Rydberg atom-based low-frequency voltage measurement signal processing system

Abstract

A Rydberg atom-based low-frequency voltage measurement signal processing system, comprising: an ADC data acquisition module used for converting EIT spectrum data into an electrical signal and performing data acquisition; a wavelet transform denoising module used for performing denoising processing on data acquired by the ADC data acquisition module; an automatic peak finding module used for performing automatic peak finding on the denoised data to obtain a peak position difference; and a voltage inversion module used for performing voltage inversion and frequency calculation on the basis of the peak position difference to complete Rydberg atom-based low-frequency voltage measurement. The ADC data acquisition module and the wavelet transform denoising module are implemented by a PL end; and the automatic peak finding module and the voltage inversion module are implemented by a PS end. The system has a simple structure, good stability, and good real-time performance, satisfies the requirements for miniaturization and low power consumption, and lays a foundation for engineering application of the Rydberg atom-based low-frequency voltage measurement system.

Description

A Rydberg Atom Low-Frequency Voltage Measurement Signal Processing System

[0001] This application claims priority to Chinese Patent Application No. 202411560445.5, filed on November 4, 2024, entitled "A Rydberg Atom Low-Frequency Voltage Measurement Signal Processing System", the entire contents of which are incorporated herein by reference. Technical Field

[0002] This invention relates to the field of electromagnetically induced transparency (EIT) spectral line analysis and processing, specifically to a Rydberg atom low-frequency voltage measurement signal processing system. Background Technology

[0003] Voltage, as a crucial electrical parameter in power systems, reflects the status of transmission lines and the operational status of electrical equipment. Research and development of non-contact, precise voltage measurement are of great significance for monitoring operational status and diagnosing faults in power systems. Rydberg atoms, with their large principal quantum numbers, have attracted considerable attention as a novel quantum sensing medium. Due to their non-metallic, non-invasive, non-contact, high-precision, and high-sensitivity measurement characteristics, quantum measurement techniques based on the Rydberg atom EIT-Stark effect have broad application prospects in power system voltage measurement. The detection and processing of EIT spectral lines is one of the important methods for analyzing the EIT-Stark effect.

[0004] Meanwhile, many voltage parameters in power systems and related fields are low-frequency voltages. Low-frequency voltage signals are also important indicators of the operating status and faults of power systems. In recent years, many research teams have gradually solved the problems of signal loading and measurement implementation for low-frequency voltage signals. However, in the Rydberg atomic quantum measurement platform in the laboratory, a data acquisition card is usually used in conjunction with a host computer with LabVIEW software to process the measurement signal. The entire signal processing system is redundant and difficult to transfer, making it difficult to adapt to engineering application scenarios outside the laboratory. This is also a point that is usually overlooked in theoretical research and in the study of measurement accuracy, sensitivity, and influencing factors.

[0005] To realize the engineering application of low-frequency voltage measurement of the Rydberg atom and make the measurement platform adaptable to more application scenarios, higher requirements must be placed on the miniaturization and embedding of the measurement signal processing system, while ensuring the efficiency and real-time performance of signal processing and computation. Summary of the Invention

[0006] To address the aforementioned shortcomings in the existing technology, the present invention provides a Rydberg atom low-frequency voltage measurement signal processing system that solves the problem that the existing Rydberg atom low-frequency voltage measurement signal processing system has redundant performance and is difficult to migrate, making it difficult to adapt to engineering application scenarios outside the laboratory.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows:

[0008] A low-frequency voltage measurement signal processing system for the Rydberg atom is provided, comprising:

[0009] The ADC data acquisition module is used to convert the EIT spectrum data from the photodetector into electrical signals and acquire the data.

[0010] The wavelet transform denoising module is used to denoise the data acquired by the ADC data acquisition module to obtain denoised data.

[0011] The automatic peak finding module is used to automatically find the peaks in the denoised data and obtain the peak position difference;

[0012] The voltage inversion module is used to perform voltage inversion and frequency calculation based on the peak position difference, and to complete the low-frequency voltage measurement of the Rydberg atom.

[0013] The host computer is used to configure the parameters of the wavelet transform denoising module and the voltage inversion module;

[0014] The ADC data acquisition module and wavelet transform denoising module are implemented by the PL terminal; the automatic peak finding module and voltage inversion module are implemented by the PS terminal.

[0015] Furthermore, it also includes:

[0016] The DDR3 memory control module is used to control the DDR3 memory and store the voltage inversion results into the specified memory space.

[0017] The Ethernet communication module is used to send data stored in a specified memory space to the host computer.

[0018] The phase-locked loop module generates a 100MHz clock for the DDR3 memory control module and a 50MHz clock for other modules.

[0019] The phase-locked loop module is implemented by the PL terminal; the DDR3 storage control module and the Ethernet communication module are implemented by the PS terminal.

[0020] Furthermore, the specific method for denoising the data acquired by the ADC data acquisition module includes the following steps:

[0021] A1. Perform low-pass filtering and high-pass filtering on the data acquired by the ADC data acquisition module to obtain the first-layer approximation coefficients and the first-layer detail coefficients respectively.

[0022] A2. The first-level approximation coefficients are decomposed into subbands through iteration;

[0023] A3. Perform an inverse transformation on each subband to obtain the reconstructed subband;

[0024] A4. Add all the reconstructed subbands together to obtain the inverse approximation coefficients;

[0025] A5. Set the noise reduction threshold, and set the first layer detail coefficients that are less than the noise reduction threshold to zero to obtain the noise-reduced detail coefficients.

[0026] A6. The inverse approximation coefficients and the denoised detail coefficients are merged by inverse wavelet transform to obtain the denoised data, namely the denoised EIT spectrum.

[0027] Furthermore, low-pass filtering and high-pass filtering are implemented using the bior6_8 wavelet function; the number of sub-bands decomposed into the first-level approximation coefficients is 4.

[0028] Furthermore, the denoising threshold is configured using a particle swarm optimization algorithm.

[0029] Furthermore, the specific method for automatically finding peaks and obtaining peak position differences in the denoised data is as follows:

[0030] A quadratic polynomial least squares method is used to fit the data points as the width of the fit. The fitted peak amplitude is compared with the amplitude threshold, and the fitted peak amplitudes that are less than the amplitude threshold are removed. The difference between the maximum position value of the peak and the position value corresponding to the maximum amplitude is taken as the peak position difference.

[0031] Furthermore, the specific method for voltage inversion and frequency calculation based on peak position difference is as follows:

[0032] According to the formula:

[0033] Obtain the inverted voltage U, which is the low-frequency voltage value of the Rydberg atom; where k′ is the fitting coefficient; Δ Stark This represents the peak position difference.

[0034] According to the formula:

[0035] Obtain the frequency f of the low-frequency voltage of the Rydberg atom; where t is the time interval between two consecutive points where the voltage amplitude reaches its maximum value.

[0036] Furthermore, the specific method for configuring the parameters of the wavelet transform denoising module and the voltage inversion module includes the following sub-steps:

[0037] B1. Conduct a test experiment to observe the peak frequency shift of the EIT spectrum using a standard source with a known electric field, and sample the ADC signal to obtain the field strength test signal;

[0038] B2. Initially configure the wavelet transform denoising module according to the standard source parameters and field strength test signal;

[0039] B3. Data is acquired through the ADC data acquisition module, and the data acquired by the ADC data acquisition module is denoised through the wavelet transform denoising module after initial configuration to obtain the denoised data.

[0040] B4. Within the peak-finding data range set by the host computer, the automatic peak-finding module automatically finds the corresponding peak value from the denoised data.

[0041] B5. Determine if the number of peaks found by the automatic peak finding module is greater than 1. If yes, proceed to step B6; otherwise, lower the amplitude threshold and return to step B4.

[0042] B6. Determine whether the number of peaks found by the automatic peak finding module is less than or equal to 3. If yes, proceed to step B7; otherwise, determine that an erroneous peak has been detected, increase the amplitude threshold, and return to step B5.

[0043] B7. Obtain the difference between the position value corresponding to the maximum value of the peak and the position value corresponding to the maximum value of the amplitude to get the peak position difference;

[0044] B7. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters;

[0045] B8. Compare the measured standard source voltage parameters with the actual standard source voltage parameters, and adjust the voltage inversion module parameters until the error between the measured standard source voltage parameters and the actual standard source voltage parameters is within the set range.

[0046] The beneficial effects of this invention are as follows:

[0047] 1. This system has a simple structure, good stability and real-time performance, and meets the requirements of miniaturization and low power consumption, laying the foundation for the engineering application of the Rydberg atomic low-frequency voltage measurement system.

[0048] 2. The wavelet transform denoising module uses a bior6_8 wavelet basis with a decomposition level of 5 for processing. The threshold is configured by the host computer through particle swarm optimization. Furthermore, the entire wavelet transform denoising process is implemented on the PL side. Due to the use of a hardware-based design approach, the algorithm executes quickly and can rapidly and effectively extract the feature information of EIT spectral lines. Attached Figure Description

[0049] Figure 1 is a structural block diagram of this system;

[0050] Figure 2 shows the changes in Stark frequency shift and transmission peak amplitude during voltage amplitude variation.

[0051] Figure 3 shows the original EIT spectrum before noise reduction and peak finding in the embodiment;

[0052] Figure 4 shows the effect of EIT spectral denoising and peak finding in the embodiment. Detailed Implementation

[0053] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0054] As shown in Figure 1, the Rydberg atom low-frequency voltage measurement signal processing system includes:

[0055] The ADC data acquisition module is used to convert the EIT spectrum data from the photodetector into electrical signals and acquire the data.

[0056] The wavelet transform denoising module is used to denoise the data acquired by the ADC data acquisition module to obtain denoised data.

[0057] The automatic peak finding module is used to automatically find the peaks in the denoised data and obtain the peak position difference;

[0058] The voltage inversion module is used to perform voltage inversion and frequency calculation based on the peak position difference, and to complete the low-frequency voltage measurement of the Rydberg atom.

[0059] The host computer is used to configure the parameters of the wavelet transform denoising module and the voltage inversion module;

[0060] The DDR3 memory control module is used to control the DDR3 memory and store the voltage inversion results into the specified memory space.

[0061] The Ethernet communication module is used to send data stored in a specified memory space to the host computer.

[0062] The phase-locked loop module generates a 100MHz clock for the DDR3 memory control module and a 50MHz clock for other modules.

[0063] In this embodiment, the PL terminal uses an FPGA chip, and the PS terminal uses an ARM chip. The PL terminal is used to implement the ADC data acquisition module, wavelet transform denoising module, and phase-locked loop module. The PS terminal is used to implement the automatic peak finding module, voltage inversion module, frequency calculation module, DDR3 memory control module, and Ethernet communication module. In addition, the conversion of the probe light signal to the EIT spectrum digital signal is achieved through a photodetector module, signal amplification module, and digital-to-analog converter module.

[0064] In practical implementation, to achieve engineering applications of Rydberg atom low-frequency voltage measurement and adapt the measurement platform to more application scenarios, higher requirements must be placed on the miniaturization and embedding of the measurement signal processing system, while ensuring the efficiency and real-time performance of signal processing and computation. This embodiment integrates an FPGA and an ARM processor into the same ZYNQ chip, taking advantage of both hardware and software co-design. The ZYNQ series integrates the software programmability of an ARM processor with the hardware programmability of an FPGA, enabling not only critical analysis, hardware acceleration, and excellent performance-to-power ratio and design flexibility, but also highly integrating CPU, DSP, ASSP, and mixed-signal functions on a single device. The advantages of using the ZYNQ chip include: high integration of the development environment; AXI4 standard bus interconnect; convenient and fast data exchange between PS and PL; saving the layout area of ​​an ARM and the resource consumption of the communication bus; and high hardware integration.

[0065] The DDR3 memory control module uses the Xilinx ZYNQ7 Processing System IP core for parameter configuration to control the DDR3 memory on the PS side. The DDR3 memory is part of the ZYNQ onboard chip and is connected to the PS side through internal wiring to transmit control signals, data signals, and address signals.

[0066] To ensure the sequentiality and stability of data stream transmission, each module can be buffered by adding a FIFO module between modules. The FIFO module can also be configured using IP cores provided by Xilinx.

[0067] The specific method for denoising the data acquired by the ADC data acquisition module is to use the Undecimated Wavelet Transform (UWT) method for signal denoising, which includes the following steps:

[0068] A1. Perform low-pass filtering and high-pass filtering on the data acquired by the ADC data acquisition module to obtain the first-layer approximation coefficients and the first-layer detail coefficients respectively.

[0069] A2. The first-level approximation coefficients are decomposed into subbands through iteration;

[0070] A3. Perform an inverse transformation on each subband to obtain the reconstructed subband;

[0071] A4. Add all the reconstructed subbands together to obtain the inverse approximation coefficients;

[0072] A5. Set the noise reduction threshold, and set the first layer detail coefficients that are less than the noise reduction threshold to zero to obtain the noise-reduced detail coefficients.

[0073] A6. The inverse approximation coefficients and the denoised detail coefficients are merged by inverse wavelet transform to obtain the denoised data, namely the denoised EIT spectrum.

[0074] In this embodiment, filtering is essentially a convolution operation. Once the wavelet basis is selected, the parameters of the low-pass and high-pass filters are determined. Specifically, the low-pass and high-pass filters are implemented using the bior6_8 wavelet function. The number of sub-bands decomposed into the first-layer approximation coefficients is 4. That is, the first-layer approximation coefficients are decomposed into the second-layer approximation coefficients and the second-layer detail coefficients through low-pass and high-pass filtering. Similarly, the second-layer approximation coefficients are decomposed into the third-layer approximation coefficients and the third-layer detail coefficients, and so on, completing 4 iterations to obtain a total of five layers of approximation coefficients and five layers of detail coefficients. The approximation coefficients and detail coefficients obtained from the second layer onwards form a set of sub-bands. The reconstructed sub-bands are the approximation coefficients and detail coefficients corresponding to the second to fifth layers.

[0075] The denoising threshold is configured by the host computer using a particle swarm optimization algorithm. Furthermore, the entire wavelet transform denoising process is implemented on the PL (Programmer) side. Due to its hardware-based design, the algorithm executes quickly and efficiently, effectively extracting the feature information of the EIT spectral lines. The wavelet transform denoising module and the automatic peak finding module are connected via an AXI bus interface.

[0076] The specific method for automatically finding peaks and obtaining peak position difference for the denoised data is as follows: using three data points as the fitting width, perform a quadratic polynomial fitting using the least squares method, compare the fitted peak amplitude with the amplitude threshold, and remove the fitted peak amplitudes that are less than the amplitude threshold; the difference between the maximum position value of the peak and the position value corresponding to the maximum amplitude is taken as the peak position difference.

[0077] The automatic peak finding module can automatically set an appropriate threshold based on the number of detected peaks. The peak finding threshold is dynamically adjusted in steps based on the mean of the sampled data and the number of peaks to be found. The amplitude of the peak obtained from each peak finding is compared with the current threshold; if it is greater than the threshold, it is retained, and the amplitude and position information are recorded.

[0078] In practical implementation, in quantum mechanics, the interaction between an atomic system and an electric field E can be represented by the effective Hamiltonian:

[0079] H eff =H0+H S =W E ;

[0080] Where H0 is the Hamiltonian of the atomic system in the absence of an electric field, and W... E It is the energy of an atomic system under the influence of an electric field. And H... S It is the Hamiltonian that describes the Stark interaction under the influence of an electric field, and its expression is as follows:

[0081] Where μ ij It is the projection component of the electric dipole moment vector between quantum states i and j along the direction of the electric field.

[0082] Expanding W using fourth-order perturbation theory E W can be obtained E The expressions for each order are given. By analyzing the relationship between energy and field strength for different principal quantum numbers, the Stark level diagram can be obtained. Due to the presence of atomic cores in hydrogen-like atoms such as cesium, the quantum defect effect of atoms must be considered. Therefore, in the electric field range before the occurrence of energy level anti-crossover, the Stark frequency shift has a quadratic relationship with the applied electric field, expressed as:

[0083] Where Δ S tark represents the Stark frequency shift; α i (f) represents the polarizability of the Rydberg state i at frequency f. For a fixed Rydberg state, the theoretical expression for its polarizability is:

[0084] Where j represents all states that Rydberg state i is coupled to through the electric field, and f ij Γ is the transition frequency between state i and state j. j Let be the spontaneous decay frequency of state j.

[0085] This embodiment employs a gas chamber-mounted electrode method for voltage loading; therefore, the relationship between the electric field and voltage is considered when the electrode spacing is fixed. The formula for Stark frequency shift inversion voltage is as follows:

[0086] However, in engineering applications, due to the difficulty in achieving perfect uniformity of the electric field between the electrodes, ionization caused by coupled lasers, and uncertainties introduced by optical devices, inversion based on theoretical formulas is quite challenging. Compensation for multiple factors in the mathematical model is necessary to achieve good inversion accuracy. Furthermore, when the mathematical model is complex, porting it to the ZYNQ chip consumes excessive resources.

[0087] Therefore, this system, through a host computer, can obtain the m-values ​​derived from the D3 / 2 state by fitting test data based on standard source experiments. j =m separated from the 3 / 2 peak and the D5 / 2 state j = Δ, the frequency axis position difference between 1 / 2 peaks Stark The inversion coefficient k′ is derived from the relationship between the voltage U applied across the plates and the inversion coefficient k′. Therefore, the voltage inversion formula is as follows:

[0088] In this embodiment, the peak position difference is the displacement of the transmission peak in the EIT spectrum, and the coordinate axis of the displacement is a frequency quantity. The shift of the transmission peak is a characteristic of the Stark effect. Therefore, the peak position difference is actually the frequency shift Δ. Stark .

[0089] This method highly refines the mathematical relationships behind the complex quantum coherence mechanism, so that the voltage inversion module only involves multiplication and square root operations. It simplifies the calculation process when accuracy allows, saves the hardware and software resources of the ZYNQ chip, improves the real-time performance of the measurement system to a certain extent, and expands the frequency measurement range of the platform.

[0090] Figure 2 shows the changes in Stark frequency shift and transmission peak amplitude during voltage amplitude variation. As the voltage increases, the Stark shift increases, while the transmission peak amplitude decreases significantly.

[0091] The frequency calculation process includes a timer and continuously acquires amplitude data output from the peak-finding module. After acquiring the amplitude data, the time to reach the maximum amplitude point is recorded each time, and the time interval t between two consecutive arrivals at the maximum amplitude point is calculated. The period calculation formula is as follows:

[0092] Where f is the frequency of the low-frequency voltage of the Rydberg atom.

[0093] Since higher transmission peak amplitudes are easier to monitor than lower transmission peak amplitudes, this frequency calculation method can determine the frequency within one cycle, exhibiting good real-time performance.

[0094] In practical applications, due to the influence of different measurement environments and the uncertainties arising from factors such as the state of optical devices, lasers, and laser frequency locking effects, parameter initialization configuration is required before each formal measurement. The specific method for configuring the parameters of the wavelet transform denoising module and the voltage inversion module via a host computer includes the following sub-steps:

[0095] B1. Conduct a test experiment to observe the peak frequency shift of the EIT spectrum using a standard source with a known electric field, and sample the ADC signal to obtain the field strength test signal;

[0096] B2. Configure the initial denoising threshold of the wavelet transform denoising module according to the standard source parameters and field strength test signal;

[0097] B3. Data is acquired through the ADC data acquisition module, and the data acquired by the ADC data acquisition module is denoised through the wavelet transform denoising module after initial configuration to obtain the denoised data.

[0098] B4. Within the peak-finding data range set by the host computer, the automatic peak-finding module automatically finds the corresponding peak value from the denoised data.

[0099] B5. Determine if the number of peaks found by the automatic peak finding module is greater than 1. If yes, proceed to step B6; otherwise, lower the amplitude threshold and return to step B4.

[0100] B6. Determine whether the number of peaks found by the automatic peak finding module is less than or equal to 3. If yes, proceed to step B7; otherwise, determine that an erroneous peak has been detected, increase the amplitude threshold, and return to step B5.

[0101] B7. Obtain the difference between the position value corresponding to the maximum value of the peak and the position value corresponding to the maximum value of the amplitude to get the peak position difference;

[0102] B7. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters;

[0103] B8. Compare the measured standard source voltage parameters with the actual standard source voltage parameters, and adjust the voltage inversion module parameter k′ until the error between the measured standard source voltage parameters and the actual standard source voltage parameters is within the set range.

[0104] In this embodiment, during the parameter configuration of the wavelet transform denoising module and the voltage inversion module, two digital-to-analog conversion ports can be reserved on the ZYNQ platform for the analog trigger signal and the backup digital trigger signal.

[0105] In one embodiment of the present invention, as shown in Figures 3 and 4, the identification of the peak with the highest amplitude and the farthest offset is based on the highest and farthest peaks within a sinusoidal period, rather than the highest or farthest peaks at a single moment. The peak height under pressure is definitely much smaller than that without pressure. However, theoretically, the highest and farthest peaks can only appear on the two sides. Therefore, the present invention identifies and outputs the positions of the peaks on both sides. The small amplitude changes of the peaks are the result of wavelet analysis. For non-Gaussian and non-stationary signals, this system can better preserve the characteristics of the EIT spectral lines, highlight the shape and position of the EIT spectral peaks, and effectively filter out non-Gaussian noise caused by factors of the photodetector itself and environmental factors, thereby reducing the possibility of subsequent stages failing to find peaks or finding the wrong peaks.

[0106] Meanwhile, the peak-finding method is suitable for finding single strong peaks, has a fast processing speed, is applicable to non-Gaussian signals, and is suitable for applications that focus on the peak with the highest amplitude and the peak with the furthest frequency shift, rather than overlapping peaks. With the help of this system, it can easily and quickly detect two or three peaks with good results, achieving excellent voltage measurement performance. The measurement accuracy and sensitivity of the Rydberg atomic low-frequency voltage measurement platform using this system surpass many non-contact voltage measurement technologies on the market, and its development and improvement prospects remain broad.

[0107] Furthermore, the specific frequency range of low-frequency signals that this system can process depends on the laser's sweep frequency and the processing speed of the ZYNQ platform. From a signal processing perspective, the time required for wavelet transform denoising, peak finding, and automatic peak threshold adjustment must be much faster than the time required for non-negligible large changes in real-time voltage. Therefore, the frequency range can be determined comprehensively through multiple experiments and equipment operation experience.

[0108] If this system needs to be applied in a power system, the Ethernet control module can be optimized to facilitate interconnection with other electrical equipment, enabling it to encapsulate measurement data into data frames that conform to IEC 61850.

[0109] If this system needs to implement multi-channel signal processing on a single ZYNQ platform, wavelet transform denoising at the PL end will consume a lot of hardware resources. The utilization rate of PL end resources can be improved by time-division multiplexing to ensure measurement efficiency and real-time performance.

[0110] In summary, this invention has a simple structure, good stability and real-time performance, and meets the requirements of miniaturization and low power consumption, laying the foundation for the engineering application of the Rydberg atom low-frequency voltage measurement system.

Claims

1. A Rydberg atom low-frequency voltage measurement signal processing system, characterized in that, include: The ADC data acquisition module is used to convert the EIT spectrum data from the photodetector into electrical signals and acquire the data. The wavelet transform denoising module is used to denoise the data acquired by the ADC data acquisition module to obtain denoised data. The automatic peak finding module is used to automatically find the peaks in the denoised data and obtain the peak position difference; The voltage inversion module is used to perform voltage inversion and frequency calculation based on the peak position difference, and to complete the low-frequency voltage measurement of the Rydberg atom. The host computer is used to configure the parameters of the wavelet transform denoising module and the voltage inversion module; The ADC data acquisition module and wavelet transform denoising module are implemented by the PL terminal; the automatic peak finding module and voltage inversion module are implemented by the PS terminal.

2. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 1, characterized in that, Also includes: The DDR3 memory control module is used to control the DDR3 memory and store the voltage inversion results into the specified memory space. The Ethernet communication module is used to send data stored in a specified memory space to the host computer. The phase-locked loop module generates a 100MHz clock for the DDR3 memory control module and a 50MHz clock for other modules. The phase-locked loop module is implemented by the PL terminal; the DDR3 storage control module and the Ethernet communication module are implemented by the PS terminal.

3. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 1, characterized in that, The specific method for denoising the data acquired by the ADC data acquisition module includes the following steps: A1. Perform low-pass filtering and high-pass filtering on the data acquired by the ADC data acquisition module to obtain the first-layer approximation coefficients and the first-layer detail coefficients respectively. A2. The first-level approximation coefficients are decomposed into subbands through iteration; A3. Perform an inverse transformation on each subband to obtain the reconstructed subband; A4. Add all the reconstructed subbands together to obtain the inverse approximation coefficients; A5. Set the noise reduction threshold, and set the first layer detail coefficients that are less than the noise reduction threshold to zero to obtain the noise-reduced detail coefficients. A6. The inverse approximation coefficients and the denoised detail coefficients are merged by inverse wavelet transform to obtain the denoised data, namely the denoised EIT spectrum.

4. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 3, characterized in that, Low-pass and high-pass filtering are implemented using the bior6_8 wavelet function; the number of sub-bands decomposed into the first-level approximation coefficients is 4.

5. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 3, characterized in that, The denoising threshold is configured using the particle swarm optimization algorithm.

6. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 1, characterized in that, The specific method for automatically finding peaks and obtaining peak position differences in the denoised data is as follows: A quadratic polynomial least squares method is used to fit the data points as the width of the fit. The fitted peak amplitude is compared with the amplitude threshold, and the fitted peak amplitudes that are less than the amplitude threshold are removed. The difference between the maximum position value of the peak and the position value corresponding to the maximum amplitude is taken as the peak position difference.

7. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 6, characterized in that, Specific methods for voltage inversion and frequency calculation based on peak position difference: According to the formula: Obtain the inverted voltage U, which is the low-frequency voltage value of the Rydberg atom; where k′ is the fitting coefficient; Δ Stark This represents the peak position difference. According to the formula: Obtain the frequency f of the low-frequency voltage of the Rydberg atom; where t is the time interval between two consecutive points where the voltage amplitude reaches its maximum value.

8. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 6, characterized in that, The specific method for configuring the parameters of the wavelet transform denoising module and the voltage inversion module includes the following sub-steps: B1. Conduct a test experiment to observe the peak frequency shift of the EIT spectrum using a standard source with a known electric field, and sample the ADC signal to obtain the field strength test signal; B2. Initially configure the wavelet transform denoising module according to the standard source parameters and field strength test signal; B3. Data is acquired through the ADC data acquisition module, and the data acquired by the ADC data acquisition module is denoised through the wavelet transform denoising module after initial configuration to obtain the denoised data. B4. Within the peak-finding data range set by the host computer, the automatic peak-finding module automatically finds the corresponding peak value from the denoised data. B5. Determine if the number of peaks found by the automatic peak finding module is greater than 1. If yes, proceed to step B6; otherwise, lower the amplitude threshold and return to step B4. B6. Determine whether the number of peaks found by the automatic peak finding module is less than or equal to 3. If yes, proceed to step B7; otherwise, determine that an erroneous peak has been detected, increase the amplitude threshold, and return to step B5. B7. Obtain the difference between the position value corresponding to the maximum value of the peak and the position value corresponding to the maximum value of the amplitude to get the peak position difference; B7. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters; B8. Compare the measured standard source voltage parameters with the actual standard source voltage parameters, and adjust the voltage inversion module parameters until the error between the measured standard source voltage parameters and the actual standard source voltage parameters is within the set range.

Images