A method and system for estimating the health of parallel capacitors
By collecting and calculating the branch current, voltage, and temperature signals of parallel capacitors, and using methods such as ripple current and temperature rise rate, the problem of early identification and accurate early warning of the health status of parallel capacitors is solved, realizing early identification of abnormal branch current distribution and early warning of high loss and temperature rise risk.
Patent Information
- Application Number
- CN202611580775.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-13
- Publication Date
- 2026-11-10
AI Technical Summary
Existing technologies cannot achieve earlier identification and more accurate health warnings for the health of parallel capacitors, especially when abnormal branch current distribution cannot be detected in time to detect high losses and temperature rise risks.
By collecting and calculating branch current signals, bus voltage signals, and temperature signals, and using the effective value of ripple current, the rate of temperature change, and complex frequency domain quantities, the equivalent series resistance and capacitance of parallel capacitors are calculated. Combined with the current balance health component, the health of the branch is judged and maintenance is carried out.
It enables early identification of abnormal current distribution in parallel capacitor branches, improves identification accuracy, and can detect high losses and temperature rise risks in advance, providing targeted maintenance suggestions.
Smart Images

Figure CN123066471A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power electronics and power capacitor health management technology, specifically to a method and system for estimating the health of parallel capacitors. Background Technology
[0002] In high-power converters, energy storage systems, and high-voltage DC support circuits, multiple large-capacity capacitors are often connected in parallel to form a mF-level capacitor bank to achieve filtering and energy storage functions. Since the equivalent series resistance (ESR) of the capacitor body increases with increasing temperature and decays more rapidly under high ripple current, parallel capacitor banks often exhibit inconsistent branch aging, necessitating an assessment of their health status.
[0003] Chinese Patent Publication No. CN121663734A discloses a control method and system for extending the lifespan of a supercapacitor system. It primarily uses capacitance decay and ESR (Electrostatic Discharge Rate) as parameters to evaluate the health status of individual capacitors. While capacitance and ESR are intrinsic parameters of a single capacitor, parallel high-current capacitor banks present a unique problem: each branch shares the busbar ripple current, but the actual distribution may be uneven. Even if the capacitance and ESR of two branches do not change significantly, if one branch bears excessive ripple current due to factors such as connection impedance, busbar position, terminal contact, or parasitic parameter changes, that branch will experience higher losses and temperature rise risks. Relying solely on capacitance and ESR ratios may not be sufficient to identify such problems in a timely manner, thus rendering the method in this patent document insufficiently accurate. Capacity decay and increased ESR are usually results of aging, while abnormal branch current distribution may occur even earlier. Therefore, this patent document cannot detect abnormal current distribution between branches in advance, failing to provide earlier health warnings.
[0004] The comparative document focuses on series-connected supercapacitor modules, where the State of Harm (SOH) calculation is based on the individual cell capacitance and ESR, focusing on the degradation of individual cell electrochemical performance and voltage balance between series-connected cells. For parallel power capacitor banks, each branch bears essentially the same bus voltage, but the ripple current of each branch is determined by the capacitor's equivalent capacitance, equivalent series resistance, equivalent series inductance, busbar connection impedance, terminal contact state, and installation location. When the connection impedance of a branch decreases, local parasitic parameters change, or the impedance of adjacent branches increases, that branch may bear an abnormally increased ripple current before its capacitance and ESR reach traditional aging criteria, resulting in additional Joule losses and local temperature rise. Therefore, relying solely on capacitance decay and ESR changes cannot fully reflect the current redistribution risk between parallel branches, nor can it distinguish between overall capacitor aging and local connection, busbar, or branch impedance anomalies. Summary of the Invention
[0005] The technical problem to be solved by this invention is that the existing parallel capacitor health estimation method cannot achieve earlier health warning and the identification is not accurate enough.
[0006] This invention solves the above-mentioned technical problems through the following technical means: a method for estimating the health of parallel capacitors, applied to a parallel power capacitor system, wherein the parallel power capacitor system includes a positive busbar, a negative busbar, and multiple capacitor branches connected in parallel between the positive busbar and the negative busbar; the method includes: Collect current signals from each branch, bus voltage signals, temperature from each branch, and ambient temperature; Calculate the effective value of the branch ripple current and the average ripple current of all branches. Based on the relative deviation between the effective value of the branch ripple current and the average ripple current of all branches, obtain the degree of deviation of the branch from the average ripple current of the parallel capacitor bank. Calculate the branch temperature rise and temperature rise rate. Based on the branch temperature rise, temperature rise rate, and effective value of branch ripple current, calculate the estimated equivalent series resistance of the branch. Based on the estimated equivalent series resistance of the branch, calculate the ESR healthy component. Complex frequency domain quantities within the same ripple frequency band are extracted from the bus voltage and branch current to calculate the equivalent capacity estimate. Based on the equivalent capacity estimate of each branch capacitor and the initial capacity reference value, the capacity health component is calculated. The current balance health component is calculated based on the degree of deviation of the branch from the average ripple current of the parallel capacitor bank. The minimum value among the ESR health component, capacity health component, and current balance health component is used as the health index of each branch. The variable corresponding to the minimum value is used as the limiting health component. The limiting health component is used to determine the abnormality type, and the health index is used to maintain each branch.
[0007] Further, the calculation of the effective value of the branch ripple current and the average ripple current of all branches includes: The acquired branch current signals are bandpass filtered from 500 to 2000 Hz to extract the ripple current component. The effective value of the branch ripple current is calculated using the ripple current component. The average ripple current of all branches is obtained by summing the effective values of the ripple current of all branches and taking the mean value.
[0008] Furthermore, the deviation of the obtained branch from the average ripple current of the parallel capacitor bank includes: The difference between the effective value of the branch ripple current and the average ripple current of all branches is divided by the average ripple current of all branches. The absolute value of the result is used to determine the degree of deviation of the branch from the average ripple current of the parallel capacitor bank.
[0009] Furthermore, the calculation of branch temperature rise and temperature rise rate includes: The branch temperature rise is obtained by subtracting the branch temperature from the ambient temperature. The rate of change of temperature rise is obtained by dividing the difference between two adjacent temperature samplings by the time interval between the two adjacent temperature samplings.
[0010] Further, the step of calculating the estimated equivalent series resistance of the branch based on the branch temperature rise, the rate of change of temperature rise, and the effective value of the branch ripple current, and calculating the ESR health component based on the estimated equivalent series resistance of the branch, includes: The branch's equivalent series resistance estimate is obtained by multiplying the branch's thermal time constant by the rate of change of temperature rise and adding the branch's temperature rise to the numerator, and multiplying the square of the branch's effective value of ripple current by the branch's thermal resistance parameter to the denominator. The ratio of the branch's initial healthy ESR reference value to the branch's equivalent series resistance estimate is compared with 1, and the minimum of the two is selected as the ESR healthy component.
[0011] Furthermore, the calculated capacity health component includes: The bus voltage signal and branch current signal are bandpass filtered or frequency-domain transformed in the range of 500–2000 Hz to obtain the complex frequency domain quantities of the bus ripple voltage and branch ripple current, respectively. The difference between the complex frequency domain quantity of the bus ripple voltage and the ripple voltage drop along the busbar path is used to obtain the complex frequency domain quantity of the connection point ripple voltage. The ratio of the complex frequency domain quantity of the connection point ripple voltage to the complex frequency domain quantity of the branch ripple current is used as the complex impedance estimate. An impedance model is constructed using the complex impedance estimate.
[0012] in, For the first The branch road is in The estimated complex impedance at each ripple frequency point Indicates the first i Branch equivalent series resistance; For the first The equivalent series inductance of the branch, For the analysis of the frequency band of 500–2000 Hz, the first One ripple frequency, It is the symbol for a complex number. For the first Branch current equivalent capacity estimate and
[0013] in, Indicates the total number of ripple frequency points; As a credibility weight, For the first Capacity estimates obtained from effective ripple frequency points and , The symbol for the imaginary part; healthy portion size , For the first Reference value for initial capacity of branch circuits.
[0014] Furthermore, the calculation of the current balancing health component includes: 1 is used as the subtrahend. The ratio of the deviation of the branch from the average ripple current of the parallel capacitor bank to the current deviation limit between branches is compared with 1. The minimum value of the two is selected as the minuend. The difference between the subtrahend and the minuend is used to obtain the current balance health component.
[0015] Furthermore, the method of determining the type of abnormality using a restricted health component includes: Limit healthy portions ,in, As variables, For resistance, capacitance and current, Indicates the healthy portion size. Indicates the ESR health component. Indicates the balanced and healthy component of the current; , and The variables corresponding to the equivalent series resistance healthy component, capacitance healthy component, and current balance healthy component are respectively used as the limiting healthy components; When multiple branches reach a preset ratio or When the synchronous current is lower than the health limit of the capacitor bank itself, and the deviation of the average ripple current of all branches relative to the parallel capacitor bank is not higher than the overall balance limit, it is judged that the capacitor bank is aging as a whole. When the preset proportion of branches When the corresponding ESR health component and capacity health component are not lower than the body health limit, it is determined to be a local current imbalance in the branch. As a side road When the temperature rise of the corresponding branch exceeds the thermal state limit, it is determined that the local loss of the branch has increased. As a side road When this occurs, it is determined to be branch capacity attenuation.
[0016] Furthermore, the maintenance of each branch using health indicators includes: Set three health thresholds and ,when At that time, the first The branch is determined to be in a normal state; when At that time, for the first Branch lines should provide health warnings and shorten monitoring cycles; when At that time, for the first Branch lines will undergo key inspections or operate at reduced capacity; when At that time, for the first Branch lines should be isolated or replaced.
[0017] The present invention also provides a parallel capacitor health estimation system, which performs the above-described method, the system comprising: The signal acquisition module is used to acquire the current signal of each branch, the bus voltage signal, the temperature of each branch, and the ambient temperature. The ripple feature extraction module is used to calculate the effective value of the branch ripple current and the average ripple current of all branches. Based on the relative deviation between the effective value of the branch ripple current and the average ripple current of all branches, the deviation of the branch from the average ripple current of the parallel capacitor bank is obtained. The parameter estimation module is used to calculate the branch temperature rise and the rate of temperature rise change. Based on the branch temperature rise, the rate of temperature rise change, and the effective value of the branch ripple current, it calculates the estimated value of the equivalent series resistance of the branch. It extracts complex frequency domain quantities within the same ripple frequency band from the bus voltage and the branch current to calculate the estimated value of the equivalent capacity. The health component calculation module is used to calculate the ESR health component based on the estimated equivalent series resistance of the branch; calculate the capacity health component based on the estimated equivalent capacitance of each branch capacitor and the initial capacitance reference value; and calculate the current balance health component based on the deviation of the branch from the average ripple current of the parallel capacitor bank. The health assessment module is used to obtain the health index of each branch by taking the minimum value among the ESR health component, capacity health component and current balance health component. The variable corresponding to the minimum value is used as the limit health component. The limit health component is used to determine the abnormality type, and the health index is used to maintain each branch.
[0018] The advantages of this invention are: (1) This invention calculates the deviation of a branch from the average ripple current of the parallel capacitor bank, thereby introducing the unique branch current distribution anomaly of the parallel capacitor bank into the calculation of the health index. This can identify the high loss and temperature rise risk caused by a branch bearing excessive ripple current. Compared with the prior art, which only relies on the capacitance ratio and ESR ratio for identification, the identification accuracy is higher. In addition, the branch current distribution anomaly may occur earlier than capacitance decay and ESR increase. Therefore, by introducing the branch current distribution anomaly into the calculation of the health index, this invention enables the early detection of the inter-branch current distribution anomaly and achieves earlier health warning.
[0019] (2) Identifying early anomalies based on the unique current distribution relationship of parallel topology: This invention utilizes the physical relationship that each branch of a parallel capacitor shares the bus ripple voltage, but the branch ripple current is determined by the combined impedance of each branch to calculate the deviation of the branch ripple current from the reference current level of the parallel capacitor bank. When the connection impedance, busbar position, terminal contact state, branch parasitic parameters, or capacitor body parameters change, the branch current distribution will change accordingly. Therefore, current distribution anomalies can reflect changes in the local branch state before the capacity significantly decreases or the ESR significantly increases. Thus, this invention can identify local current concentration, branch undercurrent, and current redistribution problems that are difficult to detect in time based solely on capacity and ESR, and expose local losses and overheating risks in advance.
[0020] (3) Online ESR Identification Using Operating Ripple and Thermal Dynamic Response: This invention establishes a thermal dynamic model based on the energy balance relationship between the power loss generated by the branch ripple current and the branch temperature rise dynamic response, and simultaneously estimates the branch ESR using the temperature rise amplitude and temperature rise rate of change. The temperature rise rate of change characterizes the heat accumulation process when the internal losses of the branch have not yet been fully converted into steady-state temperature rise, thus compensating for the thermal inertia caused by the capacitor shell, potting material, and heat dissipation structure, and reducing the lag and underestimation caused by using steady-state temperature rise to estimate ESR during the rapid temperature rise phase. This method uses the existing ripple current during normal operation of the equipment as excitation, eliminating the need to adjust the capacitor to a specific state of charge, let it stand still, and apply a dedicated 1 kHz test signal as in the patent literature in the background art, making it more suitable for online monitoring of high-voltage, large-capacity parallel power capacitor banks under continuous operation.
[0021] (4) Estimating capacity using operating ripple impedance to avoid periodic full charge-discharge calibration: This invention utilizes the complex impedance relationship between the bus ripple voltage and the ripple current of each branch to estimate the equivalent capacity of the branch after deducting the effects of the identified ESR and the equivalent inductance of the branch. This method uses the frequency band excitation during the normal operation of the parallel capacitor to complete the capacity identification, without the need for a complete charge-discharge cycle or periodic full charge-discharge calibration. The patent literature in the background technology uses piecewise current integration and periodic full charge-discharge correction of capacity, which is difficult to directly apply to filter or energy storage capacitor banks that are connected to the high-voltage DC bus for a long time and cannot be frequently taken out of operation.
[0022] (5) Differentiating between capacitor body degradation and branch local imbalance: When the capacitance and ESR health components of multiple branches decrease simultaneously, while the overall deviation of the branch current is relatively small, it indicates that the capacitor bank is closer to overall aging; when the current deviation of a certain branch increases significantly, while its capacitance and ESR have not changed proportionally, it indicates that the abnormality is more likely to originate from changes in local connection impedance, busbar path, or parasitic parameters; when current deviation, ESR increase, and temperature rise occur simultaneously, it indicates that the branch has formed a positive feedback process of current concentration—increased loss—increased temperature. Therefore, this invention not only outputs health values but also identifies abnormal branches and their main sources of risk, providing targeted basis for connection point inspection, branch derating, or capacitor replacement. Attached Figure Description
[0023] Figure 1 This is a flowchart of a method for estimating the health of parallel capacitors as disclosed in an embodiment of the present invention.
[0024] Figure 2 This is a comparison diagram of the effective values of ripple current in each branch under the condition of increased connection resistance of the fourth branch in a parallel capacitor health estimation method disclosed in an embodiment of the present invention. Figure 3 This is a comparison of the responses of different ESR estimation methods when the ESR of a branch changes abruptly in a parallel capacitor health estimation method disclosed in an embodiment of the present invention. Figure 4 This is a comparison chart of the set capacity of each branch and the estimated capacity at multiple frequency points in a parallel capacitor health estimation method disclosed in an embodiment of the present invention. Detailed Implementation
[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below in conjunction with the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, 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.
[0026] Example 1 Embodiment 1 of the present invention provides a method for estimating the health of parallel capacitors. The method is applied to a parallel power capacitor system, which includes a positive busbar, a negative busbar, and a circuit connected in parallel between the positive and negative busbars. N One capacitor branch, of which N≥2. Each capacitor branch includes at least one power capacitor, whose positive terminal is electrically connected to the positive busbar and whose negative terminal is electrically connected to the negative busbar; in actual engineering, a single branch can be composed of a single capacitor, or it can be equivalently composed of multiple capacitors connected in series and / or in parallel.
[0027] In this invention, a "branch" refers to a capacitor unit that forms an independent current path between the positive and negative busbars and is capable of independently acquiring ripple current and temperature signals. During system operation, each branch shares the ripple current on the busbar side. Due to differences in capacitance, equivalent series resistance (ESR), parasitic impedance, and aging degree among the branches, different ripple current distribution characteristics and temperature rise responses will be exhibited between the branches. This invention estimates the health status of each branch based on these differences.
[0028] This embodiment uses a filter capacitor bank consisting of 28 parallel capacitors (223×530×780 mm, horizontally stacked, with a capacitance in the mF range) in a dual-row 14+14 configuration as an example. The system voltage is 3500V, and the total current is approximately 50kA. This embodiment does not limit the specific form of the busbar material or the fastening structure. In this embodiment, the 28 capacitors are connected in parallel between the positive and negative busbars, forming 28 independently monitorable capacitor branches. The positive terminals of each branch are connected to the positive busbar, and the negative terminals of each branch are connected to the negative busbar, thus forming a multi-branch parallel energy storage and filtering topology. During operation, the ripple current on the DC bus is distributed to each capacitor branch via the busbar. Different branches experience uneven ripple current and temperature rise differences due to differences in their own parameters and connection impedances. This embodiment uses "one monitoring branch corresponding to a single capacitor" as an example for illustration; in other embodiments, a series-parallel combination unit composed of multiple capacitors can also be regarded as a monitoring branch and the same health estimation method can be used.
[0029] like Figure 1 As shown, the method for estimating the health of parallel capacitors includes the following steps: 1. Signal Acquisition Current sensors, voltage sensors, and temperature sensors are installed on each branch of the parallel power capacitor system to synchronously collect bus voltage signals and branch current signals. Temperature of each branch And ambient temperature, among which, t express t time, Indicates the first Branch circuits. The synchronous sampling frequency of branch circuit current signals and bus voltage signals is preferably not less than 20kHz; the sampling period for branch circuit temperature and ambient temperature is determined according to the thermal time constant, preferably 0.10 to 1.00 s.
[0030] 2. Ripple Extraction The acquired branch current signal was bandpass filtered in the 500~2000Hz range to extract the ripple current component. Calculate the time t. RMS value of branch ripple current :
[0031] in, Indicates the length of the time window. Represents the integral variable. t This indicates the current calculation time, which is also the end time of this integration window.
[0032] Calculate the average ripple current of all branches. :
[0033] in, This indicates the total number of branches.
[0034] And calculate the first The degree of deviation of the branch from the average ripple current of the parallel capacitor bank:
[0035] The average ripple current serves as a reference benchmark for the overall current distribution level of the parallel capacitor bank. The degree of deviation of a branch from the average ripple current of the parallel capacitor bank is used to characterize the first... The degree of deviation of a branch from the overall average level is used as the input for the current balance health component in subsequent health calculations.
[0036] 3. Temperature rise feature extraction Record number Branch operating temperature With ambient temperature Calculate the branch temperature rise and rate of change of temperature ,in, Indicates the discrete analysis window number. This indicates the time interval between two consecutive temperature samples. The temperature rise... Used to characterize the i The steady-state or quasi-steady-state heating degree of the branch under the current ripple current, the rate of temperature change. Used to characterize the i The dynamic trend of branch temperature rising or falling over time. The rate of temperature change is used to correct the thermal response hysteresis in the ESR estimation, so that the ESR estimation reflects both the current temperature rise amplitude and the temperature change trend.
[0037] 4. ESR estimation A first-order thermal balance equation is established based on the empirical relationship between ripple loss and thermal response:
[0038]
[0039] in, For the first Branch thermal resistance parameters For the first Branch thermal time constant, For ripple loss, For the first The current RMS value of the branch ripple current. Indicates the first i The equivalent series resistance of the branch in the current operating state and within the target ripple frequency band.
[0040] Within the ESR estimation window, losses, leakage losses, dielectric losses, and external thermal disturbances outside the target ripple band remain stable, negligible, or can be subtracted by the healthy baseline loss term.
[0041] Estimate the first Estimated equivalent series resistance of the branch :
[0042] in, and The parameters are obtained from thermal step tests, manufacturer thermal models, or finite element thermal simulations of healthy capacitor branches and stored as branch reference parameters under initial healthy conditions. Compared to estimating ESR solely based on the temperature rise amplitude, the above formula further incorporates the rate of temperature rise. , Used to compensate for branch thermal inertia. It can anticipate the increasing trend of branch losses during the rapid temperature rise phase, reducing the risk of ESR underestimation due to thermal response lag.
[0043] Calculate the ESR health component based on the ESR estimate:
[0044] in, For the first The ESR reference value under the initial healthy state of the branch. As ESR increases, A decrease indicates an increased risk of loss and overheating in that branch.
[0045] 5. Calculation of healthy serving sizes The bus voltage signal and the current signals of each branch are synchronously acquired, and the bus voltage signal and the branch current signal are subjected to bandpass filtering or frequency domain transformation in the range of 500 to 2000 Hz to obtain the first... Window number q The complex frequency domain quantity of the bus ripple voltage at each ripple frequency point and the branch ripple current complex frequency domain quantity .
[0046] No. The branch road is in The complex frequency domain quantity of the ripple voltage at the access point at each ripple frequency is:
[0047] in, According to the first The ripple voltage drop of the busbar path is calculated based on the branch connection location and the busbar path impedance. Since each capacitor branch is connected in parallel between the same positive and negative busbars, after ignoring or correcting for the voltage difference between busbar nodes, each branch bears the same busbar ripple voltage. Therefore, based on the busbar ripple voltage and the first... i The complex ratio between the branch ripple currents is used to calculate the first... i The branch road is in q The estimated complex impedance at the i-th ripple frequency point, the i-th The branch road is in The estimated complex impedance at each ripple frequency point is:
[0048] in, For the analysis of the frequency band of 500–2000 Hz, the first q One ripple frequency.
[0049] Within the analysis frequency band, the first i If a branch is equivalent to an impedance model formed by an equivalent series resistance, an equivalent series inductance, and an equivalent series capacitance, then:
[0050] in, For the first i The equivalent series inductance of a branch, including the internal inductance of the capacitor, the inductance of the lead terminals, and the parasitic inductance of the branch connecting conductors, is obtained by initial impedance frequency sweep, busbar parameter calculation, or finite element calibration. The estimated current equivalent capacity of the i-th branch is:
[0051] in, Indicates the total number of ripple frequency points. The confidence weight is determined based on the ripple voltage amplitude, ripple current amplitude, or frequency signal-to-noise ratio; the selected effective ripple frequency points satisfy:
[0052] For the first Capacity estimates obtained from effective ripple frequency points and , This is the symbol for the imaginary part.
[0053] Based on the estimated equivalent capacitance of each branch capacitor and the initial capacitance reference value, calculate the capacitance health component:
[0054] in, For the first The initial capacity reference value for the branch. The estimated equivalent capacity is determined by the branch ripple current, voltage ripple response, and branch impedance change. The lower the healthy capacity component, the more significant the capacity degradation of the branch.
[0055] 6. Calculation of the healthy component of current balance According to the The deviation of the branch from the average ripple current of the parallel capacitor bank is calculated to determine the current balance health component:
[0056] in, For the first Window number The degree of deviation of the branch from the average ripple current of the parallel capacitor bank. This is the limit for inter-branch current deviation. When =0, it means the first The branch ripple current is consistent with the average level. =1; when An increase indicates a rise in the degree of abnormality in the current distribution of that branch. Consequently, it decreases; when When the current balance healthy component drops to 0, it indicates that the branch has reached a state of severe current imbalance.
[0057] 7. Health Calculation No. The health indicators and the types of restrictions on health components for branch paths are as follows:
[0058]
[0059] in, As variables, For resistance, capacitance and current, , and These represent the equivalent series resistance healthy component, capacitance healthy component, and current balance healthy component as limiting healthy components, respectively.
[0060] 8. Grading and Maintenance Recommendations Set health threshold .when At that time, the first The branch is determined to be in a normal state; when At that time, for the first Branch lines should provide health warnings and shorten monitoring cycles; when At that time, for the first Branch lines will undergo key inspections or operate at reduced capacity; when At that time, for the first The branch line is isolated or replaced; when multiple maintenance conditions are met simultaneously, the maintenance action with the highest severity is performed. In this embodiment, We take values of 0.8, 0.6, and 0.4 respectively.
[0061] The above determination process cannot distinguish the specific type of abnormality when a condition is determined to be abnormal. In this embodiment, to determine the specific abnormality type, the abnormality type is determined based on the type of the restricted health component and the distribution relationship of the health components in each branch, including: When multiple branches reach a preset ratio or When the synchronous current is below the body health limit, and the deviation of all branches from the average ripple current of the parallel capacitor bank is not higher than the overall balance limit, the capacitor bank is judged to be aging as a whole. In this embodiment, the preset ratio is 70%, which corresponds to 10 branches for a capacitor bank containing 14 parallel branches. In this embodiment, the body health limit is 0.80, and the overall balance limit is 10%.
[0062] When the preset proportion of branches When the corresponding ESR health component and capacity health component are not lower than the body health limit, it is determined to be a local current imbalance in the branch. As a side road When the temperature rise of the corresponding branch exceeds the thermal state limit, it is determined that the local loss of the branch has increased; in this embodiment, the thermal state limit is 20°C.
[0063] As a side road When this occurs, it is determined to be branch capacity attenuation.
[0064] 9. Simulation Verification To verify the ability of this invention to identify the current redistribution of parallel branches, the dynamic changes in the equivalent series resistance of branches, and the capacity decay of branches, an equivalent simulation model of eight parallel capacitor branches was established. The initial equivalent capacity of each branch was 15 mF, the initial equivalent series resistance was 3 mΩ, the equivalent series inductance was 60 nH, and the initial connection resistance was 0.8 mΩ. A ripple current with a total effective value of 800 A was applied at a ripple frequency of 1 kHz. In the thermal dynamic model, the equivalent thermal resistance of the branches was taken as 0.6 K / W, and the thermal time constant was taken as 120 s. Capacity estimation was performed at three effective ripple frequencies: 600 Hz, 1000 Hz, and 1600 Hz. The above values are illustrative simulation parameters used to verify the mechanism of the method and do not constitute a limitation on the application parameters of this invention.
[0065] In the simulation of abnormal branch connection impedance, the connection resistance of the fourth branch was increased from 0.8 mΩ to 5.0 mΩ, while the capacitance, equivalent series resistance, and connection parameters of the other branches remained unchanged. Figure 2 As shown, under healthy conditions, the effective value of the ripple current in each branch is approximately 100 A. After the connection resistance increases, the effective value of the ripple current in the 4th branch decreases to 86.14 A, while the effective value in the remaining branches increases to 102.50 A. At this point, the average effective value of the ripple current in each branch is 100.45 A, with the deviation of the 4th branch from the average value being 14.25%, and the deviation of the remaining branches being approximately 2.04%. This result indicates that even without changes in the capacitor's capacitance and ESR, the deviation of the branch current can already identify local undercurrent caused by increased connection resistance and the redistribution of current carried by other branches, thus providing an early identification basis for connection point anomalies and busbar path anomalies.
[0066] In the simulation of dynamic changes in branch ESR, the effective value of the ripple current in the fourth branch is kept at 120 A, and its ESR is increased from 3 mΩ to 5 mΩ in a step at 60 s. Figure 3 As shown, when both temperature rise and the rate of temperature change are used for estimation, the ESR estimation result can follow the ESR change before the temperature rise caused by thermal inertia reaches a new steady state. At 60 s after the ESR change, the thermal dynamic estimate is 5.005 mΩ, while the estimate calculated solely based on the temperature rise amplitude is 3.800 mΩ. Under the established first-order thermal model and sampling conditions, the deviations from the 5 mΩ setpoint are approximately 0.10% and 24.01%, respectively. This result indicates that introducing the rate of temperature change can compensate for the thermal response hysteresis caused by the casing, potting material, and heat dissipation path, reducing the underestimation of the increase in ESR during the temperature rise phase.
[0067] In the branch capacity attenuation simulation, the equivalent capacity of the 6th branch was reduced from 15 mF to 12 mF, while the remaining branches maintained 15 mF. A random disturbance with a standard deviation of 0.03 mΩ was added to the complex impedance at each frequency. Equivalent series inductance compensation and comprehensive capacity estimation were performed using three ripple frequencies: 600 Hz, 1000 Hz, and 1600 Hz. Figure 4 As shown, the estimated capacity of branch 6 is 12.00 mF, while the estimated capacity of the remaining branches ranges from 14.98 mF to 15.02 mF. The maximum relative deviation of the capacity estimates for each branch in this simulation is approximately 0.14%. This result indicates that by utilizing the complex impedance formed by the bus ripple voltage and branch ripple current, and compensating for the influence of the equivalent series inductance, it is possible to locate branches experiencing capacity decay under normal operating ripple conditions.
[0068] comprehensive Figures 2 to 4 The results show that abnormal connection impedance is mainly manifested by an initial decrease in the current balance healthy component, an increase in ESR is mainly manifested by a decrease in the ESR healthy component accompanied by changes in thermal response, and capacitance decay is mainly manifested by a decrease in the capacitance healthy component. Therefore, the joint analysis of the three healthy components can distinguish between local imbalance in parallel branches, increased capacitor losses, and capacitance decay, providing a basis for branch-level anomaly location and maintenance measure selection.
[0069] Through the above technical solutions, this invention calculates the capacity health component based on the current equivalent capacity and the initial capacity reference value; calculates the current balance health component based on the current deviation between branches; and obtains the health status (SOH) of each branch by weighting the ESR health component, capacity health component, and current balance health component. This method can simultaneously reflect the degradation of the capacitor itself and the abnormal current distribution in parallel branches, realizing online identification and predictive maintenance of abnormal branches in large-current parallel capacitor banks, and reducing the risk of local overheating and sudden failure.
[0070] Example 2 Based on Embodiment 1, Embodiment 2 of the present invention also provides a parallel capacitor health estimation system, which performs the above method, and the system includes: The signal acquisition module is used to acquire the current signal of each branch, the bus voltage signal, the temperature of each branch, and the ambient temperature. The ripple feature extraction module is used to calculate the effective value of the branch ripple current and the average ripple current of all branches. Based on the relative deviation between the effective value of the branch ripple current and the average ripple current of all branches, the deviation of the branch from the average ripple current of the parallel capacitor bank is obtained. The parameter estimation module is used to calculate the branch temperature rise and the rate of temperature rise change. Based on the branch temperature rise, the rate of temperature rise change, and the effective value of the branch ripple current, it calculates the estimated value of the equivalent series resistance of the branch. It extracts complex frequency domain quantities within the same ripple frequency band from the bus voltage and the branch current to calculate the estimated value of the equivalent capacity. The health component calculation module is used to calculate the ESR health component based on the estimated equivalent series resistance of the branch; calculate the capacity health component based on the estimated equivalent capacitance of each branch capacitor and the initial capacitance reference value; and calculate the current balance health component based on the deviation of the branch from the average ripple current of the parallel capacitor bank. The health assessment module is used to obtain the health index of each branch by taking the minimum value among the ESR health component, capacity health component and current balance health component. The variable corresponding to the minimum value is used as the limit health component. The limit health component is used to determine the abnormality type, and the health index is used to maintain each branch.
[0071] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for estimating the health of parallel capacitors, characterized in that, An application to a parallel power capacitor system, the parallel power capacitor system comprising a positive busbar, a negative busbar, and multiple capacitor branches connected in parallel between the positive busbar and the negative busbar; the method comprising: Collect current signals from each branch, bus voltage signals, temperature from each branch, and ambient temperature; Calculate the effective value of the branch ripple current and the average ripple current of all branches. Based on the relative deviation between the effective value of the branch ripple current and the average ripple current of all branches, obtain the degree of deviation of the branch from the average ripple current of the parallel capacitor bank. Calculate the branch temperature rise and temperature rise rate. Based on the branch temperature rise, temperature rise rate, and effective value of branch ripple current, calculate the estimated equivalent series resistance of the branch. Based on the estimated equivalent series resistance of the branch, calculate the ESR healthy component. Complex frequency domain quantities within the same ripple frequency band are extracted from the bus voltage and branch current to calculate the equivalent capacity estimate. Based on the equivalent capacity estimate of each branch capacitor and the initial capacity reference value, the capacity health component is calculated. The current balance health component is calculated based on the degree of deviation of the branch from the average ripple current of the parallel capacitor bank. The minimum value among the ESR health component, capacity health component, and current balance health component is used as the health index of each branch. The variable corresponding to the minimum value is used as the limiting health component. The limiting health component is used to determine the abnormality type, and the health index is used to maintain each branch.
2. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The calculation of the effective value of the branch ripple current and the average ripple current of all branches includes: The acquired branch current signals are bandpass filtered from 500 to 2000 Hz to extract the ripple current component. The effective value of the branch ripple current is calculated using the ripple current component. The average ripple current of all branches is obtained by summing the effective values of the ripple current of all branches and taking the mean value.
3. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The deviation of the obtained branch from the average ripple current of the parallel capacitor bank includes: The difference between the effective value of the branch ripple current and the average ripple current of all branches is divided by the average ripple current of all branches. The absolute value of the result is used to determine the degree of deviation of the branch from the average ripple current of the parallel capacitor bank.
4. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The calculation of branch temperature rise and temperature rise rate includes: The branch temperature rise is obtained by subtracting the branch temperature from the ambient temperature. The rate of change of temperature rise is obtained by dividing the difference between two adjacent temperature samplings by the time interval between the two adjacent temperature samplings.
5. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The step of calculating the estimated equivalent series resistance of a branch based on its temperature rise, rate of change of temperature rise, and effective value of branch ripple current, and calculating the ESR health component based on the estimated equivalent series resistance of the branch, includes: The branch's equivalent series resistance estimate is obtained by multiplying the branch's thermal time constant by the rate of change of temperature rise and adding the branch's temperature rise to the numerator, and multiplying the square of the branch's effective value of ripple current by the branch's thermal resistance parameter to the denominator. The ratio of the branch's initial healthy ESR reference value to the branch's equivalent series resistance estimate is compared with 1, and the minimum of the two is selected as the ESR healthy component.
6. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The calculated capacity health component includes: The bus voltage signal and branch current signal are bandpass filtered or frequency-domain transformed in the range of 500–2000 Hz to obtain the complex frequency domain quantities of the bus ripple voltage and branch ripple current, respectively. The difference between the complex frequency domain quantity of the bus ripple voltage and the ripple voltage drop along the busbar path is used to obtain the complex frequency domain quantity of the connection point ripple voltage. The ratio of the complex frequency domain quantity of the connection point ripple voltage to the complex frequency domain quantity of the branch ripple current is used as the complex impedance estimate. An impedance model is constructed using the complex impedance estimate. in, For the first The branch road is in The estimated complex impedance at each ripple frequency point Indicates the first i Branch equivalent series resistance; For the first The equivalent series inductance of the branch, For the analysis of the frequency band of 500–2000 Hz, the first One ripple frequency, It is the symbol for a complex number. For the first Branch current equivalent capacity estimate and in, Indicates the total number of ripple frequency points; As a credibility weight, For the first Capacity estimates obtained from effective ripple frequency points and , The symbol for the imaginary part; healthy portion size , For the first Reference value for initial capacity of branch circuits.
7. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The calculation of the current balance health component includes: 1 is used as the subtrahend. The ratio of the deviation of the branch from the average ripple current of the parallel capacitor bank to the current deviation limit between branches is compared with 1. The minimum value of the two is selected as the minuend. The difference between the subtrahend and the minuend is used to obtain the current balance health component.
8. The method for estimating the health of parallel capacitors according to claim 1, characterized in that, The method of determining the type of abnormality by limiting the healthy component includes: Limit healthy portions ,in, As variables, For resistance, capacitance and current, Indicates the healthy portion size. Indicates the ESR health component. Indicates the balanced and healthy component of the current; , and The variables corresponding to the equivalent series resistance healthy component, capacitance healthy component, and current balance healthy component are respectively used as the limiting healthy components; When multiple branches reach a preset ratio or When the synchronous current is lower than the health limit of the capacitor bank itself, and the deviation of the average ripple current of all branches relative to the parallel capacitor bank is not higher than the overall balance limit, it is judged that the capacitor bank is aging as a whole. When the preset proportion of branches When the corresponding ESR health component and capacity health component are not lower than the body health limit, it is determined to be a local current imbalance in the branch. As a side road When the temperature rise of the corresponding branch exceeds the thermal state limit, it is determined that the local loss of the branch has increased. As a side road When this occurs, it is determined to be branch capacity attenuation.
9. The method for estimating the health of a parallel capacitor according to claim 8, characterized in that, The maintenance of each branch using health indicators includes: Set three health thresholds and ,when At that time, the first The branch is determined to be in a normal state; when At that time, for the first Branch lines should provide health warnings and shorten monitoring cycles; when At that time, for the first Branch lines will undergo key inspections or operate at reduced capacity; when At that time, for the first Branch lines should be isolated or replaced.
10. A parallel capacitor health estimation system, characterized in that, The system for performing the method according to any one of claims 1-9 comprises: The signal acquisition module is used to acquire the current signal of each branch, the bus voltage signal, the temperature of each branch, and the ambient temperature. The ripple feature extraction module is used to calculate the effective value of the branch ripple current and the average ripple current of all branches. Based on the relative deviation between the effective value of the branch ripple current and the average ripple current of all branches, the deviation of the branch from the average ripple current of the parallel capacitor bank is obtained. The parameter estimation module is used to calculate the branch temperature rise and the rate of temperature rise change. Based on the branch temperature rise, the rate of temperature rise change, and the effective value of the branch ripple current, it calculates the estimated value of the equivalent series resistance of the branch. It extracts complex frequency domain quantities within the same ripple frequency band from the bus voltage and the branch current to calculate the estimated value of the equivalent capacity. The health component calculation module is used to calculate the ESR health component based on the estimated equivalent series resistance of the branch; calculate the capacity health component based on the estimated equivalent capacitance of each branch capacitor and the initial capacitance reference value; and calculate the current balance health component based on the deviation of the branch from the average ripple current of the parallel capacitor bank. The health assessment module is used to obtain the health index of each branch by taking the minimum value among the ESR health component, capacity health component and current balance health component. The variable corresponding to the minimum value is used as the limit health component. The limit health component is used to determine the abnormality type, and the health index is used to maintain each branch.
Citation Information
Patent Citations
Control method and system for prolonging service life of super-capacitor system
CN121663734A
