A method capable of processing non-random coherent noise in ground electrode defect echo signal

Abstract

The present invention discloses a method capable of processing non-random coherent noise in a grounding electrode defect echo signal, comprising the following steps: step 1, transmitting an ultrasonic guided wave through an ultrasonic guided wave transmitter, and receiving and detecting the ultrasonic guided wave echo through a sensor to obtain an ultrasonic guided wave echo signal set; step 2, performing ultrasonic guided wave echo defect detection signal preprocessing on the ultrasonic guided wave echo signal set to obtain a whitening matrix; step 3, using Newton's fast iteration algorithm to optimize and calculate each column of the whitening matrix to obtain optimized data. Through the overall structure of the device, the coherent noise contained in the echo can be processed, the sensitivity of defect detection can be improved, and the signal-to-noise ratio of the reflected echo can be improved from the characteristic, and the noise caused by the structure of the tested piece and the changes in the soil environment can be filtered out as much as possible, thereby improving the accuracy of grounding flat steel corrosion defect detection in terms of ultrasonic guided wave defect signals.

Description

Technical Field

[0001] The invention relates to the technical field of ultrasonic guided wave nondestructive testing, and in particular to a method capable of processing non-random coherent noise in a ground electrode defect echo signal. Background Art

[0002] Ultrasonic guided wave testing, as a defect detection method, has been successfully applied in the field of non-destructive testing. Due to its advantages of non-contact, wide measurement range, and high accuracy, it has been widely used in the detection of corrosion defects in grounding devices of power systems. However, the grounding devices of power systems have various configurations and geometric shapes, such as overlap, bends, and welding between multiple grounding electrodes. These characteristics and changes in the soil environment will cause the detected defect echo to attenuate. At the same time, the complex and asymmetric defect shape will cause the change of the guide mode after the ultrasonic guided wave action, thereby generating background noise, which greatly reduces the signal-to-noise ratio of the ultrasonic guided wave detection signal and reduces the sensitivity of defect detection. At the same time, the rough detection surface of the flat steel may have a great impact on the directivity of the sound field, and change the time and propagation direction of the guided wave propagation, making it difficult to locate the defect. All these make it difficult to detect corrosion defects in flat steel in power systems. Therefore, we propose a method that can process non-random coherent noise in the grounding electrode defect echo signal. Summary of the invention

[0003] The purpose of the present invention is to provide a method for processing non-random coherent noise in the echo signal of a grounding electrode defect. The method does not require a large number of observation samples, does not require making a harsh assumption that the real signal is a deterministic signal, and does not require pre-positioning of the characteristic frequency band of the processing signal. The method can process the coherent noise contained in the echo, improve the sensitivity of defect detection, and characteristically improve the signal-to-noise ratio of the reflected echo, thereby improving the accuracy of grounding flat steel corrosion defect detection in terms of ultrasonic guided wave defect signals.

[0004] The present invention discloses a method for processing non-random coherent noise in a ground electrode defect echo signal, comprising the following steps:

[0005] Step 1: transmitting ultrasonic guided waves through an ultrasonic guided wave transmitter, and receiving and detecting ultrasonic guided wave echoes through a sensor to obtain an ultrasonic guided wave echo signal set;

[0006] Step 2: Preprocess the ultrasonic guided wave echo defect detection signal set to obtain a whitening matrix;

[0007] Step 3: Use Newton's fast iteration algorithm to optimize each column of the whitening matrix to obtain optimized data;

[0008] Step 4: Use the negative entropy method to determine the optimized data to be analyzed and establish a criterion for judging whether the random vector y(i) in the optimized data is independent. Use the maximum entropy principle through an optimal linear function to maximize the negative entropy and obtain a linear data set.

[0009] Step 5: According to the linear data set, the maximum value of the non-Gaussianity between the components is obtained to remove the coherent noise in the source ultrasonic guided wave.

[0010] As a preferred solution, the sensors are provided in n numbers. Assuming that n sensors are placed, each sensor performs p measurements and each sensor has p guided wave signals, the matrix A of these ultrasonic guided wave echo signal sets can be expressed as:

[0011] A=B×C

[0012] Among them, A is the data matrix of n sensors, C is the original matrix composed of independent signals, and B is called the mixing matrix, which is composed of signal weight values.

[0013] As a preferred solution, the ultrasonic guided wave echo defect detection signal preprocessing of the ultrasonic guided wave echo signal set includes de-averaging and whitening processing.

[0014] As a preferred solution, the de-averaging is to subtract the mean from the matrix A of the received ultrasonic guided wave echo signal set to ensure that the mean after processing is 0, which can reduce the complexity of calculation, and can be specifically expressed as:

[0015] AE{A}

[0016] Where E{A} represents the expectation which is often replaced by the average value in practice.

[0017] As a preferred solution, the whitening process is to convert the covariance matrix of the ultrasonic guided wave echo signal set into a diagonal matrix, so that the variables under study are linearly independent, and the correlation between the ultrasonic guided wave echo signal sets is removed, thereby simplifying the subsequent extraction process and improving the convergence of the ICA algorithm. The whitening process is performed by using the eigenvalue decomposition of the diagonal matrix to obtain the whitening matrix of the matrix A of the ultrasonic guided wave echo signal set:

[0018] F=G -1 / 2 H i

[0019] Where G and H are the characteristic matrix and eigenvalue diagonal matrix of the covariance matrix E{X,Xi} respectively.

[0020] As a preferred solution, the iterative formula for optimizing the calculation of each column of the whitening matrix using the Newton fast iterative algorithm is as follows:

[0021]

[0022] in represents the element of the Pth column of the Kth iteration result, Z is an orthogonal matrix, which can be expressed as Z=AF, g(x) is a nonlinear function with x as the variable, and g(x)=tanh(x), g'(x) represents the derivative of g(x), <> represents the averaging process, and the superscript “—” represents the normalization process.

[0023] As a preferred solution, the method of maximizing the negative entropy is to use the negative entropy method to determine the optimized data to be analyzed and to establish a criterion for determining whether the random vector y(i) in the optimized data is independent, and to use the negative entropy J(y) as a measure of the non-Gaussianity of each component, that is, assuming that the probability density of the random vector y is P(y), then its entropy is:

[0024] H(y)=-∫P(y)lgP(y)dy

[0025] Its negative entropy is:

[0026] J(y)=H(y gauss )-H(y)

[0027] where y gauss is a Gaussian random component.

[0028] As a preferred solution, the maximum value of the non-Gaussianity between the components is obtained. According to the maximum entropy principle, the negative entropy J(y) can be estimated as:

[0029] J(y i )≈c[E{G(y i )}-E{G(u)}]

[0030] where G(x) is an arbitrary non-quadratic function, c is a positive constant, u is a Gaussian variable with zero mean and unit variance, and E{} is the mathematical expectation operator.

[0031] As a preferred solution, G(x)=x 4 / 4.

[0032] As a preferred embodiment, a method capable of processing non-random coherent noise in a ground electrode defect echo signal is stored in an application program of a computer architecture and driven by a burned program. The method also includes a bus architecture, a storage device and a bus interface. The bus architecture may include any number of interconnected buses and bridges. The bus architecture links together various circuits including one or more processors represented by a processor and a memory represented by a memory. The bus architecture may also connect together various other circuits such as peripheral devices, voltage regulators and power management circuits. The bus interface provides an interface between the bus architecture and a receiver and a transmitter. The receiver and the transmitter may be the same element, namely a transceiver, which provides a unit for communicating with various other systems on a transmission medium.

[0033] The beneficial effects of a method for processing non-random coherent noise in a ground electrode defect echo signal disclosed by the present invention are:

[0034] Through the overall structure of the equipment, the method is divided into data preprocessing, optimization calculation of the whitening matrix to establish component independence criteria and the maximum value of the non-Gaussianity between the components. This method does not require a large number of observation samples, does not require the harsh assumption that the real signal is a deterministic signal, and does not require pre-positioning of the characteristic frequency band of the processing signal. It can process the coherent noise contained in the echo, improve the sensitivity of defect detection and improve the signal-to-noise ratio of the reflected echo from the characteristics, and filter out the noise caused by the structure of the test piece and the changes in the soil environment as much as possible, thereby improving the accuracy of corrosion defect detection of ground flat steel in terms of ultrasonic guided wave defect signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 A principle block diagram of an embodiment of a method for processing non-random coherent noise in a ground electrode defect echo signal of the present invention;

[0036] Figure 2 The diagram is a principle step diagram of an embodiment of a method for processing non-random coherent noise in a ground electrode defect echo signal of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described and illustrated below in conjunction with specific embodiments and accompanying drawings:

[0038] See also Figure 1-2 The present invention provides a method for processing non-random coherent noise in a ground electrode defect echo signal, comprising the following steps:

[0039] Step 1: transmitting ultrasonic guided waves through an ultrasonic guided wave transmitter, and receiving and detecting ultrasonic guided wave echoes through a sensor to obtain an ultrasonic guided wave echo signal set;

[0040] Specifically: both the ultrasonic guided wave transmitter and the sensor use existing equipment on the market, and there are multiple sensors. The sensors are used to detect ultrasonic guided wave echoes, and as the number of sensors increases, the attenuation of the guided wave signal in the soil environment can be reduced to capture the maximum changes in the original data set.

[0041] Step 2: Preprocess the ultrasonic guided wave echo defect detection signal set to obtain a whitening matrix;

[0042] Specifically, the ultrasonic guided wave echo defect detection signal preprocessing is de-averaging and whitening processing.

[0043] Step 3: Use Newton's fast iteration algorithm to optimize each column of the whitening matrix to obtain optimized data;

[0044] Specifically: Optimizing the calculation of each column of the whitening matrix can make the signal detection more accurate.

[0045] Step 4: Use the negative entropy method to determine the optimized data to be analyzed and establish a criterion for judging whether the random vector y(i) in the optimized data is independent. Use the maximum entropy principle through an optimal linear function to maximize the negative entropy and obtain a linear data set.

[0046] Specifically: find the best linear function so that the development pattern of data can be easily predicted.

[0047] Step 5: According to the linear data set, the maximum value of the non-Gaussianity between the components is obtained to remove the coherent noise in the source ultrasonic guided wave.

[0048] The sensors are provided with n numbers. Assuming that n sensors are placed, each sensor performs p measurements and each sensor has p guided wave signals, the matrix A of these ultrasonic guided wave echo signal sets can be expressed as:

[0049] A=B×C

[0050] Among them, A is the data matrix of n sensors, C is the original matrix composed of independent signals, and B is called the mixing matrix, which is composed of signal weight values.

[0051] Specifically: Each sensor generates a signal every time it measures something, and the signal for each measurement is relative to the number of measurements.

[0052] The ultrasonic guided wave echo defect detection signal preprocessing performed on the ultrasonic guided wave echo signal set includes de-averaging and whitening processing.

[0053] The de-averaging is to subtract the mean from the matrix A of the received ultrasonic guided wave echo signal set to ensure that the mean after processing is 0, which can reduce the complexity of calculation, and can be specifically expressed as:

[0054] AE{A} where E{A} represents the expectation instead of the average value which is often used in practice.

[0055] The whitening process is to transform the covariance matrix of the ultrasonic guided wave echo signal set into a diagonal matrix, so that the variables under study are linearly independent, and the correlation between the ultrasonic guided wave echo signal sets is removed, thereby simplifying the subsequent extraction process and improving the convergence of the ICA algorithm. The whitening process is performed by using the eigenvalue decomposition of the diagonal matrix to obtain the whitening matrix of the matrix A of the ultrasonic guided wave echo signal set:

[0056] F=G -1 / 2 Hi

[0057] Where G and H are the characteristic matrix and eigenvalue diagonal matrix of the covariance matrix E{X,Xi} respectively.

[0058] The iterative formula for optimizing the calculation of each column of the whitening matrix using the Newton fast iterative algorithm is as follows:

[0059]

[0060] in represents the element of the Pth column of the Kth iteration result, Z is an orthogonal matrix, which can be expressed as Z=AF, g(x) is a nonlinear function with x as the variable, and g(x)=tanh(x), g'(x) represents the derivative of g(x), <> represents the averaging process, and the superscript “—” represents the normalization process.

[0061] The method of maximizing the negative entropy is to use the negative entropy method to determine the optimized data to be analyzed and to establish a criterion for determining whether the random vector y(i) in the optimized data is independent, and to use the negative entropy J(y) as a measure of the non-Gaussianity of each component, that is, assuming that the probability density of the random vector y is P(y), then its entropy is:

[0062] H(y)=-∫P(y)lgP(y)dy

[0063] Its negative entropy is:

[0064] J(y)=H(y gauss )-H(y)

[0065] where y gauss is a Gaussian random component.

[0066] The maximum value of the non-Gaussianity between the components is obtained. According to the maximum entropy principle, the negative entropy J(y) can be estimated as:

[0067] J(y i )≈c[E{G(y i )}-E{G(u)}]

[0068] where G(x) is an arbitrary non-quadratic function, c is a positive constant, u is a Gaussian variable with zero mean and unit variance, and E{} is the mathematical expectation operator.

[0069] G(x)=x 4 / 4.

[0070] It should be noted that for non-Gaussian, higher-order statistics are needed to obtain meaningful representations. Projection pursuit is a technique that uses higher-order statistics to find interesting projections of data. Projection pursuit uses a cost function, such as differential entropy, instead of the mean squared error used in the PCA transform.

[0071] For non-Gaussian numbers, projection pursuit is an effective method. There are similarities and connections between ICA and these techniques. ICA is a special case of projection pursuit in the absence of noise. ICA can also be viewed as a non-Gaussian factor analysis. ICA must use higher-order statistics, while PCA only uses second-order statistics.

[0072] A method capable of processing non-random coherent noise in a ground electrode defect echo signal is stored in an application program of a computer architecture and driven by a burned program. The computer architecture also includes a bus architecture, a memory and a bus interface. The bus architecture may include any number of interconnected buses and bridges. The bus architecture links together various circuits including one or more processors represented by a processor and a memory represented by a memory. The bus architecture may also connect together various other circuits such as peripheral devices, voltage regulators and power management circuits. The bus interface provides an interface between the bus architecture and a receiver and a transmitter. The receiver and the transmitter may be the same element, namely a transceiver, which provides a unit for communicating with various other systems on a transmission medium.

[0073] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A system that specifies the functions of a box or multiple boxes.

[0074] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction system, which is implemented in the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0075] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps of the functions specified in a box or multiple boxes. Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they know the basic creative concept. Therefore, the attached claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0076] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, rather than to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solution of the present invention can be modified or replaced by equivalents without departing from the essence and scope of the technical solution of the present invention.

Claims

1. A method for processing non-random coherent noise in a ground electrode defect echo signal, characterized in that: The steps include: Step 1: transmitting ultrasonic guided waves through an ultrasonic guided wave transmitter, and receiving and detecting ultrasonic guided wave echoes through a sensor to obtain an ultrasonic guided wave echo signal set; Step 2: Preprocess the ultrasonic guided wave echo defect detection signal set to obtain a whitening matrix; Step 3: Use Newton's fast iteration algorithm to optimize each column of the whitening matrix to obtain optimized data; Step 4: Use the negative entropy method to determine the optimized data to be analyzed and establish a criterion for judging whether the random vector y(i) in the optimized data is independent. Use the maximum entropy principle through an optimal linear function to maximize the negative entropy and obtain a linear data set. Step 5, according to the linear data set, the maximum value of the non-Gaussianity between the components is obtained to achieve the removal of coherent noise in the source ultrasonic guided wave; the ultrasonic guided wave echo defect detection signal preprocessing of the ultrasonic guided wave echo signal set includes de-averaging and whitening processing; The de-averaging is to subtract the mean from the matrix A of the received ultrasonic guided wave echo signal set to ensure that the mean after processing is 0, which can reduce the complexity of calculation, and can be specifically expressed as: ； in In practice, it is often used to replace the expectation with the average value; The whitening process is to transform the covariance matrix of the ultrasonic guided wave echo signal set into a diagonal matrix, so that the variables under study are linearly independent, and the correlation between the ultrasonic guided wave echo signal sets is removed, thereby simplifying the subsequent extraction process and improving the convergence of the ICA algorithm. The whitening process is performed by using the eigenvalue decomposition of the diagonal matrix to obtain the whitening matrix of the matrix A of the ultrasonic guided wave echo signal set: ； Where G and H are the characteristic matrix and eigenvalue diagonal matrix of the covariance matrix E{X, Xi} respectively; The iterative formula for optimizing the calculation of each column of the whitening matrix using the Newton fast iterative algorithm is as follows: = <Zg( Z)>-< ( Z)> ; in represents the element of the Pth column of the Kth order iteration result; Z is an orthogonal matrix, which can be expressed as Z=AF, g(x) is a nonlinear function with x as the variable, and g(x)=tanh(x) can be taken. (x) represents the derivative of g(x), <> represents the mean value processing, and the superscript "—" represents the normalization processing; The method of maximizing the negative entropy is to use the negative entropy method to determine the optimized data to be analyzed and to establish a criterion for determining whether the random vector y(i) in the optimized data is independent, and to use the negative entropy J(y) as a measure of the non-Gaussianity of each component, that is, assuming that the probability density of the random vector y is P(y), then its entropy is: ； Its negative entropy is: ； in is a Gaussian random component.

2. The method for processing non-random coherent noise in a ground electrode defect echo signal according to claim 1, characterized in that: The sensors are provided with n numbers. Assuming that n sensors are placed, each sensor performs p measurements and each sensor has p guided wave signals, the matrix A of these ultrasonic guided wave echo signal sets can be expressed as: ； Among them, A is the data matrix of n sensors, C is the original matrix composed of independent signals, and B is called the mixing matrix, which is composed of signal weight values.

3. The method for processing non-random coherent noise in a ground electrode defect echo signal according to claim 1, characterized in that: The maximum value of the non-Gaussianity between the components is obtained. According to the maximum entropy principle, the negative entropy J(y) can be estimated as: ； where G(x) is an arbitrary non-quadratic function, c is a positive constant, u is a Gaussian variable with zero mean and unit variance, and E{} is the mathematical expectation operator.

4. The method for processing non-random coherent noise in a ground electrode defect echo signal according to claim 3, characterized in that: G(x)=x 4 / 4.

5. The method for processing non-random coherent noise in a ground electrode defect echo signal according to claim 1, characterized in that: The method capable of processing non-random coherent noise in the ground electrode defect echo signal is stored in an application program of a computer architecture and driven by a burned program. The computer architecture also includes a bus architecture, a memory and a bus interface. The bus architecture includes any number of interconnected buses and bridges. The bus architecture links together various circuits including one or more processors represented by a processor and a memory represented by a memory. The bus architecture can also connect various other circuits of peripheral devices, voltage regulators and power management circuits. The bus interface provides an interface between the bus architecture and a receiver and a transmitter. The receiver and the transmitter are the same element, namely a transceiver, which provides a unit for communicating with various other systems on a transmission medium.

Images