Capacitance monitoring method and system based on swept resonance point

Abstract

The application relates to the field of electronic measurement and sensing technology, and discloses a capacitive monitoring method and system based on a swept-frequency resonance point. The method comprises the following steps: obtaining an impedance spectrum by rough sweeping in an initial frequency range, and identifying a rough resonance interval; performing fine sweeping in the interval and calculating an accurate resonance frequency by using a quadratic interpolation method; calculating a capacitive value in combination with a mapping table; and dynamically adjusting a next cycle sweeping window based on a historical resonance frequency. The system comprises an excitation signal generation unit, an impedance response collection unit, a rough / fine sweeping control unit, a resonance frequency calculation unit, a capacitive calculation unit and a dynamic frequency window updating unit, and integrates a temperature compensation and an abnormality detection mechanism. By means of two-stage sweeping and an adaptive frequency window strategy, the application shortens the monitoring period to less than 1 / 5 of that of a traditional method while ensuring that the resonance frequency measurement accuracy is better than 0.01 Hz, and significantly improves real-time performance, stability and environmental adaptability.

Description

Technical Field

[0001] This invention belongs to the field of electronic measurement and sensing technology, specifically relating to a capacitance monitoring method and system based on swept frequency resonant points. Background Technology

[0002] With the increasing penetration of power electronic systems in new energy, smart grids, and industrial automation, capacitors, as key energy storage and filtering units, are crucial for system reliability due to their condition monitoring. Traditional capacitor monitoring methods mostly rely on offline measurements or impedance analysis based on fixed frequency points, which are insufficient to reflect the dynamic characteristics of capacitors under complex operating conditions.

[0003] In recent years, online monitoring technology based on frequency sweep resonant point identification has attracted widespread attention due to its non-invasiveness and high sensitivity. Its core principle is to infer the capacitance value by scanning the excitation frequency and detecting the minimum value of the circuit impedance (i.e., the resonant point). However, this method faces significant contradictions in practical applications: to ensure the accuracy of resonant point positioning, a slow linear frequency sweep is required to cover a wide frequency band, but this process is time-consuming and severely restricts real-time monitoring capabilities; conversely, if a fast frequency sweep is used to improve the response speed, the resonant point positioning will be rough and the error will be significant due to the excessively large frequency step size or insufficient sampling, especially in environments with small capacitance changes or noise interference, which can easily lead to misjudgment or missed detection.

[0004] Capacitor monitoring technology based on frequency sweep resonant points aims to achieve online diagnosis of capacitor aging, breakdown, or parameter drift by accurately capturing the resonant frequency of an LC resonant circuit and inverting the capacitance value using known inductance parameters. Theoretically, this method possesses high resolution and strong adaptability; however, in engineering deployment, it is limited by the inherent defects of the frequency sweep strategy—indiscriminate scanning across the entire frequency band wastes computational resources and cannot simultaneously meet the dual requirements of speed and accuracy.

[0005] Traditional frequency sweeping schemes using fixed step sizes or preset frequency bands lack effective utilization of historical resonance location information, requiring repeated scanning of the entire possible frequency domain for each monitoring session. Even with the introduction of a two-stage coarse-fine sweep mechanism in some schemes, redundant operations in the initial coarse sweep stage cannot be avoided. Furthermore, in scenarios with slow capacitance changes or stable environments, a significant amount of computational resources is consumed in known stable regions. Moreover, the failure to consider the resonance frequency evolution trends inherent in historical data makes it difficult for the system to predict potential resonance intervals for the next cycle, hindering adaptive focusing scanning. These problems are particularly pronounced in high-frequency, multi-channel, or embedded resource-constrained applications, necessitating a capacitance monitoring method that can integrate historical monitoring data, intelligently predict resonance intervals, and reconstruct the frequency sweeping strategy to overcome the technical bottleneck of the trade-off between real-time performance and accuracy. Summary of the Invention

[0006] This invention provides a capacitance monitoring method and system based on frequency sweep resonant points, aiming to solve the technical problems of poor real-time performance and large resonant point positioning errors caused by the mutual constraints between scanning speed and measurement accuracy in traditional linear frequency sweep technology under wide bandwidth coverage. This invention constructs a non-uniform adaptive frequency sweep strategy, combines a two-stage resonant point detection mechanism of coarse and fine sweeps, and introduces a dynamic frequency window adjustment algorithm based on historical resonant trajectory prediction to achieve rapid and high-precision locking of the resonant frequency of the measured capacitor. This significantly improves the response speed and stability of the monitoring system while ensuring measurement accuracy.

[0007] This invention provides a capacitance monitoring method based on a swept-frequency resonant point, comprising:

[0008] An excitation signal within an initial frequency range is applied to the capacitor under test, the initial frequency range covering a preset theoretical resonant frequency band;

[0009] A coarse-grained frequency sweep is performed on the initial frequency range with a first preset step size, and impedance amplitude response data corresponding to each frequency point is collected to form a coarse-sweep impedance spectrum.

[0010] Based on the coarse impedance spectrum, the approximate resonant frequency range corresponding to the minimum impedance amplitude is identified. The approximate resonant frequency range consists of a number of consecutive frequency points, and its width is an integer multiple of the first preset step size.

[0011] A fine-grained frequency sweep is performed on the coarse resonant frequency range with a second preset step size, which is smaller than the first preset step size. Impedance phase and amplitude response data corresponding to each frequency point are collected to form a finely swept impedance spectrum.

[0012] Based on the finely scanned impedance spectrum, the precise resonant frequency corresponding to the minimum impedance amplitude point is calculated using the quadratic interpolation method.

[0013] Based on the precise resonant frequency and combined with the pre-stored resonant frequency-capacitance value mapping table, the current capacitance value of the capacitor under test is determined.

[0014] The precise resonant frequency obtained from this monitoring is compared with the historical resonant frequency sequence over time to generate the initial frequency range for the next frequency sweep. The initial frequency range is centered on the moving average of the historical resonant frequencies, and its half-width is dynamically adjusted by the standard deviation of the historical resonant frequencies.

[0015] In one embodiment of the present invention, the first preset step size is 1 / 50 of the total width of the preset theoretical resonant frequency band, and the second preset step size is 1 / 10 of the first preset step size.

[0016] As one embodiment of the present invention, the process of identifying the coarse resonant frequency range specifically includes: traversing the impedance amplitude of all frequency points in the coarse sweep impedance spectrum, finding the global minimum point, and expanding two adjacent frequency points to both sides of it as the center to form a coarse resonant frequency range containing five consecutive frequency points.

[0017] As one embodiment of the present invention, the quadratic interpolation method uses three-point parabolic fitting, selects the frequency point with the smallest impedance amplitude in the finely scanned impedance spectrum and its two adjacent frequency points, constructs a quadratic function model, and solves the abscissa of the minimum point of the quadratic function as the accurate resonant frequency.

[0018] In one embodiment of the present invention, the resonant frequency-capacitance value mapping table is obtained through offline calibration. The calibration process includes: measuring the resonant frequency of a standard capacitor bank under the same test circuit conditions, and establishing a one-to-one lookup table, wherein the resolution of the lookup table is not less than 0.1 picofarads.

[0019] As one embodiment of the present invention, the historical resonant frequency sequence stores the precise resonant frequencies of the most recent 30 effective monitoring periods, the moving average is the arithmetic mean of the most recent 10 resonant frequencies, and the standard deviation is the sample standard deviation of the most recent 20 resonant frequencies relative to their mean.

[0020] As one embodiment of the present invention, the half-width of the initial frequency range of the next cycle sweep is equal to the standard deviation multiplied by 3, and the minimum value of the half-width is not less than the total width of the fine sweep frequency range, and the maximum value does not exceed 1 / 20 of the total width of the preset theoretical resonant frequency band.

[0021] This invention provides a capacitance monitoring system based on a swept-frequency resonant point, comprising:

[0022] The excitation signal generation unit is used to generate a frequency-adjustable sinusoidal excitation signal and apply it to the resonant circuit where the capacitor under test is located.

[0023] Impedance response acquisition unit is used to synchronously acquire the voltage and current signals of the resonant circuit at each excitation frequency, and calculate the impedance amplitude and phase response data.

[0024] The coarse sweep control unit is used to control the excitation signal generation unit to sweep the frequency within the initial frequency range with a first preset step size, and to receive the coarse sweep impedance spectrum output by the impedance response acquisition unit.

[0025] The coarse resonance interval identification unit is used to analyze the coarse sweep impedance spectrum and locate the coarse resonance frequency interval corresponding to the minimum impedance amplitude.

[0026] The fine sweep control unit is used to control the excitation signal generation unit to perform frequency sweeping in the coarse resonant frequency range with a second preset step size, and to receive the fine sweep impedance spectrum output by the impedance response acquisition unit.

[0027] The precise resonant frequency calculation unit is used to calculate the precise resonant frequency based on the finely scanned impedance spectrum using the quadratic interpolation method.

[0028] The capacitance value calculation unit is used to query the pre-stored resonant frequency-capacitance value mapping table according to the precise resonant frequency and output the current capacitance value of the capacitor under test.

[0029] The dynamic frequency window update unit is used to store the accurate resonant frequency obtained from this monitoring into the historical resonant frequency sequence, and calculate the initial frequency range of the next cycle of frequency sweep based on the sequence, and feed it back to the coarse sweep control unit as the start and end frequency boundaries of the next round of coarse sweep.

[0030] In one embodiment of the present invention, the excitation signal generation unit is composed of a direct digital frequency synthesizer with a frequency resolution of 0.1 Hz and a frequency switching setup time of no more than 10 microseconds.

[0031] As one embodiment of the present invention, the impedance response acquisition unit includes a high-precision differential voltage amplifier, a current sampling resistor, a dual-channel synchronous analog-to-digital converter, and a digital quadrature demodulation module. The dual-channel synchronous analog-to-digital converter has a sampling rate of 100 MHz and a resolution of not less than 16 bits. The digital quadrature demodulation module is used to extract the real and imaginary parts of the fundamental component from the sampled data, and then calculate the impedance amplitude and phase.

[0032] In one embodiment of the present invention, the coarse scan control unit and the fine scan control unit share the same frequency control bus, but have independent sweep frequency parameter registers, which respectively store the first preset step size, the initial frequency range, the second preset step size, and the coarse resonant frequency range.

[0033] As one embodiment of the present invention, the dynamic frequency window update unit has a built-in circular buffer for cyclically storing the most recent 30 accurate resonant frequencies. Its update operation is triggered after each capacitance value calculation is completed, and the calculation of the moving average and standard deviation is performed by a fixed-point coprocessor to ensure real-time performance.

[0034] Furthermore, the system also includes a temperature compensation module, which is used to monitor the ambient temperature of the resonant circuit in real time, correct the calculated accurate resonant frequency according to the pre-stored temperature-resonant frequency offset curve, and then input it to the capacitance value calculation unit.

[0035] Furthermore, the temperature compensation module uses a platinum resistance temperature sensor with a temperature measurement range of -40℃ to 125℃ and a temperature measurement accuracy of ±0.5℃. The temperature-resonance frequency offset curve is obtained by calibrating a standard capacitor at multiple temperature points in a constant temperature chamber and stored in a non-volatile memory in the form of a piecewise linear interpolation table.

[0036] Furthermore, the system also includes an anomaly detection unit, which is used to determine whether the precise resonant frequency obtained in this monitoring exceeds three times the standard deviation of the historical resonant frequency sequence. If it does, it is marked as abnormal data and a full-band rescan process is triggered, that is, a coarse scan is performed again with the initial theoretical resonant frequency band to eliminate misjudgments caused by external interference or device mutations.

[0037] Furthermore, the full-band rescan process is only initiated after two consecutive anomaly detections to avoid frequent full scans caused by occasional noise and to ensure system operating efficiency.

[0038] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0039] This invention introduces a two-stage frequency sweep architecture, which decomposes wideband search into two stages: coarse sweep and fine sweep, effectively avoiding the inherent contradiction between speed and accuracy in traditional single-stage linear frequency sweep.

[0040] The coarse scan stage rapidly narrows the search range of the resonant point with a large step size, while the fine scan stage achieves high-resolution positioning with a small step size in a local narrow band. The two work together to ensure that the resonant frequency measurement accuracy is better than 0.01 Hz, while shortening the single complete monitoring cycle to less than 1 / 5 of the traditional method.

[0041] The dynamic frequency window update mechanism based on historical resonance trajectories proposed in this invention utilizes the continuous and slow characteristics of the measured capacitance change to intelligently predict the effective search interval for the next cycle, further reducing the scanning of invalid frequency points and significantly improving the real-time performance and energy efficiency of long-term monitoring. The integration of the temperature compensation module effectively suppresses the interference of ambient temperature drift on the resonant frequency measurement, improving the stability and reliability of the system under complex operating conditions.

[0042] The anomaly detection and full-band rescan mechanism enhance the system's robustness against sudden faults or strong external interference, ensuring the reliability of monitoring results. In summary, this invention achieves comprehensive optimization of measurement accuracy, response speed, environmental adaptability, and system robustness in the field of online capacitor monitoring, making it suitable for applications requiring high reliability, such as insulation status monitoring of power equipment, industrial process control, and precision instrument calibration. Attached Figure Description

[0043] Figure 1This is a schematic diagram of the overall technical architecture of the capacitance monitoring method and system based on frequency sweep resonant points proposed in this invention;

[0044] Figure 2 This is a schematic diagram illustrating the core principle framework of the non-uniform adaptive frequency sweeping strategy and dynamic frequency window adjustment algorithm in this invention.

[0045] Figure 3 This is a flowchart illustrating the logic of the coarse and fine two-stage resonant point detection mechanism in this invention.

[0046] Figure 4 This is a flowchart illustrating the data processing flow of the impedance response acquisition and accurate resonant frequency calculation module in this invention.

[0047] Figure 5 This is a logical framework diagram of the dynamic frequency window update and anomaly detection mechanism driven by historical resonant frequency sequence in this invention.

[0048] Figure 6 This is a schematic diagram of the multi-level interaction relationship and data flow between the terminal monitoring unit and the temperature compensation and full-band rescan control modules in this invention. Detailed Implementation

[0049] Please refer to Figures 1 to 6 This invention provides a capacitance monitoring method and system based on frequency sweep resonant points, aiming to solve the technical problems of poor real-time performance and large resonant point positioning errors caused by the mutual constraint between scanning speed and measurement accuracy in traditional linear frequency sweep technology under wide bandwidth coverage. This method constructs a non-uniform adaptive frequency sweep strategy, combines a two-stage resonant point detection mechanism of coarse and fine sweeps, and introduces a dynamic frequency window adjustment algorithm based on historical resonant trajectory prediction to achieve rapid and high-precision locking of the resonant frequency of the capacitor under test. This significantly improves the response speed and stability of the monitoring system while ensuring measurement accuracy.

[0050] The method includes the following steps:

[0051] S1. Apply an excitation signal within an initial frequency range to the capacitor under test, wherein the initial frequency range covers a preset theoretical resonant frequency band;

[0052] S2. Perform a coarse-grained frequency sweep on the initial frequency range with a first preset step size, collect impedance amplitude response data corresponding to each frequency point, and form a coarse-sweep impedance spectrum.

[0053] S3. Based on the coarse sweep impedance spectrum, identify the coarse resonant frequency range corresponding to the minimum impedance amplitude. The coarse resonant frequency range consists of a number of consecutive frequency points, and its width is an integer multiple of the first preset step size.

[0054] S4. Perform fine-grained frequency sweeping on the coarse resonant frequency range with a second preset step size, where the second preset step size is smaller than the first preset step size. Collect impedance phase and amplitude response data corresponding to each frequency point to form a finely swept impedance spectrum.

[0055] S5. Based on the finely scanned impedance spectrum, the precise resonant frequency corresponding to the minimum impedance amplitude point is calculated using the quadratic interpolation method.

[0056] S6. Based on the precise resonant frequency and the pre-stored resonant frequency-capacitance value mapping table, determine the current capacitance value of the capacitor under test.

[0057] S7. Compare the precise resonant frequency obtained from this monitoring with the historical resonant frequency sequence over time to generate the initial frequency range for the next frequency sweep. The initial frequency range is centered on the moving average of the historical resonant frequencies, and its half-width is dynamically adjusted by the standard deviation of the historical resonant frequencies.

[0058] In step S1, the excitation signal generating unit generates a frequency-adjustable sinusoidal excitation signal and applies it to the resonant circuit containing the capacitor under test. The initial frequency range is a preset theoretical resonant frequency band, which is calculated based on the nominal value of the capacitor under test and its allowable deviation range, combined with the fixed inductance parameters in the resonant circuit. For example, if the nominal value of the capacitor under test is 100 picofarads, the allowable deviation is ±10%, and the circuit inductance is 10 microhenries, then the theoretical resonant frequency range is 47 MHz to 53 MHz. This initial frequency range serves as the sweep boundary for the first monitoring or full-band rescan, ensuring coverage of all possible resonant point locations.

[0059] In step S2, the coarse scan control unit drives the excitation signal generation unit to perform frequency increment scanning within the initial frequency range with a first preset step size. The first preset step size is set to 1 / 50 of the total width of the preset theoretical resonant frequency band. Taking a center frequency of 50 MHz and a bandwidth of 6 MHz as an example, the first preset step size is 120 kHz. At each frequency point, the impedance response acquisition unit synchronously acquires the voltage and current signals of the resonant circuit and calculates the impedance amplitude in real time. The calculation of the impedance amplitude is based on digital quadrature demodulation technology: the dual-channel synchronous analog-to-digital converter performs 16-bit quantization of the voltage and current signals at a sampling rate of 100 MHz, and the digital quadrature demodulation module performs fundamental component extraction on the sampled data in each excitation cycle to obtain the real and imaginary parts of the voltage and current, and then calculates the impedance amplitude as the modulus of the voltage complex number divided by the current complex number. The impedance amplitudes at all frequency points are arranged in frequency order to form a coarse scan impedance spectrum.

[0060] In step S3, the coarse resonance interval identification unit traverses all impedance amplitude data in the coarse sweep impedance spectrum to find the global minimum point. This minimum point corresponds to the lowest impedance frequency observed during the coarse sweep, i.e., the position closest to the true resonance point. To ensure that the coarse resonance frequency interval includes the true resonance point, two adjacent frequency points are extended to both sides of the frequency axis with this minimum point as the center, forming a coarse resonance frequency interval containing five consecutive frequency points. The total width of this interval is four first preset step sizes, i.e., 480 kHz. This design takes into account the resonance point offset that may be caused by a large coarse sweep step size, and ensures that the search range in the fine sweep stage completely covers the true resonance point through local expansion.

[0061] In step S4, the fine scan control unit takes over frequency control, limiting the frequency sweep range of the excitation signal generation unit to the coarse resonant frequency range determined in step S3, and performs fine-grained scanning with a second preset step size. The second preset step size is 1 / 10 of the first preset step size, i.e., 12 kHz. Within this narrow band, the system not only acquires the impedance amplitude but also simultaneously records the impedance phase. The impedance phase is calculated from the phase angle difference between the complex voltage and current, and is used for subsequent anomaly detection or to assist in verifying the authenticity of the resonant point. The data generated during the fine scan process constitutes the fine scan impedance spectrum, whose frequency resolution is significantly higher than that of the coarse scan stage, providing a foundation for high-precision resonant frequency calculation.

[0062] In step S5, the precise resonant frequency calculation unit selects the frequency point with the smallest impedance amplitude from the finely scanned impedance spectrum, and takes the two frequency points adjacent to it on the left and right, for a total of three data points, for secondary interpolation calculation. Let these three frequency points be... The corresponding impedance amplitude is A quadratic function model is constructed by fitting a three-point parabola. The coefficients a, b, and c are uniquely determined by the coordinates of three points. The x-coordinate of the minimum point of this quadratic function is... This is the precise resonant frequency, and its calculation formula is: in This is the second preset step size. The formula approximates the derivative using finite difference to solve for the vertex position of the parabola. This method, without adding extra frequency sweep points, improves the accuracy of resonant frequency positioning to a level much smaller than the step size, typically achieving an accuracy better than 0.01 Hz.

[0063] In step S6, the capacitance calculation unit receives the precise resonant frequency. The system then queries a pre-stored resonant frequency-capacitance value mapping table. This mapping table is obtained through offline calibration: a series of standard capacitors (with a resolution of no less than 0.1 picofarads) are connected to the same test circuit, and their resonant frequencies are measured under constant temperature conditions to establish a one-to-one lookup table. The lookup table is stored using linear interpolation; when the input frequency is between two calibration points, the corresponding capacitance value is obtained through linear interpolation. The calculated result is the current capacitance value of the capacitor under test, which is output to the monitoring terminal or the host system.

[0064] In step S7, the dynamic frequency window update unit stores the precise resonant frequency obtained from this monitoring into the historical resonant frequency sequence. This sequence adopts a circular buffer structure, cyclically storing the precise resonant frequencies of the most recent 30 valid monitoring periods. After each new data is written, the system calculates the arithmetic mean of the most recent 10 resonant frequencies as a moving average. Simultaneously, the sample standard deviation of the most recent 20 resonant frequencies relative to their mean was calculated. The initial frequency range for the next frequency sweep cycle is... Centered on the standard deviation W, the half-width W is set to 3 times the standard deviation. To prevent the window from being too narrow or too wide, constraints are set: the minimum value of W is not less than the total width of the fine-scan frequency range (i.e., 4 times the second preset step size, 48 kHz), and the maximum value does not exceed 1 / 20 of the total width of the preset theoretical resonant frequency band (i.e., 300 kHz). This dynamic window mechanism utilizes the slow and continuous nature of capacitance changes to significantly reduce the frequency sweep range for subsequent monitoring, thereby improving efficiency.

[0065] The system includes an excitation signal generation unit, an impedance response acquisition unit, a coarse scan control unit, a coarse resonance interval identification unit, a fine scan control unit, a precise resonance frequency calculation unit, a capacitance value calculation unit, and a dynamic frequency window update unit.

[0066] The excitation signal generation unit consists of a direct digital frequency synthesizer with a frequency resolution of 0.1 Hz and a frequency switching settling time of no more than 10 microseconds. This unit receives frequency commands from the coarse scan control unit or the fine scan control unit, generates a sine wave signal of the corresponding frequency in real time, and drives the resonant circuit through a power amplifier.

[0067] The impedance response acquisition unit includes a high-precision differential voltage amplifier, a current sampling resistor, a dual-channel synchronous analog-to-digital converter (ADC), and a digital quadrature demodulation module. The differential voltage amplifier isolates and amplifies the voltage signal across the resonant circuit, while the current sampling resistor is connected in series in the circuit to acquire current information. The dual-channel synchronous ADC simultaneously samples both voltage and current signals at a sampling rate of 100 MHz and a resolution of 16 bits. The digital quadrature demodulation module performs a discrete Fourier transform or correlation detection on the sampled data within each excitation cycle, extracting the real and imaginary parts of the fundamental component, calculating the complex form of the impedance, and outputting the amplitude and phase.

[0068] The coarse scan control unit and the fine scan control unit share the same frequency control bus, but each is equipped with an independent sweep frequency parameter register. The register of the coarse scan control unit stores the start and end frequencies of the initial frequency range and the first preset step size; the register of the fine scan control unit stores the boundary of the coarse resonant frequency range and the second preset step size. The two units coordinate their work through a state machine: after the coarse scan is completed, the coarse resonant range identification is triggered, and the identification result is written to the fine scan register to start the fine scan process.

[0069] The dynamic frequency window update unit has a built-in circular buffer with a capacity of 30 frequency values. After each capacitance value calculation, the new precise resonant frequency is written to the end of the buffer, and the oldest data is automatically overwritten. The calculation of the moving average and standard deviation is performed by a fixed-point coprocessor, avoiding the latency caused by floating-point operations, ensuring that the window update is completed within milliseconds, and the new initial frequency range is written to the register of the coarse scan control unit for use in the next round of monitoring.

[0070] The system further includes a temperature compensation module. This module uses a platinum resistance temperature sensor with a temperature measurement range of -40℃ to 125℃ and a measurement accuracy of ±0.5℃. The sensor is installed close to the resonant circuit and outputs the ambient temperature in real time. The temperature-resonant frequency offset curve is obtained by calibrating a standard capacitor at multiple temperature points in a constant temperature chamber and stored in non-volatile memory as a piecewise linear interpolation table. After the accurate resonant frequency is calculated, the temperature compensation module looks up the frequency offset in the table based on the current temperature. and to Make corrections: Corrected frequency The value is input to the capacitance calculation unit to eliminate capacitance measurement errors caused by temperature drift.

[0071] The system also includes an anomaly detection unit. This unit, after obtaining the precise resonant frequency each time, determines whether it exceeds three times the standard deviation of the historical resonant frequency sequence, i.e., whether it meets the requirements. If the conditions are met, the current data is marked as abnormal. The abnormality detection unit sets a counter, and the full-band rescan process is only triggered when two consecutive abnormality markers appear. The full-band rescan process forcibly resets the initial frequency range to the preset theoretical resonant frequency band and re-executes steps S2 to S7. This mechanism avoids misjudgments caused by occasional noise or transient interference, and also prevents frequent full scans from affecting system efficiency.

[0072] In summary, this embodiment constructs a high-precision, high-real-time, and high-robust capacitance monitoring scheme through a two-level frequency sweep architecture, dynamic frequency window prediction, temperature compensation, and an anomaly rescanning mechanism. During a single monitoring cycle, the coarse sweep stage quickly locates the resonant region, the fine sweep stage locks the resonant point with high resolution, and secondary interpolation further improves accuracy. In long-term operation, historical data-driven window adaptation significantly reduces invalid scans; temperature compensation suppresses environmental interference; and anomaly detection ensures data reliability. This scheme is suitable for applications with stringent requirements for accuracy and stability, such as online insulation monitoring of power equipment, calibration of industrial capacitance sensors, and aging assessment of precision electronic components.

Claims

1. A method of capacitance monitoring based on swept frequency resonance point, characterized in that, The method comprises the following steps: An excitation signal in an initial frequency range is applied to the measured capacitor, and the initial frequency range covers a preset theoretical resonance frequency band; A coarse-grained sweep is performed on the initial frequency range with a first preset step, and impedance amplitude response data corresponding to each frequency point is collected to form a coarse-sweep impedance spectrum; Based on the coarse-sweep impedance spectrum, a rough resonance frequency interval corresponding to the minimum impedance amplitude is identified, the rough resonance frequency interval is composed of a plurality of continuous frequency points, and the width of the rough resonance frequency interval is an integer multiple of the first preset step; A fine-grained sweep is performed on the rough resonance frequency interval with a second preset step, the second preset step is smaller than the first preset step, and impedance phase and amplitude response data corresponding to each frequency point are collected to form a fine-sweep impedance spectrum; Based on the fine-sweep impedance spectrum, a quadratic interpolation method is used to calculate an accurate resonance frequency corresponding to the minimum impedance amplitude point; According to the accurate resonance frequency, a current capacitance value of the measured capacitor is determined in combination with a pre-stored resonance frequency-capacitance value mapping table; The accurate resonance frequency obtained in this monitoring is compared with a historical resonance frequency sequence in time sequence to generate an initial frequency range for the next cycle sweep, the initial frequency range is centered on a moving average of the historical resonance frequency, and a half-width of the initial frequency range is dynamically adjusted by a standard deviation of the historical resonance frequency.

2. The method of claim 1, wherein, The first preset step is 1 / 50 of the total width of the preset theoretical resonance frequency band, and the second preset step is 1 / 10 of the first preset step.

3. The method of claim 2, wherein, The identification process of the rough resonance frequency interval specifically comprises: traversing the impedance amplitudes of all frequency points in the coarse-sweep impedance spectrum, finding a global minimum value point, and expanding two adjacent frequency points on both sides of the global minimum value point to form a rough resonance frequency interval containing five continuous frequency points.

4. The method of claim 3, wherein, The quadratic interpolation method uses a three-point parabolic fitting, selects a frequency point with the minimum impedance amplitude in the fine-sweep impedance spectrum and two adjacent frequency points on the left and right of the frequency point, constructs a quadratic function model, and solves the minimum value point of the horizontal coordinate of the quadratic function as the accurate resonance frequency.

5. The method of claim 4, wherein, The resonance frequency-capacitance value mapping table is obtained through offline calibration, and the calibration process comprises: measuring the resonance frequencies of a standard capacitor group under the same test circuit conditions to establish a one-to-one lookup table, and the resolution of the lookup table is not less than 0.1 picofarad.

6. The method of claim 5, wherein, The historical resonance frequency sequence stores the accurate resonance frequencies of the last 30 effective monitoring cycles, the moving average is an arithmetic mean of the last 10 resonance frequencies, and the standard deviation is a sample standard deviation of the last 20 resonance frequencies relative to the mean value.

7. The method of claim 6, wherein, The half-width of the initial frequency range for the next cycle sweep is equal to the standard deviation multiplied by 3, the minimum value of the half-width is not less than the total width of the fine-sweep frequency interval, and the maximum value of the half-width is not more than 1 / 20 of the total width of the preset theoretical resonance frequency band.

8. The method of claim 7, wherein, Further comprising: The environmental temperature of the resonance loop is monitored in real time, and the accurate resonance frequency calculated is corrected according to a pre-stored temperature-resonance frequency offset curve and then input to the capacitance value calculation unit.

9. The method of claim 8, wherein, Further comprising: It is judged whether the accurate resonance frequency obtained in this monitoring is beyond the range of 3 times the standard deviation of the historical resonance frequency sequence. If it is beyond for two times in succession, the full-band re-scan process is triggered, that is, the initial theoretical resonance frequency band is re-scanned.

10. A capacitance monitoring system based on swept frequency resonance point, characterized by, The method comprises the steps of: An excitation signal generation unit is configured to generate a frequency-adjustable sine wave excitation signal and apply it to a resonance circuit in which the measured capacitance is located; An impedance response acquisition unit is configured to synchronously acquire voltage and current signals of the resonance circuit at each excitation frequency and calculate impedance amplitude and phase response data; A coarse-scan control unit is configured to control the excitation signal generation unit to perform frequency sweeping in an initial frequency range with a first preset step size and receive a coarse-scan impedance spectrum output by the impedance response acquisition unit; A coarse resonance interval identification unit is configured to analyze the coarse-scan impedance spectrum and locate a coarse resonance frequency interval corresponding to the minimum impedance amplitude value; A fine-scan control unit is configured to control the excitation signal generation unit to perform frequency sweeping in the coarse resonance frequency interval with a second preset step size and receive a fine-scan impedance spectrum output by the impedance response acquisition unit; An accurate resonance frequency calculation unit is configured to calculate an accurate resonance frequency based on the fine-scan impedance spectrum by using a quadratic interpolation method; A capacitance value solving unit is configured to query a pre-stored resonance frequency-capacitance value mapping table according to the accurate resonance frequency and output a current capacitance value of the measured capacitance; A dynamic frequency window updating unit is configured to store the accurate resonance frequency obtained in this monitoring into a historical resonance frequency sequence and calculate an initial frequency range for the next period of frequency sweeping based on the sequence and feed back to the coarse-scan control unit as the starting and ending frequency boundaries for the next round of coarse scanning.

Images