Power signal filtering method based on Soc chip

By modal decomposition of the power signal in the Soc chip and analysis of the steady-state of high-frequency switching, the filter window size is dynamically determined, which solves the problem of difficulty in selecting the filter window size in the existing technology, and effectively suppresses ripple noise and Improved power supply stability.

CN118842295BActive Publication Date: 2024-12-06HUNAN PULISIMAI ELECTRONIC TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411321524.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-23
Publication Date
2024-12-06
Estimated Expiration
2044-09-23

AI Technical Summary

Technical Problem

There are difficulties in selecting the window size of the existing DC/DC power supply filters, which leads to the inability to effectively suppress ripple noise and affects the stability of the power supply.

Method used

By modally decomposing the power signal in the Soc chip, the ripple noise characteristic value and periodic stability characteristic value are obtained, combined with the high-frequency switching stable attitude, the filter window size is dynamically determined, and the window parameters of the filter are optimized to suppress ripple noise.

Benefits of technology

It effectively enhances the anti-interference ability of the power supply, improves the stability and filtering effect of the power supply, and avoids the problem of poor filtering effect when the window size is not selected properly.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118842295B_ABST
    Figure CN118842295B_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of power supply voltage signal processing, and specifically to a power supply signal filtering method based on a Soc chip, the method comprising: obtaining a voltage signal on a filter circuit of a power supply with a built-in Soc chip at each sampling moment; performing modal decomposition on a sequence composed of voltage signals at all sampling moments in a neighborhood period of each sampling moment to obtain modal components, and obtaining ripple noise characteristic values ​​according to the numerical distribution differences between the modal components; obtaining a periodic stability characteristic value according to the numerical fluctuation changes of the ripple noise characteristic value; obtaining a high-frequency switch steady state by combining the ripple noise characteristic value and the periodic stability characteristic value at each sampling moment; determining the length of the filter window according to the high-frequency switch steady state; substituting the length of the filter window into the filter to filter all voltage signals on the filter circuit. The present application can improve the filtering effect of the voltage signal and enhance the stability of the power supply.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of power supply voltage signal processing, and in particular to a power supply signal filtering method based on a Soc chip. Background Art

[0002] With the continuous development of digital circuits, the integration of chips is getting higher and higher, especially Soc chips, which integrate a variety of practical technologies into one control management. In the application of DC / DC power supply design, Soc chips can effectively improve the conversion efficiency and reliability of the power supply system. However, due to the complexity of power supply application scenarios, DC / DC power supplies are highly sensitive to the environment, and the power supply is prone to ripple noise, which affects the performance and function of the power supply when used for a long time, making the stability of the power supply poor.

[0003] Therefore, filtering technology is needed to suppress the ripple noise generated by the power supply and enhance the anti-interference ability of the DC / DC power supply. Common filtering technologies, most filters are a filtering method based on local polynomial least squares fitting in the time domain. However, the window size of the filter in the power supply is often difficult to select. When the window is large, the filter has poor ability to capture details, and when the window is small, the smoothing effect of the filter is poor, and the ripple noise in the power supply cannot be effectively suppressed, resulting in poor filtering effect on the ripple noise generated in the power supply, and the stability of the power supply cannot be effectively enhanced. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides a power signal filtering method based on a Soc chip to solve the existing problems.

[0005] The power signal filtering method based on the Soc chip of this application adopts the following technical solutions:

[0006] An embodiment of the present application provides a power signal filtering method based on a Soc chip, the method comprising the following steps:

[0007] S1, obtaining the voltage signal on the power filter circuit of the built-in Soc chip at each sampling moment;

[0008] S2, performing modal decomposition on a sequence of voltage signals at all sampling moments in a neighborhood period of each sampling moment to obtain modal components, and obtaining a ripple noise characteristic value at each sampling moment based on the difference in numerical distribution between the modal components;

[0009] S3, obtaining the periodic stability characteristic value of each sampling moment according to the numerical fluctuation change of the ripple noise characteristic value of all sampling moments in the neighborhood time period of each sampling moment;

[0010] S4, combining the ripple noise characteristic value and the period stability characteristic value at each sampling moment, obtaining the conversion loss degree; according to the numerical distribution discreteness of the conversion loss degree at all sampling moments, obtaining the high-frequency switching stability;

[0011] S5, determining the length of the filter window according to the high-frequency switch steady state; substituting the length of the filter window into the filter, and filtering all voltage signals on the filter circuit of the power supply.

[0012] Furthermore, the method for obtaining the ripple noise characteristic value at each sampling moment includes:

[0013] The sequence composed of voltage signals at all sampling moments within a preset time window centered at each sampling moment is recorded as a local voltage sequence at each sampling moment;

[0014] For the local voltage sequence at each sampling moment, the local voltage sequence is modally decomposed to obtain each modal component; the sequence composed of the signal strength of each modal component at all sampling moments is arranged in time order and recorded as the signal strength sequence of each modal component;

[0015] Record a sequence consisting of differences between elements at the same position of the signal strength sequences of any two modal components as a signal difference sequence between the any two modal components;

[0016] The ripple noise characteristic value is obtained according to the difference in numerical changes between the signal difference sequences of all modal components.

[0017] Further, obtaining the ripple noise characteristic value according to the numerical change difference between the signal difference sequences of all modal components includes:

[0018] According to the differences of all adjacent elements in the signal difference sequence of all modal components, the difference change distance at each sampling moment is obtained;

[0019] The information entropy average value of the signal difference sequence of all modal components is obtained; and the product of the information entropy average value and the difference change distance at each sampling moment is used as the ripple noise characteristic value at each sampling moment.

[0020] Furthermore, the method for obtaining the difference change distance at each sampling moment includes:

[0021] The mean of the differences of all adjacent elements in the signal difference sequence is calculated and recorded as the average difference between any two modal components; the mean of the average differences of all the modal components in pairs is taken as the difference change distance at each sampling moment.

[0022] Further, the method for obtaining the periodic stable characteristic value at each sampling moment includes:

[0023] Recording a sequence of ripple noise characteristic values ​​at all sampling moments within a preset time window centered at each sampling moment as a ripple noise characteristic sequence at each sampling moment; obtaining an extreme value in the ripple noise characteristic sequence; and recording a sequence composed of all extreme values ​​arranged in time order as a ripple period characteristic sequence at each sampling moment;

[0024] According to the fluctuation of the difference between all adjacent elements in the ripple period characteristic sequence at each sampling moment, the periodic stability characteristic value is obtained.

[0025] Furthermore, the periodic stability characteristic value is obtained according to the fluctuation of the difference between all adjacent elements in the ripple period characteristic sequence at each sampling moment, including:

[0026] The variance of all elements in the first-order difference sequence of the ripple periodic characteristic sequence is obtained; and the reverse normalized value of the variance is used as the periodic stable characteristic value at each sampling moment.

[0027] Furthermore, the conversion loss calculation method includes: recording the conversion loss at sampling time t as , , where is the ripple noise characteristic value at sampling time t; is the periodic stable eigenvalue at sampling time t; The default value is greater than 0.

[0028] Furthermore, the method for obtaining the steady-state of the high-frequency switch includes:

[0029] The average of the conversion loss at all sampling moments is obtained, recorded as the loss mean; the discrete degree of the conversion loss at all sampling moments is obtained; the product of the discrete degree and the loss mean is calculated, and the inverse normalized value of the product is used as the high-frequency switch steady state.

[0030] Further, the method for determining the length of the filter window according to the steady state of the high-frequency switch includes: recording the length of the filter window as , , where is the ceiling function, For high frequency switching stability, is an exponential function with the natural constant e as base, Adjust parameters for the preset.

[0031] Furthermore, the method for filtering all voltage signals on the filter circuit of the power supply includes:

[0032] The length of the filter window is substituted into the SG filter in the power supply, and the SG filter is used to filter the sequence composed of the voltage signals at all sampling moments on the filter circuit in the power supply.

[0033] This application has at least the following beneficial effects:

[0034] The present application aims to solve the problem that when the filter window is not properly selected, the ripple noise in the power supply cannot be effectively suppressed, resulting in poor filtering effect on the power supply voltage data; first, by analyzing the influence of the ripple noise generated by the power supply with a built-in Soc chip on the change of the power supply voltage data, the numerical distribution difference between the modal components is obtained according to the sequence decomposition composed of the voltage signal, and the ripple noise characteristic value is obtained, which reflects the difference between the modal components and effectively evaluates the degree of influence of the power supply on the ripple noise; further, in the power supply design based on the Soc chip, by analyzing the high-frequency switching steady-state characteristics of the power supply during the DC conversion process, the periodic stability characteristic value is obtained, which reflects the stability of the periodic change of the ripple noise characteristics and accurately evaluates the stability of the DC conversion of the power supply; comprehensively considering the ripple noise characteristic value and the periodic stability characteristic value, the high-frequency switching steady-state is obtained, and the degree of influence of the power supply on the ripple noise is accurately evaluated according to the change of the voltage signal; the filtering window size of the filter in the power supply is determined according to the high-frequency switching steady-state, so as to avoid the problem of poor filtering effect when the window size is not properly selected, thereby effectively suppressing the ripple noise in the power supply and effectively enhancing the stability of the power supply. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 A flowchart of the steps of the power signal filtering method based on the Soc chip provided in this application;

[0037] Figure 2 A block diagram for obtaining ripple noise characteristic values ​​provided in one embodiment of the present application. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0039] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0040] The specific scheme of the power signal filtering method based on the Soc chip provided in the present application is described in detail below with reference to the accompanying drawings.

[0041] An embodiment of the present application provides a power signal filtering method based on a Soc chip. Specifically, the following power signal filtering method based on a Soc chip is provided. Please refer to Figure 1 , the method comprises the following steps:

[0042] S1, obtain the voltage on the power filter circuit of the built-in Soc chip at each sampling moment.

[0043] The specific implementation scenario of this application is the power data optimization processing scenario based on the Soc chip. The main purpose is to optimize the selection of the window size in the filter to avoid the influence on the filtering effect when the window size is not selected properly, thereby effectively suppressing the ripple noise in the power supply, improving the effect of filtering the ripple noise generated in the power supply, and effectively enhancing the stability of the power supply.

[0044] First, the Soc integrated chip is used as the power management chip of the power supply. In this embodiment, the power supply includes a high-frequency switching converter, an SG filter, a filtering circuit, a voltage stabilizing circuit, and an output rectifier filtering circuit; wherein the high-frequency switching converter is used to switch the power supply on and off; the SG filter is used to filter the voltage of the filtering circuit in the power supply to suppress the ripple noise in the power supply; the voltage stabilizing circuit is used to protect the input and output power supplies from over-voltage and under-voltage to ensure that the voltage is at a steady-state level; the output rectifier filtering circuit is used to output a pure voltage after filtering; the power management chip includes an FPGA module, an ARM module, and a filtering module; the FPGA module is responsible for driving and abnormal processing of the power supply; the ARM module is responsible for the implementation of the control algorithm and the implementation of the communication function, and controls and adjusts the output end parameters through the high-frequency switching converter, and performs performance testing on the power supply; the filtering module is used to filter the voltage on the filtering circuit using the SG filter to suppress the ripple noise in the power supply;

[0045] In order to optimize the window size selection method in the SG filter in the Soc chip-based power supply, thereby selecting a window of appropriate size and using the SG filter to filter the voltage on the filter circuit, the present application collects the voltage signal of the input filter circuit through a voltage sensor. In this embodiment, the collection time is 1 minute, the sampling rate is 20Hz, and the voltage signal at each sampling moment is obtained.

[0046] In this embodiment, the high-frequency switching steady-state characteristics of the power supply during the DC conversion process are analyzed mainly based on the voltage signal, and the size of the filter window parameter of the SG filter in the power supply is improved according to the high-frequency switching steady-state characteristics.

[0047] S2, performing modal decomposition on a sequence composed of voltage signals at all sampling moments in a neighborhood period of each sampling moment to obtain modal components, and obtaining the ripple noise characteristic value at each sampling moment based on the difference in numerical distribution between the modal components.

[0048] The power supply is also called a DC-to-DC power supply. It is a power supply that converts DC to DC by controlling the on and off of the high-frequency switching converter. However, when the power supply is in a high-frequency switching state, the high-frequency switching conversion is prone to generate high-frequency ripple noise, which affects the stability of the power supply.

[0049] In order to analyze the characteristics of ripple noise generated by high-frequency switching changes at different sampling moments, a sequence composed of voltage signals at all sampling moments within a preset time window centered on each sampling moment is recorded as a local voltage sequence at each sampling moment; in this embodiment, the length of the preset time window T=41, that is, it includes 41 sampling moments.

[0050] In order to analyze the characteristics of high-frequency ripple noise, the local voltage sequence at each sampling moment is modally decomposed to obtain each modal component; the sequence composed of the signal strength of each modal component at all sampling moments is recorded as the signal strength sequence of each modal component. The signal strength is arranged in chronological order.

[0051] In this embodiment, the VMD variational modal decomposition algorithm is adopted, the mode number in the algorithm parameters is set to 2, the penalty coefficient is set to 2000, the convergence tolerance is set to 2e-6, and e is a natural constant; the output of the VMD variational modal decomposition algorithm is used as the two modal components at the sampling time t, wherein the VMD variational modal decomposition algorithm is a well-known technology, and the specific process will not be repeated here.

[0052] The ripple noise generated by the switching on and off of the high-frequency switching converter will lead to large differences between the modal components and large changes in the differences. In order to reflect the differences between the modal components, the sequence composed of the differences between the elements at the same position of the signal strength sequences of any two modal components is recorded as the signal difference sequence between the any two modal components. Specifically, the signal strength sequences of the two modal components at the sampling time t are subtracted, that is, the signal strengths of all the same positions in the two sequences are subtracted to obtain the signal difference sequence. The signal difference sequence reflects the changes in the differences between the two modal components at the sampling time t. The greater the change in the difference and the more chaotic the difference, the greater the characteristics of the power supply being affected by the ripple noise.

[0053] Calculate the mean of the differences of all adjacent elements in the signal difference sequence, and record it as the average difference between any two modal components; take the mean of the average differences of all the modal components in pairs as the difference change distance at each sampling moment; obtain the information entropy average value of the signal difference sequence of all modal components; take the product of the information entropy average value and the difference change distance at each sampling moment as the ripple noise characteristic value at each sampling moment.

[0054] The greater the difference in changes between elements at adjacent moments, the greater the difference between the two modal components at the acquisition moment, the greater the influence of ripple noise on the power supply, and the greater the difference change distance. The calculation of information entropy is a well-known technology, and the specific process is not repeated. The greater the difference change distance and the greater the information entropy, to a certain extent, it shows that the difference between the modal components is greater and the difference is more chaotic, which means that the power supply is affected by ripple noise at this time, and the greater the ripple noise characteristic value.

[0055] The block diagram for obtaining the ripple noise characteristic value is as follows: Figure 2 shown.

[0056] S3, obtaining the periodic stability characteristic value of each sampling moment according to the numerical fluctuation change of the ripple noise characteristic value of all sampling moments in the neighborhood time period of each sampling moment.

[0057] Normally, a high-frequency switching converter switches on and off at a fixed switching frequency. The smaller the periodic change of the ripple noise during the switching process, the more stable the periodic change of the DC conversion of the power supply is, and the high-frequency switching converter can achieve an ideal switching function.

[0058] In order to analyze the stable characteristics of the periodic changes of the DC conversion of the power supply, a sequence composed of the ripple noise characteristic values ​​at all sampling moments within a preset time window centered on each sampling moment is recorded as the ripple noise characteristic sequence at each sampling moment; the extreme values ​​in the ripple noise characteristic sequence are obtained; and a sequence composed of all the extreme values ​​arranged in chronological order is recorded as the ripple period characteristic sequence at each sampling moment.

[0059] Specifically, the ripple noise feature sequence at sampling time t is used as the input of the extreme point detection algorithm, and the output of the extreme point detection algorithm is used as all extreme values ​​in the ripple noise feature sequence at sampling time t, wherein the extreme point detection algorithm is a well-known technology and the specific process is not repeated here.

[0060] The variance of all elements in the first-order difference sequence of the ripple periodic characteristic sequence is obtained; and the reverse normalized value of the variance is used as the periodic stable characteristic value at each sampling moment.

[0061] The reverse normalization value represents the result after reverse normalization and the result before reverse normalization. There is a negative correlation relationship that the larger the result before reverse normalization, the smaller the result after reverse normalization, and the result after reverse normalization is between (0,1]. As an embodiment of the present application, the periodic stable characteristic value at sampling time t , where is the variance of the first-order difference sequence of the ripple period characteristic sequence at sampling time t.

[0062] Among them, the first-order difference sequence of the ripple periodic characteristic sequence reflects the periodic change of the ripple noise during the on-off process of the switch. The smaller the discreteness of all differential values ​​in the first-order difference sequence, to a certain extent, it shows that the periodicity of the DC conversion of the power supply is more stable. The smaller the discreteness of all differential values ​​in the first-order difference sequence, the more stable the periodic change of the ripple noise characteristics, which is more conducive to the stability of the DC conversion of the power supply, and the larger the periodic stability characteristic value.

[0063] S4, combining the ripple noise characteristic value and the periodic stability characteristic value at each sampling moment, obtains the conversion loss degree; according to the numerical distribution discreteness of the conversion loss degree at all sampling moments, obtains the high-frequency switching steady-state.

[0064] Generally speaking, the more unstable the periodic characteristics of the power supply during DC conversion, and the greater the ripple noise characteristics generated by the switching on and off of the high-frequency switching converter, the greater the loss of the power supply during DC conversion.

[0065] Based on the above analysis, the conversion loss at sampling time t is calculated. , , where is the ripple noise characteristic value at sampling time t; is the periodic stable eigenvalue at sampling time t; The default value is greater than 0.

[0066] The smaller the periodic stability characteristic value and the larger the ripple noise characteristic value when the power supply performs DC conversion, the more it can be explained that the DC conversion process is interfered by ripple noise, which will cause higher losses during energy conversion, and the greater the conversion loss.

[0067] Generally speaking, the more unstable the conversion loss is during the DC conversion process of the power supply and the higher the average level of the conversion loss is, the more it indicates that the high-frequency switch cannot be in a steady state and the more likely it is to generate ripple noise.

[0068] Specifically, the mean of the conversion loss at all sampling moments is obtained, recorded as the loss mean; the discrete degree of the conversion loss at all sampling moments is obtained; the product of the discrete degree and the loss mean is calculated, and the inverse normalized value of the product is used as the high-frequency switch steady state.

[0069] In this embodiment, the high frequency switch is steady state ; In the formula, is the loss mean; is the discrete degree of conversion loss at all sampling moments.

[0070] It should be noted that the degree of dispersion of a set of data indicates the degree of difference between all values ​​in the set of data and the average value. In this embodiment, the dispersion coefficient is used for calculation; as other embodiments of the present application, the standard deviation can also be used for calculation. The calculation of the dispersion coefficient is a well-known technology, and the specific process will not be repeated here.

[0071] The smaller the mean value of the conversion loss and the smaller the coefficient of dispersion of the conversion loss, to a certain extent, it indicates that the high-frequency switch is more likely to be in a steady state during the DC conversion process, and is less likely to generate ripple noise, and the higher the stability of the high-frequency switch.

[0072] S5, determining the length of the filter window according to the high-frequency switch steady state; substituting the length of the filter window into the filter, and filtering all voltage signals on the filter circuit of the power supply.

[0073] The present application utilizes the high-frequency switching steady state of the DC conversion process to select the window size of the SG filter in the power supply. When the high-frequency switching steady state is larger, it is less likely to generate ripple noise, that is, the voltage signal changes more normally. At this time, it is suitable to select a smaller filter window to improve the search capability for details. Conversely, when the high-frequency switching steady state is smaller, it is more likely to generate ripple noise, that is, the voltage signal changes more abnormally. At this time, it is suitable to select a larger filter window to improve the smoothing effect of the SG filter.

[0074] Specifically, calculate the length of the filter window , , where is the ceiling function, For high frequency switching stability, is an exponential function with the natural constant e as base, is a preset adjustment parameter. In this embodiment, =10. The calculation result of the filter window parameter is The odd number between them, and the larger the high-frequency switching steady state, the smaller the filter window parameter in the SG filter.

[0075] Furthermore, the SG filter is used to filter the voltage on the filter circuit based on the filter window parameter to suppress the ripple noise generated by the power supply. Specifically, the length of the filter window is substituted into the SG filter in the power supply, and the SG filter is used to filter the voltage on the filter circuit in the power supply to suppress the ripple noise generated by the power supply and enhance the anti-interference ability of the power supply. The filtering process of the SG filter is a well-known technology and will not be described in detail.

[0076] Through the above description of the implementation method in combination with the accompanying drawings, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0077] The above contents are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be covered by the protection scope of the present application.

Claims

1. A power signal filtering method based on Soc chip, characterized in that: The method comprises the following steps: S1, obtaining the voltage signal on the power filter circuit of the built-in Soc chip at each sampling moment; S2, performing modal decomposition on a sequence of voltage signals at all sampling moments in a neighborhood period of each sampling moment to obtain modal components, and obtaining a ripple noise characteristic value at each sampling moment based on the difference in numerical distribution between the modal components; S3, obtaining the periodic stability characteristic value of each sampling moment according to the numerical fluctuation change of the ripple noise characteristic value of all sampling moments in the neighborhood time period of each sampling moment; S4, combining the ripple noise characteristic value and the period stability characteristic value at each sampling moment, obtaining the conversion loss degree; according to the numerical distribution discreteness of the conversion loss degree at all sampling moments, obtaining the high-frequency switching stability; S5, determining the length of the filter window according to the high-frequency switch steady state; substituting the length of the filter window into the filter, and filtering all voltage signals on the filter circuit of the power supply; The method for calculating the conversion loss degree comprises: recording the conversion loss degree at sampling time t as , , where is the ripple noise characteristic value at sampling time t; is the periodic stable eigenvalue at sampling time t; It is a parameter preset to be greater than 0; The method for obtaining the steady-state of the high-frequency switch comprises: Obtaining the mean of the conversion loss at all sampling moments, recorded as the loss mean; obtaining the discreteness of the conversion loss at all sampling moments; calculating the product of the discreteness and the loss mean, and taking the inverse normalized value of the product as the high-frequency switch steady state; The method for determining the length of the filter window according to the steady state of the high-frequency switch comprises: recording the length of the filter window as , , where is the ceiling function, For high frequency switching stability, is an exponential function with the natural constant e as base, Adjust parameters for the preset.

2. The power signal filtering method based on Soc chip according to claim 1, characterized in that: The method for obtaining the ripple noise characteristic value at each sampling moment includes: The sequence composed of voltage signals at all sampling moments within a preset time window centered at each sampling moment is recorded as a local voltage sequence at each sampling moment; For the local voltage sequence at each sampling moment, the local voltage sequence is modally decomposed to obtain each modal component; the sequence composed of the signal strength of each modal component at all sampling moments is arranged in time order and recorded as the signal strength sequence of each modal component; Record a sequence consisting of differences between elements at the same position of the signal strength sequences of any two modal components as a signal difference sequence between the any two modal components; The ripple noise characteristic value is obtained according to the difference in numerical changes between the signal difference sequences of all modal components.

3. The power signal filtering method based on Soc chip as claimed in claim 2, characterized in that: The step of obtaining the ripple noise characteristic value according to the numerical change difference between the signal difference sequences of all modal components includes: According to the differences of all adjacent elements in the signal difference sequence of all modal components, the difference change distance at each sampling moment is obtained; The information entropy average value of the signal difference sequence of all modal components is obtained; and the product of the information entropy average value and the difference change distance at each sampling moment is used as the ripple noise characteristic value at each sampling moment.

4. The power signal filtering method based on Soc chip as claimed in claim 3, characterized in that: The method for obtaining the difference change distance at each sampling moment includes: The mean of the differences of all adjacent elements in the signal difference sequence is calculated and recorded as the average difference between any two modal components; the mean of the average differences of all the modal components in pairs is taken as the difference change distance at each sampling moment.

5. The power signal filtering method based on Soc chip as claimed in claim 1, characterized in that: The method for obtaining the periodic stable characteristic value at each sampling moment includes: Recording a sequence of ripple noise characteristic values ​​at all sampling moments within a preset time window centered at each sampling moment as a ripple noise characteristic sequence at each sampling moment; obtaining an extreme value in the ripple noise characteristic sequence; and recording a sequence composed of all extreme values ​​arranged in time order as a ripple period characteristic sequence at each sampling moment; According to the fluctuation of the difference between all adjacent elements in the ripple period characteristic sequence at each sampling moment, the periodic stability characteristic value is obtained.

6. The power signal filtering method based on Soc chip as claimed in claim 5, characterized in that: The periodic stability characteristic value is obtained according to the fluctuation of the difference between all adjacent elements in the ripple period characteristic sequence at each sampling moment, including: The variance of all elements in the first-order difference sequence of the ripple periodic characteristic sequence is obtained; and the reverse normalized value of the variance is used as the periodic stable characteristic value at each sampling moment.

7. The power signal filtering method based on Soc chip according to claim 1, characterized in that: The method for filtering all voltage signals on the filter circuit of the power supply comprises: The length of the filter window is substituted into the SG filter in the power supply, and the SG filter is used to filter the sequence composed of the voltage signals at all sampling moments on the filter circuit in the power supply.