupling feature parameter module, an operation mode module, and a correction module.

[0184] Adaptive State Feature Vector Module: At a preset sampling time in the high-voltage capacitor power extraction system, the operating parameters of the high-voltage capacitor power extraction system are collected, and continuous time series data is constructed by combining the timestamps. A preset time window is set, and based on the continuous time series data, the change features of the operating parameters within the time window are extracted to construct an adaptive state feature vector.

[0185] The adaptive state feature vector module includes a changing feature unit and a unit for constructing the adaptive state feature vector:

[0186] The variation feature unit includes voltage variation features, load variation features, fluctuation features, and time features. The voltage variation feature is calculated based on voltage parameters in continuous time series data, calculating the voltage difference between adjacent sampling times in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and an average value is calculated, which is then used as the voltage variation feature. The load variation feature is calculated based on load parameters in continuous time series data, calculating the load difference between adjacent sampling times in chronological order. The load change is obtained by taking the difference. Within a preset time window, the load change is statistically analyzed, and the average and standard deviation are calculated. The average and standard deviation are used together as the load change characteristic. The fluctuation characteristic calculation process is as follows: within the preset time window, the maximum and minimum values ​​of the voltage parameter are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic.

[0187] Constructing adaptive state feature vector units: Normalize different types of change features respectively. According to the preset weights of each change feature, calculate the weighted sum of each normalized change feature and its corresponding weight to obtain the adaptive state feature of each change feature. Summarize the adaptive state features of each change feature to construct an adaptive state feature vector.

[0188] Coupling feature parameter module: Based on the state feature vector, calculate the coupling strength of the time window, and based on the coupling strength of the historical records, calculate the coupling feature parameters;

[0189] The coupling characteristic parameter module includes a unit for calculating coupling strength and a unit for calculating coupling characteristic parameters.

[0190] Calculate the coupling strength unit: In the adaptive state feature vector, extract the voltage adaptive state feature, load adaptive state feature, and fluctuation adaptive state feature according to the time window. Calculate the change of the voltage adaptive state feature, load adaptive state feature, and fluctuation adaptive state feature within the time window. The change is the difference between the maximum and minimum values ​​in the adaptive state feature. Calculate the mean and standard deviation of the historical time window changes, calculate the standardized change, summarize the standardized changes of voltage, load, and fluctuation within the time window, and calculate the coupling strength of each time window.

[0191] Calculate coupling characteristic parameters: Within a continuous time window, acquire the coupling strength of each time window, calculate the difference in coupling strength between adjacent time windows, and divide it by the time length of the time window to calculate the coupling change rate; acquire historical records under stable operating conditions. A stable operating condition is determined when the following conditions are met simultaneously: the average change of voltage change characteristics within each time window does not exceed a preset threshold; the standard deviation of load change characteristics does not exceed a preset threshold; the change in fluctuation characteristics does not exceed a preset threshold, and the rate of change of peak frequency and zero-crossing frequency does not exceed a preset threshold; the absolute difference in coupling change rate between adjacent time windows does not exceed a preset threshold; when the above conditions are met within N consecutive time windows, the operating state corresponding to the time period is determined as a stable operating condition. Collect the coupling strength of the historical records and calculate the average value of the coupling strength. Calculate the coupling characteristic parameters by weighted summing of the average value, the coupling strength of the current time window, and the coupling change rate.

[0192] Operating mode module: Based on the coupled characteristic parameters of multiple consecutive time windows, the operating mode of the system is identified.

[0193] The operation mode module includes a coupled feature parameter change unit and a preset operation mode unit:

[0194] Coupled feature parameter change unit: Coupled feature parameters from multiple consecutive time windows are arranged in chronological order to obtain a set of coupled feature parameters. Based on this set, the slope and magnitude of the changes in the coupled feature parameters are calculated. The slope calculation process involves extracting the coupled feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupled feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupled feature parameters per unit time. This rate of change is used as the slope. The magnitude calculation process involves calculating the absolute value of the difference between the coupled feature parameters between adjacent time windows, and this absolute value is used as the magnitude.

[0195] Preset operating mode unit: In the high-voltage capacitor power supply system, an operating mode is preset based on the time evolution sequence of the coupling characteristic parameters and the adaptive state characteristics of the current time window. The operating modes include accelerated change mode, oscillation evolution mode and critical instability mode.

[0196] Correction module: Calculates the comprehensive safety level based on the operating mode, operating parameters and coupling characteristic parameters, generates a hierarchical control strategy based on the comprehensive safety level, calculates the control lead based on the difference change, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety level, and performs advance correction based on the control lead.

[0197] The correction module includes a unit for calculating the overall safety factor and a unit for calculating the control lead time.

[0198] The comprehensive safety unit calculates the following: Within the current time window, it obtains the difference between the coupling characteristic parameter and the critical instability threshold, subtracts the current coupling characteristic parameter from the critical instability threshold, and calculates the coupling margin. The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting transition stage data during the transition from stable to unstable modes, performing statistical analysis based on the transition stage data, calculating the mean and standard deviation of the coupling characteristic parameter, calculating the critical instability threshold, obtaining the adaptive state feature vector of the current time window, calculating the average adaptive state feature vector corresponding to the stable operating condition, calculating the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, and normalizing the difference. The normalized difference is then weighted and summed to obtain the state deviation. Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the coupling margin, state deviation, and operating mode assignments, they are weighted and summed according to preset weights to calculate the comprehensive safety.

[0199] The control lead calculation unit calculates the difference in voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety factor based on the changing trend of voltage parameters within a continuous time window. After normalization, the normalized difference in voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety factor are weighted and summed to obtain the control lead.

[0200] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A high-voltage capacitor power extraction method based on safety monitoring, characterized in that, The methods include: Step S100: At a preset sampling time in the high-voltage capacitor power supply system, the operating parameters of the high-voltage capacitor power supply system are collected. The operating parameters include voltage parameters, current parameters and load parameters. Continuous time series data is constructed by combining the data with timestamps. A preset time window is established. Based on the continuous time series data, the change characteristics of the operating parameters within the time window are extracted. The change characteristics include voltage change characteristics, load change characteristics, fluctuation characteristics and time characteristics. An adaptive state feature vector is constructed. Step S200: Based on the adaptive state feature vector, calculate the coupling strength of the time window, and based on the coupling strength of the historical records, calculate the coupling feature parameters; The calculation of coupling characteristic parameters in step S200 includes the following steps: In the adaptive state feature vector, voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features are extracted according to time windows. The changes of voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features within the time window are calculated respectively. The change is the difference between the maximum and minimum values ​​of the adaptive state features. The mean and standard deviation of the changes in the historical time windows are calculated. The standardized change is calculated as A = (BC) / D, where B represents the change in the current time window, C represents the mean of the historical time windows, and D represents the standard deviation of the historical time windows. Summarize the standardized changes in voltage, load, and fluctuation within the time window, and calculate the coupling strength K = a × A1 × A2 + b × A3 for each time window, where A1, A2, and A3 represent the standardized changes in voltage, load, and fluctuation, respectively, a represents the weights of voltage and load, and b represents the weights of fluctuation. Within a continuous time window, the coupling strength of each time window is obtained, the difference in coupling strength between adjacent time windows is calculated, and then divided by the time length of the time window to calculate the coupling change rate. Obtain historical records under stable operating conditions, collect the coupling strength of the historical records, and calculate the average value of the coupling strength. Then, perform a weighted summation of the average value, the coupling strength of the current time window, and the coupling change rate to calculate the coupling characteristic parameters. Step S300: Identify the system operation mode based on the coupling characteristic parameters of multiple consecutive time windows; The step S300 of identifying the operating mode includes the following steps: The coupling feature parameters of multiple consecutive time windows are arranged in chronological order to obtain a set of coupling feature parameters. Based on the set of coupling feature parameters, the slope and magnitude of the change of the coupling feature parameters are calculated. In a high-voltage capacitor power extraction system, an operating mode is preset based on the time evolution sequence of coupling characteristic parameters and the adaptive state characteristics of the current time window. The operating modes include an accelerated change mode, an oscillation evolution mode, and a critical instability mode. Step S400: Calculate the comprehensive safety level based on the operating mode, operating parameters and coupling characteristic parameters, generate a hierarchical control strategy based on the comprehensive safety level, calculate the control lead based on the difference change, the rate of change of coupling characteristic parameters and the rate of change of comprehensive safety level, and make advance corrections based on the control lead. Step S400 includes the following steps: Within the current time window, the difference between the coupling characteristic parameter and the critical instability threshold is obtained, and the current coupling characteristic parameter is subtracted from the critical instability threshold to calculate the coupling margin. Obtain the adaptive state feature vector of the current time window, calculate the average adaptive state feature vector under stable operating conditions, calculate the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, normalize the difference, and sum the normalized difference by weight to obtain the state deviation. Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the values ​​of coupling margin, state deviation, and operating mode, the values ​​are weighted and summed according to preset weights to calculate the comprehensive safety degree.

2. The high-voltage capacitor power extraction method based on safety monitoring according to claim 1, characterized in that, The construction of the adaptive state feature vector in step S100 includes the following steps: The voltage change feature calculation process is based on the voltage parameters in the continuous time series data. The voltage difference between adjacent sampling times is calculated in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and the average value is calculated. The average value is used as the voltage change feature. The load change characteristic calculation process is based on the load parameters in the continuous time series data. The load difference between adjacent sampling times is calculated in chronological order to obtain the load change amount. Within a preset time window, the load change amount is statistically analyzed, and the average value and standard deviation are calculated. The average value and standard deviation are used together as the load change characteristic. The fluctuation characteristic calculation process is as follows: within a preset time window, the maximum and minimum values ​​of the voltage parameter within the preset time window are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation amount. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the time length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation amount, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic. Normalize the different types of change features separately. According to the preset weights of each change feature, calculate the weighted sum of each normalized change feature and its corresponding weight to obtain the adaptive state feature of each change feature. Summarize the adaptive state features of each change feature to construct an adaptive state feature vector.

3. The high-voltage capacitor power extraction method based on safety monitoring according to claim 2, characterized in that, Step S200 includes the following steps: The stable operating condition is determined when the following conditions are met simultaneously: the average value of the voltage change characteristics within each time window does not exceed a preset threshold; the standard deviation of the load change characteristics does not exceed a preset threshold; the fluctuation amount of the fluctuation characteristics does not exceed a preset threshold, and the rate of change of the peak frequency and the zero-crossing frequency does not exceed a preset threshold; the absolute difference of the coupling rate of change between adjacent time windows does not exceed a preset threshold; when the above conditions are met in N consecutive time windows, the operating state corresponding to the time period is determined as a stable operating condition.

4. The high-voltage capacitor power extraction method based on safety monitoring according to claim 2, characterized in that, Step S300 includes the following steps: The slope calculation process involves extracting the coupling feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupling feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupling feature parameters per unit time, using the rate of change as the slope. The amplitude calculation process involves calculating the absolute value of the difference between the coupling feature parameters between adjacent time windows, using the absolute value as the amplitude. The accelerated change mode is defined as follows: when the slope of the change of the coupled feature parameters is greater than a preset threshold, and at least two operating parameters in the adaptive state features exceed the judgment range of the stable operating condition, the system is determined to enter the accelerated change mode. The oscillation evolution mode is defined as follows: when the change amplitude of the coupling characteristic parameter increases compared with the coupling characteristic parameter under the previous stable condition and the fluctuation amount of the fluctuation characteristic is greater than a preset threshold, the system is determined to enter the oscillation evolution mode. The critical instability mode is defined as the system being in a critical instability mode when the difference between the coupling characteristic parameters of the system and the coupling characteristic parameters of the stable operating condition exceeds a preset threshold and the rate of change exceeds the preset threshold.

5. The high-voltage capacitor power extraction method based on safety monitoring according to claim 1, characterized in that, Step S400 includes the following steps: The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting transition stage data during the transition from stable to unstable mode, performing statistical analysis based on the transition stage data, calculating the mean g and standard deviation h of the coupling characteristic parameters, and calculating the critical instability threshold L = g + h × j, where j represents the preset safety factor. Based on the changing trend of voltage parameters within a continuous time window, the difference change of voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are calculated and normalized. The normalized difference change, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are then weighted and summed to calculate the control lead. The system obtains the overall safety level from historical records, calculates the mean n and standard deviation m of the overall safety level, and divides the system operating state into three safety intervals based on the range of the overall safety level. The first safety interval is [nm×u1, n+m×u1], the second safety interval is [nm×u2, nm×u1)∪(n+m×u1, n+m×u2], and the third safety interval is [-∞, nm×u2)∪(n+m×u2, +∞]. u1 and u2 are adjustable coefficients, and are matched with corresponding hierarchical control strategies. The hierarchical control strategy includes performing gradual adjustment control when the overall safety level is in the first safety interval. When the overall safety level is in the second safety range, output limit control is executed; When the overall safety level S is in the third safety range, the cutoff control is executed; When executing the hierarchical control strategy, the triggering time of the control command is adjusted in advance according to the control lead amount.

6. A high-voltage capacitor power extraction system based on safety monitoring, used to implement the high-voltage capacitor power extraction method based on safety monitoring as described in any one of claims 1-5, characterized in that, The system includes an adaptive state feature vector module, a coupled feature parameter module, an operating mode module, and a correction module; The adaptive state feature vector module: At a preset sampling time in the high-voltage capacitor power supply system, the operating parameters of the high-voltage capacitor power supply system are collected. The operating parameters include voltage parameters, current parameters, and load parameters. Continuous time series data is constructed by combining the parameters with timestamps. A preset time window is set. Based on the continuous time series data, the change features of the operating parameters within the time window are extracted. The change features include voltage change features, load change features, fluctuation features, and time features. An adaptive state feature vector is constructed. The coupling feature parameter module calculates the coupling strength of the time window based on the adaptive state feature vector, and calculates the coupling feature parameters based on the coupling strength of historical records. The operation mode module identifies the system operation mode based on the coupled feature parameters of multiple consecutive time windows. The correction module calculates the overall safety level based on the operating mode, operating parameters, and coupling characteristic parameters, generates a hierarchical control strategy based on the overall safety level, calculates the control lead based on the difference change, the rate of change of coupling characteristic parameters, and the rate of change of overall safety level, and performs advance correction based on the control lead.

7. The high-voltage capacitor power extraction system based on safety monitoring according to claim 6, characterized in that, The adaptive state feature vector module includes a changing feature unit and an adaptive state feature vector construction unit: The change feature unit comprises voltage change features, load change features, fluctuation features, and time features. The voltage change feature is calculated based on voltage parameters in continuous time series data, calculating the voltage difference between adjacent sampling times in chronological order. Based on the voltage difference corresponding to each sampling time segment, the change in difference between adjacent sampling time segments is calculated. The change in difference within a preset time window is summarized and an average value is calculated, which is then used as the voltage change feature. The load change feature is calculated based on load parameters in continuous time series data, calculating the load difference between adjacent sampling times in chronological order to obtain the load change. Within a preset time window, the load change is statistically analyzed, and an average value and standard deviation are calculated. The average value and standard deviation are then used together as the load change feature. The fluctuation characteristic calculation process is as follows: within a preset time window, the maximum and minimum values ​​of the voltage parameter within the preset time window are obtained, and the difference between the maximum and minimum values ​​is calculated as the fluctuation amount. The number of peaks and zero crossings of the voltage parameter within the preset time window are counted. The number of peaks and zero crossings are divided by the time length of the preset time window to calculate the peak frequency and zero crossing frequency. The fluctuation amount, peak frequency, and zero crossing frequency are used together as the fluctuation characteristic. The adaptive state feature vector unit is constructed by: normalizing different types of change features respectively; calculating the weight of each normalized change feature according to the preset weight of each change feature to obtain the adaptive state feature of each change feature; and summing the adaptive state features of each change feature to construct the adaptive state feature vector.

8. The high-voltage capacitor power extraction system based on safety monitoring according to claim 7, characterized in that, The coupling characteristic parameter module includes a unit for calculating coupling strength and a unit for calculating coupling characteristic parameters: The unit for calculating coupling strength involves extracting voltage adaptive state features, load adaptive state features, and fluctuation adaptive state features from the adaptive state feature vector according to time windows. It then calculates the changes in these features within the time window, where each change is the difference between the maximum and minimum values ​​of the adaptive state features. The unit also calculates the mean and standard deviation of the changes within historical time windows, calculates the standardized changes, summarizes the standardized changes in voltage, load, and fluctuation within the time window, and calculates the coupling strength for each time window. The unit for calculating coupling characteristic parameters: within a continuous time window, acquires the coupling strength of each time window, calculates the difference in coupling strength between adjacent time windows, and divides it by the time length of the time window to calculate the coupling change rate; acquires historical records under stable operating conditions, wherein a stable operating condition is determined to be a stable operating condition when the following conditions are met simultaneously: the change amplitude of the average value of the voltage change characteristic within each time window does not exceed a preset threshold; the standard deviation of the load change characteristic does not exceed a preset threshold. The fluctuation amount in the fluctuation characteristics does not exceed a preset threshold, and the rate of change of the peak frequency and the zero-crossing frequency does not exceed a preset threshold; the absolute difference of the coupling change rate between adjacent time windows does not exceed a preset threshold; when the above conditions are met in N consecutive time windows, the operating state corresponding to the time period is determined as a stable operating condition, the coupling strength of the historical records is collected, and the average value of the coupling strength is calculated. The average value, the coupling strength of the current time window, and the coupling change rate are weighted and summed to calculate the coupling characteristic parameters.

9. The high-voltage capacitor power extraction system based on safety monitoring according to claim 6, characterized in that, The operation mode module includes a coupled feature parameter change unit and a preset operation mode unit: The coupling feature parameter change unit: Arranges the coupling feature parameters of multiple consecutive time windows in chronological order to obtain a set of coupling feature parameters. Based on this set, calculates the slope and magnitude of the change of the coupling feature parameters. The slope calculation process involves extracting the coupling feature parameter values ​​corresponding to adjacent time windows, calculating the difference between the coupling feature parameters between two consecutive time windows, and combining this with the corresponding time interval to obtain the rate of change of the coupling feature parameters per unit time. This rate of change is used as the slope. The magnitude calculation process involves calculating the absolute value of the difference between the coupling feature parameters between adjacent time windows, and using this absolute value as the magnitude. The preset operation mode unit: In the high-voltage capacitor power supply system, an operation mode is preset based on the time evolution sequence of the coupling characteristic parameters and the adaptive state characteristics of the current time window. The operation mode includes an accelerated change mode, an oscillation evolution mode, and a critical instability mode.

10. The high-voltage capacitor power extraction system based on safety monitoring according to claim 6, characterized in that, The correction module includes a unit for calculating the overall safety factor and a unit for calculating the control lead time. The comprehensive safety calculation unit: within the current time window, obtains the difference between the coupling characteristic parameter and the critical instability threshold, subtracts the current coupling characteristic parameter from the critical instability threshold to calculate the coupling margin. The critical instability threshold is obtained by collecting coupling characteristic parameter data under various operating modes, labeling the data as stable and unstable modes, extracting transition stage data during the transition from stable to unstable modes, performing statistical analysis based on the transition stage data, calculating the mean and standard deviation of the coupling characteristic parameter, calculating the critical instability threshold, obtaining the adaptive state feature vector of the current time window, calculating the average adaptive state feature vector corresponding to the stable operating condition, calculating the difference between the adaptive state feature of each operating parameter and the corresponding average adaptive state feature, normalizing the difference, and weighting and summing the normalized difference to obtain the state deviation. Different operating modes are assigned values ​​to obtain the operating mode of the current time window. After normalizing the coupling margin, state deviation, and operating mode assignments, they are weighted and summed according to preset weights to calculate the comprehensive safety. The control advance calculation unit calculates the difference change of voltage parameters, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety based on the changing trend of voltage parameters within a continuous time window, and performs normalization processing. The normalized difference change, the rate of change of coupled characteristic parameters, and the rate of change of comprehensive safety are then weighted and summed to calculate the control advance.

Citation Information

Patent Citations

  • Power station voltage-frequency coupling characteristic evaluation method

    CN121561386A

  • New energy power station multistage safety early warning method based on abnormal mode recognition

    CN121961224A