Rydberg atom low-frequency voltage measurement signal processing system

Abstract

The invention discloses a Rydberg atom low-frequency voltage measurement signal processing system, and the system comprises an ADC data collection module which is used for converting EIT spectrum data into an electric signal and carrying out the data collection; the wavelet transform denoising module is used for carrying out denoising processing on the data acquired by the ADC data acquisition module; the automatic peak searching module is used for performing automatic peak searching on the de-noised data to obtain a peak position difference; and the voltage inversion module is used for performing voltage inversion and frequency calculation according to the peak position difference and completing Rydberg atom low-frequency voltage measurement. Wherein the ADC data acquisition module and the wavelet transform denoising module are realized by a PL end; and the automatic peak searching module and the voltage inversion module are realized by a PS end. The system is simple in structure and good in stability and real-time performance, meets the requirements of miniaturization and low power consumption, and lays a foundation for engineering application of a Rydberg atom low-frequency voltage measurement system.

Description

Technical Field

[0001] The invention relates to the field of electromagnetically induced transparency (EIT) spectral line analysis and processing, and in particular to a Rydberg atom low-frequency voltage measurement signal processing system. Background Art

[0002] Voltage is an important electrical parameter in the power system, which can reflect the state of the transmission line and the operating status of electrical equipment. The research and development of non-contact accurate measurement of voltage is of great significance to the operating status monitoring and fault diagnosis in the power system. Rydberg atoms with a large principal quantum number have attracted the attention of many researchers as a new type of quantum sensing medium. Due to its non-metallic, non-invasive, non-contact, high-precision and high-sensitivity measurement characteristics, quantum measurement technology based on the EIT-Stark effect of Rydberg atoms has broad application prospects in power system voltage measurement. The detection and processing of EIT spectral line signals is one of the important means to analyze its EIT-Stark effect.

[0003] At the same time, many voltage parameters in power systems and related fields are low-frequency voltages. Low-frequency voltage signals are also an important reflection 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 signal measurement. However, in the Rydberg atom quantum measurement platform in the laboratory, an acquisition card is usually used in conjunction with a host computer installed with LabVIEW software to process the measurement signal. The performance of the entire signal processing system is redundant and difficult to migrate, making it difficult to adapt to engineering application scenarios outside the laboratory. This is also a point that is often overlooked in theoretical research and research on measurement accuracy, sensitivity, and influencing factors.

[0004] In order to realize the engineering application of low-frequency voltage measurement of Rydberg atoms and make the measurement platform adapt to more application scenarios, higher requirements must be made on the miniaturization and embeddedness of the measurement signal processing system, while ensuring the high efficiency and real-time performance of signal processing and calculation. Summary of the invention

[0005] In view of the above-mentioned deficiencies in the prior art, a Rydberg atom low-frequency voltage measurement signal processing system provided by the present invention 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.

[0006] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0007] A Rydberg atom low-frequency voltage measurement signal processing system is provided, comprising:

[0008] ADC data acquisition module, used to convert EIT spectrum data from the photodetector into electrical signals and perform data acquisition;

[0009] The wavelet transform denoising module is used to denoise the data collected by the ADC data acquisition module to obtain denoised data;

[0010] Automatic peak search module, used to automatically search for peaks in the denoised data and obtain peak position differences;

[0011] Voltage inversion module, used to perform voltage inversion and frequency calculation based on peak position difference, and complete the low-frequency voltage measurement of Rydberg atoms;

[0012] The host computer is used to configure the parameters of the wavelet transform denoising module and the voltage inversion module;

[0013] The ADC data acquisition module and wavelet transform denoising module are implemented by the PL side; the automatic peak search module and voltage inversion module are implemented by the PS side.

[0014] Furthermore, it also includes:

[0015] DDR3 storage control module, used to control DDR3 memory and store the voltage inversion result into the specified memory space;

[0016] Ethernet communication module, used to send the data stored in the specified memory space to the host computer;

[0017] Phase-locked loop module, used to generate 100M clock for DDR3 storage control module and 50M clock for other modules;

[0018] The phase-locked loop module is implemented by the PL side; the DDR3 storage control module and the Ethernet communication module are implemented by the PS side.

[0019] Furthermore, the specific method for denoising the data collected by the ADC data acquisition module includes the following steps:

[0020] A1. Perform low-pass filtering and high-pass filtering on the data collected by the ADC data acquisition module, respectively, to obtain the first-layer approximate coefficients and the first-layer detail coefficients;

[0021] A2, decomposing the first layer approximation coefficients into subbands through iteration;

[0022] A3. Perform inverse transformation on each sub-band to obtain a reconstructed sub-band;

[0023] A4. Add all reconstructed subbands to obtain the inverse approximation coefficients;

[0024] A5. Set a denoising threshold, set the first layer detail coefficient that is smaller than the denoising threshold to zero, and obtain the denoised detail coefficient;

[0025] A6. The inverse approximation coefficients and the denoised detail coefficients are combined by inverse wavelet transform to obtain denoised data, namely, denoised EIT spectrum lines.

[0026] Furthermore, low-pass filtering and high-pass filtering are implemented by bior6_8 wavelet function; the first layer of approximate coefficients is decomposed into 4 sub-band layers.

[0027] Furthermore, the denoising threshold is configured through the particle swarm algorithm.

[0028] Furthermore, the specific method for automatically searching the peaks of the denoised data and obtaining the peak position difference is as follows:

[0029] A quadratic polynomial fitting of the least squares method is performed with three data points as the fitting width, the fitted peak amplitude is compared with the amplitude threshold, and the fitted peak amplitude less than the amplitude threshold is removed; the difference between the position maximum of the peak value and the position value corresponding to the amplitude maximum is taken as the peak position difference.

[0030] Furthermore, the specific method of voltage inversion and frequency calculation based on the peak position difference is as follows:

[0031] According to the formula:

[0032]

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

[0034] According to the formula:

[0035]

[0036] The frequency f of the Rydberg atom low-frequency voltage is obtained; wherein t is the time interval between two consecutive points where the voltage amplitude reaches the maximum value.

[0037] Furthermore, the specific method for configuring the parameters of the wavelet transform denoising module and the voltage inversion module includes the following sub-steps:

[0038] B1. Conduct a test experiment of observing the EIT spectrum peak frequency shift with a known electric field using a standard source, and perform ADC signal sampling to obtain a field strength test signal;

[0039] B2. Initially configure the wavelet transform denoising module according to the standard source parameters and the field strength test signal;

[0040] B3, collecting data through the ADC data collection module, and denoising the data collected by the ADC data collection module through the initially configured wavelet transform denoising module to obtain denoised data;

[0041] B4. Automatically search for peaks on the denoised data within the peak search data range set by the host computer through the automatic peak search module to obtain the corresponding peak value;

[0042] B5, determine whether the number of peaks found by the automatic peak finding module is greater than 1, if so, proceed to step B6; otherwise, reduce the amplitude threshold and return to step B4;

[0043] B6, determine whether the number of peaks found by the automatic peak finding module is less than or equal to 3, if so, proceed to step B7; otherwise, determine that an error peak is detected, increase the amplitude threshold, and return to step B5;

[0044] B7, obtaining the difference between the position value corresponding to the maximum value of the peak value and the position value corresponding to the maximum value of the amplitude, and obtaining the peak position difference;

[0045] B8. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters;

[0046] B9. Compare the measured voltage parameters of the standard source with the real voltage parameters of the standard source, and adjust the parameters of the voltage inversion module until the error between the measured voltage parameters of the standard source and the real voltage parameters of the standard source is within the set range.

[0047] The beneficial effects of the present invention are:

[0048] 1. This system has a simple structure, good stability and real-time performance, meets the requirements of miniaturization and low power consumption, and lays the foundation for the engineering application of Rydberg atomic low-frequency voltage measurement system.

[0049] 2. The wavelet transform denoising module uses the bior6_8 wavelet basis with a decomposition layer of 5 for processing, and the threshold is configured by the host computer through the particle swarm optimization algorithm. In addition, the wavelet transform denoising is all implemented on the PL side. Due to the trial use of the hardware language-based design method, the algorithm execution speed is fast, and the characteristic information of the EIT spectrum line can be quickly and effectively extracted. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 This is the structural diagram of this system;

[0051] Figure 2 is the change of the Stark frequency shift and the amplitude of the transmission peak during the change of voltage amplitude;

[0052] Figure 3 This is the original EIT spectrum before denoising and peak searching in the embodiment;

[0053] Figure 4 This is a diagram showing the effect of EIT spectrum line denoising and peak searching in the embodiment. DETAILED DESCRIPTION

[0054] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0055] like Figure 1 As shown, the Rydberg atom low-frequency voltage measurement signal processing system includes:

[0056] ADC data acquisition module, used to convert EIT spectrum data from the photodetector into electrical signals and perform data acquisition;

[0057] The wavelet transform denoising module is used to denoise the data collected by the ADC data acquisition module to obtain denoised data;

[0058] Automatic peak search module, used to automatically search for peaks in the denoised data and obtain peak position differences;

[0059] Voltage inversion module, used to perform voltage inversion and frequency calculation based on peak position difference, and complete the low-frequency voltage measurement of Rydberg atoms;

[0060] The host computer is used to configure the parameters of the wavelet transform denoising module and the voltage inversion module;

[0061] DDR3 storage control module, used to control DDR3 memory and store the voltage inversion result into the specified memory space;

[0062] Ethernet communication module, used to send the data stored in the specified memory space to the host computer;

[0063] The phase-locked loop module is used to generate a 100M clock for the DDR3 storage control module and a 50M clock for other modules.

[0064] In this embodiment, the PL end uses an FPGA chip and the PS end uses an ARM chip. The PL end is used to implement the ADC data acquisition module, the wavelet transform denoising module and the phase-locked loop module. The PS end is used to implement the automatic peak search module, the voltage inversion module, the frequency calculation module, the DDR3 storage control module, and the Ethernet communication module. In addition, the conversion of the detection light signal to the EIT spectrum digital signal is also realized through the photoelectric detector module, the signal amplification module, and the digital-to-analog converter module.

[0065] In the specific implementation process, in order to realize the engineering application of Rydberg atom low-frequency voltage measurement and make the measurement platform adapt to more application scenarios, higher requirements must be made for the miniaturization and embedding of the measurement signal processing system, while ensuring the efficiency and real-time performance of signal processing and calculation. This embodiment integrates FPGA and ARM processors into the same ZYNQ chip, taking into account the advantages of hardware and software co-design. The ZYNQ series integrates the software programmability of ARM processors and the hardware programmability of FPGAs, which can not only realize important analysis, hardware acceleration, excellent performance-to-power ratio and design flexibility, but also highly integrates CPU, DSP, ASSP and mixed signal functions on a single device. The advantages of using ZYNQ chips are: large integration of development environment; AXI4 standard bus interconnection; convenient data exchange between PS and PL and fast communication speed; saving the layout area of ​​an ARM and the resource consumption of the communication bus, and high hardware integration.

[0066] The DDR3 memory control module is configured with parameters through the IP core ZYNQ7 Processing System provided by Xilinx 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 achieve the transmission of control signals, data signals, and address signals.

[0067] In order to ensure the order and stability of data flow transmission, each module can add a FIFO module between modules for buffering. The configuration of the FIFO module can also be implemented through the IP core provided by Xilinx.

[0068] The specific method for denoising the data collected by the ADC data acquisition module is to use the undecimated wavelet transform (UWT) method to denoise the signal, which specifically includes the following steps:

[0069] A1. Perform low-pass filtering and high-pass filtering on the data collected by the ADC data acquisition module, respectively, to obtain the first-layer approximate coefficients and the first-layer detail coefficients;

[0070] A2, decomposing the first layer approximation coefficients into subbands through iteration;

[0071] A3. Perform inverse transformation on each sub-band to obtain a reconstructed sub-band;

[0072] A4. Add all reconstructed subbands to obtain the inverse approximation coefficients;

[0073] A5. Set a denoising threshold, set the first layer detail coefficient that is smaller than the denoising threshold to zero, and obtain the denoised detail coefficient;

[0074] A6. The inverse approximation coefficients and the denoised detail coefficients are combined by inverse wavelet transform to obtain denoised data, namely, denoised EIT spectrum lines.

[0075] In this embodiment, filtering is essentially a convolution operation. When the wavelet basis is selected, the parameters of low-pass filtering and high-pass filtering are also determined. Specifically, low-pass filtering and high-pass filtering are implemented by bior6_8 wavelet function; the number of sub-bands into which the first-layer approximate coefficients are decomposed is 4, that is, the first-layer approximate coefficients are decomposed into the second-layer approximate coefficients and the second-layer detail coefficients through low-pass filtering and high-pass filtering, and the second-layer approximate coefficients are similarly decomposed into the third-layer approximate coefficients and the third-layer detail coefficients, and so on. After 4 iterations, a total of five layers of approximate coefficients and five layers of detail coefficients are obtained. The approximate coefficients and detail coefficients obtained starting from the second layer are a group of sub-bands, and the reconstructed sub-bands are the approximate coefficients and detail coefficients corresponding to the second to fifth layers.

[0076] The denoising threshold is configured by the host computer through the particle swarm algorithm. And the wavelet transform denoising is all implemented on the PL side. Because it is designed based on hardware language, the algorithm executes quickly and can quickly and effectively extract the characteristic information of the EIT spectrum line. The wavelet transform denoising module is connected to the automatic peak search module through the AXI bus interface.

[0077] The specific method for automatically searching for peaks in the denoised data and obtaining the peak position difference is as follows: performing quadratic polynomial fitting of the least squares method with three data points as the fitting width, comparing the fitted peak amplitude with the amplitude threshold, and removing the fitted peak amplitude that is less than the amplitude threshold; the difference between the position maximum of the peak value and the position value corresponding to the amplitude maximum is taken as the peak position difference.

[0078] The automatic peak search module can automatically set the appropriate threshold value according to the number of detected peaks. The peak search threshold is dynamically adjusted in a certain step size according to the mean of the sampled data and the number of peaks to be searched. The peak amplitude obtained from each peak search is compared with the current threshold. If it is greater than the threshold, it is retained and the amplitude and position information are recorded.

[0079] In the specific implementation process, in quantum mechanics theory, the interaction between the atomic system and the electric field E can be expressed by the effective Hamiltonian:

[0080] H eff =H 0 +H S =W E

[0081] Among them, H 0is the Hamiltonian of the atomic system without electric field, W E is the energy of the atomic system under the action of the electric field. S is the Hamiltonian describing the Stark interaction under the action of electric field, and its expression is as follows:

[0082]

[0083] 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.

[0084] Using the fourth-order perturbation theory to expand W E , we can get W E The expressions of each order. By analyzing the relationship between energy and field strength of different principal quantum numbers, the Stark energy level diagram can be obtained. Due to the existence of atomic nuclei in hydrogen-like atoms such as cesium atoms, the quantum loss effect of atoms must be considered. Therefore, within the electric field range before the energy level anti-crossing phenomenon occurs, the Stark frequency shift has a quadratic relationship with the external electric field, and the expression is:

[0085]

[0086] where Δ S tark represents the Stark frequency shift; α i (f) represents the polarizability of Rydberg state i at frequency f. For a fixed Rydberg state, the theoretical expression of its polarizability is:

[0087]

[0088] Where j represents all states to which Rydberg state i is coupled through the electric field, and f ij is the transition frequency between state i and state j, Γ j is the spontaneous decay frequency of state j.

[0089] The voltage loading method adopted in this embodiment is the method of placing the plates in the air chamber, so the relationship between the electric field and the voltage under the condition of a fixed plate spacing is considered. The formula for the Stark frequency shift inversion voltage is as follows:

[0090]

[0091] However, in engineering applications, it is difficult to perform inversion based on theoretical formulas due to the difficulty of achieving completely uniform electric fields between the plates, ionization caused by coupled lasers, and uncertainties introduced by optical devices. It is necessary to compensate for multiple factors in the mathematical model to achieve better inversion accuracy. Moreover, when the mathematical model is more complex, transplanting it to the ZYNQ chip will also take up too many resources.

[0092] Therefore, this system can use the host computer to fit the m separated by D3 / 2 state according to the test data based on the standard source experiment. j = m separated from 3 / 2 peak and D5 / 2 state j = Frequency axis position difference Δ between 1 / 2 peaks Stark The inversion coefficient k' is obtained from the relationship between the square of the voltage U applied across the plate. The voltage inversion formula is as follows:

[0093]

[0094] 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 movement of the transmission peak is also a representation of the Stark effect. Therefore, the peak position difference is actually the frequency shift Δ Stark .

[0095] This method highly refines the mathematical relationship behind the complex quantum coherence mechanism, so that the voltage inversion module only involves multiplication and square root operations, simplifies the calculation process when the accuracy permits, saves the software and hardware 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.

[0096] In the process of voltage amplitude change, the Stark frequency shift and the amplitude of the transmission peak change as shown in Figure 2 As shown in Figure 2, as the voltage increases, the Stark shift increases and the transmission peak amplitude decreases significantly.

[0097] The frequency calculation process includes a timer, and the amplitude data output by the peak-finding module is obtained in real time. After obtaining the amplitude data, the time of reaching the maximum amplitude point each time is recorded, and the time interval t between two consecutive points of reaching the maximum amplitude is calculated. The cycle calculation formula is as follows:

[0098]

[0099] Where f is the frequency of the Rydberg atom low-frequency voltage.

[0100] Since a higher transmission peak amplitude is easier to monitor than a lower transmission peak amplitude, the frequency calculation method can determine the frequency within one cycle and has good real-time performance.

[0101] In actual application, due to the influence of different measurement environments and the uncertainty of factors such as the state of optical devices, the state of lasers, and the effect of laser frequency locking, parameter initialization configuration is required before each formal measurement. The specific method of configuring the parameters of the wavelet transform denoising module and the voltage inversion module through the host computer includes the following sub-steps:

[0102] B1. Conduct a test experiment of observing the EIT spectrum peak frequency shift with a known electric field using a standard source, and perform ADC signal sampling to obtain a field strength test signal;

[0103] B2. Initially configure the denoising threshold of the wavelet transform denoising module according to the standard source parameters and the field strength test signal;

[0104] B3, collecting data through the ADC data collection module, and denoising the data collected by the ADC data collection module through the initially configured wavelet transform denoising module to obtain denoised data;

[0105] B4. Automatically search for peaks on the denoised data within the peak search data range set by the host computer through the automatic peak search module to obtain the corresponding peak value;

[0106] B5, determine whether the number of peaks found by the automatic peak finding module is greater than 1, if so, proceed to step B6; otherwise, reduce the amplitude threshold and return to step B4;

[0107] B6, determine whether the number of peaks found by the automatic peak finding module is less than or equal to 3, if so, proceed to step B7; otherwise, determine that an error peak is detected, increase the amplitude threshold, and return to step B5;

[0108] B7, obtaining the difference between the position value corresponding to the maximum value of the peak value and the position value corresponding to the maximum value of the amplitude, and obtaining the peak position difference;

[0109] B8. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters;

[0110] B9. Compare the measured voltage parameters of the standard source with the real voltage parameters of the standard source, and adjust the voltage inversion module parameter k′ until the error between the measured voltage parameters of the standard source and the real voltage parameters of the standard source is within the set range.

[0111] In this embodiment, during the process of configuring parameters of the wavelet transform denoising module and the voltage inversion module, two digital-to-analog conversion ports may be reserved on the ZYNQ platform for analog trigger signals and spare digital trigger signals.

[0112] In one embodiment of the present invention, Figure 3 and Figure 4As shown, the peak with the highest amplitude and the farthest offset is identified for the highest and farthest in a sinusoidal cycle rather than the highest or farthest at a single moment. The height of the peak under pressure is definitely much smaller than that without pressure. However, the highest and farthest peaks will theoretically only appear on the two sides, so the present invention will identify and output the positions of the peaks on both sides. The small changes in the peak amplitude are the results of wavelet analysis. For non-Gaussian and non-stationary signals, this system can better retain the characteristics of the EIT spectrum line, highlight the shape and position of the EIT spectrum peak, and effectively filter out the non-Gaussian noise caused by the photodetector's own factors and environmental factors, so as to reduce the possibility of failing to find the peak or finding the wrong peak in the subsequent links.

[0113] At the same time, the comparative peak search method is suitable for finding a single strong peak, and has a fast processing speed. It is suitable for non-Gaussian signals and for application scenarios that focus on the peak with the highest amplitude and the peak with the farthest frequency shift rather than overlapping peaks. With the help of this system, two or three peaks can be detected simply and quickly, and good voltage measurement results can be achieved. The measurement accuracy and sensitivity of the Rydberg atomic low-frequency voltage measurement platform using this system can lead many non-contact voltage measurement technologies on the market, and the prospects for development and improvement are still broad.

[0114] In addition, the specific frequency range of low-frequency signals that this system can process depends on the sweep frequency of the laser and the processing speed of the ZYNQ platform. From the perspective of signal processing, the time required for wavelet transform denoising, comparative peak finding, and automatic adjustment of peak finding thresholds must be much faster than the time it takes for the real-time voltage to change significantly and cannot be ignored. Therefore, the frequency range can be comprehensively determined through multiple experiments and equipment operation experience.

[0115] If this system needs to be applied in the power system, in order to facilitate interconnection with other electrical equipment, the Ethernet control module can also be optimized to enable it to have the function of encapsulating the measurement data into data frames that comply with IEC61850.

[0116] If this system needs to implement multi-channel signal processing on a single ZYNQ platform, the wavelet transform denoising on the PL side will occupy a lot of hardware resources. The utilization rate of the PL side resources can be improved through time-division multiplexing to ensure the efficiency and real-time performance of the measurement.

[0117] In summary, the present invention has a simple structure, good stability and real-time performance, meets the requirements of miniaturization and low power consumption, and lays a 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: ADC data acquisition module, used to convert EIT spectrum data from the photodetector into electrical signals and perform data acquisition; The wavelet transform denoising module is used to denoise the data collected by the ADC data acquisition module to obtain denoised data; Automatic peak search module, used to automatically search for peaks in the denoised data and obtain peak position differences; Voltage inversion module, used to perform voltage inversion and frequency calculation based on peak position difference, and complete the low-frequency voltage measurement of Rydberg atoms; 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 side; the automatic peak search module and voltage inversion module are implemented by the PS side.

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

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 collected by the ADC data acquisition module includes the following steps: A1. Perform low-pass filtering and high-pass filtering on the data collected by the ADC data acquisition module, respectively, to obtain the first-layer approximate coefficients and the first-layer detail coefficients; A2, decomposing the first layer approximation coefficients into subbands through iteration; A3. Perform inverse transformation on each sub-band to obtain a reconstructed sub-band; A4. Add all reconstructed subbands to obtain the inverse approximation coefficients; A5. Set a denoising threshold, set the first layer detail coefficient that is smaller than the denoising threshold to zero, and obtain the denoised detail coefficient; A6. The inverse approximation coefficients and the denoised detail coefficients are combined by inverse wavelet transform to obtain denoised data, namely, denoised EIT spectrum lines.

4. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 3, characterized in that: Low-pass filtering and high-pass filtering are implemented by bior6_8 wavelet function; the first layer of approximate coefficients is decomposed into 4 sub-band layers.

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

6. The Rydberg atom low-frequency voltage measurement signal processing system according to claim 1, characterized in that: The specific method for automatically searching for peaks in the denoised data and obtaining the peak position difference is as follows: A quadratic polynomial fitting of the least squares method is performed with three data points as the fitting width, the fitted peak amplitude is compared with the amplitude threshold, and the fitted peak amplitude less than the amplitude threshold is removed; the difference between the position maximum of the peak value and the position value corresponding to the amplitude maximum 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 method for voltage inversion and frequency calculation based on peak position difference: According to the formula: Obtain the inverted voltage U, that is, the low-frequency voltage value of the Rydberg atom; where k′ is the fitting coefficient; Δ Stark is the peak position difference; According to the formula: The frequency f of the Rydberg atom low-frequency voltage is obtained; wherein t is the time interval between two consecutive points where the voltage amplitude reaches the 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 parameters of the wavelet transform denoising module and the voltage inversion module includes the following sub-steps: B1. Conduct a test experiment of observing the EIT spectrum peak frequency shift with a known electric field using a standard source, and perform ADC signal sampling to obtain a field strength test signal; B2. Initially configure the wavelet transform denoising module according to the standard source parameters and the field strength test signal; B3, collecting data through the ADC data collection module, and denoising the data collected by the ADC data collection module through the initially configured wavelet transform denoising module to obtain denoised data; B4. Automatically search for peaks on the denoised data within the peak search data range set by the host computer through the automatic peak search module to obtain the corresponding peak value; B5, determine whether the number of peaks found by the automatic peak finding module is greater than 1, if so, proceed to step B6; otherwise, reduce 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 so, proceed to step B7; otherwise, determine that an error peak is detected, increase the amplitude threshold, and return to step B5; B7, obtaining the difference between the position value corresponding to the maximum value of the peak value and the position value corresponding to the maximum value of the amplitude, and obtaining the peak position difference; B8. Perform voltage inversion and frequency calculation based on the peak position difference to obtain the measured standard source voltage parameters; B9. Compare the measured voltage parameters of the standard source with the real voltage parameters of the standard source, and adjust the parameters of the voltage inversion module until the error between the measured voltage parameters of the standard source and the real voltage parameters of the standard source is within the set range.

Images