Adaptive extraction and separation method and device for hybrid signals

By delaying the signal and adding it to the undelayed signal, the high frequency signal is gradually eliminated, the 4X frequency double signal is extracted and the original phase and amplitude is restored, which solves the problem of low signal extraction accuracy in the prior art and achieves high Accurate and fast signal separation effect, suitable for complex signal processing.

CN119004058APending Publication Date: 2024-11-22ZHEJIANG UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411150691.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

The existing signal extraction technology has low accuracy in noisy environments, especially insensitive to high-frequency noise and low-frequency interference, and has limited nonlinear and time-varying signal processing capabilities, which limits its application in complex signal processing.

Method used

通过延迟信号并与未延时信号相加,逐步消除高频次信号,提取出4X倍频信号,并通过信号复位模块恢复原始相位和幅值,实现信号的自适应提取与分离。

Benefits of technology

It realizes high-precision and fast signal extraction and separation, is suitable for complex signal processing, improves the stability and control accuracy of the system, and has a wide range of applications.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004058A_ABST
    Figure CN119004058A_ABST
Patent Text Reader

Abstract

The invention discloses a hybrid signal adaptive extraction and separation method and device, and the method comprises the steps: carrying out the delay T of a signal A with a plurality of even-frequency hybrid signals, adding the delayed signal with A, and then taking half, thereby obtaining a 4X-frequency mixed signal; wherein X is a positive integer; the method comprises the following steps of: sequentially eliminating higher-frequency signals from high to low according to the frequency in a mixed signal with 4X frequency, sequentially delaying non-even times of time of half cycles of the eliminated signals from high to low, overlapping with a non-delayed quantity and taking half until the last quantity is left, extracting to obtain a single-frequency signal, and outputting the single-frequency signal. If the phase and / or amplitude in the single-frequency signal changes, recovering the original phase and / or amplitude; and subtracting the extracted single-frequency signals, repeating the process for a plurality of times for the residual mixed signals with 4X frequencies, and extracting a plurality of original single-frequency signals in sequence. According to the method, the 4X frequency multiplication signals can be rapidly and accurately extracted from the mixed signals, and effective separation of the signals is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of signal extraction, and in particular to a method and device for adaptively extracting and separating mixed signals. Background Art

[0002] In recent years, signal extraction technology has been widely used and deeply studied in various fields. The problem of extraction accuracy is one of the hot topics in the field of signal processing. Factors such as noise interference, sampling frequency, and extraction algorithm will affect the extraction accuracy, which directly affects the stability and control accuracy of the system. Effectively achieving the accuracy of signal extraction and signal separation methods is crucial to improving system performance, reducing errors, and improving control accuracy.

[0003] The disadvantages of traditional filtering methods mainly include the following aspects. First, low-pass filters are not effective in removing high-frequency noise, especially in high-noise environments, where their performance may be significantly reduced. Second, high-pass filters are insensitive to interference from low-frequency signals, which means that when low-frequency interference is severe, their filtering effect may be limited. In addition, bandpass filters and band-stop filters need to determine the frequency range in advance in applications, and have high requirements for the specific frequency of the signal, which may limit their applicability in situations where the frequency changes are large or the frequency is uncertain. In addition, these traditional filtering methods usually require manual design of filter parameters, so more expertise and experience may be required in practical applications. In summary, traditional filtering methods have the disadvantages of limited frequency selectivity and weak processing capabilities for nonlinear and time-varying signals, which affect their application effects and scope of application in complex signal processing.

[0004] In recent years, many methods, including those based on deep learning and adaptive filtering, have been proposed for signal separation and extraction. These methods have overcome the limitations of traditional filtering methods to a certain extent, but they also have some disadvantages. First, methods based on deep learning require a large amount of data for training, and the design of the network structure and the selection of parameters require high expertise, which increases the complexity of implementation. Secondly, adaptive filtering methods have limited processing capabilities for the time-varying and nonlinear characteristics of signals and may perform poorly in complex environments. Actual signals are often complex mixed signals, so additional prior information may be required to improve the separation effect. In summary, although the new signal separation and extraction methods have improved the processing efficiency and separation accuracy to a certain extent, they still rely on the amount of data and the design of the network structure, which limits their wide application and promotion in practical applications. Summary of the invention

[0005] The purpose of the present invention is to address the deficiencies of the prior art and provide a method and device for adaptively extracting and separating mixed signals, which can quickly and accurately extract 4X (X is a positive integer) frequency-multiplied signals from mixed signals and achieve effective separation between these signals.

[0006] The objective of the present invention is achieved through the following technical solution: A method for adaptively extracting and separating mixed signals, comprising the following steps:

[0007] (1) A signal A having multiple even frequencies is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then halved to obtain a mixed signal with a frequency of 4X; wherein X is a positive integer;

[0008] (2) Eliminate the higher frequency signal from high to low according to the frequency in the mixed signal of 4X frequency, delay the eliminated signal by non-even multiples of half cycle from high to low, and add it to the undelayed amount and take half of it until the last amount remains, then extract a single frequency signal, and if the phase and / or amplitude of the single frequency signal changes, restore its original phase and / or amplitude;

[0009] (3) Subtract the extracted single-frequency signal, and repeat step (2) several times for the remaining 4X frequency mixed signal to extract several original single-frequency signals in turn.

[0010] Furthermore, the step (1) is specifically as follows:

[0011] Let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″″Y , represents the amplitude and phase of k″Y frequency signal respectively, and t represents time. Delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

[0012] Furthermore, in step (2), a single frequency signal is extracted as follows:

[0013] The mixed signal of 4X frequency is represented by Bn, where n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use (m is any positive integer) to achieve the elimination of 4n components, and the same goes for other quantities until the last quantity remains.

[0014] Furthermore, if the phase and / or amplitude of the single-frequency signal changes, the original phase and / or amplitude is restored, specifically:

[0015] Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1 B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; the single-frequency signal extracted in step (2) is expressed as Among them, i, Represents 1 B The distortion of signal amplitude and phase, i, k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

[0016] Furthermore, the original single-frequency signals extracted in step (2) and step (3) are extracted from the mixed signal of 4X frequency in order from low to high frequency.

[0017] The present invention also provides an adaptive extraction and separation device for mixed signals, comprising a signal extraction module and a signal separation module;

[0018] The signal extraction module, whose input is a mixed signal A containing multiple even frequencies, is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then halved to obtain a mixed signal with a frequency of 4X; wherein X is a positive integer;

[0019] The signal separation module includes a circulation module;

[0020] The circulation module, whose input is a mixed signal of 4X frequency, outputs a number of single-frequency signals with original phase and amplitude;

[0021] The circulation module includes several circulation submodules; the input of each circulation submodule is the difference between the input and output of the previous circulation submodule, and the output is a single-frequency signal with original phase and amplitude;

[0022] The circulation submodule at least includes a sequential delay cancellation module;

[0023] The sequential delay cancellation module has an input that is the difference between the input and output of the previous cyclic submodule, that is, the remaining 4X frequency mixed signal; according to the frequency in the remaining 4X frequency mixed signal, the higher frequency signal is eliminated from high to low, and the eliminated signal is sequentially delayed from high to low by a non-even multiple of the half cycle, and the delay is superimposed with the undelayed amount and halved to output a single-frequency signal. If the phase and / or amplitude in the single-frequency signal changes, the cyclic submodule also includes a signal reset module, which is used to restore the original phase and / or amplitude of the output result with a changed phase and / or amplitude.

[0024] Furthermore, the signal extraction module is specifically:

[0025] Let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″Y , represents the amplitude and phase of k″Y frequency signal respectively, and t represents time. Delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

[0026] Furthermore, the sequential delay cancellation module in the first loop submodule has a mixed signal B of 4X frequency as input. n , which outputs a single-frequency signal as follows:

[0027] n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use Elimination of 4n components is achieved, and the same is true for other quantities until the last quantity remains.

[0028] Furthermore, if the phase and / or amplitude of a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule changes, the first cyclic submodule further includes a signal resetting module for restoring its original phase and / or amplitude, specifically:

[0029] Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule is expressed as Among them, i, Represents 1 B The distortion of the signal amplitude and phase, where i and k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

[0030] Furthermore, the original single-frequency signal extracted in the signal separation module is extracted from the mixed signal of 4X frequency in order from low to high frequency.

[0031] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0032] The 4X (X is a positive integer) frequency-multiplied signals of multiple signals can be quickly extracted from a set of mixed quantities of even frequencies, and effective separation between these frequency-multiplied signals can be achieved. The present invention has the following advantages:

[0033] High-precision control: Through the signal separation and extraction method, the patented method can accurately extract the required 4X frequency-multiplied signal, thereby achieving precise control of current harmonics and making the motor run more stable.

[0034] Wide range of applications: This separation method is not only suitable for harmonic extraction of specific types of motors, but also for signal extraction and separation in other different scenarios.

[0035] High reliability: The signal extraction and separation method proposed in this patent has high reliability and stability, can maintain long-term stability during operation, and is not easily affected by external interference.

[0036] Efficient separation: The signal extraction and separation method proposed in this patent can quickly and effectively extract and separate special frequency-multiplied signals, thereby improving the processing efficiency and response speed of the system. This is crucial for real-time control and feedback regulation.

[0037] In summary, the present invention provides an advanced signal processing method, which can effectively extract a 4X frequency signal from a multi-frequency even-order mixed signal and separate the frequency signals in the 4X frequency signal, providing strong support for the technical development in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative labor.

[0039] Figure 1 A schematic flow chart of a method for adaptively extracting and separating mixed signals provided by an embodiment of the present invention;

[0040] Figure 2 A schematic flow chart of a signal extraction method provided by an embodiment of the present invention;

[0041] Figure 3 A schematic diagram of a signal separation module provided by an embodiment of the present invention;

[0042] Figure 4 A schematic diagram of a circulation submodule provided in an embodiment of the present invention

[0043] Figure 5 This is a diagram showing an application example of an adaptive extraction and separation method for mixed signals provided in Example 1 of the present invention. DETAILED DESCRIPTION

[0044] The present invention is described in detail below in conjunction with the accompanying drawings. In the absence of conflict, the features of the following embodiments and implementations can be combined with each other.

[0045] In the fields of industry and communications, although signal extraction technology has made significant progress, there are still many challenges in processing complex mixed signals containing multiple even frequencies, such as current harmonic extraction in power systems and multipath interference in communication systems. The current technical means of signal extraction and separation are often inefficient and inaccurate due to the high complexity of signal processing, which limits its application scope and system performance.

[0046] In view of this, the main purpose of the embodiments of the present invention is to provide a new signal separation and extraction method, aiming to solve the following key technical problems:

[0047] 1. High-precision and high-efficiency signal extraction: The present invention introduces an innovative signal extraction module, which can quickly and accurately extract a specific 4X frequency-multiplied signal (X is a positive integer), such as 4, 8, 12, etc., from a mixed signal containing multiple even frequencies. Such extraction not only reduces the error caused by signal mixing, but also greatly improves the extraction accuracy and efficiency.

[0048] 2. Effective separation of complex signals: The present invention can effectively and sequentially separate even-frequency signals from mixed signals through an innovative signal separation module. By applying sequential delay cancellation and signal resetting technology, the signal separation error is reduced, while the original phase and amplitude of the signal are restored, thereby improving the overall performance of the system.

[0049] 3. Wide technical applicability: The method of the present invention is applicable to the extraction of motor current harmonics and other fields requiring high-precision signal separation. This versatility and flexibility provide broad prospects for the application of this technology in multiple key fields.

[0050] In summary, the present invention aims to promote the advancement of signal processing technology and meet the demand for high-performance signal processing systems in modern industry and electronic technology by providing a more accurate, efficient and reliable signal extraction and separation method.

[0051] The present invention provides a method for adaptively extracting and separating mixed signals, such as Figure 1 As shown, the following steps are included:

[0052] (1) Figure 2 As shown, a signal A with multiple even frequencies is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then half is taken, resulting in a mixed signal of 4X frequency; wherein X is a positive integer;

[0053] Specifically, let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″Y , represents the amplitude and phase of k″Y frequency signal respectively, and t represents time. Delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

[0054] (2) Eliminate the higher frequency signal from high to low according to the frequency in the mixed signal of 4X frequency, delay the eliminated signal by non-even multiples of half cycle from high to low, and add it to the undelayed amount and take half of it until the last amount remains, then extract a single frequency signal, and if the phase and / or amplitude of the single frequency signal changes, restore its original phase and / or amplitude;

[0055] Among them, a single frequency signal is extracted as follows:

[0056] The mixed signal of 4X frequency is represented by Bn, where n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use Elimination of 4n components is achieved, and the same is true for other quantities until the last quantity remains.

[0057] If the phase and / or amplitude of the single frequency signal changes, the original phase and / or amplitude is restored, specifically:

[0058] Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1 B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; the single-frequency signal extracted in step (2) is expressed as Among them, i, Represents 1 B The distortion of the signal amplitude and phase, where i and k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

[0059] (3) Subtract the extracted single-frequency signal, and repeat step (2) several times for the remaining 4X frequency mixed signal to extract several original single-frequency signals in turn.

[0060] Specifically, the original single-frequency signals extracted in step (2) and step (3) are extracted from the mixed signal of 4X frequency in order from low to high frequency. B , 2 B , 3 B …n B ; 1 B It is the first quantity in B with the frequency from low to high.

[0061] It should be noted that when there are only two mixed signals of the remaining 4X frequency, step (2) is repeated again for separation and extraction to obtain (n-1) B , the last one left is n B or B ′,n B ' indicates that the phase and / or amplitude changes B; if n B ′, then we need to restore its original phase and / or amplitude to obtain n B .

[0062] The present invention also provides an adaptive extraction and separation device for mixed signals, comprising a signal extraction module and a signal separation module;

[0063] The signal extraction module, whose input is a mixed signal A containing multiple even frequencies, is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then halved to obtain a mixed signal with a frequency of 4X; wherein X is a positive integer;

[0064] The signal extraction module is specifically:

[0065] Let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″Y , represents the amplitude and phase of k″Y frequency signal respectively, and t represents time. Delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

[0066] The signal separation module includes a circulation module;

[0067] The circulation module, whose input is a mixed signal of 4X frequency, outputs a number of single-frequency signals with original phase and amplitude;

[0068] like Figure 3 As shown, the cycle module includes several cycle submodules; the input of each cycle submodule is the difference between the input and output of the previous cycle submodule, and the output is a single-frequency signal with original phase and amplitude;

[0069] The circulation submodule at least includes a sequential delay cancellation module;

[0070] The sequential delay cancellation module, whose input is the difference between the input and output of the previous cyclic submodule, that is, the remaining 4X frequency mixed signal; according to the frequency in the remaining 4X frequency mixed signal, the higher frequency signal is eliminated from high to low, and the non-even multiples of the half cycle of the eliminated signal are sequentially delayed from high to low, and the half is superimposed with the undelayed amount and then taken, and a single-frequency signal is output. If the phase and / or amplitude in the single-frequency signal changes, the cyclic submodule also includes a signal reset module, such as Figure 4 As shown, the signal resetting module is used to restore the original phase and / or amplitude of the output result with a changed phase and / or amplitude.

[0071] Take the sequential delay cancellation module in the first loop submodule as an example, its input is a mixed signal B with a frequency of 4X n , which outputs a single-frequency signal as follows:

[0072] n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use Elimination of 4n components is achieved, and the same is true for other quantities until the last quantity remains.

[0073] If the phase and / or amplitude of a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule changes, the first cyclic submodule further includes a signal resetting module for restoring its original phase and / or amplitude (the same is true for the other cyclic submodules, which will not be described one by one), specifically:

[0074] Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1 B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule is expressed as Among them, i, Represents 1 B The distortion of the signal amplitude and phase, where i and k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

[0075] The input of the sequential delay cancellation module is the difference between the input and output of the previous cyclic submodule. The input of the first cyclic submodule is B n , the output is 1 B, the difference between its input and output is B n-1 ; B n-1 This is the input of the second loop submodule (and also the input of the sequential delay cancellation module in the second loop submodule), and the output is 2 B ; Cycle in sequence, no more details.

[0076] It should be noted that the input of the last loop submodule is B 2 , the output is (n-1) B , then the last one left is n B or B ′,n B ' indicates that the phase and / or amplitude changes B ; if n B ′, the circulation module also includes a signal reset module for restoring its original phase and / or amplitude to obtain n B .

[0077] Example 1

[0078] This embodiment is a method for separating and controlling current harmonics of a permanent magnet synchronous motor. Taking a dual three-phase motor as an example, its position and function in the dual three-phase permanent magnet synchronous motor control system are as follows: Figure 5 This embodiment takes the 5th, 7th, 11th, 13th, 17th, 19th, 23rd, and 25th time harmonics as examples to illustrate the method of the present invention.

[0079] The phase current in the natural coordinate system is transformed by Clark transformation and Prak coordinate transformation, and the phase current in the rotating coordinate system is a mixture of 6th, 12th, 18th, and 24th current harmonics.

[0080] The multi-frequency delay extraction module includes a signal extraction module and a signal separation module. First, according to the signal extraction module, the non-even multiple times of the 4-fold frequency signal cycle in the delayed mixed signal are calculated, that is, (m is a positive integer. To reduce the delay error, m is set to 1, that is, ), the delayed amount is added or subtracted from the undelayed amount and the result is halved to achieve the separation of the 12th harmonic and the 24th harmonic.

[0081]

[0082] The above B n As input, it is sent to the sequential delay cancellation module in the first loop submodule, according to B n The frequencies in the sequence are high to low, and the higher frequency signals are eliminated in turn. The signal is delayed from high to low for an even multiple of half the cycle of the eliminated signal, and the delay is added to half of the delay. At this time, B n To achieve the mixing of 12th harmonic and 24th harmonic, the corresponding delay time is (m is a positive integer. To reduce the delay error, m is set to 1, that is, ):

[0083] Delay

[0084]

[0085] Since the mixed signal frequency is small in this example, and It is half a cycle of the 12th harmonic, so the sequential delay cancellation module does not affect the phase and amplitude of the 12th harmonic. Therefore, the signal reset module does not perform any calculation and delay at this time, and the input quantity can be directly output to separate the original 12th harmonic.

[0086] After the original 12th harmonic is separated, the original signal B n Subtract the 12th harmonic obtained by separation, that is:

[0087]

[0088]

[0089] In this example, the frequency of the mixed signal is relatively low, and only the 24th harmonic remains after one cycle. Moreover, the phase and amplitude of the 24th harmonic do not change, so the original 24th harmonic can be separated by direct output.

[0090] After extracting the harmonic components in the phase current, the controller is used to track the harmonic and fundamental components of each frequency separately to achieve independent control of each component and avoid the disadvantages of mixed control. Finally, after PWM modulation, it passes through the voltage inverter and is output to the motor end to provide power speed and torque.

[0091] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.

Claims

1. An adaptive extraction and separation method for mixed signals, It is characterized in that The following steps are involved: (1) A signal A having multiple even frequencies is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then halved to obtain a mixed signal with a frequency of 4X; wherein X is a positive integer; (2) Eliminate the higher frequency signal from high to low according to the frequency in the mixed signal of 4X frequency, delay the eliminated signal by non-even multiples of half cycle from high to low, and add it to the undelayed amount and take half of it until the last amount remains, then extract a single frequency signal, and if the phase and / or amplitude of the single frequency signal changes, restore its original phase and / or amplitude; (3) Subtract the extracted single-frequency signal, and repeat step (2) several times for the remaining 4X frequency mixed signal to extract several original single-frequency signals in turn.

2. The method according to claim 1, It is characterized in that The step (1) is specifically: Let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″Y , represents the amplitude and phase of k″Y frequency signals respectively, t represents time; delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

3. The method according to claim 1, It is characterized in that In the step (2), a single frequency signal is extracted as follows: The mixed signal of 4X frequency is represented by Bn, where n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use (m is any positive integer) to achieve the elimination of 4n components, and the same goes for other quantities until the last quantity remains.

4. The method according to claim 1, It is characterized in that If the phase and / or amplitude of the single frequency signal changes, the original phase and / or amplitude is restored, specifically: Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1 B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; the single-frequency signal extracted in step (2) is expressed as in, Represents 1 B The distortion of signal amplitude and phase, i, k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

5. The method according to claim 1, It is characterized in that The original single-frequency signals extracted in step (2) and step (3) are extracted from the mixed signal of 4X frequency in order from low to high frequency.

6. An adaptive extraction and separation device for mixed signals, It is characterized in that It includes a signal extraction module and a signal separation module; The signal extraction module, whose input is a mixed signal A containing multiple even frequencies, is delayed by T, where T is a non-even multiple of the period of the 4-fold frequency signal in the mixed signal A, and the delayed signal is added to A and then halved to obtain a mixed signal with a frequency of 4X; wherein X is a positive integer; The signal separation module includes a circulation module; The circulation module, whose input is a mixed signal of 4X frequency, outputs a number of single-frequency signals with original phase and amplitude; The circulation module includes several circulation submodules; the input of each circulation submodule is the difference between the input and output of the previous circulation submodule, and the output is a single-frequency signal with original phase and amplitude; The circulation submodule at least includes a sequential delay cancellation module; The sequential delay cancellation module has an input that is the difference between the input and output of the previous cyclic submodule, that is, the remaining 4X frequency mixed signal; according to the frequency in the remaining 4X frequency mixed signal, the higher frequency signal is eliminated from high to low, and the eliminated signal is sequentially delayed from high to low by a non-even multiple of the half cycle, and the delay is superimposed with the undelayed amount and halved to output a single-frequency signal. If the phase and / or amplitude in the single-frequency signal changes, the cyclic submodule also includes a signal reset module, which is used to restore the original phase and / or amplitude of the output result with a changed phase and / or amplitude.

7. The device according to claim 6, It is characterized in that The signal extraction module is specifically: Let the time domain expression of A be Where (n+q) is the number of signals contained in the mixed signal A, where n represents the number of 4X frequency signals, q represents the number of k″Y frequency signals, n and q are both non-zero integers; k″ is a non-zero positive even number and k″≠4j, j is a positive integer; X and Y are both positive integers, where X∈(1,n); Y∈(1,q); I 4X , Respectively represent the amplitude and phase of the 4X frequency signal; I k″Y , represents the amplitude and phase of k″Y frequency signals respectively, t represents time; delay (m is an arbitrary positive integer) and then added to the amount before the delay and halved to achieve 4X times the extraction of the mixed signal.

8. The device according to claim 6, It is characterized in that The sequential delay cancellation module in the first loop submodule has a mixed signal B with a frequency of 4X as input. n , which outputs a single-frequency signal as follows: n is the number of mixed signals; B n The time domain model is expressed as: The highest frequency in the mixed signal is the 4n-th signal. First, eliminate the 4n-th signal and use Elimination of 4n components is achieved, and the same is true for other quantities until the last quantity remains.

9. The device according to claim 6, It is characterized in that If the phase and / or amplitude of a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule changes, the first cyclic submodule further includes a signal resetting module for restoring its original phase and / or amplitude, specifically: Assumption B n Middle 1 B The time domain expression of the secondary signal is expressed as Among them, 1 B The original amplitude and phase of the secondary signal are A v , It is the vth harmonic; a single-frequency signal output by the sequential delay cancellation module in the first cyclic submodule is expressed as in, Represents 1 B The distortion of the signal amplitude and phase, where i and k' are non-zero real numbers; at this time, the delay and multiply by the amplitude factor m′ is a positive integer, and the original 1 can be restored B secondary signal.

10. The device according to claim 6, It is characterized in that The original single-frequency signal extracted in the signal separation module is extracted from the mixed signal of 4X frequency in order from low to high frequency.