Method, device and readable storage medium for extracting multi-scale periodic regularity of velocity data based on improved Morlet wavelet

Abstract

The present invention proposes a method, device, and readable storage medium for extracting multi-scale periodic patterns from flow velocity data based on an improved Morlet wavelet. This method addresses the problems of insufficient time-domain localization, imbalanced high- and low-frequency resolution, and noise interference in non-stationary flow velocity data, as observed in traditional methods. The method is implemented through the following steps: The original flow velocity data is preprocessed using 5-point Gaussian interpolation and normalization; a Morlet wavelet function whose center frequency adaptively changes with scale is constructed, and parameters are dynamically optimized based on a period identification contrast index using an improved whale optimization algorithm; a wavelet transform is performed on the preprocessed data and the power spectrum is calculated; and key periods are extracted through local maxima of the energy spectral density and threshold screening. The present invention significantly improves period identification accuracy, enhances the signal-to-noise ratio, and achieves accurate and stable extraction of multi-scale periodic features. The method is suitable for non-stationary signal analysis scenarios such as hydrological flow velocity status monitoring.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing and hydrological information analysis, and in particular to a method and device for extracting multi-scale periodicity from velocity data based on an improved Morlet wavelet, and a readable storage medium thereof. Background Art

[0002] In natural environments such as rivers and lakes, flow velocity data is influenced by multiple factors, including tides and wind, and exhibits non-stationary and multi-scale periodic fluctuations. Traditional analysis methods, such as the Fast Fourier Transform (FFT), lack time-domain localization capabilities and are difficult to identify non-stationary periodic variations. The Morlet wavelet in the standard Continuous Wavelet Transform (CWT) uses a fixed center frequency, which cannot achieve resolution for both high- and low-frequency periods. Furthermore, flow velocity signals often contain high noise, short-period perturbations, and long-period trends. Traditional methods are prone to missing weak periodic features or misidentifying the dominant periodic frequency band. Therefore, an adaptive, multi-scale period extraction method is urgently needed to improve detection accuracy, stability, and time-frequency resolution. Summary of the Invention

[0003] The embodiments of the present invention provide a method, device and readable storage medium for extracting multi-scale periodic patterns from flow velocity data based on an improved Morlet wavelet, which addresses the problems existing in current technologies in the multi-scale periodic analysis of non-stationary flow velocity data, such as insufficient time domain positioning, imbalance in high and low frequency resolution, susceptibility to noise interference leading to loss of weak periods or misidentification of main periods.

[0004] The core technology of this invention is to construct an improved Morlet wavelet function whose center frequency changes adaptively with scale, and dynamically optimize the wavelet parameters based on the period recognition contrast index and the improved whale optimization algorithm to achieve high-precision extraction of multi-scale periodic features of flow velocity data.

[0005] In a first aspect, the present invention provides a method for extracting multi-scale periodic patterns from velocity data based on an improved Morlet wavelet, the method comprising the following steps:

[0006] Preprocess the original flow velocity data to generate a continuous time series signal;

[0007] A Morlet wavelet function with a center frequency that changes adaptively with scale is constructed, and the wavelet function parameters are dynamically determined through an optimization algorithm based on the period identification contrast index.

[0008] Perform wavelet transform on the pre-processed velocity data and calculate the wavelet power spectrum;

[0009] Identify periodic peaks based on wavelet power spectrum and extract key periodic factors;

[0010] The period recognition contrast index is the ratio of the average power of the target period interval to the average power of the non-target period interval.

[0011] Furthermore, the preprocessing includes: using a 5-point Gaussian interpolation method to complete the missing data, the 5-point Gaussian interpolation method builds a regression model based on 4 frames of historical data at adjacent moments, and calculates the estimated value of the missing moment.

[0012] Furthermore, the Morlet wavelet function expression of which the center frequency changes adaptively with the scale is:

[0013]

[0014] Among them, the center frequency It satisfies the linear relationship with the scale factor a:

[0015] ; is the basic frequency, is the adaptive coefficient.

[0016] Furthermore, the optimization algorithm is an improved whale optimization algorithm, which determines the optimal adaptive coefficient by maximizing the contrast index. .

[0017] Furthermore, the improved whale optimization algorithm includes: initializing a group of whale individuals, each of which corresponds to an adaptive coefficient ;

[0018] Calculate individual fitness, i.e. contrast index, and record the best individual;

[0019] Update individual positions through either a prey encirclement mechanism or a spiral approach mechanism;

[0020] Iterate until the convergence condition is met and output the optimal parameters, that is, the optimal adaptive coefficients ;

[0021] Among them, the convergence factor b of the convergence condition is dynamically adjusted according to the cosine function with the number of iterations.

[0022] Furthermore, the scale set of wavelet transform grows exponentially, and the calculation formula is:

[0023]

[0024] in, is the minimum scale, is the logarithmic step size, j is the scale series, and J is the upper limit of the scale series.

[0025] Furthermore, the period peak identification step includes: calculating the energy spectrum density of each scale, screening the local maximum value and combining the threshold to eliminate noise, and extracting the key period.

[0026] In a second aspect, the present invention provides a device for extracting multi-scale periodic patterns from velocity data based on an improved Morlet wavelet, comprising:

[0027] A preprocessing module is used to preprocess the original flow velocity data to generate a continuous time series signal;

[0028] Adaptive Morlet wavelet construction module, used to construct a Morlet wavelet function whose center frequency changes adaptively with scale;

[0029] Parameter optimization module, which is used to dynamically determine the wavelet function parameters through optimization algorithm based on period identification contrast index;

[0030] Improved wavelet transform calculation module, used to perform wavelet transform on pre-processed velocity data and calculate wavelet power spectrum;

[0031] The periodic component extraction and feature analysis module is used to identify periodic peaks based on the wavelet power spectrum and extract key periodic factors.

[0032] In a third aspect, the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the above-mentioned method for extracting multi-scale periodic laws from velocity data based on the improved Morlet wavelet.

[0033] In a fourth aspect, the present invention provides a readable storage medium, in which a computer program is stored. The computer program includes a program code for controlling a process to execute a process, and the process includes the above-mentioned multi-scale periodic law extraction method of flow velocity data based on the improved Morlet wavelet.

[0034] The main contributions and innovations of the present invention are as follows:

[0035] 1. Dynamically balanced multi-scale resolution: By associating the Morlet wavelet center frequency with the scale factor, the problem of "poor low-frequency frequency resolution and poor high-frequency time resolution" caused by the traditional fixed center frequency is solved, so that good time-frequency analysis capabilities can be maintained at different scales.

[0036] 2. Adaptive parameter optimization improves recognition accuracy: Based on the period recognition contrast index and the improved whale optimization algorithm, the dynamic optimization parameter \alpha is significantly improved to distinguish the energy between target and non-target periods. The tidal period recognition error is reduced from 1.6% to 0.8% compared to traditional methods, and the daily period error is reduced from 5.1% to 0.9%.

[0037] 3. Enhanced noise resistance: Through energy spectral density threshold screening and adaptive parameter optimization, the signal-to-noise ratio (SNR) is improved from 9.06dB in traditional methods to 9.54dB, effectively suppressing the interference of high noise and short-period disturbances on weak periodic features and avoiding misidentification of the main period.

[0038] 4. Improved analysis stability and interpretability: Through quantitative analysis of multi-scale time-frequency spectra and energy concentration scales, accurate and stable extraction of the periodic patterns of flow velocity data is achieved. This method is suitable for multi-scale periodic analysis scenarios of non-stationary signals such as hydrological monitoring.

[0039] The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below so that other features, objects, and advantages of the invention are more readily apparent. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0041] Figure 1 Flowchart of a method for extracting multi-scale periodic regularity from velocity data based on improved Morlet wavelet according to an embodiment of the present invention;

[0042] Figure 2 is a graph of flow velocity data of a river according to an embodiment of the present invention;

[0043] Figure 3 This is the analysis result of traditional Morlet wavelet;

[0044] Figure 4 is a diagram of analysis results using the method of the present invention according to an embodiment of the present invention;

[0045] Figure 5 FIG. 4 is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0046] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The implementations described in the following exemplary embodiments are not intended to represent all implementations consistent with one or more embodiments of this specification. Rather, they are merely examples of apparatuses and methods consistent with certain aspects of one or more embodiments of this specification, as detailed in the appended claims.

[0047] It should be noted that in other embodiments, the steps of the corresponding method are not necessarily performed in the order shown and described in this specification. In some other embodiments, the method may include more or fewer steps than those described in this specification. In addition, a single step described in this specification may be broken down into multiple steps for description in other embodiments, and multiple steps described in this specification may be combined into a single step for description in other embodiments.

[0048] In natural water flow velocity monitoring, traditional methods (such as FFT and standard Morlet wavelet transform) have significant defects:

[0049] 1. The Fourier transform (FFT) cannot capture the time-varying periodic characteristics of non-stationary signals;

[0050] 2. The standard Morlet wavelet uses a fixed center frequency, which leads to an imbalance between the time resolution of the high-frequency band (small scale) and the frequency resolution of the low-frequency band (large scale);

[0051] 3. Velocity data contains noise, short-period disturbances, and long-period trends. Traditional methods are prone to missing weak periods or misidentifying the main period (e.g., the measured error reaches 5.1%).

[0052] Based on this, the present invention solves the problems existing in the prior art by dynamically adjusting the center frequency of the Morlet wavelet.

[0053] Example 1

[0054] The present invention aims to propose a method for extracting multi-scale periodic regularities from velocity data based on an improved Morlet wavelet. By dynamically adjusting the Morlet wavelet center frequency and combining it with an improved whale optimization algorithm based on periodic contrast, the method can achieve adaptive and high-precision extraction of multi-scale periodic features.

[0055] Specifically, the embodiment of the present invention provides a method for extracting multi-scale periodic regularity of velocity data based on improved Morlet wavelet. Specifically, referring to Figure 1 , the method comprises the following steps:

[0056] Step 1: Data preprocessing

[0057] The velocity data obtained by the original Doppler flow meter are supplemented by 5-point Gaussian interpolation and normalized according to the sampling time of each data to obtain a continuous velocity time series signal.

[0058] In this embodiment, for data that changes slowly, such as water flow velocity, the data correlation between adjacent moments is greater, while the data correlation between moments with large differences is smaller. Therefore, a 5-point Gaussian interpolation method is used to complete the missing data. Specifically:

[0059] First, we use the velocity data of the complete time series to establish a Gaussian function regression model:

[0060]

[0061] in is the parameter coefficient obtained by fitting the complete historical data (or weight coefficient, which reflects the contribution of data at different times to the median value). is the estimated measurement value (missing value) at the intermediate moment, is the 4 frames of data before and after the middle data (adjacent to time t). After establishing the Gaussian function regression model, substitute the 4 frames of data before and after the missing data to calculate the missing measurement data.

[0062] Essentially, it utilizes the strong temporal correlation and slow-changing characteristics of velocity data to predict missing values ​​through weighted linear combination of four adjacent frames of data, rather than simple equal-weighted interpolation (such as linear interpolation), thereby more accurately preserving data trends.

[0063] It is worth mentioning that the "Gaussian" here does not refer to the traditional Gaussian function (exponential decay form), but emphasizes the weighting strategy based on data correlation. In flow rate data, the correlation between adjacent moments (such as t-1, t+1) is usually stronger than that of distant moments (such as t-2, t+2), which is similar to the characteristic of "higher weights for nearby points" in Gaussian distribution.

[0064] Step 2: Adaptive Morlet wavelet construction

[0065] Construct the Morlet wavelet function, whose mathematical expression is:

[0066]

[0067] in express time, Indicates setting the center frequency of the wavelet, It is an imaginary unit, which makes the wavelet function have a complex form and can simultaneously analyze the amplitude and phase information of the signal. It is the complex exponential modulation part, which is essentially a sine wave oscillation. It determines the center frequency of the wavelet function and is used to extract the frequency characteristics of the signal. The Gaussian window function, shaped like a bell curve, is used to localize the complex exponential signal in the time domain, limiting the range of the wavelet function on the time axis and ensuring the locality of the time-frequency analysis. Overall effect: The product of the complex exponential and the Gaussian window makes the Morlet wavelet a time-frequency localized "bandpass filter", which can sensitively capture the changes in the time domain of signal components near specific frequencies.

[0068] in , is the basic frequency, is an adjustable scaling factor, The original Morlet wavelet adopts a fixed center frequency, which will lead to uneven resolution at different scales. Therefore, the present invention takes the scale factor into account in the calculation of the center frequency, so that the low frequency ( When the frequency is large, the frequency resolution is kept high to avoid blurring; when the frequency is high ( Small), it has good time resolution.

[0069] From the above analysis, we can get the proportional factor The value of is critical to the calculation of the center frequency. To this end, an adaptive Morlet wavelet parameter optimization method based on period recognition contrast is proposed. This method uses the contrast between the target period energy and the non-target period energy as the optimization index and adaptively adjusts the center frequency control parameter in the wavelet function. , thereby improving the ability to identify key periodic components in flow rate data.

[0070] Specifically, the center frequency of the adaptive Morlet wavelet is set as a function that varies with scale:

[0071]

[0072] in, is the basic center frequency, which determines the initial frequency of the minimum scale (highest frequency); a is the scale parameter, It is an adjustable adaptive coefficient that controls the rate at which the center frequency changes with scale.

[0073] In the classic Morlet wavelet transform, a fixed center frequency is used For high-frequency (small-scale) information, the time resolution is good, but the frequency resolution is poor; for low-frequency (large-scale) information, the frequency resolution is good, but the time resolution is poor. Fixed center frequency It is impossible to take into account the resolution of both frequency bands, and the performance is degraded when analyzing multi-scale mixed signals. Therefore, the scale factor a is introduced to achieve dynamic balance and solve the problem of unbalanced resolution between large and small scales in the classic Morlet wavelet.

[0074] For each candidate Value, perform continuous wavelet transform and calculate the power spectrum, and then calculate the average power in the target period interval Average power in non-target period , construct the cycle identification contrast index:

[0075]

[0076] Among them, ϵ is a small positive number that prevents the denominator from being zero. By maximizing this contrast index, the optimal value( ), realizes adaptive adjustment of wavelet transform parameters, so that the periodic components appear as more concentrated high-energy peaks in the power spectrum, improving the accuracy and resolution of period extraction.

[0077] Preferably, the optimization algorithm is an improved whale optimization algorithm to find the optimal parameters , the improved whale optimization algorithm is used for global optimization. The specific steps are as follows:

[0078] 2-1) Initialize the whale group, each individual position Corresponding to a parameter ;

[0079] 2-2) Calculate individual fitness, that is, indicator function ;

[0080] here The formula is as follows:

[0081]

[0082] in The parameter value is The average power within the target period is The parameter value is The average power of the non-target period interval when ϵ is a small positive number to prevent the denominator from being zero. When the value is the maximum, the parameter can be guaranteed to be When the key period parameters are extracted, the best extraction capability can be guaranteed.

[0083] 2-3) Compare within the group and record the best position , and its corresponding parameter is the current optimal parameter.

[0084] 2-4) Update individual positions, including two mechanisms:

[0085] Each individual calculates its own random number to update the parameters ,when , the prey encirclement mechanism is used for updating, otherwise, the spiral approach mechanism is used for updating.

[0086] Among them, the mechanism of surrounding prey:

[0087]

[0088] in , convergence factor It decreases linearly over time. The absolute value of A, |A|, determines the step size of the individual approaching the optimal solution. |A|>1 means that the individual position update range is expanded, enhancing the global exploration ability. |A|<1 means that the individual shrinks near the optimal solution and focuses on local development. is a random vector, which introduces randomness to avoid premature convergence. C adjusts the distance weight between the current individual and the optimal solution, so that the individual can adaptively adjust its dependence on the optimal solution during the update. is the individual position at the current iteration number t (corresponding to the candidate solution of the parameter α to be optimized); is the global optimal individual position in the current iteration (current optimal parameter ); is the position of the individual at the next iteration t+1 (the updated parameter candidate solution); A and C are coefficient vectors that control the "exploration" and "exploitation" modes of the search behavior. This formula is the core iteration rule of the Whale Optimization Algorithm (WOA), which is used to simulate the behavior of humpback whales "surrounding prey." WOA searches for the optimal solution by iteratively updating the positions of individuals (whales) in the population. This paper uses an improved whale optimization algorithm to improve optimization efficiency by dynamically adjusting parameters, and is used to solve the optimal parameters of the adaptive Morlet wavelet. .

[0089] Among them, the spiral approach mechanism:

[0090]

[0091] in is a constant that determines the shape of the spiral (e.g., when c=1, it is a standard logarithmic spiral; Control the spiral shape by The randomness of the spiral path achieves diversity; represents the exponential function term, which controls the contraction or expansion rate of the spiral; Indicates the spiral radius after exponential scaling of the distance between the current individual and the optimal solution; is the cosine function term, passing through random angles Controls the rotation direction and position of the spiral in polar coordinates. This formula is the core rule for simulating the humpback whale's "spiraling up to approach prey" behavior in the Whale Optimization Algorithm (WOA), and together with the "encircling prey mechanism" constitutes two modes of individual position update. In this invention, this mechanism is used to optimize the adaptive coefficient α of the Morlet wavelet, and through the combination of global search and local development, the period recognition contrast index is improved. The maximum value of .

[0092] When c>0, e cl Followl Increases exponentially, and the spiral radius expands (exploration mode); when c<0, e cl Exponential decay, spiral radius shrinks (development mode).

[0093] Preferably, the encirclement of the prey mechanism is a random number in [0,1], is the convergence factor and satisfies In order to further increase the diversity of individuals in the population and enhance the local optimization ability, the present invention also proposes an improved convergence factor The iterative calculation method is as follows:

[0094]

[0095] in Distribution The maximum and minimum values ​​of ; (0≤t≤T) is the current number of iterations; T is the maximum number of iterations. For example: parameter , , , number of iterations , to avoid excessive consumption of computing resources.

[0096] In this way, by adjusting b nonlinearly and dynamically, the oscillation characteristics of the cosine function are simulated, the position diversity of individuals in the population during the iteration process is increased, and at the same time, the local search accuracy is enhanced in the later stage, thereby improving the efficiency of parameter optimization.

[0097] 2-5) Iterate and update the above steps until the maximum number of iterations is reached or the convergence condition is met.

[0098] By combining the spiral approach mechanism with the prey encirclement mechanism, the improved whale optimization algorithm of the present invention achieves:

[0099] 1. Higher parameter optimization accuracy: As shown in subsequent experimental data, the tidal cycle identification error is reduced from 1.6% to 0.8% in traditional methods, and the daily cycle error is reduced from 5.1% to 0.9%;

[0100] 2. Enhanced robustness: The signal-to-noise ratio (SNR) has been improved from 9.06dB to 9.54dB, effectively suppressing noise interference on weak periodic features.

[0101] 3. Wider applicability: It can automatically adapt to the periodic analysis of flow velocity data in different hydrological scenarios (such as tides and rainfall effects) without the need for manual parameter adjustment.

[0102] This mechanism is the key innovation of the present invention in deeply integrating heuristic algorithms with signal processing technology, providing an efficient optimization tool for multi-scale feature extraction of non-stationary time series data.

[0103] Step 3: Improve wavelet transform calculation

[0104] Perform wavelet transform on the pre-processed velocity data. , time interval , wavelet scale set , the wavelet transform is defined as:

[0105]

[0106] Among them, the input signal: is the preprocessed continuous velocity time series (missing values ​​have been filled in and normalized using 5-point Gaussian interpolation).

[0107] Wavelet function: The complex conjugate of the Morlet wavelet is used to perform time-frequency localized correlation operations with the input signal. The introduction of the complex conjugate allows the wavelet coefficients to contain phase information, which can be used to analyze the phase changes of periodic components.

[0108] Scale parameter s: controls the "stretching degree" of the wavelet function. The larger s is, the wider the wavelet function is in the time domain, corresponding to the analysis of low-frequency (long-period) signals; the smaller s is, the narrower the wavelet function is, corresponding to the analysis of high-frequency (short-period) signals. is a normalization factor to ensure energy conservation at different scales.

[0109] Time parameters : Represents the translation position of the wavelet function on the time axis, which is used to locate the moment when the periodic component appears in the time domain.

[0110] Thus, the wavelet transform transforms the velocity signal into Match with Morlet wavelets of different scales and positions to calculate the wavelet coefficients The modulus of the coefficient reflects the signal in scale s and time The energy intensity at the scale is larger, and the larger the modulus value, the more significant the periodic component at that scale.

[0111] scale and analysis frequency There is an inverse relationship between them, and the approximate formula is:

[0112]

[0113] in Corresponding scale frequency; is the center frequency of the wavelet; is the sampling interval. The scale set usually grows exponentially to obtain a geometrically distributed frequency resolution.

[0114] In this way, by adjusting the scale s, the full period range of the velocity data can be systematically scanned to ensure that no periodic components of any scale (such as daily cycle, tidal cycle, etc.) are missed.

[0115]

[0116] in is the minimum scale (corresponding to the maximum frequency); is the logarithmic step size of the scale interval, which is 0.1 in this embodiment. is the upper limit of the scale series (by the maximum scale The scale grows exponentially, making the frequency resolution uniform on a logarithmic scale.

[0117] Preferably, to more precisely describe The value logic is based on the M2 tidal cycle , take the sampling interval of 300 seconds as an example. Assume the center frequency ,but:

[0118]

[0119]

[0120] The obtained scale , which means that when the wavelet scale is set to about 142, the Morlet wavelet transform is most sensitive to the 12.42-hour periodic component. Therefore, when the actual periodic data is unknown, it can be set ,use structure scales, select , and then perform the calculation.

[0121] based on , calculate the wavelet power spectrum:

[0122]

[0123] Among them, the power spectrum The larger the value of , the more concentrated the energy of the velocity data at that scale (period) and time point is, corresponding to a significant periodic component. Visualization as a two-dimensional image (time on the horizontal axis, period on the vertical axis, with darker colors indicating higher energy) allows for intuitive identification of the temporal distribution and intensity variations of periodic components. For example, a tidal period (12.42 hours) appears in the power spectrum as a continuous high-energy band at approximately 12.42 hours on the vertical axis, while a diurnal period (24 hours) appears as an energy peak at 24 hours on the vertical axis.

[0124] Step 4: Periodic component extraction and feature analysis

[0125] Based on the energy spectrum density, the periodic peak is identified, the key periodic factors are extracted, and the periodic analysis results are obtained.

[0126] At each scale, integrate over time to obtain the scale energy spectral density:

[0127]

[0128] is the wavelet power spectrum , indicating scale s, time Energy density at (energy per unit time); It is the integral of the power spectrum on the time axis, that is, the total energy under scale s, which reflects the energy concentration corresponding to the scale (period) in the velocity data. It can then be determined at which scales (periods) the overall energy is concentrated. By finding By further filtering the local maximum of the threshold, the obvious periodic information can be extracted.

[0129] Specifically, is a local maximum condition, then it satisfies:

[0130] and

[0131] By traversing each When taking value The value can be obtained local maximum of .

[0132] Preferably, in order to eliminate the interference of weak noise signals, a threshold method may be used to eliminate the period value obtained by analyzing the noise information, so as to obtain the final analyzed period information.

[0133] If the threshold method is not used to analyze and eliminate noise information without considering the interference of weak noise signals, multiple false periodic values ​​may appear. Therefore, the energy spectrum density expectation can be used as the screening threshold. Only when the local maximum value is higher than the threshold, it is considered to be valid periodic information extracted. Otherwise, it is considered to be false local maximum value information caused by signal interference such as noise, and is not valid periodic information.

[0134] For example, the data of the flow velocity of a river is used for processing. The model of the Doppler flow meter used for sampling is the acoustic Doppler flow meter developed by Kaihong Fluid Technology, model 0D6X-T, and the sampling interval is 5 minutes.

[0135] The flow velocity data of a river is as follows: Figure 2As shown, after analyzing this data and combining it with the records of the hydrological observation station, it can be found that there are two groups of flow velocity data cycles in this period, namely 12.42 hours (tidal cycle) and 24 hours (daily cycle).

[0136] Among them, the analysis results of traditional Morlet wavelet are as follows: Figure 3 As shown, it can be identified:

[0137] Number of cycles: 2;

[0138] Period value (hours): [12.6225.23];

[0139] The cycle identification errors are: 1.6%, 5.1%;

[0140] Signal-to-noise ratio (SNR) (dB): 9.06;

[0141] The analysis results of the method of the present invention are as follows Figure 4 As shown, it can be identified:

[0142] Number of cycles: 2;

[0143] Recognized period value (hours): [12.5224.23];

[0144] The cycle identification errors are: 0.8%, 0.9%;

[0145] Signal-to-noise ratio (SNR) (dB): 9.54;

[0146] Comparing the structures of the traditional Morlet wavelet and the method of the present invention, it can be seen that the method of the present invention can significantly improve the accuracy of periodic data recognition.

[0147] Example 2

[0148] Based on the same concept, the present invention also proposes a device for extracting multi-scale periodic patterns from velocity data based on an improved Morlet wavelet, comprising:

[0149] A preprocessing module is used to preprocess the original flow velocity data to generate a continuous time series signal;

[0150] Adaptive Morlet wavelet construction module, used to construct a Morlet wavelet function whose center frequency changes adaptively with scale;

[0151] Parameter optimization module, which is used to dynamically determine the wavelet function parameters through optimization algorithm based on period identification contrast index;

[0152] Improved wavelet transform calculation module, used to perform wavelet transform on pre-processed velocity data and calculate wavelet power spectrum;

[0153] The periodic component extraction and feature analysis module is used to identify periodic peaks based on the wavelet power spectrum and extract key periodic factors.

[0154] Example 3

[0155] This embodiment also provides an electronic device, referring to Figure 5 , includes a memory 404 and a processor 402, wherein the memory 404 stores a computer program, and the processor 402 is configured to run the computer program to perform the steps in any of the above method embodiments.

[0156] Specifically, the processor 402 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits for implementing the embodiments of the present invention.

[0157] Memory 404 may include a large-capacity memory 404 for data or instructions. By way of example, and not limitation, memory 404 may include a hard disk drive (HDD), a floppy disk drive, a solid-state drive (SSD), flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 404 may include removable or non-removable (or fixed) media. Where appropriate, memory 404 may be internal or external to the data processing device. In certain embodiments, memory 404 is non-volatile memory. In certain embodiments, memory 404 includes read-only memory (ROM) and random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM) or a flash memory (FLASH), or a combination of two or more of these. In appropriate circumstances, the RAM may be a static random access memory (SRAM) or a dynamic random access memory (DRAM), wherein the DRAM may be a fast page mode dynamic random access memory 404 (FPMDRAM), an extended data output dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0158] The memory 404 may be used to store or cache various data files required for processing and / or communication, as well as possible computer program instructions executed by the processor 402 .

[0159] The processor 402 reads and executes computer program instructions stored in the memory 404 to implement any one of the methods for extracting multi-scale periodicity from velocity data based on the improved Morlet wavelet in the above embodiments.

[0160] Optionally, the electronic device may further include a transmission device 406 and an input / output device 408 , wherein the transmission device 406 is connected to the processor 402 , and the input / output device 408 is connected to the processor 402 .

[0161] Transmission device 406 can be used to receive or transmit data via a network. Specific examples of such networks may include wired or wireless networks provided by the electronic device's communications provider. In one embodiment, the transmission device includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, transmission device 406 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0162] The input and output devices 408 are used to input or output information.

[0163] Example 4

[0164] This embodiment also provides a readable storage medium, which stores a computer program. The computer program includes program code for controlling a process to execute a process. The process includes the multi-scale periodic law extraction method of flow velocity data based on the improved Morlet wavelet according to the first embodiment.

[0165] It should be noted that the specific examples in this embodiment can refer to the examples described in the above embodiments and optional implementation modes, and this embodiment will not be repeated here.

[0166] In general, various embodiments may be implemented in hardware or dedicated circuitry, software, logic, or any combination thereof. Some aspects of the invention may be implemented in hardware, while other aspects may be implemented in firmware or software executed by a controller, microprocessor, or other computing device, but the invention is not limited thereto. Although various aspects of the invention may be shown and described as block diagrams, flow charts, or using some other graphical representation, it should be understood that, as non-limiting examples, the blocks, devices, systems, techniques, or methods described herein may be implemented in hardware, software, firmware, dedicated circuitry or logic, general-purpose hardware or a controller or other computing device, or some combination thereof.

[0167] The embodiments of the present invention may be implemented by computer software that is executable by a data processor of a mobile device, such as in a processor entity, or by hardware, or by a combination of software and hardware. Computer software or programs (also referred to as program products) including software routines, applets and / or macros may be stored in any device-readable data storage medium, and they include program instructions for performing specific tasks. A computer program product may include one or more computer executable components that are configured to perform an embodiment when the program is run. One or more computer executable components may be at least one software code or a portion thereof. In addition, it should be noted at this point that, for example, Figure 1 Any block of the logic flow in the program may represent program steps, or interconnected logic circuits, blocks and functions, or a combination of program steps and logic circuits, blocks and functions. The software may be stored on physical media such as memory chips or memory blocks implemented within the processor, magnetic media such as hard disks or floppy disks, and optical media such as, for example, DVDs and their data variants, CDs, etc. Physical media are non-transitory media.

[0168] Those skilled in the art should understand that the technical features of the above embodiments can be combined arbitrarily. In order to make the description concise, 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.

[0169] The above embodiments merely illustrate several embodiments of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make various modifications and improvements without departing from the spirit of the present invention, all of which fall within the scope of the present invention. Therefore, the scope of the present invention shall be determined by the appended claims.

Claims

1. A method for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet, characterized in that: The following steps are involved: Preprocess the original flow velocity data to generate a continuous time series signal; A Morlet wavelet function with a center frequency that changes adaptively with scale is constructed, and the wavelet function parameters are dynamically determined through an optimization algorithm based on the period identification contrast index. Perform wavelet transform on the pre-processed velocity data and calculate the wavelet power spectrum; Identify periodic peaks based on wavelet power spectrum and extract key periodic factors; The period recognition contrast index is the ratio of the average power of the target period interval to the average power of the non-target period interval; The Morlet wavelet function expression of the center frequency adaptively changing with scale is: Among them, the center frequency It satisfies the linear relationship with the scale factor a: ; is the fundamental frequency, is the adaptive coefficient.

2. The method for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet according to claim 1, characterized in that: The preprocessing includes: using a 5-point Gaussian interpolation method to fill in the missing data, the 5-point Gaussian interpolation constructs a regression model based on 4 frames of historical data at adjacent moments, and calculates an estimated value of the missing moment.

3. The method for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet according to claim 1, characterized in that: The optimization algorithm is an improved whale optimization algorithm, which determines the optimal adaptive coefficient by maximizing the contrast index. .

4. The method for extracting multi-scale periodic regularity from velocity data based on improved Morlet wavelet according to claim 3, characterized in that: The improved whale optimization algorithm includes: initializing a group of whale individuals, each of which corresponds to an adaptive coefficient ; Calculate individual fitness, i.e. contrast index, and record the best individual; Update individual positions through either a prey encirclement mechanism or a spiral approach mechanism; Iterate until the convergence condition is met and output the optimal parameters, that is, the optimal adaptive coefficient ; Among them, the convergence factor b of the convergence condition is dynamically adjusted according to the cosine function with the number of iterations.

5. The method for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet according to claim 1, wherein: The scale set of the wavelet transform grows exponentially, and the calculation formula is: in, is the minimum scale, is the logarithmic step size, j is the scale series, and J is the upper limit of the scale series.

6. A method for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet according to any one of claims 1 to 5, characterized in that: The cycle peak identification step includes: calculating the energy spectrum density of each scale, screening the local maximum value and combining the threshold to eliminate noise, and extracting the key cycle.

7. A device for extracting multi-scale periodic patterns from velocity data based on improved Morlet wavelet, characterized in that: include: A preprocessing module is used to preprocess the original flow velocity data to generate a continuous time series signal; The adaptive Morlet wavelet construction module is used to construct a Morlet wavelet function whose center frequency changes adaptively with the scale; wherein the expression of the Morlet wavelet function whose center frequency changes adaptively with the scale is: Among them, the center frequency It satisfies the linear relationship with the scale factor a: ; is the fundamental frequency, is the adaptive coefficient; Parameter optimization module, which is used to identify contrast indicators based on periodicity and dynamically determine wavelet function parameters through optimization algorithms; Improved wavelet transform calculation module, used to perform wavelet transform on pre-processed velocity data and calculate wavelet power spectrum; The periodic component extraction and feature analysis module is used to identify periodic peaks based on the wavelet power spectrum and extract key periodic factors.

8. An electronic device comprising a memory and a processor, characterized in that: The memory stores a computer program, and the processor is configured to run the computer program to execute the method for extracting multi-scale periodicity from velocity data based on improved Morlet wavelet according to any one of claims 1 to 6.

9. A readable storage medium, characterized in that: The readable storage medium stores a computer program, which includes a program code for controlling a process to execute a process, wherein the process includes the multi-scale periodic law extraction method for flow velocity data based on the improved Morlet wavelet according to any one of claims 1 to 6.

Images