Anti-interference method and anti-interference system for mainboard

Abstract

The invention provides an anti-interference method and an anti-interference system for a mainboard, and relates to the technical field of mainboards, and the method comprises the steps: obtaining a mixed signal in an environment where the mainboard is located, carrying out the multi-scale decomposition of the mixed signal, and obtaining signal components under different frequency scales, the mixed signal comprising an interference signal and a useful signal; performing signal classification processing based on the spectrogram of the signal component of each frequency scale to obtain a spectrogram of the interference signal under each frequency scale; according to the frequency grade of the frequency scale to which the spectrogram of the interference signal belongs, a corresponding mask strategy is matched from a preset strategy library, and mask processing is carried out; performing inverse transformation on the spectrogram of the interference signal subjected to mask processing to obtain a denoised interference signal; and performing multi-scale weighted reconstruction on the useful signal and the de-noised interference signal to obtain a de-noised mixed signal, thereby realizing effective suppression of the complex interference signal in the environment where the mainboard is located, and improving the stability and anti-interference capability of the mainboard.

Description

Technical Field

[0001] The present invention relates to the technical field of motherboards, and in particular to an anti-interference method and an anti-interference system for a motherboard. Background Art

[0002] In modern electronic systems, the motherboard is the core component of the system, and its stability and anti-interference ability have a vital impact on the overall performance. With the rapid development of electronic technology, the working environment of the motherboard is becoming increasingly complex, and problems such as electromagnetic interference and noise pollution are becoming increasingly prominent. These interference signals are often mixed into the useful signals received by the motherboard, resulting in a decrease in signal quality and even causing system errors or failures.

[0003] Traditional anti-interference methods mostly rely on hardware filtering or simple digital signal processing algorithms. These methods often have poor effects when dealing with complex and changeable interference signals, and it is difficult to effectively extract pure useful signals. Especially at different frequency scales such as high frequency, medium frequency and low frequency, the characteristics and impact of interference signals are different. Traditional single anti-interference strategies are difficult to deal with this multi-scale interference problem.

[0004] Therefore, it is necessary to provide an anti-interference method and an anti-interference system for a mainboard to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides an anti-interference method and an anti-interference system for a mainboard. Through multi-scale decomposition, spectrum analysis, signal classification, mask processing, multi-scale reconstruction and post-processing, it realizes effective suppression of complex interference signals in the environment where the mainboard is located, and improves the stability and anti-interference ability of the mainboard.

[0006] The present invention provides an anti-interference method for a mainboard, the anti-interference method comprising the following steps:

[0007] Acquire a mixed signal in the environment where the mainboard is located, and perform multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal;

[0008] Perform signal classification processing based on the frequency spectrum of the signal component at each frequency scale to obtain the frequency spectrum of the interference signal at each frequency scale;

[0009] According to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, matching the corresponding mask strategy from a preset strategy library and performing mask processing, wherein the frequency level includes high frequency, medium frequency and low frequency;

[0010] Inversely transform the spectrum of the interference signal after mask processing to obtain the interference signal after denoising;

[0011] Multi-scale weighted reconstruction is performed on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

[0012] Preferably, the acquiring of a mixed signal in the environment where the mainboard is located, and performing multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, includes:

[0013] The mixed signals in the environment are collected in real time by a sensor array installed on the mainboard, wherein the sensor array includes at least one sensor for detecting electromagnetic interference and one sensor for detecting the operating state of the mainboard;

[0014] Preprocessing the collected mixed signal, wherein the preprocessing includes signal amplification, filtering and analog-to-digital conversion;

[0015] The variational mode decomposition (VMD) method is used to perform multi-scale decomposition on the preprocessed mixed signal to obtain signal components containing different frequency components, each of which corresponds to a frequency scale.

[0016] Preferably, the signal classification processing is performed based on the spectrum diagram of the signal component at each frequency scale to obtain the spectrum diagram of the interference signal at each frequency scale, including:

[0017] Applying fast Fourier transform to convert the signal component at each frequency scale from the time domain to the frequency domain, obtaining a frequency spectrum of the signal component at each frequency scale, and determining the amplitude of each frequency point in the frequency spectrum;

[0018] Based on the signal statistics of the signal components at each frequency scale, a dynamic threshold is determined for each frequency point in combination with context awareness;

[0019] The amplitude of each frequency point in the spectrum diagram is compared with the corresponding dynamic threshold, and the frequency points whose amplitude exceeds the dynamic threshold are determined as interference signals, and the remaining frequency points are determined as useful signals, so as to obtain the spectrum diagram of the interference signal at each frequency scale.

[0020] Preferably, the signal statistics of the signal components at each frequency scale are based on the context perception to determine the dynamic threshold for each frequency point, including:

[0021] The short-time energy and entropy of the signal component at each frequency scale are calculated, wherein the calculation formulas of the short-time energy and entropy are respectively:

[0022]

[0023] In the formula, E j represents the short-time energy at scale j, N represents the time window length, s j(i) represents the signal value at the i-th time point on scale j, H represents the entropy at scale j, and p k represents the probability density of the signal on frequency band k, where K represents the number of frequency bands;

[0024] Calculate the statistics of the adjacent frequency points of each frequency point, the statistics include the median and standard deviation of the adjacent frequency points, wherein the calculation formulas of the median and standard deviation are respectively:

[0025] m=median(D f )

[0026] σ=std(D f )

[0027] Where D f The formula is:

[0028] D f ={S(f-1), S(f+1)}

[0029] Where m is the median, σ is the standard deviation, S(f-1) and S(f+1) are the amplitudes of the frequency points adjacent to frequency point f, S(f) is the amplitude of frequency point f in the spectrum, and D f Represents the amplitude distribution of adjacent frequency points of frequency point f;

[0030] The dynamic threshold is calculated by combining the short-time energy, entropy and statistics of adjacent frequency points of the signal component at each frequency scale, wherein the calculation formula of the dynamic threshold T is:

[0031] T=α·E j +β·H j +γ·m+δ·σ

[0032] Among them, α, β, γ and δ are respectively set first weight coefficients.

[0033] Preferably, matching a corresponding mask strategy from a preset strategy library and performing mask processing according to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs includes:

[0034] Dividing the frequency spectrum of the interference signal at each frequency scale into frequency levels, and determining the frequency level to which each frequency scale belongs, wherein the frequency levels include high frequency, medium frequency and low frequency;

[0035] According to the determined frequency level, a corresponding mask strategy is selected from a set of mask strategies pre-designed and stored in a strategy library, wherein the mask strategies in the strategy library are preset for different frequency levels;

[0036] The selected masking strategy is applied to perform masking processing on the frequency spectrum of the interference signal to suppress the influence of the interference signal.

[0037] Preferably, the frequency level division step is:

[0038] According to the preset frequency range threshold, each frequency scale of the signal component is primary graded;

[0039] Combining the short-time energy and entropy of the signal components at each frequency scale, the division boundaries of the preliminary divided frequency levels are dynamically adjusted to obtain the final frequency levels.

[0040] Preferably, the step of dynamically adjusting the division boundaries of the preliminarily divided frequency levels in combination with the short-time energy and entropy of the signal components at each frequency scale to obtain the final frequency levels includes:

[0041] Based on the short-time energy and entropy, combined with a preset second weight coefficient, a comprehensive evaluation index of the signal component at each frequency scale is calculated, wherein the comprehensive evaluation index characterizes the properties of the signal component at different frequency scales;

[0042] According to the comprehensive evaluation index, the boundaries of the frequency levels initially divided are adjusted;

[0043] The final frequency level division result is obtained through iterative optimization, wherein the iterative optimization includes adjusting the second weight coefficient according to the current division result and recalculating the comprehensive evaluation index until a preset stop condition is met.

[0044] Preferably, the inverse transformation of the spectrum diagram of the interference signal after mask processing to obtain the denoised interference signal is specifically:

[0045] An inverse fast Fourier transform is applied to the spectrum of the masked interference signal to convert the frequency domain signal back to the time domain signal, thereby obtaining the time domain representation of the denoised interference signal.

[0046] Preferably, performing multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal includes:

[0047] Determine a third weight coefficient of the signal component at each frequency scale in the reconstruction process according to the final frequency level;

[0048] Using the third weight coefficient, weighted summing the useful signal and the denoised interference signal at each frequency scale is performed to perform multi-scale weighted reconstruction;

[0049] The reconstructed signal is post-processed, wherein the post-processing includes signal smoothing and normalization processing to obtain a final denoised mixed signal.

[0050] The present invention also provides an anti-interference system for a mainboard, which is used to execute an anti-interference method for a mainboard. The anti-interference system comprises:

[0051] A signal decomposition module, used for acquiring a mixed signal in the environment where the mainboard is located, and performing multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal;

[0052] A signal classification module, used to perform signal classification processing based on the frequency spectrum of the signal component at each frequency scale to obtain the frequency spectrum of the interference signal at each frequency scale;

[0053] A mask processing module, used to match a corresponding mask strategy from a preset strategy library and perform mask processing according to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, wherein the frequency level includes high frequency, medium frequency and low frequency;

[0054] A frequency domain to time domain conversion module, used for inversely transforming the spectrum of the interference signal after mask processing to obtain the interference signal after denoising;

[0055] The signal reconstruction module is used to perform multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

[0056] Compared with the related art, the anti-interference method and anti-interference system of a mainboard provided by the present invention have the following beneficial effects:

[0057] The present invention uses multi-scale decomposition technology to decompose the mixed signal in the environment where the motherboard is located into signal components at different frequency scales, thereby achieving effective separation of interference signals and useful signals. Furthermore, the present invention also uses spectrum analysis and signal classification technology to perform refined processing on the signal components at each frequency scale, accurately identify the interference signal, and match the corresponding mask strategy from the preset strategy library based on its frequency level for mask processing.

[0058] In the mask processing process, the present invention not only considers the frequency characteristics of the interference signal, but also combines its short-time energy, entropy and other statistical information, as well as the statistics of adjacent frequency points, and dynamically adjusts the division boundary of the frequency level to achieve accurate suppression of the interference signal. Finally, the present invention uses multi-scale reconstruction technology to perform weighted summation of the useful signal and the denoised interference signal to obtain the final denoised mixed signal, and further improves the signal quality through post-processing steps such as signal smoothing and normalization. BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Figure 1 A flow chart of an anti-interference method for a mainboard provided by the present invention;

[0060] Figure 2 A module structure diagram of an anti-interference system of a mainboard provided by the present invention. DETAILED DESCRIPTION

[0061] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the present invention, rather than to limit the present invention. It should also be noted that, for ease of description, only the parts related to the present invention, rather than all structures, are shown in the accompanying drawings. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other without conflict.

[0062] It should also be noted that, for ease of description, only the part relevant to the present invention but not all the content is shown in the accompanying drawings. It should be mentioned before discussing the exemplary embodiments in more detail that some exemplary embodiments are described as the processing or method described as a flow chart. Although the flow chart describes each operation (or step) as being processed sequentially, many operations therein can be implemented in parallel, concurrently or simultaneously. In addition, the order of each operation can be rearranged. When its operation is completed, the processing can be terminated, but it can also have additional steps not included in the accompanying drawings. The processing can correspond to a method, a function, a procedure, a subroutine, a subprogram, etc.

[0063] Embodiment 1

[0064] The present invention provides an anti-interference method for a mainboard, and the anti-interference method comprises the following steps:

[0065] S1: Acquire a mixed signal in the environment where the mainboard is located, and perform multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal.

[0066] Specifically, step S1 includes the following steps:

[0067] S11: The mixed signals in the environment are collected in real time through a sensor array installed on the mainboard, wherein the sensor array includes at least one sensor for detecting electromagnetic interference and one sensor for detecting the operating status of the mainboard.

[0068] In modern electronic devices, motherboards, as core components, face a complex electromagnetic environment. Electromagnetic interference (EMI) not only comes from the external environment, such as wireless communication devices, power lines, etc., but may also come from the mutual interference between other electronic components inside the motherboard. In order to ensure that the motherboard can operate stably, real-time monitoring and analysis of mixed signals in the environment is crucial for anti-interference design.

[0069] When designing for anti-interference, an array of multiple sensors is installed on the mainboard, including electromagnetic interference sensors and mainboard operating status sensors. Electromagnetic interference sensors refer to devices that can detect changes in electromagnetic field strength, including but not limited to fluxgate sensors and Hall effect sensors. These sensors are used to capture electromagnetic fluctuations in the environment.

[0070] The motherboard operation status sensors include but are not limited to temperature sensors, current sensors and voltage sensors, which are used to monitor the operation status of the motherboard itself. For example, excessive temperature may cause signal distortion.

[0071] S12: Preprocessing the collected mixed signal, wherein the preprocessing includes signal amplification, filtering, and analog-to-digital conversion.

[0072] In this embodiment, since the signal received by the sensor may contain a large amount of noise and the signal strength varies, it must be preprocessed before performing more complex analysis to improve the signal-to-noise ratio and make the signal adapt to the requirements of subsequent algorithms.

[0073] Among them, preprocessing usually includes the following aspects:

[0074] Signal amplification: Use a preamplifier to increase signal strength, making weak signals visible.

[0075] Filtering: Use filters to remove unnecessary frequency components and retain useful signals. Filters include low-pass filters, high-pass filters and band-pass filters.

[0076] Analog-to-digital conversion (ADC): Converts a continuous analog signal into a discrete digital signal so that a computer can process it.

[0077] S13: The preprocessed mixed signal is decomposed into multiple scales using the variational mode decomposition (VMD) method to obtain signal components containing different frequency components, each of which corresponds to a frequency scale.

[0078] In practical applications, mixed signals contain a large number of frequency components, some of which come from useful signals and some from interference. Traditional single-scale processing methods are difficult to effectively separate these components, while multi-scale analysis techniques, especially variational mode decomposition (VMD), can decompose mixed signals into multiple modal components, each of which corresponds to a different frequency range, thus facilitating subsequent signal processing.

[0079] In this application, the specific working process of variational mode decomposition is:

[0080] First, determine the target number of modes to be decomposed, which is set according to the complexity of the signal and the expected resolution, and set the convergence criterion, penalty factor (used to control the bandwidth of the modal component) and number of inner loop iterations of the VMD algorithm according to requirements.

[0081] Next, the Lagrangian function is constructed. The VMD algorithm finds the optimal modal components by constructing a Lagrangian function. Given a signal, it is hoped to find a set of modal components so that their Hilbert transforms are concentrated as much as possible around a center frequency in the frequency domain.

[0082] Then, the alternating direction method of multipliers (ADMM) is used to minimize the above Lagrangian function. ADMM is a method for solving constrained optimization problems, which gradually approaches the optimal solution by iteratively updating the modal components and center frequencies, as well as the Lagrangian multipliers and dual variables.

[0083] In each iteration, the center frequency is first fixed and the modal components are updated, and then the modal components are fixed and the center frequency is updated.

[0084] Finally, check whether the updated modal components meet the convergence conditions. If the predetermined error limit or number of iterations is reached, the iteration is stopped. Once the algorithm converges, a set of modal components is obtained. These components are sorted according to their respective center frequencies to form a multi-scale signal representation.

[0085] Through VMD's multi-scale decomposition, complex mixed signals can be split into several relatively simple modal components, each with different frequency characteristics. Such decomposition not only helps with subsequent signal classification processing, but also better identifies and suppresses interference signals while retaining useful signal components.

[0086] S2: Perform signal classification processing based on the spectrum diagram of the signal component at each frequency scale to obtain the spectrum diagram of the interference signal at each frequency scale.

[0087] Specifically, step S2 includes the following steps:

[0088] S21: Apply fast Fourier transform to convert the signal component at each frequency scale from the time domain to the frequency domain, obtain the spectrum diagram of the signal component at each frequency scale, and determine the amplitude of each frequency point in the spectrum diagram.

[0089] Fast Fourier Transform (FFT) is an efficient algorithm for converting signals in the time domain to the frequency domain so that different components of the signal can be analyzed in the frequency domain. In the anti-interference method, FFT converts signal components at different frequency scales into a spectrum diagram, making it easier to distinguish interference signals from useful signals.

[0090] In implementation, an FFT algorithm is applied to the signal components at each frequency scale. FFT can quickly calculate the discrete Fourier transform (DFT), thereby converting the signal from the time domain to the frequency domain.

[0091] Calculate the spectrum of each signal component, that is, get the complex value at each frequency point. The amplitude of these complex values ​​represents the intensity of the corresponding frequency component, and the phase reflects the phase information of the signal.

[0092] The amplitude of each frequency point is calculated from the complex number result obtained by FFT. The amplitude can be obtained by calculating the modulus length of the complex number.

[0093] Through FFT conversion, we can intuitively see the distribution of signals at different frequencies, thus providing a basis for the next step of signal classification. In addition, by calculating the amplitude, we can highlight those frequency components with higher intensity, which is particularly important for identifying interference signals.

[0094] S22: Determine a dynamic threshold for each frequency point based on the signal statistics of the signal components at each frequency scale and in combination with context perception.

[0095] In this embodiment, in order to accurately distinguish interference signals from useful signals, a dynamic threshold needs to be determined based on signal statistics and context perception. This dynamic threshold helps to determine which frequency components belong to interference signals and which belong to useful signals.

[0096] During implementation, the dynamic threshold is calculated by calculating the short-time energy and entropy of the signal components at each frequency scale, and calculating the statistics of the adjacent frequency points of each frequency point, including the median and standard deviation, and combining the short-time energy, entropy and statistics.

[0097] Wherein, step S22 specifically includes the following steps:

[0098] S22a: Calculate the short-time energy and entropy of the signal component at each frequency scale, where the calculation formulas for short-time energy and entropy are:

[0099]

[0100] In the formula, E j represents the short-time energy at scale j, N represents the time window length, s j (i) represents the signal value at the i-th time point on scale j, H j represents the entropy at scale j, p k represents the probability density of the signal on frequency band k, where K represents the number of frequency bands.

[0101] In signal processing, short-time energy and entropy are important indicators for measuring signal strength and randomness. Short-time energy can help determine the activity of a signal, while entropy reflects the complexity or uncertainty of a signal. When identifying interference signals, these two indicators can provide important information about signal characteristics.

[0102] During implementation, an appropriate time window length is selected according to requirements. This length should be long enough to capture the characteristics of the signal, but not too long to lose instantaneous changes. For each signal value at a frequency scale, the short-time energy is calculated within the selected time window.

[0103] In the process of entropy calculation, the probability density of the signal can be estimated by the histogram method, and then the entropy is calculated using the probability density.

[0104] S22b: Calculate the statistics of the adjacent frequency points of each frequency point. The statistics include the median and standard deviation of the adjacent frequency points. The calculation formulas of the median and standard deviation are:

[0105] m=median(D f )

[0106] σ=std(D f )

[0107] Where D f The formula is:

[0108] D f ={S(f-1), S(f+1)}

[0109] Where m is the median, σ is the standard deviation, S(f-1) and S(f+1) are the amplitudes of the frequency points adjacent to frequency point f, S(f) is the amplitude of frequency point f in the spectrum, and D f Represents the amplitude distribution of adjacent frequency points of frequency point f.

[0110] In the frequency domain, the statistical characteristics of adjacent frequency points help to understand the changing trend of the signal. By calculating the median and standard deviation of adjacent frequency points, we can understand the local changes of the signal, which is especially important for identifying abnormal interference signals.

[0111] S22c: Calculate the dynamic threshold by combining the short-time energy, entropy and statistics of adjacent frequency points of the signal component at each frequency scale, where the calculation formula of the dynamic threshold T is:

[0112] T=α·E j +β·H j +γ·m+δ·σ

[0113] Among them, α, β, γ and δ are respectively set first weight coefficients.

[0114] In this embodiment, the dynamic threshold is a threshold dynamically adjusted according to the statistical characteristics of the signal, and is used to distinguish the interference part from the useful part of the signal. The combination of short-time energy, entropy and statistics of adjacent frequency points can more accurately identify interference signals.

[0115] Set four first weight coefficients α, β, γ and δ, which are used to adjust the contribution ratio of short-time energy, entropy, median and standard deviation. Combine short-time energy, entropy, median and standard deviation, and calculate the dynamic threshold through the above formula. The dynamic threshold can more accurately identify which frequency points are affected by interference, adapt to signal changes, avoid the misjudgment problem caused by fixed thresholds, and thus improve the accuracy of interference identification.

[0116] S23: Compare the amplitude of each frequency point in the spectrum diagram with the corresponding dynamic threshold, determine the frequency point whose amplitude exceeds the dynamic threshold as an interference signal, and determine the remaining frequency points as useful signals, so as to obtain the spectrum diagram of the interference signal at each frequency scale.

[0117] After the dynamic threshold is determined, the signals are classified according to the relationship between the amplitude of the frequency point and the threshold, that is, which are interference signals and which are useful signals are identified.

[0118] During implementation, the amplitude of each frequency point is first compared with the corresponding dynamic threshold. If the amplitude exceeds the dynamic threshold, the frequency point is considered to be an interference signal. If the amplitude is lower than or equal to the dynamic threshold, the frequency point is considered to be a useful signal.

[0119] According to the above determination results, a spectrum diagram of the interference signal at each frequency scale is generated.

[0120] S3: According to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, a corresponding mask strategy is matched from a preset strategy library and mask processing is performed, wherein the frequency level includes high frequency, medium frequency and low frequency.

[0121] Specifically, step S3 includes the following steps:

[0122] S31: dividing the frequency spectrum of the interference signal at each frequency scale into frequency levels, and determining the frequency level to which each frequency scale belongs, where the frequency levels include high frequency, medium frequency, and low frequency.

[0123] In the process of combating interference, different interference signals may exist in different frequency ranges. For example, high-frequency interference may come from fast switching circuits, while low-frequency interference may come from power supply ripple or other slowly changing signals. In order to effectively deal with these interferences, it is necessary to classify the signal components according to the frequency characteristics so as to take targeted treatment measures.

[0124] When implemented, the primary classification is first performed, and the signal components are divided into three levels: high frequency, medium frequency and low frequency according to the preset frequency range thresholds. Usually, these thresholds can be adjusted according to the actual application scenario. For example, high frequency may be defined as greater than 100kHz, medium frequency as 10kHz to 100kHz, and low frequency as less than 10kHz.

[0125] A preliminary partitioning can be done by setting fixed frequency limits and using histogram analysis to determine the main frequency distribution areas of the signal.

[0126] Then, the boundaries of the preliminarily divided frequency levels are dynamically adjusted by combining the short-time energy and entropy of the signal components at each frequency scale. The short-time energy reflects the strength of the signal in a short period of time, while the entropy measures the randomness or uncertainty of the signal.

[0127] The dynamic adjustment is achieved by calculating the comprehensive evaluation index of each frequency component. The comprehensive evaluation index is a value calculated based on short-time energy and entropy, which can be used to evaluate the "interference degree" of the signal component.

[0128] The calculation formula of the comprehensive evaluation index is to perform weighted summation of short-time energy and entropy according to a preset second weight coefficient.

[0129] Finally, through multiple iterations, the second weight coefficient is continuously adjusted and the comprehensive evaluation index is recalculated until the preset stop condition is met. The stop condition is reaching the set number of iterations.

[0130] S32: According to the determined frequency level, a corresponding mask strategy is selected from a set of mask strategies pre-designed and stored in a strategy library, wherein the mask strategies in the strategy library are preset for different frequency levels.

[0131] After completing the division of the spectrum graph of the interference signal, the next step is to select the corresponding mask strategy according to the frequency level. The mask strategy refers to a method of processing the signal in the frequency domain. By masking the spectrum graph of the interference signal, the influence of the interference signal can be effectively suppressed.

[0132] In implementation, according to the frequency level classification in step S31, the spectrum of the interference signal is divided into three levels: high frequency, medium frequency and low frequency. This step has determined which frequency components belong to high frequency interference, medium frequency interference and low frequency interference.

[0133] A set of mask strategies pre-designed and stored in the strategy library. These strategies are preset for different frequency levels, namely high-frequency mask strategy, medium-frequency mask strategy and low-frequency mask strategy. Specifically:

[0134] High frequency mask strategy:

[0135] For high-frequency interference signals, a high-frequency masking strategy (narrowband masking strategy) is adopted. Specifically, a Gaussian function or a Hamming window function is used as a masking function. The characteristics of these functions are that they have a strong suppression effect near the center frequency and gradually recover at positions far away from the center frequency. By applying these functions to the spectrum diagram of the high frequency band, high-frequency interference signals can be effectively suppressed.

[0136] IF mask strategy:

[0137] For the intermediate frequency interference signal, a wider band masking strategy is adopted. Specifically, a rectangular window function or a triangular window function is used as a masking function. The characteristics of these functions are that the signal is completely suppressed within a certain frequency range, while the signal is completely retained within other frequency ranges. By applying these functions to the spectrum diagram of the intermediate frequency band, the intermediate frequency interference signal can be effectively suppressed.

[0138] Low frequency mask strategy:

[0139] For low-frequency interference signals, a broadband masking strategy is adopted. Specifically, a Hanning window function or a Heiman window function is used as a masking function. The characteristics of these functions are to provide a smooth suppression effect in the entire low-frequency band to cover the interference components in the low-frequency range. By applying these functions to the spectrum diagram of the low-frequency band, the low-frequency interference signal can be effectively suppressed.

[0140] S33: Apply the selected masking strategy to perform masking processing on the frequency spectrum of the interference signal to suppress the influence of the interference signal.

[0141] After determining the masking strategies, the next step is to apply these strategies to process the interference signals. The purpose of masking is to weaken or eliminate the interference signals in the frequency domain, thereby improving the signal quality.

[0142] When implemented, the selected mask strategy is applied to process the spectrum of the interference signal, specifically modifying the amplitude of the signal in the frequency domain and reducing the signal strength within the specified frequency range by multiplying it by the mask function.

[0143] The mask function can be a 0 / 1 binary function or an attenuation function, depending on the selected mask strategy.

[0144] Through mask processing, the influence of interference signals can be significantly reduced, thereby improving the quality and reliability of signals. This method is very effective in practical applications, especially in situations where precise control of signal characteristics is required.

[0145] S4: Inversely transform the spectrum of the interference signal after mask processing to obtain a denoised interference signal.

[0146] In the previous steps, the interference signal has been identified and suppressed through multi-scale decomposition of the mixed signal, signal classification processing and mask processing. However, the signal is still in the frequency domain at this time. In order to return the signal to the original time domain signal form for further processing or transmission, the signal in the frequency domain needs to be converted back to the time domain. This process is usually completed through the Inverse Fast Fourier Transform (IFFT). After IFFT, the time domain signal with the interference component removed is obtained. Through the inverse transform, the signal in the frequency domain is converted back to the time domain, realizing the complete recovery of the signal.

[0147] S5: Perform multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

[0148] Specifically, step S5 includes the following steps:

[0149] S51: Determine a third weight coefficient of the signal component at each frequency scale during the reconstruction process according to the final frequency level.

[0150] After the interference signal is identified and suppressed, the next step is to reconstruct the signal, that is, to recombine the useful signal with the signal from which the interference component has been removed. In this process, the weights of the signal components at different frequency scales need to be determined to ensure the quality of the reconstructed signal.

[0151] During implementation, a third weight coefficient is assigned to each signal component at each frequency scale according to the frequency level determined in the previous step. The determination of the third weight coefficient can be adjusted according to the importance and reliability of the signal. Low-frequency signals contain more useful information and have higher weights, while high-frequency signals are more affected by interference and have lower weights.

[0152] Specifically, the third weight coefficient is a normalized value of the comprehensive evaluation index of the signal component at each frequency scale.

[0153] S52: Using the third weight coefficient, weighted summation is performed on the useful signal and the denoised interference signal at each frequency scale to perform multi-scale weighted reconstruction.

[0154] In the field of signal processing, multi-scale reconstruction is a commonly used signal recovery technique. By adding signal components at different frequency scales according to weights, a new signal that is closer to the original useful signal can be obtained.

[0155] During implementation, the third weight coefficient determined in step S51 is used to perform weighted summation on the useful signal and the denoised interference signal at each frequency scale. Specifically, the signal components at each frequency scale are multiplied by their corresponding third weight coefficients, and then all the components are added together, while gradually adding higher frequency components starting from the lowest frequency until all components are added.

[0156] S53: Post-processing the reconstructed signal, wherein the post-processing includes signal smoothing and normalization processing to obtain a final denoised mixed signal.

[0157] In this embodiment, although most interference has been removed from the reconstructed signal, some edge effects or discontinuities may still exist, so further post-processing is required to improve the signal quality.

[0158] During implementation, a smoothing filter is used to eliminate spikes and discontinuities in the signal to make the signal curve smoother, and the reconstructed signal is normalized to keep its amplitude within a fixed range, thus avoiding problems caused by signals that are too strong or too weak.

[0159] Embodiment 2

[0160] The present invention also provides an anti-interference system for a mainboard, which is used to execute an anti-interference method for a mainboard, referring to Figure 2 As shown, the anti-interference system includes:

[0161] The signal decomposition module 100 is used to obtain a mixed signal in the environment where the mainboard is located, and perform multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal.

[0162] The signal classification module 200 is used to perform signal classification processing based on the frequency spectrum of the signal component at each frequency scale to obtain the frequency spectrum of the interference signal at each frequency scale.

[0163] The mask processing module 300 is used to match the corresponding mask strategy from a preset strategy library and perform mask processing according to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, wherein the frequency level includes high frequency, medium frequency and low frequency.

[0164] The frequency domain to time domain conversion module 400 is used to perform inverse transformation on the spectrum diagram of the interference signal after mask processing to obtain a denoised interference signal.

[0165] The signal reconstruction module 500 is used to perform multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

[0166] The present application is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0167] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0168] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device that includes a series of elements includes not only those elements, but also other elements that are not explicitly listed, or includes elements that are inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device that includes the elements.

Claims

1. A method for anti-interference of a mainboard, characterized in that: The anti-interference method comprises the following steps: Acquire a mixed signal in the environment where the mainboard is located, and perform multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal; Perform signal classification processing based on the frequency spectrum of the signal component at each frequency scale to obtain the frequency spectrum of the interference signal at each frequency scale; According to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, matching the corresponding mask strategy from a preset strategy library and performing mask processing, wherein the frequency level includes high frequency, medium frequency and low frequency; Inversely transform the spectrum of the interference signal after mask processing to obtain the interference signal after denoising; Multi-scale weighted reconstruction is performed on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

2. The anti-interference method of a mainboard according to claim 1, characterized in that: The obtaining of a mixed signal in the environment where the mainboard is located, and performing multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, includes: The mixed signals in the environment are collected in real time by a sensor array installed on the mainboard, wherein the sensor array includes at least one sensor for detecting electromagnetic interference and one sensor for detecting the operating state of the mainboard; Preprocessing the collected mixed signal, wherein the preprocessing includes signal amplification, filtering and analog-to-digital conversion; The variational mode decomposition (VMD) method is used to perform multi-scale decomposition on the preprocessed mixed signal to obtain signal components containing different frequency components, each of which corresponds to a frequency scale.

3. The anti-interference method of a mainboard according to claim 2, characterized in that: The signal classification process is performed based on the spectrum diagram of the signal component at each frequency scale to obtain the spectrum diagram of the interference signal at each frequency scale, including: Applying fast Fourier transform to convert the signal component at each frequency scale from the time domain to the frequency domain, obtaining a frequency spectrum of the signal component at each frequency scale, and determining the amplitude of each frequency point in the frequency spectrum; Based on the signal statistics of the signal components at each frequency scale, a dynamic threshold is determined for each frequency point in combination with context awareness; The amplitude of each frequency point in the spectrum diagram is compared with the corresponding dynamic threshold, and the frequency points whose amplitude exceeds the dynamic threshold are determined as interference signals, and the remaining frequency points are determined as useful signals, so as to obtain the spectrum diagram of the interference signal at each frequency scale.

4. The anti-interference method of a mainboard according to claim 3, characterized in that: The signal statistics of the signal components at each frequency scale are determined based on context perception for each frequency point, including: The short-time energy and entropy of the signal component at each frequency scale are calculated, wherein the calculation formulas of the short-time energy and entropy are respectively: In the formula, E j represents the short-time energy at scale j, N represents the time window length, s j (i) represents the signal value at the i-th time point on scale j, H j represents the entropy at scale j, p k represents the probability density of the signal on frequency band k, where K represents the number of frequency bands; Calculate the statistics of the adjacent frequency points of each frequency point, the statistics include the median and standard deviation of the adjacent frequency points, wherein the calculation formulas of the median and standard deviation are respectively: m=median(D f ) σ=std(D f ) Where D f The formula is: Df={S(f-1),S(f+1)} Where m is the median, σ is the standard deviation, S(f-1) and S(f+1) are the amplitudes of the frequency points adjacent to frequency point f, S(f) is the amplitude of frequency point f in the spectrum, and D f Represents the amplitude distribution of adjacent frequency points of frequency point f; The dynamic threshold is calculated by combining the short-time energy, entropy and statistics of adjacent frequency points of the signal component at each frequency scale, wherein the calculation formula of the dynamic threshold T is: T=α·E j +β·H j +γ·m+δ·σ Among them, α, β, γ and δ are respectively set first weight coefficients.

5. The anti-interference method of a mainboard according to claim 4, characterized in that: The matching of a corresponding mask strategy from a preset strategy library and performing mask processing according to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs includes: Dividing the frequency spectrum of the interference signal at each frequency scale into frequency levels, and determining the frequency level to which each frequency scale belongs, wherein the frequency levels include high frequency, medium frequency and low frequency; According to the determined frequency level, a corresponding mask strategy is selected from a set of mask strategies pre-designed and stored in a strategy library, wherein the mask strategies in the strategy library are preset for different frequency levels; The selected masking strategy is applied to perform masking processing on the frequency spectrum of the interference signal to suppress the influence of the interference signal.

6. The anti-interference method of a mainboard according to claim 5, characterized in that: The frequency level division steps are as follows: According to the preset frequency range threshold, each frequency scale of the signal component is primary graded; Combining the short-time energy and entropy of the signal components at each frequency scale, the division boundaries of the preliminary divided frequency levels are dynamically adjusted to obtain the final frequency levels.

7. The anti-interference method of a mainboard according to claim 6, characterized in that: The method combines the short-time energy and entropy of the signal components at each frequency scale to dynamically adjust the division boundaries of the preliminary divided frequency levels to obtain the final frequency levels, including: Based on the short-time energy and entropy, combined with a preset second weight coefficient, a comprehensive evaluation index of the signal component at each frequency scale is calculated, wherein the comprehensive evaluation index characterizes the properties of the signal component at different frequency scales; According to the comprehensive evaluation index, the boundaries of the frequency levels initially divided are adjusted; The final frequency level division result is obtained through iterative optimization, wherein the iterative optimization includes adjusting the second weight coefficient according to the current division result and recalculating the comprehensive evaluation index until a preset stop condition is met.

8. The anti-interference method of a mainboard according to claim 7, characterized in that: The inverse transformation of the spectrum diagram of the interference signal after mask processing to obtain the denoised interference signal is specifically: An inverse fast Fourier transform is applied to the spectrum of the masked interference signal to convert the frequency domain signal back to the time domain signal, thereby obtaining the time domain representation of the denoised interference signal.

9. The anti-interference method of a mainboard according to claim 7, characterized in that: The performing multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal includes: Determine a third weight coefficient of the signal component at each frequency scale in the reconstruction process according to the final frequency level; Using the third weight coefficient, weighted summing the useful signal and the denoised interference signal at each frequency scale is performed to perform multi-scale weighted reconstruction; The reconstructed signal is post-processed, wherein the post-processing includes signal smoothing and normalization processing to obtain a final denoised mixed signal.

10. An anti-interference system for a mainboard, used to execute an anti-interference method for a mainboard as claimed in any one of claims 1 to 9, characterized in that: The anti-interference system comprises: A signal decomposition module, used for acquiring a mixed signal in the environment where the mainboard is located, and performing multi-scale decomposition on the mixed signal to obtain signal components at different frequency scales, wherein the mixed signal includes an interference signal and a useful signal; A signal classification module, used to perform signal classification processing based on the frequency spectrum of the signal component at each frequency scale to obtain the frequency spectrum of the interference signal at each frequency scale; A mask processing module, used to match a corresponding mask strategy from a preset strategy library and perform mask processing according to the frequency level of the frequency scale to which the spectrum diagram of the interference signal belongs, wherein the frequency level includes high frequency, medium frequency and low frequency; A frequency domain to time domain conversion module, used for inversely transforming the spectrum of the interference signal after mask processing to obtain the interference signal after denoising; The signal reconstruction module is used to perform multi-scale weighted reconstruction on the useful signal and the denoised interference signal to obtain a denoised mixed signal.

Images