A method for constructing an imaging algorithm of an ultrasonic testing device for power transmission hardware

Abstract

The application discloses a kind of transmission hardware ultrasonic detection device imaging algorithm construction methods, including, through wedge probe generation and optimization ultrasonic wave evanescent wave signal, the evanescent wave signal and propagation wave signal of transmission hardware are collected, using anti-interference algorithm to the signal collected is carried out noise reduction processing;To the evanescent wave and propagation wave signal after noise reduction is carried out multimodal fusion processing, based on deep learning model signal feature is classified and predicted with defect;Through super-resolution reconstruction algorithm realizes the three-dimensional imaging of hardware inside, and the three-dimensional imaging result is carried out regional feature extraction, generates layered data, to layered data is carried out intelligent segmentation and defect labeling, simultaneously generated detection result is uploaded to remote server, realizes scene analysis.The application method can be in complex environment in the field efficiently detects, also can be through remote data processing and analysis, provides technical support for the state monitoring and fault prediction of transmission hardware.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission hardware ultrasonic detection, and particularly relates to a construction method of an imaging algorithm of a power transmission hardware ultrasonic detection device. BACKGROUND

[0002] With the rapid development of the power industry, power transmission hardware plays a key role in connecting and conducting in the power grid, and its quality directly affects the safe operation of the power system. However, the existing power transmission hardware detection technology mainly relies on appearance detection, grip test or X-ray detection, which has certain limitations. For example, appearance detection cannot reflect the internal structure state of the hardware; grip test is a destructive test, which is only suitable for batch sampling before installation; and X-ray detection equipment is heavy and complex to operate, and is difficult to apply in the field high-altitude working environment. In addition, the imaging of traditional detection methods is mostly two-dimensional planar images, which cannot fully reflect the three-dimensional structural characteristics of the internal defects of the hardware, limiting the accuracy and refinement level of defect analysis. Therefore, it is particularly important to develop an efficient three-dimensional imaging method suitable for internal structure detection of power transmission hardware. SUMMARY

[0003] In view of the above existing problems, the present application is proposed.

[0004] Therefore, the present application provides a construction method of an imaging algorithm of a power transmission hardware ultrasonic detection device, which proposes a construction method of an imaging algorithm based on an ultrasonic detection device to solve the following technical problems in view of the technical limitations of the existing power transmission hardware detection method: first, how to realize high-resolution detection of the complex internal structure of the power transmission hardware through optimized ultrasonic evanescent wave and propagation wave signal acquisition technology; second, how to improve the quality and reliability of the signal through an anti-interference signal processing algorithm to eliminate noise interference in a complex environment; third, how to realize accurate classification and extended prediction of defects through a multi-modal fusion algorithm and a deep learning model; and finally, how to comprehensively reconstruct the internal structure model of the power transmission hardware and label and locate the defects through super-resolution three-dimensional imaging and intelligent layering technology to meet the application requirements of on-site detection and remote collaboration.

[0005] To solve the above technical problems, the present application provides the following technical scheme, a construction method of an imaging algorithm of a power transmission hardware ultrasonic detection device, comprising:

[0006] The wedge-shaped probe generates and optimizes the ultrasonic evanescent wave signal, acquires the evanescent wave signal and the propagation wave signal of the power transmission hardware, and performs noise reduction processing on the acquired signals using an anti-interference algorithm;

[0007] The multi-modal fusion processing is performed on the noise-reduced evanescent wave and propagation wave signals, and the deep learning model is used to classify and predict the signal features.

[0008] The three-dimensional imaging inside the fitting is realized by a super-resolution reconstruction algorithm, regional feature extraction is performed on the three-dimensional imaging result, layered data is generated, intelligent segmentation and defect labeling are performed on the layered data, the generated detection result is uploaded to a remote server, and scene analysis is realized.

[0009] As a preferred scheme of the construction method of the imaging algorithm of the power transmission fitting ultrasonic detection device, the generated and optimized ultrasonic evanescent wave signal comprises using an adaptive wedge-shaped probe to dynamically adjust the ultrasonic incident angle to the critical angle to form an evanescent wave field according to the detection environment.

[0010] The frequency and pulse width of the ultrasonic wave are adjusted, and a nine-square format electromagnetic induction array collector is combined to collect the evanescent wave signal of the fitting surface near field with high sensitivity.

[0011] As a preferred scheme of the construction method of the imaging algorithm of the power transmission fitting ultrasonic detection device, the collection of the evanescent wave signal and the propagation wave signal of the power transmission fitting comprises designing a sensor array for multi-frequency signal collection, simultaneously acquiring evanescent wave signals and propagation wave signals of different frequencies, using high-viscosity coupling materials between the signal collection device and the power transmission fitting to ensure stable signal transmission, and calibrating the collection time of the evanescent wave signal and the propagation wave signal through a signal synchronization device to maintain signal synchronization.

[0012] As a preferred scheme of the construction method of the imaging algorithm of the power transmission fitting ultrasonic detection device, the anti-interference algorithm comprises using sparse representation decomposition technology to split the input ultrasonic signal into sparse components and interference components, dynamically adjusting the frequency range through an adaptive band-pass filter to filter non-target noise signals, and using multi-dimensional wavelet transform technology to separate and suppress interference signals from time domain, frequency domain and direction domain.

[0013] As a preferred scheme of the construction method of the imaging algorithm of the power transmission fitting ultrasonic detection device, the multi-modal fusion processing comprises constructing a multi-input convolutional neural network model, inputting signal data of the evanescent wave and the propagation wave into the neural network, setting a dynamic weighting mechanism in the network structure, adjusting the fusion ratio of the evanescent wave and the propagation wave signal according to the signal type and characteristics, and extracting time sequence features and spatial distribution features in the signal through a space-time domain feature extraction module.

[0014] As a preferred scheme of the construction method of the imaging algorithm of the power transmission hardware ultrasonic detection device, the defect classification and prediction of the signal characteristics comprises: constructing a deep learning classification model by using a transfer learning technology, pre-training the model through industrial ultrasonic detection data, generating different types of defect simulation samples by using a generative adversarial network, expanding the training data set of the classification model, and predicting the development trend of the defect by using a recurrent neural network to analyze the time sequence characteristics contained in the ultrasonic signal.

[0015] As a preferred scheme of the construction method of the imaging algorithm of the ultrasonic detection device for power transmission hardware, the super-resolution reconstruction algorithm comprises: optimizing the spatial resolution of the ultrasonic wave signal based on the eigenvalue decomposition of the time reversal matrix, and cutting the depth information of the ultrasonic signal into two-dimensional image slices, each slice representing signal data at a specific depth, and reconstructing the slice image into a complete three-dimensional structure model of the power transmission hardware through a layer-by-layer superposition technology.

[0016] As a preferred scheme of the construction method of the imaging algorithm of the ultrasonic detection device for power transmission hardware, the generation of the layered data comprises: using a deep learning segmentation algorithm to segment the generated three-dimensional model in a set region, and dividing the three-dimensional model into multiple independent layers according to the material properties and signal characteristics, including a metal substrate layer, a crack layer and a corrosion layer.

[0017] The layered results are subjected to preliminary data structure processing to form a data model for defect analysis.

[0018] A computer device comprises a memory and a processor, and the memory stores a computer program, characterized in that the processor implements the steps of the construction method of the imaging algorithm of the ultrasonic detection device for power transmission hardware when executing the computer program.

[0019] A computer readable storage medium stores a computer program, characterized in that the computer program is executed by a processor to implement the steps of the construction method of the imaging algorithm of the ultrasonic detection device for power transmission hardware.

[0020] The present application provides an efficient, accurate and intelligent power transmission fitting ultrasonic detection device imaging algorithm construction method, which realizes high-resolution detection of fitting internal defects through innovative evanescent wave signal optimization and multi-modal fusion technology; combined with anti-interference signal processing and deep learning algorithm, the quality of the signal and the accuracy of defect classification are significantly improved; based on super-resolution three-dimensional imaging technology, the internal structure of the fitting can be completely reconstructed, and through intelligent layering and defect labeling technology, accurate identification and classification of complex defects are realized. The method is simple to operate and has strong adaptability, which can not only perform efficient detection in complex outdoor environments, but also provide technical support for power transmission fitting state monitoring and fault prediction through remote data processing and analysis, significantly improving the safety and reliability of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can be obtained without creative labor on the basis of these drawings.

[0022] Figure 1 A flowchart of a construction method of a power transmission fitting ultrasonic detection device imaging algorithm provided by an embodiment of the present application. DETAILED DESCRIPTION

[0023] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings of the specification. Obviously, the described embodiments are only a part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should be within the scope of protection of the present application.

[0024] In the following description, many specific details are set forth in order to provide a thorough understanding of the present application, but the present application can also be implemented in other ways different from those described herein, and those skilled in the art can make similar generalizations without departing from the connotation of the present application, therefore the present application is not limited to the specific embodiments disclosed below.

[0025] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. "In one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment that excludes other embodiments.

[0026] The application is described in detail in combination with the schematic diagram. In the detailed description of the embodiments of the application, the cross-sectional view of the device structure is partially enlarged without the general proportion for the convenience of illustration, and the schematic diagram is only an example, which should not limit the scope of protection of the application herein. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in actual production.

[0027] Meanwhile, in the description of the application, it should be noted that the terms "upper, lower, inner and outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the application. In addition, the terms "first, second or third" are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0028] In the application, unless otherwise explicitly specified and limited, the terms "mounting, connection, connection" should be understood broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the application can be understood according to the specific circumstances.

[0029] Embodiment 1, refer to Figure 1 As the first embodiment of the application, the embodiment provides a construction method of imaging algorithm of power transmission hardware ultrasonic detection device, comprising:

[0030] S1: generating and optimizing ultrasonic evanescent wave signals through a wedge-shaped probe, collecting evanescent wave signals and propagation wave signals of the power transmission hardware, and using an anti-interference algorithm to process the collected signals.

[0031] S2: performing multi-modal fusion processing on the noise-reduced evanescent wave and propagation wave signals, and classifying and predicting the signal features based on a deep learning model.

[0032] S3: realizing three-dimensional imaging of the hardware inside through a super-resolution reconstruction algorithm, extracting regional features of the three-dimensional imaging results, generating layered data, intelligently segmenting and defect labeling the layered data, uploading the generated detection results to a remote server, and realizing scene analysis.

[0033] The generating and optimizing of the ultrasonic evanescent wave signals includes using an adaptive wedge-shaped probe, dynamically adjusting the ultrasonic incident angle to the critical angle to form an evanescent wave field according to the detection environment;

[0034] By adjusting the frequency and pulse width of the ultrasonic waves and combining them with a nine-square electromagnetic induction array collector, the evanescent wave signal in the near field of the hardware surface is acquired with high sensitivity.

[0035] The acquisition of evanescent wave and propagation wave signals from power transmission fittings includes: a sensor array designed for multi-band signal acquisition, which simultaneously acquires evanescent wave and propagation wave signals of different frequencies; the use of a high-viscosity coupling material between the signal acquisition device and the power transmission fittings to ensure stable signal transmission; and the calibration of the acquisition time of the evanescent wave and propagation wave signals through a signal synchronization device to maintain signal synchronization.

[0036] The anti-interference algorithm includes using sparse representation decomposition technology to split the input ultrasonic signal into sparse components and interference components, dynamically adjusting the frequency range through an adaptive bandpass filter to filter non-target noise signals, and using multidimensional wavelet transform technology to separate and suppress interference signals in the time domain, frequency domain, and direction domain.

[0037] The multimodal fusion processing includes constructing a multi-input convolutional neural network model, inputting the signal data of evanescent wave and propagating wave into the neural network respectively, setting a dynamic weighting mechanism in the network structure, adjusting the fusion ratio of evanescent wave and propagating wave signals according to the signal type and characteristics, and extracting the time series features and spatial distribution features in the signal through a spatiotemporal domain feature extraction module.

[0038] The defect classification and prediction of signal features includes: constructing a deep learning classification model using transfer learning technology; pre-training the model using industrial ultrasonic testing data; generating different types of defect simulation samples using generative adversarial networks; expanding the training dataset of the classification model; and using recurrent neural networks to analyze the time series features contained in the ultrasonic signals to predict the development trend of defects.

[0039] The super-resolution reconstruction algorithm includes optimizing the spatial resolution of the ultrasonic signal based on the eigenvalue decomposition of the time-reversed matrix, dividing the depth information of the ultrasonic signal into two-dimensional image slices, each slice representing signal data at a specific depth, and reconstructing the sliced ​​images into a complete three-dimensional structural model of the power transmission fittings through a layer-by-layer stacking technique.

[0040] The generation of layered data includes using a deep learning segmentation algorithm to segment a defined region of the generated 3D model, and dividing the 3D model into multiple independent layers based on material properties and signal characteristics, including a metal substrate layer, a crack layer, and a corrosion layer.

[0041] The stratified results undergo preliminary data structuring to form a data model for defect analysis.

[0042] According to the geometric characteristics of the three-dimensional model, the boundary information of the defect area is extracted and labeled; based on the signal characteristics of the defects, a classification algorithm is used to classify the labeled area, and type labels of cracks, pores or corrosion and other defects are generated; the spatial coordinate information of the defect area is attached to the three-dimensional model to form a complete defect labeling model.

[0043] Through the communication module, the three-dimensional imaging data and its labeling information are transmitted to the remote server; data analysis is performed on the server side, and a visual inspection report including the three-dimensional model and the defect labeling is generated; a remote collaboration interface is provided to allow operators to perform further analysis and processing based on the uploaded data, and decision support data is generated.

[0044] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the present application, and they should be covered in the scope of the claims of the present application.

[0045] Embodiment 2, the second embodiment of the present application, which is different from the previous embodiment:

[0046] If the function is realized in the form of a software function unit and sold or used as an independent product, it can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the embodiments of the present application. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various media that can store program codes.

[0047] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems), and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices produce a machine that implements the functions described in the flowcharts and / or block diagrams. Figure 1one or more processes and / or blocks Figure 1 an apparatus for performing the functions specified in the flowchart

[0048] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the flowchart Figure 1 one or more processes and / or blocks Figure 1 an apparatus for performing the functions specified in the flowchart

[0049] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the flowchart Figure 1 one or more processes and / or blocks Figure 1 an apparatus for performing the functions specified in the flowchart

[0050] While the preferred embodiments of the application have been described, additional variations and modifications can be made to the embodiments by those skilled in the art once they learn of the basic inventive concepts. Therefore, the appended claims are intended to cover all such additional variations and modifications as fall within the scope of the application.

[0051] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described.

Claims

1. A method for constructing an imaging algorithm for an ultrasonic testing device for power transmission fittings, characterized in that: include, The ultrasonic evanescent wave signal is generated and optimized by a wedge probe, and the evanescent wave signal and propagation wave signal of the power transmission fitting are collected. The collected signal is then denoised using an anti-interference algorithm. Multimodal fusion processing is performed on denoised evanescent wave and propagating wave signals, and defect classification and prediction of signal features are performed based on a deep learning model. The super-resolution reconstruction algorithm is used to realize three-dimensional imaging of the interior of the fittings, and regional features are extracted from the three-dimensional imaging results to generate layered data. The layered data is intelligently segmented and defect labeled, and the generated detection results are uploaded to a remote server to realize scene-based analysis. The multimodal fusion processing includes constructing a multi-input convolutional neural network model, inputting the signal data of evanescent wave and propagating wave into the neural network respectively, setting a dynamic weighting mechanism in the network structure, adjusting the fusion ratio of evanescent wave and propagating wave signals according to the signal type and characteristics, and extracting the time series features and spatial distribution features in the signal through a spatiotemporal domain feature extraction module. The defect classification and prediction of signal features includes: constructing a deep learning classification model using transfer learning technology; pre-training the model using industrial ultrasonic testing data; generating different types of defect simulation samples using generative adversarial networks; expanding the training dataset of the classification model; and using recurrent neural networks to analyze the time series features contained in the ultrasonic signals to predict the development trend of defects. The super-resolution reconstruction algorithm includes optimizing the spatial resolution of the ultrasonic signal based on the eigenvalue decomposition of the time-reversal matrix, dividing the depth information of the ultrasonic signal into two-dimensional image slices, each slice representing signal data at a specific depth, and reconstructing the sliced ​​images into a complete three-dimensional structural model of the power transmission fittings through a layer-by-layer stacking technique. The generation of layered data includes using a deep learning segmentation algorithm to segment a defined region of the generated 3D model, and dividing the 3D model into multiple independent layers based on material properties and signal characteristics, including a metal substrate layer, a crack layer, and a corrosion layer. The stratified results undergo preliminary data structuring to form a data model for defect analysis.

2. The method for constructing an imaging algorithm for an ultrasonic testing device for power transmission fittings as described in claim 1, characterized in that: The generation and optimization of the ultrasonic evanescent wave signal includes using an adaptive wedge probe to dynamically adjust the ultrasonic incident angle to a critical angle according to the detection environment to form an evanescent wave field. By adjusting the frequency and pulse width of the ultrasonic waves and combining them with a nine-square electromagnetic induction array collector, the evanescent wave signal in the near field of the hardware surface is acquired with high sensitivity.

3. The method for constructing an imaging algorithm for an ultrasonic testing device for power transmission fittings as described in claim 2, characterized in that: The acquisition of evanescent wave and propagation wave signals from power transmission fittings includes: a sensor array designed for multi-band signal acquisition, which simultaneously acquires evanescent wave and propagation wave signals of different frequencies; the use of a high-viscosity coupling material between the signal acquisition device and the power transmission fittings to ensure stable signal transmission; and the calibration of the acquisition time of the evanescent wave and propagation wave signals through a signal synchronization device to maintain signal synchronization.

4. The method for constructing an imaging algorithm for an ultrasonic testing device for power transmission fittings as described in claim 3, characterized in that: The anti-interference algorithm includes using sparse representation decomposition technology to split the input ultrasonic signal into sparse components and interference components, dynamically adjusting the frequency range through an adaptive bandpass filter to filter non-target noise signals, and using multidimensional wavelet transform technology to separate and suppress interference signals in the time domain, frequency domain, and direction domain.

5. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 4.

6. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 4.

Images