CN114079503B - CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer - Google Patents
CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer Download PDFInfo
- Publication number
- CN114079503B CN114079503B CN202111252061.3A CN202111252061A CN114079503B CN 114079503 B CN114079503 B CN 114079503B CN 202111252061 A CN202111252061 A CN 202111252061A CN 114079503 B CN114079503 B CN 114079503B
- Authority
- CN
- China
- Prior art keywords
- wavelet
- signal
- imf component
- noise
- imf
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/07—Arrangements for monitoring or testing transmission systems; Arrangements for fault measurement of transmission systems
- H04B10/071—Arrangements for monitoring or testing transmission systems; Arrangements for fault measurement of transmission systems using a reflected signal, e.g. using optical time domain reflectometers [OTDR]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
- G06F17/148—Wavelet transforms
Abstract
The invention relates to the technical field of signal denoising, in particular to a CEEMDAN-wavelet threshold denoising method, a device, equipment and an optical time domain reflectometer, wherein the method comprises the steps of decomposing an original signal added with self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors of layers 1 to n, taking the IMF component with small order as the IMF component with dominant signal, and taking the IMF component with large order as the IMF component with dominant noise; determining a critical point of the IMF component dominated by the noise and the IMF component dominated by the signal, and removing the IMF component dominated by the noise; the method comprises the steps of presetting a wavelet base, a decomposition layer number, a threshold and a threshold function of a wavelet threshold algorithm, and denoising IMF components dominated by signals based on the preset wavelet threshold algorithm. The problem that the existing denoising method is poor in denoising effect is solved. The method has the effects of improving the denoising effect and facilitating detection of smaller event points such as non-reflection events and the like.
Description
Technical Field
The invention relates to the technical field of signal denoising, in particular to a CEEMDAN-wavelet threshold denoising method, a device and equipment and an optical time domain reflectometer.
Background
OTDR (Optical Time-Domain Reflectometer) is an instrument that uses the backscattered light signals generated when pulsed light is transmitted through an Optical fiber to characterize the transmission characteristics of the fiber. The OTDR, as a non-destructive optical fiber measuring device, can measure the length of an optical fiber, the transmission attenuation and fault location of the optical fiber, and is widely applied in the production, construction, maintenance, and other aspects of optical fiber cables, and is an indispensable test instrument in the field of optical fiber communication.
The OTDR curve may reflect the loss profile of the back-scattered light along the fiber. In the OTDR test curve, the types of events that are involved are mainly non-reflection events, and fiber ends, such as fiber splices, fusion splices, bends, breaks, etc., which correspond to losses at various locations in the fiber. For reflection events, the OTDR curve has higher signal-to-noise ratio under normal conditions, and is easy to detect and identify; however, for smaller event points such as non-reflection events, the noise is easily buried, and when the curve is seriously polluted by the noise, the events in the curve are difficult to identify, so that denoising processing of the signal is essential in the signal analysis process.
The EMD (Empirical Mode Decomposition) is a commonly used signal denoising method at present, and is widely applied to nonlinear and non-stationary signal denoising, but the method itself has problems of modal aliasing, end effect and the like, and affects the denoising effect. Other traditional denoising methods also cannot meet the requirements of detecting smaller event points such as non-reflection events.
For the related technologies, the inventor thinks that the existing denoising method has a defect of poor denoising effect.
Disclosure of Invention
In order to improve the denoising effect, the invention provides a CEEMDAN-wavelet threshold denoising method, a device and equipment and an optical time domain reflectometer.
In a first aspect, the invention provides a CEEMDAN-wavelet threshold denoising method, which has the characteristic of improving the denoising effect.
The invention is realized by the following technical scheme:
a CEEMDAN-wavelet threshold denoising method comprises the following steps:
based on CEEMDAN decomposition algorithm, the method adopts the original signal added with self-adaptive white noiseDecomposing by EMD decomposition algorithm to obtain IMF components and residual r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
determining a critical point of a noise dominant IMF component and a signal dominant IMF component, and removing the noise dominant IMF component;
the method comprises the steps of presetting a wavelet base, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and denoising IMF components dominated by signals based on the preset wavelet threshold value algorithm.
The present invention in a preferred example may be further configured to: the step of determining the critical points of the noise dominated IMF component and the signal dominated IMF component comprises:
calculating correlation coefficient R (x, I) between each layer IMF component and original signal h ) And obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) A difference of (d);
according to the correlation coefficient R (x, I) of each layer IMF component h ) Obtaining a difference curve;
and selecting a first peak point larger than zero as a critical point of the IMF component dominated by noise and the IMF component dominated by signals based on the difference curve.
The present invention in a preferred example may be further configured to: correlation coefficient R (x, I) between each layer of IMF component and original signal h ) The calculation formula (2) includes:
wherein x (n) is the original signal,
is the average of the original signal>
Is the average of the IMF components of the h-th layer.
The present invention in a preferred example may be further configured to: the preset wavelet threshold algorithm comprises the following steps:
adopting sym4 wavelet as wavelet base;
presetting 5 decomposition layers;
set the threshold value to
Wherein j is the number of decomposition layers, λ j Is the wavelet coefficient threshold of the j-th layer, N is the length of the signal, sigma j The standard deviation of noise contained in the j-th layer wavelet coefficient;
setting a threshold function to
Wherein j is the number of decomposition layers, w j,k For the kth wavelet coefficient at the jth scale of the decomposition, W j,k Is w j,k Estimated wavelet coefficients; lambda [ alpha ] j A threshold value for layer j; a. b is a positive integer and is such that a =2, b =20.
The invention in a preferred example may be further configured to: further comprising the steps of:
and adding and summing the IMF components which are dominated by the denoised signal, and reconstructing the signal.
By adopting the technical scheme, based on the CEEMDAN decomposition algorithm, the self-adaptive white noise is introduced into the original signal, and then EMD decomposition is carried out to obtain the IMF component dominated by the signal and the IMF component dominated by the noise, so that improvement is carried out on the basis of the EMD algorithm, the white noise is introduced to improve the problems of modal aliasing and end point effect, and the self-adaptability of the denoising method is stronger; the correlation degree between the component and the original signal is measured by using the correlation coefficient, a difference curve is obtained according to the difference value of the correlation coefficient of each adjacent IMF component, and a first peak point which is larger than zero is selected as a critical point of the IMF component which is dominant by noise and the IMF component which is dominant by the signal, so that the IMF component which is dominant by the noise is removed according to the critical point, and the signal is purer; adopting sym4 wavelet as wavelet base, presetting decomposition layer number as 5 layers, presetting threshold and threshold function, and denoising IMF component dominated by signals based on improved wavelet threshold algorithm, so as to reduce residual white noise of reconstructed signals; adding and summing IMF components dominated by the denoised signal, and reconstructing the signal to obtain a reconstructed signal so as to verify the condition of residual white noise; furthermore, the CEEMDAN-wavelet threshold denoising method can effectively improve the denoising effect of the signal, the denoised signal has higher signal-to-noise ratio and smaller mean square error, the high-quality denoised signal can be obtained, the event OTDR event characteristics are highlighted, and the method is easier to detect smaller event points such as non-reflection events and the like.
In a second aspect, the invention provides a CEEMDAN-wavelet threshold denoising device, which has the characteristic of improving the denoising effect.
The invention is realized by the following technical scheme:
a CEEMDAN-wavelet threshold denoising device comprises:
a decomposition module for decomposing the original signal added with the self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
the noise component removing module is used for determining the critical points of the IMF component dominated by noise and the IMF component dominated by signals and removing the IMF component dominated by noise;
and the denoising module is used for presetting a wavelet basis, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and denoising the IMF component dominated by the signal based on the preset wavelet threshold value algorithm.
The invention in a preferred example may be further configured to: the noise component removing module includes:
a correlation analysis submodule for calculating a correlation coefficient R (x, I) h ) Measuring the degree of correlation between each layer of IMF component and the original signal, and obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) Difference and difference curve.
The present invention in a preferred example may be further configured to: further comprising:
and the signal reconstruction module is used for adding and summing the IMF components dominated by the denoised signal to reconstruct the signal.
By adopting the technical scheme, the decomposition module introduces self-adaptive white noise into the original signal based on a CEEMDAN decomposition algorithm and then carries out EMD decomposition to obtain an IMF component dominated by the signal and an IMF component dominated by the noise, so that improvement is carried out on the basis of the EMD algorithm, the white noise is introduced to improve the problems of modal aliasing and end point effect, and the self-adaptability of the denoising method is stronger; the correlation degree analysis sub-module measures the correlation degree between the component and the original signal by using the correlation coefficient, obtains a difference curve according to the difference value of the correlation coefficients of all adjacent IMF components, and then the noise component removing module selects a first peak point larger than zero as a critical point of the IMF component dominated by noise and the IMF component dominated by signal so as to remove the IMF component dominated by noise according to the critical point, so that the signal is purer; the denoising module adopts sym4 wavelet as wavelet basis, the number of preset decomposition layers is 5, a preset threshold and a threshold function, and denoises IMF components dominated by signals based on an improved wavelet threshold algorithm, so that residual white noise of reconstructed signals is reduced; the signal reconstruction module adds and sums IMF components dominated by the denoised signal to reconstruct the signal so as to verify the residual white noise condition of the reconstructed signal; furthermore, the CEEMDAN-wavelet threshold denoising device can effectively improve the denoising effect of the signal, so that the denoised signal has higher signal-to-noise ratio and smaller mean square error, the high-quality denoised signal can be obtained, the event OTDR event characteristics are highlighted, and the device is easier to detect smaller event points such as non-reflection events and the like.
In a third aspect, the present invention provides a computer device, which has a feature of improving a denoising effect.
The invention is realized by the following technical scheme:
a computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the CEEMDAN-wavelet threshold denoising method described above when executing the computer program.
In a fourth aspect, the present invention provides a computer-readable storage medium having a feature of improving a denoising effect.
The invention is realized by the following technical scheme:
a computer-readable storage medium, having stored thereon a computer program which, when being executed by a processor, carries out the steps of a CEEMDAN-wavelet threshold denoising method as described above.
In a fifth aspect, the present invention provides an optical time domain reflectometer, which has a characteristic of improving a denoising effect.
The invention is realized by the following technical scheme:
an optical time domain reflectometer, comprising:
the laser emission module is used for generating electric pulses corresponding to the optical pulses required by OTDR measurement, converting the generated electric pulses into pulse light sources and outputting the pulse light sources through optical fibers;
the optical fiber coupler module is used for receiving the pulse light source output by the laser emission module, outputting the pulse light source to an optical fiber to be tested through the optical fiber, and receiving and outputting light reflected by the optical fiber to be tested;
the optical detector module is used for receiving the light reflected by the optical fiber to be detected and output by the optical fiber coupler module, and performing analog-to-digital conversion on the light and outputting the light;
the controller module is used for storing and reading the data of the optical signal reflected by the optical fiber to be detected and output by the optical detector module;
and the denoising module is used for receiving the output signal of the controller module and denoising the signal based on the steps of the CEEMDAN-wavelet threshold denoising method.
In summary, the invention includes at least one of the following beneficial technical effects:
1. based on the CEEMDAN decomposition algorithm, the method is improved on the basis of the EMD algorithm, and the problems of modal aliasing and end point effect are improved by introducing white noise, so that the self-adaptability of the denoising method is stronger; denoising IMF components dominated by signals based on an improved wavelet threshold algorithm to reduce residual white noise of reconstructed signals; the denoising effect of the signal is effectively improved, so that the denoised signal has higher signal-to-noise ratio and smaller mean square error, the high-quality denoised signal can be obtained, the event OTDR event characteristics are highlighted, and the detection of smaller event points such as non-reflection events and the like is easier;
2. determining a critical point of the IMF component dominated by noise and the IMF component dominated by the signal by using the correlation coefficient to measure the degree of correlation between the components and the original signal, so as to remove the IMF component dominated by the noise according to the critical point, thereby enabling the signal to be purer;
3. and adding and summing IMF components dominated by the denoised signal, and reconstructing the signal to verify the residual white noise condition of the reconstructed signal.
Drawings
Fig. 1 is a schematic overall flow chart of a CEEMDAN-wavelet threshold denoising method according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of a CEEMDAN-wavelet threshold denoising apparatus according to an embodiment of the present invention.
Fig. 3 is a block diagram of an optical time domain reflectometer according to an embodiment of the present invention.
Detailed Description
The present embodiment is only for explaining the present invention, and it is not limited to the present invention, and those skilled in the art can make modifications without inventive contribution to the present embodiment as required after reading the present specification, but all of them are protected by patent law within the scope of the present invention.
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The OTDR can reflect the state change of the optical fiber to be detected aiming at various mutation forms during signal detection so as to accurately detect the optical fiber state information. When a sudden change occurs, strong noise of partial aliasing is usually mixed in a high-frequency part of a detection signal, an EMD algorithm can decompose an inherent modal component according to characteristic information of the signal to realize multi-resolution analysis, but the problems of modal aliasing, end point effect and the like exist.
The traditional wavelet threshold denoising function has a hard threshold function and a soft threshold function, but the two methods have respective defects, and the original OTDR signal is often lost while the strong noise signal is removed, so that the detection fails.
Specifically, the hard threshold function:
soft threshold function:
wherein, w j,k For the kth wavelet coefficient at the jth scale of the decomposition, W j,k Is w j,k Estimated wavelet coefficients; the lambda is a threshold value and is used as the threshold value,
n is the length of the signal, σ n Is the standard deviation, σ, of the noise contained in layer 1 n ＝median(|w 1,j |)/0.6745。
The hard threshold function is discontinuous at the threshold, so that the reconstructed signal has oscillation and pseudo Gibbs effect; although the soft threshold function is continuous at the threshold, a constant deviation exists between the processed wavelet coefficient and the real wavelet coefficient, so that the reconstruction precision of the wavelet coefficient is reduced, the distortion condition exists, and the denoising effect is poor.
The method is based on the CEEMDAN decomposition algorithm, and introduces white noise to improve the problems of modal aliasing and end effect, so that the self-adaptability of the denoising method is stronger; denoising based on an improved wavelet threshold algorithm to reduce residual white noise of a reconstructed signal; the method effectively improves the denoising effect of the signal, enables the denoised signal to have higher signal-to-noise ratio and smaller mean square error, can obtain a high-quality denoised signal, highlights the event OTDR event characteristics, and is easier to detect smaller event points such as non-reflection events and the like.
The embodiments of the present invention will be described in further detail with reference to the drawings attached hereto.
Referring to fig. 1, an embodiment of the present invention provides a CEEMDAN-wavelet threshold denoising method, and main steps of the method are described as follows.
S1, decomposing the original signal added with the self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
s2, determining a critical point of the IMF component dominated by the noise and the IMF component dominated by the signal, and removing the IMF component dominated by the noise;
and S3, presetting a wavelet base, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and denoising the IMF component dominated by the signal based on the preset wavelet threshold value algorithm.
Specifically, S1, EMD decomposition is carried out on an original signal after Adaptive Noise is added based on a CEEMDAN (Complete EEMD with Adaptive Noise Complete set empirical mode) decomposition algorithm.
Wherein the original signal can be represented as
Adding i groups of adaptive white noise to the acquired original signal, wherein the ith group of signals can be expressed as X i (n)＝x(n)+e i (n), where x (n) is the original signal collected and i is the number of added groups (i =1, 2.. Eta., T), e) i (n) is the added noise signal.
Signal X with noise added to each group by EMD decomposition algorithm i (n) decomposing to obtain 1 st order modal component I of each group 1,i I.e. the IMF component of the first layer, and then immediately carrying out summation average calculation on the IMF component of the first layer to obtain the 1 st modal component I 1 ：
Subtracting the 1 st modal component from the signal x (n) to obtain a residual error r 1 (n)＝x(n)-I 1 Adding i groups of white noise e into the residual error i (n) forming a new signal to be decomposed R 1,i (n)＝r 1 (n)+e i (n) then to R 1,i (n) EMD decomposition to obtain 2 nd order modal component I 2,i I.e. second layer IMF components, and summing to take the average value I 2 (ii) a And analogizing in sequence to obtain IMF components of the 1 st to the n th layers and a residual error r h And (n), wherein the IMF components with small orders are low-frequency components, and the IMF components with large orders are high-frequency components.
In this embodiment, the standard deviation of the added adaptive white noise is 0.2, and the number of groups is 500, so that the original signal is added with the adaptive white noise, such as w i (n) Gaussian white noise which obeys normal distribution, and based on an EMD decomposition algorithm, the maximum screening iteration number of the decomposed IMF components is allowed to be 5000 times, so that the noise-containing original signals are decomposed by CEEMDAN to obtain 14 IMF components.
And taking the IMF component with a small order as the IMF component with the dominant signal, and taking the IMF component with a large order as the IMF component with the dominant noise. Aiming at the EEMD residual noise problem, the CEEMDAN decomposition algorithm is improved on the basis of the EMD algorithm, and modal aliasing and end point effects are improved due to the addition of adaptive noise.
Further, the step of determining a critical point of the noise dominated IMF component and the signal dominated IMF component S2 comprises:
s21, calculating correlation coefficients R (x, I) between each layer of IMF component and the original signal h ) And obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) The difference of (c). In particular, the correlation coefficient R (x, I) between each layer IMF component and the original signal h ) The calculation formula of (a) is as follows:
wherein x (n) is the original signal,
is the average of the original signal>
Is the average of the IMF components of the h-th layer. And calculating the correlation coefficient of each layer of IMF component to measure the correlation degree between each layer of IMF component and the original signal.
By passing
Calculating the difference value of the adjacent IMF components to obtain the correlation coefficient R (x, I) of each layer of IMF component h ) The difference of (c).
S22, according to the correlation coefficient R (x, I) of each layer IMF component h ) To obtain a difference curve.
And S23, selecting a first peak point larger than zero as a critical point of the IMF component dominated by noise and the IMF component dominated by signals based on the difference curve. And then based on the critical point, removing the IMF component before the point, namely removing the IMF component with dominant noise.
In this embodiment, the preset wavelet threshold algorithm includes:
s31, adopting sym4 wavelet as wavelet base;
s32, presetting the number of decomposition layers to be 5;
s33 setting the threshold value to
Wherein j is the number of decomposition layers, λ j Is the wavelet coefficient threshold of the j-th layer, N is the length of the signal, sigma j The standard deviation of noise contained in the j-th layer wavelet coefficient;
s34, setting a threshold function as
Wherein j is the number of decomposition layers, w j,k For the kth wavelet coefficient at the jth scale of the decomposition, W j,k Is w j,k Estimated wavelet coefficients; lambda [ alpha ] j A threshold value for layer j; a. b is an adjustment factor of a positive integer, and adjustment factors of the threshold function are set to 2 and 20, respectively.
The preset wavelet threshold denoising method selects a wavelet basis and a decomposition layer number according to the characteristics of signals, and performs wavelet decomposition on the noise-containing signals; with the increase of the scale, the amplitude of the wavelet coefficient of the noise is smaller and smaller, and the amplitude of the wavelet coefficient of the signal is larger and larger, so that the threshold is selected for the noise distribution condition of different decomposition layer numbers, namely the threshold is selected according to the wavelet coefficient, so that the threshold can adapt to the noise distribution condition of each layer; the wavelet coefficient is subjected to preset threshold function processing, so that the threshold function is continuous at a threshold, and the processed wavelet coefficient tends to a true value along with the increase of the wavelet coefficient, so that constant deviation is reduced, the precision of processing a noise signal is improved, and the quality of a denoised signal is improved; the improved wavelet threshold algorithm enables the threshold to be adaptive to the noise distribution condition of each IMF component by selecting a proper wavelet basis, decomposition layer number, threshold and threshold function, has strong adaptability, and overcomes the defects of the traditional soft threshold denoising method and the traditional hard threshold denoising method.
And further S3, based on a preset wavelet threshold algorithm, after the IMF component which is dominant by the signal is subjected to denoising processing, the method also comprises the following steps:
s4, signal reconstruction is carried out on IMF components dominated by the denoised signals, and processed IMF components and residual errors r of each layer are utilized h And (n) reconstructing signals, adding and summing the IMF components which are dominated by the denoised signals so as to reconstruct the wavelet coefficients and obtain the reconstructed signals.
Denoising and comparing the reconstructed signal based on the CEEMDAN-improved wavelet threshold denoising method with the reconstructed signal of the hard threshold denoising method, the reconstructed signal of the CEEMDAN-hard threshold denoising method and the reconstructed signal of the improved wavelet threshold denoising method respectively. Preferably, the invention adopts the signal-to-noise ratio and the root mean square error after removing the useless signals at the tail end as the evaluation index of the denoising performance. The calculation formulas of the signal-to-noise ratio and the root mean square error are respectively as follows:
The results of the denoising comparison are shown in the following table.
TABLE 1 De-noising Performance index comparison
| Denoising method | SNR/db | RMSE |
| Hard threshold denoising | 42.189 | 0.0052 |
| CEEMDAN-HARD-THRESHOLD DE-NOISE | 46.746 | 0.0031 |
| Improved wavelet threshold denoising | 48.241 | 0.0026 |
| CEENDAN-improved wavelet threshold denoising | 50.068 | 0.0021 |
The noise signal is received after the tail end event in the original signal based on the OTDR, and the data comparison in the table 1 can find that the signal-to-noise ratio is highest and the root mean square error is minimum based on the CEEMDAN-improved wavelet threshold denoising method, which shows that the method can better keep the original signal while effectively removing the noise; and the signal after denoising reconstruction becomes smoother and more real, the high-frequency noise is obviously reduced, and the analysis and the detection of subsequent events are easier.
Therefore, the CEEMDAN-based improved wavelet threshold denoising method has better performance for denoising the OTDR original signal, and compared with the traditional denoising method, the method can obtain higher signal-to-noise ratio and smaller mean square error, and improves the quality of the denoised signal; meanwhile, noise can be better suppressed, a better denoising effect is achieved, details of OTDR signals can be highlighted, detection of smaller event points such as non-reflection events is easier, and the method has important significance for practical application of OTDR.
It should be understood that, the sequence numbers of the steps in the foregoing embodiments do not imply an execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present invention.
Referring to fig. 2, an embodiment of the present invention further provides a CEEMDAN-wavelet threshold denoising device, where the CEEMDAN-wavelet threshold denoising device corresponds to the CEEMDAN-wavelet threshold denoising method in the foregoing embodiment one to one. The CEEMDAN-wavelet threshold denoising device comprises,
a decomposition module for decomposing the original signal added with the self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
the noise component removing module is used for determining the critical points of the IMF component dominated by the noise and the IMF component dominated by the signal and removing the IMF component dominated by the noise;
the noise component removing module includes a correlation analyzing sub-module for calculating a correlation coefficient R (x, I) h ) Measuring the degree of correlation between each layer of IMF component and the original signal, and obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) Difference and difference curve;
the de-noising module is used for presetting a wavelet basis, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and de-noising the IMF component which is dominant by the signal based on the preset wavelet threshold value algorithm;
and the signal reconstruction module is used for adding and summing IMF components dominated by the denoised signal to reconstruct the signal.
The specific definition of the CEEMDAN-wavelet threshold denoising device can be referred to the above definition of the CEEMDAN-wavelet threshold denoising method, and is not described herein again. The modules in the CEEMDAN-wavelet threshold denoising apparatus can be wholly or partially implemented by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
Referring to fig. 3, in one embodiment, there is provided an optical time domain reflectometer, comprising:
the laser emission module is used for generating electric pulses corresponding to the optical pulses required by OTDR measurement, converting the generated electric pulses into pulse light sources and outputting the pulse light sources through optical fibers;
the optical fiber coupler module is used for receiving the pulse light source output by the laser emission module, outputting the pulse light source to the optical fiber to be tested through the optical fiber, and receiving and outputting light reflected by the optical fiber to be tested;
the optical detector module is used for receiving the light reflected by the optical fiber to be detected and output by the optical fiber coupler module, and performing analog-to-digital conversion on the light and outputting the light;
the controller module is used for storing and reading the data of the optical signal reflected by the optical fiber to be detected and output by the optical detector module;
and the denoising module is used for receiving the output signal of the controller module and denoising the signal based on the steps of the CEEMDAN-wavelet threshold denoising method.
In one embodiment, a computer device is provided, which may be a server. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operating system and the computer program to run on the non-volatile storage medium. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to implement the steps of the CEEMDAN-wavelet threshold denoising method described above.
In one embodiment, a computer-readable storage medium is provided, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program:
s1, decomposing the original signal added with the self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
s2, determining a critical point of the IMF component dominated by the noise and the IMF component dominated by the signal, and removing the IMF component dominated by the noise;
and S3, presetting a wavelet base, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and denoising the IMF component dominated by the signal based on the preset wavelet threshold value algorithm.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, databases, or other media used in embodiments provided herein may include non-volatile and/or volatile memory.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-mentioned division of the functional units and modules is illustrated, and in practical applications, the above-mentioned function distribution may be performed by different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to perform all or part of the above-mentioned functions.
Claims (10)
1. A CEEMDAN-wavelet threshold denoising method is characterized in that the method is applied to OTDR detection signals of small event points of non-reflection events and comprises the following steps:
based on CEEMDAN decomposition algorithm, the original signal added with the self-adaptive white noise is decomposed by EMD decomposition algorithm to obtain IMF components and residual r of the 1 st to the n th layers h (n) and the order is reducedThe IMF component is used as the IMF component dominated by the signal, and the IMF component with a large order is used as the IMF component dominated by the noise;
determining a critical point of a noise dominant IMF component and a signal dominant IMF component, and removing the noise dominant IMF component;
presetting a wavelet basis, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and denoising an IMF component which is dominant by the signal based on the preset wavelet threshold value algorithm;
the preset wavelet threshold algorithm comprises the following steps:
adopting sym4 wavelet as wavelet base;
presetting 5 decomposition layers;
set the threshold value to
Wherein j is the number of decomposition layers, λ j Is the wavelet coefficient threshold of the j-th layer, N is the length of the signal, sigma j The standard deviation of noise contained in the j-th layer wavelet coefficient;
setting a threshold function to
Wherein j is the number of decomposition layers, w j,k For the kth wavelet coefficient at the jth scale of the decomposition, W j,k Is w j,k Estimated wavelet coefficients; lambda [ alpha ] j A threshold value for layer j; a. b is a positive integer and is such that a =2, b =20.
2. The CEEMDAN-wavelet threshold denoising method of claim 1, wherein the step of determining the critical points of the noise-dominated IMF component and the signal-dominated IMF component comprises:
calculating correlation coefficient R (x, I) between each layer IMF component and original signal h ) And obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) A difference of (d);
according to the correlation coefficient R (x, I) of each layer IMF component h ) Obtaining a difference curve;
and selecting a first peak point larger than zero as a critical point of the IMF component dominated by noise and the IMF component dominated by signals based on the difference curve.
3. The CEEMDAN-wavelet threshold denoising method of claim 2, wherein the correlation coefficient R (x, I) between each layer of IMF components and the original signal h ) The calculation formula (2) includes:
wherein x (n) is the original signal,
is the average of the original signal>
Is the average of the IMF components of the h-th layer.
4. A CEEMDAN-wavelet threshold denoising method according to any one of claims 1-3, further comprising the steps of:
and adding and summing the IMF components which are dominated by the denoised signal, and reconstructing the signal.
5. A CEEMDAN-wavelet threshold denoising device is characterized in that an OTDR detection signal applied to a small event point of a non-reflection event comprises:
a decomposition module for decomposing the original signal added with the self-adaptive white noise by adopting an EMD decomposition algorithm based on a CEEMDAN decomposition algorithm to obtain IMF components and residual errors r of the 1 st to the n th layers h (n), the IMF component with small order is taken as the IMF component with signal dominance, and the IMF component with large order is taken as the IMF component with noise dominance;
the noise component removing module is used for determining the critical points of the IMF component dominated by noise and the IMF component dominated by signals and removing the IMF component dominated by noise;
the de-noising module is used for presetting a wavelet basis, a decomposition layer number, a threshold value and a threshold value function of a wavelet threshold value algorithm, and de-noising the IMF component dominated by the signal based on the preset wavelet threshold value algorithm; the preset wavelet threshold algorithm comprises the following steps: adopting sym4 wavelet as wavelet base; presetting 5 decomposition layers; set the threshold value to
Wherein j is the number of decomposition layers, λ j Is the wavelet coefficient threshold of the j-th layer, N is the length of the signal, sigma j The standard deviation of noise contained in the j-th layer wavelet coefficient;
setting a threshold function to
Wherein j is the number of decomposition layers, w j,k For the kth wavelet coefficient at the jth scale of the decomposition, W j,k Is w j,k Estimated wavelet coefficients; lambda j A threshold value for layer j; a. b is an adjustment of a positive integerFactor, and let a =2, b =20.
6. The CEEMDAN-wavelet threshold denoising apparatus of claim 5, wherein the noise component removal module comprises:
a correlation analysis submodule for calculating a correlation coefficient R (x, I) h ) Measuring the degree of correlation between each layer of IMF component and the original signal, and obtaining the correlation coefficient R (x, I) of each layer of IMF component h ) Difference and difference curve.
7. The CEEMDAN-wavelet threshold denoising apparatus of any one of claims 5-6, further comprising:
and the signal reconstruction module is used for adding and summing IMF components dominated by the denoised signal to reconstruct the signal.
8. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the CEEMDAN-wavelet threshold denoising method of any one of claims 1-4 when executing the computer program.
9. A computer-readable storage medium, characterized in that it stores a computer program which, when executed by a processor, implements the steps of the CEEMDAN-wavelet threshold denoising method of any one of claims 1-4.
10. An optical time domain reflectometer, comprising:
the laser emission module is used for generating electric pulses corresponding to the optical pulses required by OTDR measurement, converting the generated electric pulses into pulse light sources and outputting the pulse light sources through optical fibers;
the optical fiber coupler module is used for receiving the pulse light source output by the laser emission module, outputting the pulse light source to the optical fiber to be tested through the optical fiber, and receiving and outputting light reflected by the optical fiber to be tested;
the optical detector module is used for receiving the light reflected by the optical fiber to be detected and output by the optical fiber coupler module, and performing analog-to-digital conversion on the light and outputting the light;
the controller module is used for storing and reading the data of the optical signal reflected by the optical fiber to be detected and output by the optical detector module;
a de-noising module for receiving the output signal of the controller module and de-noising the signal based on the steps of the CEEMDAN-wavelet threshold de-noising method as claimed in any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111252061.3A CN114079503B (en) | 2021-10-26 | 2021-10-26 | CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111252061.3A CN114079503B (en) | 2021-10-26 | 2021-10-26 | CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114079503A CN114079503A (en) | 2022-02-22 |
| CN114079503B true CN114079503B (en) | 2023-03-28 |
Family
ID=80283874
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111252061.3A Active CN114079503B (en) | 2021-10-26 | 2021-10-26 | CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114079503B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114992033B (en) * | 2022-05-25 | 2023-04-07 | 武汉理工大学 | NLM-CEEMDAN-based hydroelectric generating set signal denoising method |
| CN115255405B (en) * | 2022-09-23 | 2022-12-23 | 相国新材料科技江苏有限公司 | Intelligent control method and system of additive manufacturing equipment |
| CN116299562B (en) * | 2023-05-26 | 2023-08-04 | 中国海洋大学 | Altimeter ranging ionosphere error correction filtering processing method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110702149A (en) * | 2019-11-05 | 2020-01-17 | 南京申威光电技术研究院有限公司 | Method and system for positioning disturbance signal of optical fiber distributed disturbance sensing system |
-
2021
- 2021-10-26 CN CN202111252061.3A patent/CN114079503B/en active Active
Non-Patent Citations (1)
| Title |
|---|
| 基于CEEMDAN与小波自适应阈值的去噪新方法;张建文等;《电测与仪表》(第10期);第20-24页，第39页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114079503A (en) | 2022-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114079503B (en) | CEEMDAN-wavelet threshold denoising method, device and equipment and optical time domain reflectometer | |
| JP4769411B2 (en) | Polarized lightwave reflectometry and system | |
| EP2998751A1 (en) | Detecting, localizing and ranking copper connectors | |
| CN112380934B (en) | Cable partial discharge signal self-adaptive wavelet denoising method based on wavelet entropy and sparsity | |
| CN108828015B (en) | Method and device for positioning local weak physical damage of cable | |
| CN110717472B (en) | Fault diagnosis method and system based on improved wavelet threshold denoising | |
| CN115169409B (en) | Sliding window-based bridge structure self-vibration frequency identification and early warning method and equipment | |
| US10359465B2 (en) | Method for characterizing a soft fault in a cable | |
| CN117332205B (en) | High-precision automatic optimization method and device for temperature compensation resistance of piezoresistor | |
| CN113624447A (en) | Composite denoising method for fiber bragg grating sensor network | |
| TWI428581B (en) | Method for identifying spectrum | |
| CN111079591B (en) | Bad data restoration method and system based on improved multi-scale principal component analysis | |
| CN113556169A (en) | Wavelet threshold denoising method and optical time domain reflectometer based on same | |
| CN115561310A (en) | Method for processing non-random coherent noise in grounding electrode defect echo signal | |
| CN113221692B (en) | Continuous variational modal decomposition DWT denoising method for optical fiber sensing | |
| CN117478221B (en) | Method for locating fault point of long-distance transmission optical cable with high spatial resolution | |
| CN106797231A (en) | A kind of method and apparatus for compensating optical time domain reflectometer originating terminal signal error | |
| CN115423221B (en) | Facility operation trend prediction method | |
| CN113031074B (en) | Logging curve denoising method, device, equipment and storage medium | |
| CN118013191A (en) | Denoising method based on LSM filtering and CEEMDAN combined wavelet detection insulator | |
| CN116996117A (en) | Line fault detection method and device suitable for communication optical cable | |
| CN117614525A (en) | ICEEMDAN and NLM-based OTDR signal denoising method | |
| CN117251676A (en) | Phase signal noise reduction method and system for optical cable vibration signal | |
| CN117478214A (en) | Event detection method and system for optical time domain transmitter | |
| CN109977350B (en) | Ultrasonic eddy current signal self-adaptive radial basis function neural network cancellation method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |