Radar signal-to-noise ratio time sequence data denoising method, device and equipment and storage medium

Abstract

The present application relates to a radar signal-to-noise ratio time series data denoising method, device, equipment and storage medium. The method comprises: adding adaptive white noise to the original signal to generate an adaptive noise-added signal; searching for extreme points in the adaptive noise-added signal; selecting key points from non-extreme points according to an adaptive selection strategy based on the extreme points; constructing an envelope line according to the extreme points and the key points, decomposing the adaptive noise-added signal based on the envelope line to obtain a set of intrinsic mode components; further decomposing the intrinsic mode components in the set of intrinsic mode components to obtain secondary intrinsic mode components; calculating the scale index of the secondary intrinsic mode components, and superimposing and reconstructing the secondary intrinsic mode components with all scale indexes greater than a noise threshold to obtain a final denoised signal. The method provided by the present application can reduce the degree of information loss, realize more complete decomposition of the signal, and improve the overall performance and practicality of signal denoising.

Description

Technical Field

[0001] This invention relates to the field of radar signal processing technology, and in particular to a method, apparatus, device, and storage medium for denoising radar signal-to-noise ratio time-series data. Background Technology

[0002] In the complex maritime environment, radar systems are affected by inter-pulse interference and sea clutter during missions, causing high-frequency oscillations in the signal-to-noise ratio (SNR) timing data. These interference factors often prevent the radar SNR data from accurately reflecting the true signal situation, posing significant challenges to radar system performance evaluation and target identification.

[0003] Currently, various methods have been proposed and applied for denoising radar signal-to-noise ratio (SNR) data. Common signal decomposition methods include Empirical Mode Decomposition (EMD), Variational Mode Decomposition (VMD), and Wavelet Transform (WT). Among these, Empirical Mode Decomposition suffers from mode aliasing, which may result in the loss of some useful signals or the retention of some noise during denoising. To address the mode aliasing problem, Complete Set Empirical Mode Decomposition (CEEMD) eliminates noise interference by adding complementary white noise.

[0004] However, the principle of CEEMDAN signal decomposition is based on constructing the envelope from extrema. When dealing with complex, non-stationary radar signal-to-noise ratio signals, constructing the envelope solely from a small number of extrema inevitably leads to information loss, resulting in unstable envelope fitting and consequently affecting the signal decomposition accuracy. Furthermore, after CEEMDAN decomposition, the first few eigenmode components still exhibit a small amount of mode aliasing. Summary of the Invention

[0005] Therefore, it is necessary to provide a radar signal-to-noise ratio time-series data denoising method, apparatus, device, and storage medium that can solve the problem of insufficient signal decomposition accuracy in existing systems and improve the overall performance and practicality of signal denoising.

[0006] A method for denoising radar signal-to-noise ratio time-series data, the method comprising: Acquire the original radar signal, add adaptive white noise to the original signal to generate an adaptive noise-added signal; Search for extreme points in the adaptive noise-adding signal; based on the extreme points, select key points from non-extreme points according to the adaptive selection strategy; An envelope is constructed based on the extreme points and the key points. The adaptive noise signal is decomposed based on the envelope to obtain a set of intrinsic mode components. The intrinsic mode components in the intrinsic mode component set are decomposed again to obtain the second-order intrinsic mode components; Calculate the scaling exponent of the quadratic intrinsic mode components, and superimpose and reconstruct all quadratic intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

[0007] On the other hand, a radar signal-to-noise ratio time-series data denoising device is also provided, the device comprising: An adaptive noise generation module is used to acquire the original radar signal and add adaptive white noise to the original signal to generate an adaptive noise signal. The key point selection module is used to search for extreme points in the adaptive noise signal; based on the extreme points, key points are selected from non-extreme points according to the adaptive selection strategy. The intrinsic mode component calculation module is used to construct an envelope based on the extreme points and the key points, and decompose the adaptive noise signal based on the envelope to obtain a set of intrinsic mode components. The secondary decomposition module is used to further decompose the intrinsic mode components in the intrinsic mode component set to obtain secondary intrinsic mode components. The denoising signal acquisition module is used to calculate the scaling exponent of the second-order intrinsic mode components, and to superimpose and reconstruct all second-order intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

[0008] In another aspect, a computer device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the radar signal-to-noise ratio time-series data denoising method.

[0009] In another aspect, a computer-readable storage medium is also provided, on which a computer program is stored, wherein the computer program, when executed by a processor, implements the steps of the radar signal-to-noise ratio timing data denoising method.

[0010] Compared with existing technologies, the radar signal-to-noise ratio time-series data denoising method, apparatus, device, and storage medium provided by this invention have the following beneficial effects: 1. By using an adaptive selection strategy, key points are selected from non-extreme points. Both key points and extreme points contain important fluctuation information in the sequence. By using extreme points and key points, the accuracy of the envelope can be improved, thereby reducing the degree of loss of important information in the sequence and enhancing the ability to decompose the signal into signals with different characteristic time scales.

[0011] 2. To address the residual noise and signal aliasing, the intrinsic mode components are further decomposed, achieving further separation of noise and information, resulting in higher signal decomposition accuracy and more thorough signal decomposition.

[0012] 3. By calculating the scaling exponent of the second-order intrinsic mode components and excluding second-order intrinsic mode components with smaller scaling exponents through a noise threshold, the sequence can be made smoother, further reducing noise components.

[0013] 4. The method proposed in this invention can significantly reduce the mode aliasing problem in signal decomposition, improve the overall performance and practicality of signal denoising, and can be applied not only to radar signal denoising but also to other complex non-stationary signals, achieving good denoising effect. Attached Figure Description

[0014] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0015] Figure 1 This is a flowchart illustrating the radar signal-to-noise ratio time-series data denoising method provided in Example 1; Figure 2 This is a schematic diagram of the envelope construction based on extreme points provided in Example 1; Figure 3 This is a schematic diagram of the key point calculation provided in Example 1; Figure 4 This is a schematic diagram showing the location of key points provided in Example 1, wherein, Figure 4 (a) is a schematic diagram of the first case of key point location. Figure 4 (b) is a schematic diagram of the first scenario involving the location of the key points. Figure 4 (c) is a schematic diagram of the second scenario involving the location of key points. Figure 4 (d) is a schematic diagram of the second type of key point location; Figure 5 This is a schematic diagram of the envelope constructed based on the upper key point, lower key point, maximum point, and minimum point provided in Example 1; Figure 6 This is a structural block diagram of the radar signal-to-noise ratio timing data denoising device provided in Example 2; Figure 7 This is an internal structural diagram of a computer device in one embodiment.

[0016] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0018] It is understood that the technical solutions of the various embodiments of the present invention can be combined with each other, but only if they are based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0019] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0020] Example 1 like Figure 1 As shown, the radar signal-to-noise ratio time-series data denoising method provided in this embodiment includes the following steps: Step 201: Obtain the original radar signal, add adaptive white noise to the original signal, and generate an adaptive noise-added signal.

[0021] Step 202: Search for extreme points in the adaptive noise signal; based on the extreme points, select key points from the non-extreme points according to the adaptive selection strategy.

[0022] Step 203: Construct an envelope based on the extreme points and key points, and decompose the adaptive noise signal based on the envelope to obtain the set of intrinsic mode components.

[0023] Step 204: Decompose the intrinsic mode components in the intrinsic mode component set again to obtain the second-order intrinsic mode components.

[0024] Step 205: Calculate the scaling exponent of the second-order intrinsic mode components, and superimpose and reconstruct all second-order intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

[0025] In the specific implementation of step 201, the length is... raw radar signal Multiple sets of adaptively noised signals are obtained by adding different adaptive white noise multiple times. Among them, the first... The addition of white noise can be represented as: ; in, Indicates the first The adaptive noise-adding signal obtained by adding white noise in stages; Indicates the first The white noise added later; It represents the number of times white noise was added.

[0026] Adaptive white noise The mean is 0, and the standard deviation is 0. Based on the standard deviation and noise figure of the signal to be decomposed The decision is made jointly, and the expression is: ; In the formula, This represents the mean of the signal.

[0027] In the specific implementation of step 202, for each group of adaptive noise-adding signals, extreme points in the signal are searched. Then, based on the extreme points, key points are selected from the non-extreme points according to the adaptive selection strategy, including the following steps: Step 301: Draw straight line segments between adjacent extreme points, and regard non-extreme points located between adjacent extreme points as turning points; Step 302: Calculate the vertical distance from each turning point to the straight line segment, and determine whether the largest vertical distance satisfies the adaptive selection strategy; if so, include the corresponding turning point as a key point in the extreme point to obtain the updated extreme point. Step 303: Based on the updated extreme points, repeat steps 301 to 302 until the maximum vertical distance no longer satisfies the adaptive selection strategy, thus completing the selection of key points.

[0028] The adaptive selection strategy expression is as follows: ; In the formula, It represents the vertical distance from the turning point to the straight line segment; Indicates the threshold coefficient; Represents extreme points of coordinate; Represents extreme points of coordinate.

[0029] It is understandable that traditional methods of fitting complex non-stationary signals by establishing envelopes based on extreme points have low accuracy. Inaccurate envelopes can cause deviations in residual calculations during signal decomposition, thus affecting the accuracy of the entire decomposition process. Consequently, the main features and detailed information of the signal cannot be accurately separated and extracted. Figure 2 By constructing an envelope based on the extreme points, it can be seen that there is significant information loss in the envelope within the signal in the yellow region.

[0030] To improve the accuracy of envelope fitting, this invention introduces key points and constructs the envelope based on both extrema and key points. Key points refer to non-extrema points in the signal that exhibit significant fluctuations; they are a series of points in the sequence that reflect important fluctuations and transitions in the signal. Constructing the envelope based on key points and extrema points can reduce the degree of loss of sequence information.

[0031] The extreme points obtained from the search include maxima. and the minimum point When selecting key points, the first iteration calculation is performed based on adjacent minimum and maximum points.

[0032] like Figure 3 As shown, extreme points The point is the minimum point, the extreme point. A local maximum point; connect adjacent local maximum points. and minimum point This forms a straight line segment. Simultaneously, spatially, it will be located at adjacent maxima. and minimum point Non-extreme points between these points are considered inflection points. Then calculate each turning point. Vertical distance between the line segment and the line segment It can be seen that the turning point It can reflect the degree of volatility and turning points of the sequence. The greater the turning point, the longer the length before and after the turning point, then the turning point... The more critical, the more corresponding to The larger.

[0033] Calculate all turning points vertical distance Then, find the maximum vertical distance. Determine the maximum vertical distance Does it satisfy the adaptive selection strategy? If so, then determine the corresponding inflection point. By incorporating key points into the extreme points, we obtain updated extreme points, thus completing the first iteration calculation.

[0034] Then, based on the updated extreme points, steps 301 to 302 are repeated for the second iteration. In subsequent iterations, the minimum and maximum points are no longer distinguished; calculations are performed directly using adjacent extreme points until no key point satisfying the adaptive selection strategy is found among all adjacent extreme points, at which point the iteration ends.

[0035] It is worth noting that too many key points introduce redundant information, while too few may result in the loss of key features, both of which affect envelope accuracy. Based on this, this embodiment proposes an adaptive selection strategy. As can be seen from the formula, the threshold coefficient... The larger the threshold, the more drastic the fluctuations at key points. Therefore, by adjusting the threshold coefficient... This allows for the precise selection of an appropriate number of key points.

[0036] Threshold coefficient Adaptive adjustment is achieved based on the Euclidean distance between the two extreme points. The larger the Euclidean distance between the two extreme points, the sparser the number of fitted points in the sequence, and the distance threshold coefficient needs to be reduced. To increase the number of key points; conversely, the smaller the Euclidean distance between the two extreme points, the denser the fitted points, and the threshold coefficient should be increased. This reduces the number of key points, preventing an excessive number of key points from affecting fitting efficiency and accuracy.

[0037] However, if the key points are not distinguished and are treated as both maxima and minima when constructing the envelope, obvious overfitting and underfitting will occur.

[0038] Based on this, the key points are further divided into upper key points and lower key points. An upper envelope is constructed based on the upper key points and the maximum points; a lower envelope is constructed based on the lower key points and the minimum points.

[0039] like Figure 4 As shown, assuming a turning point If a point is considered a key point, then there are four possible scenarios. Among them, Figure 4 (a) and Figure 4 In (d), the key point is located above the straight line segment and is considered as the upper key point; Figure 4 (b) and Figure 4 In (c), the key point is located below the straight line segment and is considered the lower key point.

[0040] Specifically, the upper and lower key points are determined by calculating the first slope of the line segment connecting the left extreme point and the key point. And calculate the second slope of the line segment connecting the key point and the right-hand extreme point. .like Figure 4 As shown in (a), the upward trend of the sequence slows down, and the signal enters a relatively flat period of increase. Therefore, this key point is the upper key point; or as... Figure 4 As shown in (d), the downward trend of the sequence increases, and the signal ends the relatively flat period of decline. Therefore, this key point is also a top key point.

[0041] Conversely, such as Figure 4As shown in (b), the downward trend of the sequence slows down, and the signal enters a relatively flat period of decline. This key point is the next key point; or as follows: Figure 4 As shown in (c), the upward trend is strengthening, and the signal indicates the end of a relatively flat upward phase. If so, then that key point is the next key point.

[0042] In the specific implementation of step 203, when constructing the envelope based on extreme points and key points, the introduction of key points increases the number of envelope fitting points. Therefore, using Hermite interpolation to construct the envelope effectively avoids overfitting. Figure 5 As shown, an upper envelope is constructed based on the upper keypoints and maxima, and a lower envelope is constructed based on the lower keypoints and minima. The red dots in the figure represent the upper keypoints, and the blue dots represent the lower keypoints. It can be observed that the envelope constructed after introducing keypoints better fits the original signal, without significant information loss.

[0043] Then, based on the constructed envelope, all adaptively noise-adding signals are... Decomposed into a set of intrinsic mode components and minimal residual energy .

[0044] Specifically, for all adaptive noise-adding signals After the decomposition is completed, the first intrinsic mode component is obtained by overall averaging. The expression is: ; Adaptive noise signal Remove Obtain residual energy The expression is: ; Will As the signal to be decomposed, further decomposition yields... and Repeat the above steps until the residual energy is reached. Empirical mode decomposition cannot be performed. Therefore, the original signal... Decomposed into and residual energy The expression is: .

[0045] In the specific implementation of step 204, the intrinsic mode components in the intrinsic mode component set are decomposed again. This process can be accomplished by VMD, or by using wavelet transform and wavelet basis functions to decompose the signal into wavelet coefficients of different frequencies.

[0046] In this embodiment, VMD is preferably used for signal decomposition. VMD is essentially a process of constructing and solving a variational problem, with a rigorous mathematical solution process, which can decompose a signal into a specified number of eigenmode components.

[0047] The constrained variational problem constructed by VMD can be expressed as: ; ; In the formula, Represents the original signal; Represents the mode function; This represents the actual center frequency of the mode; This indicates the estimated center frequency of the analytic signal.

[0048] Use secondary penalty factor and Lagrange multipliers The constrained variational problem is transformed into an unconstrained variational problem, and then the optimal solution of the expression is found. The augmented Lagrange expression is: ; The Alternating Direction Multiplier (ADMM) algorithm is used for iterative updates. , and The saddle point of the equation is obtained, which is equivalent to obtaining the second-order eigenmode components of the eigenmode components. .

[0049] In the specific implementation of step 205, all second-order intrinsic mode components can be calculated using DFA. scaling index To identify the noise level in a signal, permutation entropy and multi-scale permutation entropy can be used to calculate the entropy value of the sequence, reflecting the degree of disorder in the sequence, to identify the noise level in the signal. Alternatively, the correlation coefficient and Euclidean distance between IMFs and the original signal can be used to measure the similarity between the two; the higher the similarity, the more information is contained and the less noise, thus identifying the noise level in the signal.

[0050] In this embodiment, all second-order intrinsic mode components are mainly calculated using DFA. scaling index .

[0051] Specifically, given a length of time series Construct new deviation time series The expression is: ; In the formula, This represents the mean of the signal.

[0052] Will Divided into A length of The non-overlapping subintervals, i.e. The u-th subinterval can be represented as: ; In the formula, Represents the u-th subinterval; Indicates the sub-interval index; This represents the last value in each sub-interval.

[0053] The local trend of the sub-interval is fitted using the polynomial least squares method. Taking l=3 as an example, the fitting function is... for: ; In the formula, Represents the coefficient of the quadratic term; Denotes the coefficient of the linear term; Represents variables; Indicates a constant.

[0054] Calculate the elimination trend sequence for each sub-interval The expression is: ; The detrending oscillation function for each sub-interval is expressed as follows: ; The overall root mean square fluctuation is obtained by averaging over all subintervals. The expression is: ; Change the length of the subinterval Repeat the above steps to draw to... Follow A changing double logarithmic curve. For different The value, the slope obtained from the fitted curve. This is the scaling index. With scaling index The following relationship must be satisfied: ; It can be seen that the scaling exponent reflects the roughness of the curve. The larger the value, the smoother the sequence and the fewer the noise components.

[0055] Set noise threshold If Then discard the corresponding quadratic eigenmode components. , retain satisfaction of Each quadratic eigenmode component .

[0056] All quadratic eigenmode components that satisfy the conditions Overlay reconstruction, the expression is: ; Then all The signal is then reconstructed by superposition and then smoothed by median filtering to obtain the final denoised signal, expressed as: .

[0057] The method provided by this invention, in order to further reduce the mode aliasing problem in signal decomposition, introduces additional key points that can reflect important fluctuations in the sequence, and constructs the envelope of the sequence together with key points and extreme points, thereby reducing the degree of information loss and improving the ability to decompose the signal into signals with different characteristic time scales.

[0058] To address the residual noise and signal aliasing problem that exists after adding adaptive white noise, the intrinsic mode components are further decomposed to separate the noise and information components, resulting in higher signal decomposition accuracy and more thorough signal decomposition.

[0059] Meanwhile, using scaling exponents to measure the noise level of a signal does not require setting parameters, and the noise threshold can be determined based on existing literature, thus making the evaluation more accurate.

[0060] This invention is particularly suitable for denoising complex non-stationary signals, and can significantly reduce the mode mixing problem in signal decomposition, thereby achieving signal denoising and facilitating subsequent signal-to-noise ratio data analysis.

[0061] Although this embodiment Figure 1 The steps are shown sequentially as indicated by the arrows, but they are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order in which these steps are performed; they can be executed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0062] Example 2 Based on the radar signal-to-noise ratio (SNR) time-series data denoising method in Embodiment 1, this embodiment discloses a radar SNR time-series data denoising device, such as... Figure 6 As shown, the radar signal-to-noise ratio time-series data denoising device includes: an adaptive noise generation module 401, a key point selection module 402, an intrinsic mode component calculation module 403, a secondary decomposition module 404, and a denoised signal acquisition module 405, wherein: The adaptive noise generation module 401 is used to acquire the original radar signal, add adaptive white noise to the original signal, and generate an adaptive noise signal. The key point selection module 402 is used to search for extreme points in the adaptive noise signal; based on the extreme points, key points are selected from non-extreme points according to the adaptive selection strategy; The intrinsic mode component calculation module 403 is used to construct an envelope based on the extreme points and the key points, and decompose the adaptive noise signal based on the envelope to obtain a set of intrinsic mode components. The secondary decomposition module 404 is used to decompose the intrinsic mode components in the intrinsic mode component set again to obtain secondary intrinsic mode components. The denoising signal acquisition module 405 is used to calculate the scaling exponent of the second-order intrinsic mode components, and to superimpose and reconstruct all second-order intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

[0063] In this embodiment, the specific working process and working principle of the adaptive noise generation module 401, key point selection module 402, intrinsic mode component calculation module 403, secondary decomposition module 404, and noise reduction signal acquisition module 405 are the same as those in Embodiment 1, and therefore will not be described again in this embodiment. Each unit module can be implemented entirely or partially through software, hardware, or a combination thereof. Each unit module can be embedded in or independent of the processor in the computer device in hardware form, or it can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above unit modules.

[0064] Example 3 like Figure 7 The diagram illustrates a terminal device disclosed in this embodiment, comprising a transmitter, a receiver, a memory, and a processor. The transmitter transmits instructions and data, the receiver receives instructions and data, the memory stores computer-executed instructions, and the processor executes the computer-executed instructions stored in the memory to implement the method described in Embodiment 1 above.

[0065] It is important to note that the aforementioned memory can be either standalone or integrated with the processor. When the memory is set up independently, the terminal device also includes a bus for connecting the memory and the processor.

[0066] Example 4 This embodiment discloses a computer-readable storage medium storing computer-executable instructions. When a processor executes the computer-executable instructions, it implements the method in Embodiment 1 above.

[0067] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0068] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0069] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for denoising radar signal-to-noise ratio time-series data, characterized in that, The method includes: Acquire the original radar signal, add adaptive white noise to the original signal to generate an adaptive noise-added signal; Search for extreme points in the adaptive noise-adding signal; based on the extreme points, select key points from non-extreme points according to the adaptive selection strategy; An envelope is constructed based on the extreme points and the key points. The adaptive noise signal is decomposed based on the envelope to obtain a set of intrinsic mode components. The intrinsic mode components in the intrinsic mode component set are decomposed again to obtain the second-order intrinsic mode components; Calculate the scaling exponent of the quadratic intrinsic mode components, and superimpose and reconstruct all quadratic intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

2. The radar signal-to-noise ratio time-series data denoising method according to claim 1, characterized in that, Based on the extreme points, key points are selected from non-extreme points according to an adaptive selection strategy, including: Step 301: Draw straight line segments between adjacent extreme points, and regard non-extreme points located between adjacent extreme points as turning points; Step 302: Calculate the vertical distance from each turning point to the straight line segment, and determine whether the largest vertical distance satisfies the adaptive selection strategy; if so, include the corresponding turning point as a key point in the extreme point to obtain the updated extreme point. Step 303: Based on the updated extreme points, repeat steps 301 to 302 until the maximum vertical distance no longer satisfies the adaptive selection strategy, thus completing the selection of key points.

3. The radar signal-to-noise ratio time-series data denoising method according to claim 2, characterized in that, The adaptive selection strategy expression is: ； In the formula, It represents the vertical distance from the turning point to the straight line segment; Indicates the threshold coefficient; Represents extreme points of coordinate; Represents extreme points of coordinate.

4. The radar signal-to-noise ratio time-series data denoising method according to claim 3, characterized in that, The extreme points include maximum points and minimum points; the key points include upper key points and lower key points. Construct an upper envelope based on the key points and the maximum points; Construct a lower envelope based on the lower key point and the lower minimum point.

5. The radar signal-to-noise ratio time-series data denoising method according to claim 4, characterized in that, The upper key point and the lower key point are determined as follows: Calculate the first slope of the line segment connecting the left extreme point and the key point. And calculate the second slope of the line segment connecting the key point and the right extreme point. ; like , Then the key point mentioned above is the upper key point; like , Then the key point mentioned above is the next key point.

6. The radar signal-to-noise ratio time-series data denoising method according to any one of claims 1 to 5, characterized in that, When constructing the envelope based on the extreme points and the key points, Hermite interpolation is used to construct the envelope.

7. The radar signal-to-noise ratio time-series data denoising method according to any one of claims 1 to 5, characterized in that, After superimposing and reconstructing all quadratic intrinsic mode components with scaling exponents greater than the noise threshold, the process also includes: performing median filtering smoothing on the superimposed and reconstructed signal to obtain the final denoised signal.

8. A radar signal-to-noise ratio time-series data denoising device, characterized in that, The device includes: An adaptive noise generation module is used to acquire the original radar signal and add adaptive white noise to the original signal to generate an adaptive noise signal. The key point selection module is used to search for extreme points in the adaptive noise signal; based on the extreme points, key points are selected from non-extreme points according to the adaptive selection strategy. The intrinsic mode component calculation module is used to construct an envelope based on the extreme points and the key points, and decompose the adaptive noise signal based on the envelope to obtain a set of intrinsic mode components. The secondary decomposition module is used to further decompose the intrinsic mode components in the intrinsic mode component set to obtain secondary intrinsic mode components. The denoising signal acquisition module is used to calculate the scaling exponent of the second-order intrinsic mode components, and to superimpose and reconstruct all second-order intrinsic mode components with scaling exponents greater than the noise threshold to obtain the final denoised signal.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, The memory stores a computer program, and when the processor executes the computer program, it implements the steps of the radar signal-to-noise ratio time-series data denoising method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, It stores a computer program, which, when executed by a processor, implements the steps of the radar signal-to-noise ratio time-series data denoising method according to any one of claims 1 to 7.

Images