Capacitance value online monitoring method and system based on thin film capacitor self-healing characteristic perception

Abstract

The invention is suitable for the technical field of circuit monitoring, and provides a capacitance value online monitoring method and system based on thin film capacitor self-healing feature perception, and the method comprises the steps: synchronously collecting a voltage signal and a current signal of a to-be-detected thin film capacitor based on a preset sampling period, and constructing a voltage sequence and a current sequence with a unified time sequence; determining a candidate transient time window based on a preset sampling period and the voltage sequence; in the candidate transient time window, a time sequence of a self-healing event is determined based on the voltage sequence and the current sequence, and the self-healing event comprises an arcing stage, an arc quenching stage and a charging recovery stage; and based on the initial voltage, the final voltage and the effective charging current in the charging recovery stage, determining an effective capacitance estimation value of the to-be-tested thin-film capacitor. According to the technical scheme provided by the invention, the self-healing behavior of the thin-film capacitor is accurately identified, so that online monitoring of the capacitance value of the thin-film capacitor is realized, and the capacitance value monitoring error is effectively reduced.

Description

Technical Field

[0001] This application belongs to the field of circuit monitoring technology, and in particular relates to an online capacitance monitoring method and system based on the sensing of the self-healing characteristics of thin-film capacitors. Background Technology

[0002] Currently, flexible DC transmission projects, as one of the core technical solutions for large-capacity, long-distance power transmission in new power systems, are increasingly being applied in inter-regional interconnection, offshore wind power grid connection, and large power grid support scenarios. Modular multilevel converters (MMCs) are a key component of flexible DC transmission systems, and their numerous internal thin-film capacitor submodules play crucial roles in energy support, voltage stability, and harmonic suppression, making them critical components for ensuring the safe and stable operation of converter valves. Thin-film capacitors are numerous and widely distributed, accounting for a very high proportion of converter valve equipment, and their health directly affects the reliability, economy, and safety of the DC transmission system.

[0003] However, during the long-term operation of flexible DC systems, thin-film capacitors are subjected to a combination of stresses caused by high-frequency pulse width modulation (PWM), including current ripple, pulsating voltage, harmonic distortion, and transient overvoltage. This can easily induce localized dielectric breakdown, leading to metallization layer ablation, a phenomenon known as "self-healing." Extensive operational experience shows that when thin-film capacitors exhibit phenomena such as metallization layer perforation, reduced effective area, and excessive arc energy during the self-healing process, their capacitance value will continuously decrease over time. If this capacitance decrease is not detected in time, it will lead to deterioration of submodule voltage balance, increased voltage ripple, and excessive stress on converter valve components. In severe cases, it may cause submodule shutdown or even converter valve failure, resulting in significant hidden dangers such as DC system power fluctuations and reduced grid stability.

[0004] Currently, the protection of thin-film capacitors largely relies on offline detection after power outages, indirect monitoring of the operating environment, and estimation methods based on equivalent parameters. However, these methods generally suffer from drawbacks such as response lag, significant susceptibility to harmonic and current ripple interference, lack of awareness of self-healing transients, and inability to track capacitance changes online. Particularly in thin-film capacitors, the duration of the self-healing pulse is typically only tens of microseconds, and the signal is superimposed with numerous switching harmonics and arc pulses, making it difficult for traditional capacitance detection methods to effectively extract direct evidence of capacitance changes.

[0005] Therefore, accurately identifying the self-healing behavior of film capacitors to reduce capacitance monitoring errors and achieve online capacitance monitoring has become an urgent problem to be solved. Summary of the Invention

[0006] This application provides a method and system for online capacitance monitoring based on the sensing of the self-healing characteristics of thin-film capacitors, aiming to accurately identify the self-healing behavior of thin-film capacitors, so as to reduce capacitance monitoring errors and realize online capacitance monitoring.

[0007] In a first aspect, embodiments of this application provide a method for online capacitance monitoring based on the self-healing characteristics of thin-film capacitors, the method comprising: S1, based on a preset sampling period, synchronously acquire the voltage and current signals of the thin-film capacitor under test, and construct a voltage sequence and current sequence with a unified time series; S2, Based on the preset sampling period and the voltage sequence, determine the candidate transient time window; S4, within the candidate transient time window, based on the voltage sequence and the current sequence, determine the time sequence of the self-healing event, the self-healing event including the arcing stage, the arc extinguishing stage and the recharging stage; S5. Based on the initial voltage, termination voltage, and effective charging current of the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

[0008] In one possible implementation, step S2 includes: S21, Calculate the first-order voltage difference sequence based on the preset sampling period and the voltage sequence; S22, in the first-order voltage difference sequence, determine the start time and end time corresponding to the first-order voltage change rate continuously exceeding the preset voltage change rate threshold, and after the end time, the first-order voltage change rate recovers to within the preset voltage change rate threshold and remains stable, then the time interval corresponding to the start time and the end time is determined as the candidate transient time window.

[0009] In one possible implementation, step S4 includes: S41, within the candidate transient time window, based on the first-order difference sequence, the second-order difference sequence of the voltage sequence and the current sequence, determine the self-healing trigger time of the self-healing event; S42, within the candidate transient time window, starting from the self-healing trigger moment, the time sequence of the self-healing event is determined based on the first-order difference sequence and second-order difference sequence of the voltage sequence and the current pulse structure of the current sequence.

[0010] In one possible implementation, step S41 includes: S411, within the candidate transient time window, the interval in which the first-order rate of change of voltage in the first-order difference sequence of the voltage sequence is less than or equal to a preset voltage drop amplitude threshold, and the absolute value of the second-order rate of change of voltage in the second-order difference sequence of the voltage sequence is greater than or equal to a preset curvature threshold, is determined as the voltage rapid drop interval. S412, within the voltage rapidly decreasing range, based on the current sequence, determine the set of current peak points, the set of current zero-crossing points, and the current singularity points; S413, based on the set of current peak points, the set of current zero-crossing points, and the current singular points, the voltage rapid drop interval in which current singular points and abnormal current peak points appear simultaneously, and the interval between any two adjacent current zero-crossing points is less than a preset threshold, is determined as the target voltage rapid drop interval. The target voltage rapid drop interval contains transient pulse current generated by self-healing breakdown, and the abnormal current peak points are current peak points in the set of current peak points that exceed the preset range. S414, within the candidate transient time window, based on the first-order difference sequence of the voltage sequence, determine the first moment corresponding to the minimum first-order voltage change rate; within the target voltage rapid decrease interval, based on the current sequence, determine the second moment corresponding to the maximum current pulse value; based on the first moment and the second moment, determine the self-healing trigger moment of the self-healing event.

[0011] In one possible implementation, step S414, determining the self-healing trigger time of the self-healing event based on the first time and the second time, includes: When the absolute value of the difference between the first time point and the second time point is less than or equal to a preset time difference threshold, the minimum value between the first time point and the second time point is determined as the self-healing trigger time of the self-healing event.

[0012] In one possible implementation, step S42 includes: S421, within the candidate transient time window, the self-healing trigger time is taken as the start time of the arcing stage, and the end time of the arcing stage is determined based on the first-order differential sequence and second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset arcing stage identification rules. S422, within the candidate transient time window, the end time of the arcing stage is taken as the start time of the arc extinguishing stage. Based on the first-order differential sequence and second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset arc extinguishing stage identification rules, the end time of the arc extinguishing stage is determined. S423, within the candidate transient time window, the end time of the arc extinction phase is taken as the start time of the recovery charging phase. Based on the first-order differential sequence and second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset recovery charging phase identification rules, the end time of the recovery charging phase is determined. The preset arcing stage identification rule is the voltage signal u. cap (t) satisfies du cap (t) / dt≤0 and d 2 u cap (t) / dt 2 <0, current signal i s (t) satisfies Effective current i c (t) is negative, i arc (t) represents the arc current, i rip (t) represents the ripple current; The preset arc extinction stage identification rule is the voltage signal u. cap (t) satisfies 0 < du cap (t) / dt<△ max And the current signal i s (t) satisfies i c (t) is a positive value, △ max The preset maximum charging slope; The preset recovery charging phase identification rule is the voltage signal u. cap (t) satisfies du cap (t) / dt=△ max , and i arc (t)=0, current signal i s (t) satisfies .

[0013] In one possible implementation, step S5 includes: S51, acquire the initial voltage, termination voltage and current signals of the recovery charging phase; S52, perform wavelet packet decomposition or bandpass filtering on the current signal of the recovery charging stage to obtain the effective charging current of the recovery charging stage; S53, calculate the effective charging charge amount of the recovery charging phase based on the effective charging current of the recovery charging phase; S54, based on the initial voltage, termination voltage and effective charging charge of the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

[0014] In one possible implementation, prior to step S4, the method further includes: S3, within the candidate transient time window, based on the voltage sequence and the current sequence, determine whether a self-healing event has occurred; if so, proceed to steps S4 to S5. Step S3 includes: S31, within the candidate transient time window, perform time-frequency joint analysis on the voltage signal and the current signal respectively to obtain the voltage time-frequency energy distribution and the current time-frequency energy distribution; S32, perform integration and aggregation based on the voltage time-frequency energy distribution and the current time-frequency energy distribution to construct a real-time feature vector; S33. Based on the real-time feature vector and the preset template feature vector, determine the similarity index. The preset template feature vector includes the feature vector corresponding to any one or more operating switch transient states of the thin film capacitor under test in normal tuning, latch-up, cut-in, and cut-out. S34, when the similarity index is less than the preset similarity threshold, it is determined that a self-healing event has occurred.

[0015] In one possible implementation, after step S5, the method further includes: S6. Based on the estimated effective capacitance value of the thin-film capacitor under test and the time series of the self-healing event, a degradation trend analysis is performed on the thin-film capacitor under test.

[0016] On the other hand, this application also provides a computer system including a memory and a processor; the memory stores program code executable by the processor; the program code is used to perform the method as described in the first aspect or any of the implementations thereof.

[0017] The advantages of this application compared to existing technologies are as follows: Based on a preset sampling period, the voltage and current signals of the thin-film capacitor under test are simultaneously acquired, and a voltage and current sequence with a unified time series is constructed; based on the preset sampling period and voltage sequence, candidate transient time windows are determined; within the candidate transient time window, based on the voltage and current sequences, the time series of self-healing events is determined, including the arcing stage, the arc extinguishing stage, and the recovery charging stage; based on the initial voltage, termination voltage, and effective charging current of the recovery charging stage, the effective capacitance estimate of the thin-film capacitor under test is determined. Compared with relying on offline detection after power outage, indirect monitoring of the operating environment, and estimation methods based on equivalent parameters for capacitance estimation of thin-film capacitors, this application achieves online capacitance monitoring of thin-film capacitors by accurately identifying the self-healing behavior of the thin-film capacitors under the condition that the converter valve does not stop, effectively reducing capacitance monitoring errors and ensuring the safe and stable operation of the converter valve.

[0018] It is understood that the computer system provided in this application has the same beneficial effects as the above-described online capacitance monitoring method based on the self-healing characteristics of thin-film capacitors, and will not be repeated here. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in this application, the accompanying drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating an online capacitance monitoring method based on the self-healing characteristics of a thin-film capacitor, provided in an embodiment of this application; Figure 2 A flowchart illustrating another online capacitance monitoring method based on the self-healing feature sensing of a thin-film capacitor provided in an embodiment of this application; Figure 3 A topology of a flexible DC transmission converter valve and its self-healing identification principle are provided in an embodiment of this application; Figure 4 A schematic diagram of voltage waveforms of a thin-film capacitor under test at different stages of a self-healing event, provided as an embodiment of this application; Figure 5 A schematic diagram of the current waveform of a thin-film capacitor under test at different stages of a self-healing event, provided as an embodiment of this application; Figure 6 This is a schematic diagram of the capacitance waveform of a thin-film capacitor under test at different stages of a self-healing event, provided as an embodiment of this application. Detailed Implementation

[0021] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of this application. However, those skilled in the art will understand that this application may be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0022] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0023] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0024] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0025] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0026] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0027] The online capacitance monitoring method and system based on the self-healing characteristics of thin-film capacitors provided in this application are applicable to various types of thin-film capacitors, especially to thin-film capacitors in flexible DC transmission converter valve submodules. For ease of understanding, the technical solution of this application will be described in detail below with reference to the accompanying drawings.

[0028] Figure 1 This is a flowchart illustrating an online capacitance monitoring method based on the self-healing characteristics of a thin-film capacitor, provided in one embodiment of this application. Figure 2 This is a flowchart illustrating another online capacitance monitoring method based on the self-healing feature sensing of a thin-film capacitor, provided in one embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The method provided in this embodiment includes the following steps: S1, based on a preset sampling period, synchronously acquires the voltage and current signals of the thin-film capacitor under test, and constructs a voltage sequence and current sequence with a unified time series.

[0029] Optionally, based on a preset sampling period, the voltage signal u of the thin-film capacitor under test is acquired synchronously. cap (t) and current signal i s (t), ensuring that the sampling time deviation is less than the preset synchronization threshold; aligning the sampling timestamps to a unified reference clock and removing outliers to form a voltage sequence and a current sequence with a unified time series.

[0030] For example, when the thin-film capacitor under test is any thin-film capacitor in the flexible DC transmission converter valve submodule, the voltage signal u is triggered by the flow valve submodule drive signal. cap (t) and current signal i s Synchronous acquisition of (t).

[0031] S2, based on the preset sampling period and voltage sequence, determines the candidate transient time window.

[0032] In one possible implementation, step S2 may include: S21, Calculate the first-order difference sequence of voltage based on the preset sampling period and voltage sequence.

[0033] As an example, when self-healing occurs, the voltage drops abruptly within a very short time. The first-order difference sequence of the voltage is calculated based on the voltage sequence. ; where u cap (t k Let u be the capacitor voltage at the k-th sampling time. cap (t k-1 T represents the capacitor voltage at the next sampling time after the k-th sampling time. s This is the preset sampling period.

[0034] S22, in the first-order voltage difference sequence, determine the start time and end time corresponding to the first-order voltage change rate continuously exceeding the preset voltage change rate threshold, and after the end time, the first-order voltage change rate recovers to within the preset voltage change rate threshold and remains stable, then the time interval corresponding to the start time and end time is determined as the candidate transient time window.

[0035] As an example, when the time series (t) c , t m )satisfy When, the time interval (t) c , t m ) is determined as a candidate transient time window. Where, θ u This is a preset voltage change rate threshold. That is, when the first-order voltage change rate is within the time interval (t... c , t m The voltage change rate continuously exceeds the preset threshold θ within a certain period of time. u and in t mWhen the voltage returns to within the preset voltage change rate threshold and remains stable, the time interval (t) is... c , t m This is determined as a candidate transient time window.

[0036] Preferred, , where μ is the statistical average of the voltage change rate under normal operating conditions; is the corresponding standard deviation; v is the threshold coefficient, used to characterize the degree of abnormal change in the rate of voltage change.

[0037] S4. Within the candidate transient time window, the time series of self-healing events is determined based on the voltage and current sequences. The self-healing events include the arcing stage, the arc extinguishing stage, and the recharging stage.

[0038] In one possible implementation, step S4 may include: S41, within the candidate transient time window, determine the self-healing trigger time of the self-healing event based on the first-order difference sequence, second-order difference sequence, and current sequence of the voltage sequence.

[0039] As an example, step S41 may optionally include: S411, within the candidate transient time window, the interval in which the first-order rate of change of voltage in the first-order difference sequence of the voltage sequence is less than or equal to a preset voltage drop amplitude threshold, and the absolute value of the second-order rate of change of voltage in the second-order difference sequence of the voltage sequence is greater than or equal to a preset curvature threshold, is determined as the voltage rapid drop interval.

[0040] For example, extracting the voltage sequence u cap (t) within the candidate transient time window (t) c , t m First-order difference sequence within) With second-order difference sequence And within the candidate transient time window (t c , t m The internal identification satisfies the joint condition. The continuous interval is taken as the voltage rapid decrease interval (t) start , t end ), where η u ξ is the preset voltage drop threshold. u This is a preset curvature threshold.

[0041] S412, within the voltage rapidly decreasing range, determines the set of current peak points, the set of current zero-crossing points, and current singularities based on the current sequence.

[0042] For example, in the voltage drop interval (t start , t end Within ) will satisfy The current signal is used as the current peak point to obtain the set of current peak points {(t)}. pk,k , I pk,k )}, where k is the peak point time.

[0043] For example, satisfying The current signal is taken as the current zero-crossing point, and the set of current zero-crossing points {(t)} is obtained. pk,j , I pk,j )}, where j is the zero-crossing time.

[0044] For example, satisfying The current signal is taken as the current singularity, where η i This is a preset threshold for the magnitude of current drop.

[0045] S413, based on the set of current peak points, the set of current zero-crossing points, and current singular points, the voltage fast drop interval in which current singular points and abnormal current peak points appear simultaneously, and the interval between any two adjacent current zero-crossing points is less than a preset threshold, is determined as the target voltage fast drop interval. There is a transient pulse current generated by self-healing breakdown in the target voltage fast drop interval. The abnormal current peak points are the current peak points in the set of current peak points that exceed the preset range.

[0046] Preferably, the interval containing both current singularities and abnormal current peaks, and where the difference between the interval between any two adjacent current zero-crossing points and the ripple main period is less than a preset threshold (or the interval is significantly less than the ripple main period), is defined as a voltage rapid decrease interval (t) where both current singularities and abnormal current peaks occur simultaneously, and the difference between the interval between any two adjacent current zero-crossing points and the ripple main period is less than a preset threshold (or the interval is significantly less than the ripple main period). start , t end The target voltage rapidly decreases within a certain range, and transient pulse currents caused by self-healing breakdown are identified within this range. Abnormal current peaks can be selected as those significantly larger or smaller than other current peaks.

[0047] S414, within the candidate transient time window, based on the first-order difference sequence of the voltage sequence, determine the first moment corresponding to the minimum first-order voltage change rate; within the target voltage rapid decrease interval, based on the current sequence, determine the second moment corresponding to the maximum current pulse value; based on the first moment and the second moment, determine the self-healing trigger moment of the self-healing event.

[0048] Preferably, in step S414, determining the self-healing trigger time of the self-healing event based on the first time and the second time may include: when the absolute value of the difference between the first time and the second time is less than or equal to a preset time difference threshold, determining the minimum value between the first time and the second time as the self-healing trigger time of the self-healing event.

[0049] For example, in the candidate transient time window (t c , tm Within a given range, based on the first-order difference sequence of the voltage sequence, the first moment corresponding to the minimum first-order voltage change rate, i.e., the initiation moment of the voltage change, is defined as follows: ; in the target voltage rapidly decreasing range (t start , t end Within this timeframe, based on the current sequence, the second moment corresponding to the maximum value of the current pulse is determined, i.e., the moment corresponding to the main peak of the current pulse. When the two satisfy the temporal coupling relationship When the time is right, the self-healing trigger time is determined to be t1 = min(t u,0 , t i,0 ), where τ ui The preset time difference threshold is (preferably considering the setting to accommodate sampling delay and arc establishment delay).

[0050] S42, within the candidate transient time window, starting from the self-healing trigger moment, determine the time sequence of the self-healing event based on the first-order difference sequence of the voltage sequence, the second-order difference sequence of the voltage sequence, and the current pulse structure of the current sequence.

[0051] In one possible implementation, step S42 may include: S421, within the candidate transient time window, the self-healing trigger moment is taken as the start moment of the arcing stage. Based on the current pulse structure of the first-order differential sequence of the voltage sequence, the second-order differential sequence of the voltage sequence, and the current sequence, as well as the preset arcing stage identification rules, the end moment of the arcing stage is determined. S422, within the candidate transient time window, the end time of the arcing stage is taken as the start time of the arc extinguishing stage. Based on the current pulse structure of the first-order differential sequence of the voltage sequence, the second-order differential sequence of the current sequence, and the preset arc extinguishing stage identification rules, the end time of the arc extinguishing stage is determined. S423, within the candidate transient time window, the end time of the arc extinction phase is used as the start time of the recovery charging phase. Based on the current pulse structure of the first-order differential sequence of the voltage sequence, the second-order differential sequence of the current sequence, and the preset recovery charging phase identification rules, the end time of the recovery charging phase is determined. Typically, the end time of the recovery charging phase is the end time t of the candidate transient time window. m ; Among them, such as Figures 3 to 6 As shown, the preset arcing stage identification rule is the voltage signal u. cap (t) satisfies du cap (t) / dt≤0 and d 2 u cap (t) / dt 2 <0, current signal i s (t) satisfies Effective current i s(t) is negative (discharge), arc current i arc (t) A high-amplitude pulse appears, i rip (t) represents the ripple current; The preset rule for identifying the arc extinction stage is the voltage signal u. cap (t) satisfies 0 < du cap (t) / dt<△ max And the current signal i s (t) satisfies i s (t) is a positive value (charging), i arc (t) has not yet been extinguished, △ max The preset maximum charging slope; The preset rule for recognizing the recovery charging phase is the voltage signal u. cap (t) satisfies du cap (t) / dt=△ max That is, the first-order rate of change of voltage reaches the preset maximum charging slope Δ max And the arc current i arc (t)=0, current signal i s (t) satisfies .

[0052] For example, the final determined time series of the self-healing event is as follows: the time series of the arc ignition phase is (t1, t2), the time series of the arc extinction phase is (t2, t3), and the time series of the recharging phase is (t3, t4). m ).

[0053] Among them, the current signal (i.e., the sampling current) i s (t) is divided into: arc current i arc (t), transient discharge current generated by self-healing channel breakdown; ripple current i rip (t), the periodic harmonic current generated by PWM modulation, arm current variation and submodule voltage switching, from the sampled current i s The effective current i is obtained by performing bandpass filtering or wavelet packet multi-resolution decomposition in (t); c (t), where i is the sampling current. s (t) minus i arc (t) and i rip The current after (t).

[0054] Optionally, the operating conditions at the moment of self-healing directly affect the charging capacity of the capacitor during the recovery charging phase. Therefore, a preset maximum charging slope Δ is used to distinguish between the arc extinction phase and the recovery charging phase. maxThe system adaptively adjusts based on at least one of the following: the current terminal voltage of the capacitor, the statistical characteristics of historical self-healing events, or the current operating condition of the submodule (such as the arm current amplitude, modulation state, or submodule activation state). Preferably, the adjustment is based on the converter valve arm current i. arm Sure , where t c Let η be the start time of the candidate transient time window, η be the operating condition correction factor, and C be the value of C. ref This is the reference capacitance value for the film capacitor under test.

[0055] In another possible implementation, step S4 may include: constructing and training a deep learning model (e.g., a convolutional neural network model, a recurrent neural network model, or a combination thereof); using time segments of the voltage sequence and current sequence as input to the deep learning model, and having the model directly output the identification results of the self-healing transient, including the self-healing trigger time, the time sequence of the arcing stage, the time sequence of the arc extinguishing stage, and the time sequence of the recovery charging stage.

[0056] S5. Based on the initial voltage, termination voltage, and effective charging current of the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

[0057] In one possible implementation, step S5 may include: S51 acquires the initial voltage, termination voltage, and current signals of the recovery charging phase.

[0058] As an example, based on the time series of the recovery charging phase (t3,t... m Extract t3 to t from the voltage sequence and current sequence respectively. m The voltage and current signal segments at time t3 are shown. The voltage value at time t3 in the voltage signal segment is the initial voltage during the recovery charging phase. m The voltage value at that moment is the termination voltage of the recovery charging phase.

[0059] S52 performs wavelet packet decomposition or bandpass filtering on the current signal during the recovery charging phase to obtain the effective charging current during the recovery charging phase.

[0060] As an example, wavelet packet decomposition or bandpass filtering is performed on the current signal segment corresponding to the recovery charging stage to separate and remove ripple current, obtain the effective charging current of the recovery charging stage, and construct the trajectory of the effective charging current as the main recovery period waveform of the thin film capacitor under test after self-healing.

[0061] S53, calculate the effective charging charge during the recovery charging phase based on the effective charging current during the recovery charging phase.

[0062] As an example, the effective charging current during the recovery charging phase is integrated to obtain the effective charging charge during the recovery charging phase.

[0063] S54. Based on the initial voltage, termination voltage, and effective charging charge during the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

[0064] As an example, the formula for calculating the effective capacitance estimate is: ,in, To restore the effective charging charge during the charging phase, To restore the starting voltage of the charging phase, u cap (t m This is the termination voltage for the recovery charging phase.

[0065] In one possible implementation, such as Figure 2 As shown, before step S4, the method further includes: S3. Within the candidate transient time window, determine whether a self-healing event has occurred based on the voltage sequence and current sequence. If so, proceed to steps S4 to S5.

[0066] As an example, step S3 may optionally include: S31. Within the candidate transient time window, perform time-frequency joint analysis on the voltage signal and the current signal respectively to obtain the voltage time-frequency energy distribution and the current time-frequency energy distribution.

[0067] For example, in determining the candidate transient time window (t) c , t m After that, the voltage signal u of the thin-film capacitor under test is respectively... cap (t) and current signal i s (t) Perform joint time-frequency analysis (such as S-transform or Hilbert-Huang transform) to obtain the voltage time-frequency energy distribution E. u (f,t) and the time-frequency energy distribution of the current E i (f,t).

[0068] S32, based on the voltage time-frequency energy distribution and the current time-frequency energy distribution, performs integral aggregation to construct a real-time feature vector.

[0069] For example, within a preset frequency band, the effects of switching transients, electromagnetic interference, or sampling anomalies are mainly concentrated near the submodule switching frequency (e.g., 2kHz~5kHz) and its odd harmonics; due to their randomness and suddenness, the entropy value of the time-frequency distribution of the self-healing pulse is significantly higher than that of the regularly occurring switching signal, and its energy is mostly concentrated in the high-frequency band (>10kHz). The voltage and current time-frequency energy distributions are integrated and aggregated to construct a real-time feature vector. ,in, and These represent the low-frequency energy characteristics of voltage and current signals, respectively. and These represent the high-frequency energy characteristics of voltage and current signals, respectively. The time-frequency entropy characteristics are calculated based on the time-frequency energy distribution of the current.

[0070] S33. Based on the real-time feature vector and the preset template feature vector, determine the similarity index. The preset template feature vector includes the feature vector corresponding to any one or more operating switch transient states of the thin film capacitor under test in normal tuning, latch-up, cut-in, and cut-out.

[0071] For example, a pre-set graph library is provided, which stores the feature vectors X corresponding to typical operating switching transients of the thin-film capacitor under test, such as normal tuning, latch-up, cut-in, and cut-out. temp The real-time obtained voltage and current signal feature vector X SH Compared with the preset template feature vector X in the spectral library temp Calculate cosine similarity , where m is the number of feature vectors, and the cosine similarity is used as the similarity index.

[0072] S34. When the similarity index is less than the preset similarity threshold, a self-healing event is determined to have occurred.

[0073] For example, the calculated similarity index is compared with a preset similarity threshold. When the similarity index is greater than or equal to the preset similarity threshold, it is determined to be a pseudo-pulse segment, and the system maintains steady-state monitoring without executing steps S4 to S5. If the similarity index is lower than the preset similarity threshold, it is determined to be a "self-healing event" and steps S4 to S5 are triggered.

[0074] In one possible implementation, such as Figure 2 As shown, after step S5, the method further includes: S6. Based on the estimated effective capacitance value of the thin-film capacitor under test and the time series of self-healing events, a degradation trend analysis is performed on the thin-film capacitor under test.

[0075] As an example, step S6 may optionally include: S61. Based on the estimated effective capacitance value of the thin-film capacitor under test and the time series of self-healing events, calculate the capacitance value decline characteristics, including any one or more of the following: single self-healing capacitance loss, cumulative capacitance decay, and capacitance decline slope.

[0076] For example, the formula for calculating the loss of self-healing capacity in a single instance is: , where C k-1C is the estimated effective capacitance value of the thin-film capacitor under test, calculated in step S5 after the (k-1)th self-healing event. k This is the estimated effective capacitance value of the thin-film capacitor under test, calculated in step S5 after the k-th self-healing event.

[0077] For example, the formula for calculating the cumulative capacitance decay is: Where C0 is the effective capacitance value of the film capacitor under test during the initial operation phase.

[0078] For example, the formula for calculating the capacitance decrease slope is: Where C is the estimated effective capacitance of the film capacitor under test at the current moment, and N... SH This represents the number of self-healing cycles of the thin-film capacitor under test up to the current time.

[0079] S62, construct a capacitor degradation state model, input the capacitance value decline characteristics into the capacitor degradation state model, and obtain the predicted value of the remaining service life of the thin film capacitor under test.

[0080] Optionally, the capacitor degradation state model is preferably at least one of the following: a fitting model based on a physical failure model or an exponential decay model; a probabilistic lifetime model based on a statistical distribution (such as the Weibull distribution); or a time series prediction model based on a machine learning regression algorithm (such as a long short-term memory network or a gated recurrent unit).

[0081] For example, the capacitance value decline characteristics, the estimated effective capacitance value of the film capacitor under test, and the time series of self-healing events are input into the capacitor degradation state model to extrapolate and predict the capacitance degradation process of the film capacitor under test, determine the failure time or number of self-healing failures when its effective capacitance value reaches the preset failure threshold, and thereby calculate the predicted remaining service life of the film capacitor under test.

[0082] S63, based on the characteristics of capacitance decrease and the predicted value of remaining service life, performs degradation trend analysis on the thin film capacitor under test.

[0083] Optionally, the control strategy of the flexible DC transmission converter valve submodule can be dynamically adjusted or a graded early warning signal can be generated based on the capacitance degradation characteristics and the predicted remaining service life. The control strategy includes controlling at least one of the following: submodule modulation participation, voltage or current stress level, and submodule deployment sequence. The graded early warning signal characterizes the degradation level based on the predicted remaining service life or capacitance degradation characteristics of the thin-film capacitor under test, providing a basis for operation control or maintenance decisions.

[0084] The technical solution provided in this application synchronously acquires the voltage and current signals of the thin-film capacitor under test based on a preset sampling period, and constructs a voltage and current sequence with a unified time series. Based on the preset sampling period and voltage sequence, candidate transient time windows are determined. Within the candidate transient time window, the time series of self-healing events is determined based on the voltage and current sequences. The self-healing events include the arcing stage, the arc extinguishing stage, and the recovery charging stage. Based on the initial voltage, termination voltage, and effective charging current of the recovery charging stage, the effective capacitance estimate of the thin-film capacitor under test is determined. Compared with offline detection after power outage, indirect monitoring of the operating environment, and estimation methods based on equivalent parameters for thin-film capacitor capacitance estimation, this solution achieves online capacitance monitoring of the thin-film capacitor by accurately identifying the self-healing behavior of the thin-film capacitor under the condition that the converter valve does not stop, effectively reducing capacitance monitoring errors and ensuring the safe and stable operation of the converter valve.

[0085] On the other hand, this application also provides a computer system, including a memory and a processor; the memory stores program code that can be executed by the processor; the program code is used to execute any of the above-mentioned online capacitance monitoring methods based on the self-healing characteristics of thin-film capacitors.

[0086] For example, program code can be divided into one or more modules / units, which are stored in memory and executed by a processor to complete the present invention. One or more modules / units can be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the program code in a computer system.

[0087] The computer system can be a desktop computer, laptop, handheld computer, cloud server, or other computing device, and may optionally include, but is not limited to, processors and memory. Those skilled in the art will understand that it may also include input / output devices, network access devices, buses, etc.

[0088] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0089] Memory can be an internal storage unit of a terminal device, such as a hard drive or RAM. Memory can also be an external storage device of a terminal device, such as a plug-in hard drive, Smart MediaCard (SMC), Secure Digital (SD) card, or Flash Card. Furthermore, memory can include both internal and external storage units of a computer system. Memory is used to store program code and other programs and data required by the computer system. Memory can also be used to temporarily store data that has been output or will be output.

[0090] The above-mentioned computer system is created based on the above-mentioned online capacitance monitoring method based on the self-healing feature sensing of thin film capacitors. Its technical function and beneficial effects will not be repeated here. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0091] The above embodiments merely illustrate several implementation methods of this application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the protection scope of this invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for online capacitance monitoring based on the self-healing characteristics of thin-film capacitors, characterized in that, The method includes: S1, based on a preset sampling period, synchronously acquire the voltage and current signals of the thin-film capacitor under test, and construct a voltage sequence and current sequence with a unified time series; S2, Based on the preset sampling period and the voltage sequence, determine the candidate transient time window; S4, within the candidate transient time window, based on the voltage sequence and the current sequence, determine the time sequence of the self-healing event, the self-healing event including the arcing stage, the arc extinguishing stage and the recharging stage; S5. Based on the initial voltage, termination voltage, and effective charging current of the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

2. The method according to claim 1, characterized in that, Step S2 includes: S21, Calculate the first-order voltage difference sequence based on the preset sampling period and the voltage sequence; S22, in the first-order voltage difference sequence, determine the start time and end time corresponding to the first-order voltage change rate continuously exceeding the preset voltage change rate threshold, and after the end time, the first-order voltage change rate recovers to within the preset voltage change rate threshold and remains stable, then the time interval corresponding to the start time and the end time is determined as the candidate transient time window.

3. The method according to claim 1, characterized in that, Step S4 includes: S41, within the candidate transient time window, based on the first-order difference sequence, the second-order difference sequence of the voltage sequence, and the current sequence, determine the self-healing trigger time of the self-healing event; S42, within the candidate transient time window, starting from the self-healing trigger moment, the time sequence of the self-healing event is determined based on the first-order difference sequence and second-order difference sequence of the voltage sequence and the current pulse structure of the current sequence.

4. The method according to claim 3, characterized in that, Step S41 includes: S411, within the candidate transient time window, the interval in which the first-order rate of change of voltage in the first-order difference sequence of the voltage sequence is less than or equal to a preset voltage drop amplitude threshold, and the absolute value of the second-order rate of change of voltage in the second-order difference sequence of the voltage sequence is greater than or equal to a preset curvature threshold, is determined as the voltage rapid drop interval. S412, within the voltage rapidly decreasing range, based on the current sequence, determine the set of current peak points, the set of current zero-crossing points, and the current singularity points; S413, based on the set of current peak points, the set of current zero-crossing points, and the current singular points, the voltage rapid drop interval in which current singular points and abnormal current peak points appear simultaneously, and the interval between any two adjacent current zero-crossing points is less than a preset threshold, is determined as the target voltage rapid drop interval. The target voltage rapid drop interval contains transient pulse current generated by self-healing breakdown, and the abnormal current peak points are current peak points in the set of current peak points that exceed the preset range. S414, within the candidate transient time window, based on the first-order difference sequence of the voltage sequence, determine the first moment corresponding to the minimum first-order voltage change rate; within the target voltage rapid decrease interval, based on the current sequence, determine the second moment corresponding to the maximum current pulse value; based on the first moment and the second moment, determine the self-healing trigger moment of the self-healing event.

5. The method according to claim 4, characterized in that, In step S414, determining the self-healing trigger time of the self-healing event based on the first time and the second time includes: When the absolute value of the difference between the first time point and the second time point is less than or equal to a preset time difference threshold, the minimum value between the first time point and the second time point is determined as the self-healing trigger time of the self-healing event.

6. The method according to claim 3, characterized in that, Step S42 includes: S421, within the candidate transient time window, the self-healing trigger time is taken as the start time of the arcing stage, and the end time of the arcing stage is determined based on the first-order differential sequence and the second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset arcing stage identification rules. S422, within the candidate transient time window, the end time of the arcing stage is taken as the start time of the arc extinguishing stage. Based on the first-order differential sequence and second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset arc extinguishing stage identification rules, the end time of the arc extinguishing stage is determined. S423, within the candidate transient time window, the end time of the arc extinction phase is taken as the start time of the recovery charging phase. Based on the first-order differential sequence and second-order differential sequence of the voltage sequence, the current pulse structure of the current sequence, and the preset recovery charging phase identification rules, the end time of the recovery charging phase is determined. The preset arcing stage identification rule is the voltage signal u. cap (t) satisfies du cap (t) / dt≤0 and d 2 u cap (t) / dt 2 <0, current signal i s (t) satisfies Effective current i c (t) is negative, i arc (t) represents the arc current, i rip (t) represents the ripple current; The preset arc extinction stage identification rule is the voltage signal u. cap (t) satisfies 0 < du cap (t) / dt<△ max And the current signal i s (t) satisfies i c (t) is a positive value, △ max The preset maximum charging slope; The preset recovery charging phase identification rule is the voltage signal u. cap (t) satisfies du cap (t) / dt=△ max , and i arc (t)=0, current signal i s (t) satisfies .

7. The method according to claim 1, characterized in that, Step S5 includes: S51, acquire the initial voltage, termination voltage and current signals of the recovery charging phase; S52, perform wavelet packet decomposition or bandpass filtering on the current signal of the recovery charging stage to obtain the effective charging current of the recovery charging stage; S53, calculate the effective charging charge amount of the recovery charging phase based on the effective charging current of the recovery charging phase; S54, based on the initial voltage, termination voltage and effective charging charge of the recovery charging phase, determine the estimated effective capacitance value of the thin-film capacitor under test.

8. The method according to claim 1, characterized in that, Before step S4, the method further includes: S3, within the candidate transient time window, based on the voltage sequence and the current sequence, determine whether a self-healing event has occurred; if so, proceed to steps S4 to S5. Step S3 includes: S31, within the candidate transient time window, perform time-frequency joint analysis on the voltage signal and the current signal respectively to obtain the voltage time-frequency energy distribution and the current time-frequency energy distribution; S32, perform integration and aggregation based on the voltage time-frequency energy distribution and the current time-frequency energy distribution to construct a real-time feature vector; S33. Based on the real-time feature vector and the preset template feature vector, determine the similarity index. The preset template feature vector includes the feature vector corresponding to any one or more operating switch transient states of the thin film capacitor under test in normal tuning, latch-up, cut-in, and cut-out. S34, when the similarity index is less than the preset similarity threshold, it is determined that a self-healing event has occurred.

9. The method according to any one of claims 1 to 8, characterized in that, After step S5, the method further includes: S6. Based on the estimated effective capacitance value of the thin-film capacitor under test and the time series of the self-healing event, a degradation trend analysis is performed on the thin-film capacitor under test.

10. A computer system, characterized in that, It includes a memory and a processor; the memory stores program code that can be executed by the processor; the program code is used to execute the online capacitance monitoring method based on the self-healing feature sensing of thin-film capacitors as described in any one of claims 1 to 9.

Images