Insulator pollution degree detection method and device, storage medium and electronic equipment

Through the fusion of infrared images and LIBS spectral information, convolutional neural networks are used to generate insulator filth data, which solves the problem of infrared imaging technology being affected by the weather and has low accuracy, and achieves accurate judgment of insulator filth levels.

CN119024112APending Publication Date: 2024-11-26GUANGDONG POWER GRID CO LTD +1

Patent Information

Application Number
CN202411125139.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-16
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

In the prior art, infrared imaging technology is used to detect that the insulator pollution level is greatly affected by the weather, and there is a lack of comprehensive judgment, resulting in low accuracy of the judgment results.

Method used

The detection method of multi-source information fusion of infrared images and LIBS spectral fusion is adopted, and the infrared image information of insulators is obtained through infrared imaging technology and the laser-induced breakdown spectral technology is obtained to obtain atomic spectral information, and the deep learning model of convolutional neural networks is used to inform the two. Extract and fuse, generate grayscale data and atomic data, and then determine the level of filth.

Benefits of technology

Through the fusion of the physical information of the image and the chemical composition of the elements, the accurate judgment of the filth level is improved, the problems of weather impact and comprehensive judgment difficulty are solved, and the accuracy of the detection results is improved.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119024112A_ABST
    Figure CN119024112A_ABST
Patent Text Reader

Abstract

The invention provides an insulator pollution degree detection method and device, a storage medium and electronic equipment. The method comprises the following steps: acquiring infrared image information of an insulator at the current moment by adopting an infrared imaging technology; acquiring spectral information of the insulator by adopting a laser-induced breakdown spectroscopy technology, wherein the spectral information at least comprises atomic spectral information of the insulator at the current moment; information extraction and fusion are carried out on the infrared image information and the spectral information through a deep learning model of a convolutional neural network, insulator pollution data are obtained, and the insulator pollution data comprise gray data of an insulator and atomic data of the insulator; according to the method, the pollution level of the insulator is determined according to pollution data of the insulator, the pollution degree of the insulator is in direct proportion to the pollution level of the insulator, two kinds of information of two-dimensional image information and element spectrum information can be fused, and the pollution level of the insulator is determined from two kinds of characteristics of physical information of an image and chemical composition of elements. And the pollution grades can be classified more accurately.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the field of insulator contamination detection, and specifically, to a method for detecting insulator contamination based on multi-source information fusion of infrared images and LIBS spectra, a device for detecting insulator contamination based on multi-source information fusion of infrared images and LIBS spectra, a computer-readable storage medium and an electronic device. Background Art

[0002] Insulators are a special insulating material, usually made of ceramic or glass. Insulators are fixed between wires and towers to increase the distance between wires and the ground and prevent current from flowing from the tower to the ground. Due to the working environment of insulators, they are easily contaminated, which leads to a decrease in insulation performance and safety hazards. Therefore, in order to ensure the normal operation and safety of the power system, it is necessary to conduct regular quality and contamination tests on insulators to detect and deal with problems in a timely manner to ensure that they have good insulation performance and stability.

[0003] Currently, equivalent salt density and equivalent ash density are universal standard methods for evaluating the pollution level of insulators. However, the detection method requires climbing and sampling, laboratory analysis, and the accuracy of the analysis results needs to be improved. Therefore, in recent years, image recognition methods based on infrared imaging have been applied to the identification of the degree of contamination of insulators. In the drone-mounted mode, on-site level judgment of transmission lines can be achieved. However, infrared imaging is still greatly affected by weather, sunlight, etc., and it is still difficult to accurately judge the pollution level. Therefore, the detection technology based on a single data information source lacks comprehensive judgment when detecting insulator contamination, and the detection accuracy of the contamination level is limited. Summary of the invention

[0004] The main purpose of the present application is to provide a method for detecting the contamination level of an insulator based on the fusion of multi-source information of infrared images and LIBS spectra, a device for detecting the contamination level of an insulator based on the fusion of multi-source information of infrared images and LIBS spectra, a computer-readable storage medium and an electronic device, so as to at least solve the problem that the method of detecting the pollution level of an insulator using infrared imaging technology in the prior art is greatly affected by the weather, lacks comprehensive judgment, and has low accuracy of judgment results.

[0005] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a method for detecting the contamination degree of an insulator based on the fusion of multi-source information of infrared images and LIBS spectra is provided, comprising: using infrared imaging technology to obtain infrared image information of the insulator at the current moment; using laser induced breakdown spectroscopy technology to obtain spectral information of the insulator, wherein the spectral information at least includes atomic spectral information of the insulator at the current moment; extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, wherein the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0006] Optionally, the infrared image information and the spectral information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, including: generating an image information data array of the insulator according to the infrared image information; generating a spectral data array of the insulator according to the spectral information; and fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, wherein the contamination data array is the insulator contamination data.

[0007] Optionally, generating an image information data array of the insulator based on the infrared image information includes: performing feature extraction on the infrared image information to obtain grayscale image information of the insulator; and generating an image information data array of the insulator based on the grayscale image information.

[0008] Optionally, the image information data array and the spectral data array are fused through a deep learning model of a convolutional neural network to obtain a pollution data array of the insulator, including: using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array to respectively obtain first characteristic data and second characteristic data; performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; using a feature hierarchical method to process the first processed data and the second processed data to obtain a first optimal characteristic subset and a second optimal characteristic subset; using two-dimensional correlation analysis to fuse the first optimal characteristic subset and the second optimal characteristic subset to obtain the pollution data array of the insulator.

[0009] Optionally, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, the method further includes: performing data preprocessing on the infrared image information and the spectral information using a preprocessing algorithm, the preprocessing algorithm including at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

[0010] Optionally, determining the pollution level of the insulator according to the pollution data of the insulator includes: combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm including at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0011] Optionally, after inputting the pollution data array of the insulator into the fusion classification model and obtaining the output of the fusion classification model as the pollution level of the insulator, the method further includes: obtaining a target mapping relationship, wherein the target mapping relationship is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; determining the target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; and adjusting the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model to obtain an updated classification model, wherein the updated classification model corresponds one-to-one to the target geographical range.

[0012] According to another aspect of the present application, a device for detecting the contamination degree of an insulator based on multi-source information fusion of infrared images and LIBS spectra is provided, comprising: a first acquisition unit, used to acquire infrared image information of the insulator at the current moment by using infrared imaging technology; a second acquisition unit, used to acquire spectral information of the insulator by using laser induced breakdown spectroscopy technology, wherein the spectral information at least includes atomic spectral information of the insulator at the current moment; a fusion unit, used to extract and fuse the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, wherein the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; a determination unit, used to determine the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0013] According to another aspect of the present application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein when the program is run, the device where the computer-readable storage medium is located is controlled to execute any one of the insulator contamination detection methods based on multi-source information fusion of infrared images and LIBS spectra.

[0014] According to another aspect of the present application, an electronic device is provided, comprising: one or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing any one of the above-mentioned insulator contamination detection methods based on multi-source information fusion of infrared images and LIBS spectra.

[0015] By applying the technical solution of the present application, the above-mentioned insulator contamination degree detection method based on infrared image and LIBS spectrum multi-source information fusion first uses infrared imaging technology to obtain the infrared image information of the insulator at the current moment; then uses laser induced breakdown spectroscopy technology to obtain the spectral information of the insulator, and the spectral information at least includes the atomic spectral information of the insulator at the current moment; then, through the deep learning model of the convolutional neural network, the infrared image information and the spectral information are extracted and fused to obtain the insulator contamination data, and the insulator contamination data includes the grayscale data of the insulator and the atomic data of the insulator; finally, the contamination level of the insulator is determined according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator. The method can fuse two types of information, namely, two-dimensional image information and element spectrum information, and more accurately classify the pollution level from the two types of characteristics, namely, the physical information of the image and the chemical composition of the elements. By superimposing and fusing the two-dimensional image and the pollution element spectrum, a multi-source data source is formed, and accurate judgment of the pollution level is achieved with the help of multi-source information. This solves the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings constituting part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0017] Figure 1 A hardware structure block diagram of a mobile terminal for performing a method for detecting the contamination degree of an insulator provided in an embodiment of the present application is shown;

[0018] Figure 2 A schematic flow chart of a method for detecting the contamination degree of an insulator provided in an embodiment of the present application is shown;

[0019] Figure 3 A structural block diagram of a device for detecting the contamination level of an insulator provided according to an embodiment of the present application is shown.

[0020] The above drawings include the following reference numerals:

[0021] 102, processor; 104, memory; 106, transmission device; 108, input and output devices. DETAILED DESCRIPTION

[0022] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] As introduced in the background technology, in the prior art, equivalent salt density and equivalent ash density are universal standard methods for evaluating the pollution level of insulators, but the detection method requires climbing and sampling, laboratory analysis, and the accuracy of the analysis results needs to be improved. In order to solve the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results, the embodiments of the present application provide a method for detecting the pollution level of insulators based on multi-source information fusion of infrared images and LIBS spectra, a device for detecting the pollution level of insulators based on multi-source information fusion of infrared images and LIBS spectra, a computer-readable storage medium and an electronic device.

[0026] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0027] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a mobile terminal of a method for detecting the contamination degree of an insulator according to an embodiment of the present invention. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations are shown.

[0028] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the detection method of the insulator contamination degree in the embodiment of the present invention. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, the above method is implemented. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above networks include but are not limited to the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof. The transmission device 106 is used to receive or send data via a network. The specific examples of the above networks may include a wireless network provided by a communication provider of the mobile terminal. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0029] In this embodiment, a method for detecting the contamination degree of an insulator running on a mobile terminal, a computer terminal or a similar computing device is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0030] Figure 2 FIG. 1 is a flow chart of a method for detecting the degree of insulator contamination according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0031] Step S201, using infrared imaging technology to obtain infrared image information of the insulator at the current moment;

[0032] Specifically, the wavelength of infrared imaging is an order of magnitude larger than that of visible light. Infrared thermal imaging generally works at a wavelength of 8 to 14 μm. Compared with visible light (0.4 to 0.8 μm wavelength), it has better smoke penetration ability and is not affected by strong light and glare. Compared with ultraviolet light sources, it is not easy to attenuate when propagating in the air. When interacting with pollution, it has a better penetration depth and can obtain more comprehensive pollution information.

[0033] Among them, the grayscale conversion of the infrared image is performed to generate a data array containing image features.

[0034] Step S202, using laser induced breakdown spectroscopy to obtain spectrum information of the insulator, wherein the spectrum information at least includes atomic spectrum information of the insulator at the current moment;

[0035] Specifically, laser-induced breakdown spectroscopy technology is highly sensitive to the atomic spectrum information of insulators and can detect the presence of contaminants, which helps to promptly discover contamination and defects on the surface of insulators. The two-dimensional image is superimposed and fused with the spectrum of contaminant elements to form a multi-source data source, which can help to accurately judge the contamination level with the help of multi-source information.

[0036] Step S203, extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, where the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator;

[0037] Specifically, through the deep learning model of convolutional neural network (CNN), the two types of data sources, image information data array and spectral data array, are fused to form a new data array. Through large-scale data training and learning, the efficiency of data processing and analysis is improved, saving manpower and time costs.

[0038] The deep learning model of the convolutional neural network is used to extract and fuse the infrared image information and the spectral information to obtain the insulator contamination data, including the following steps:

[0039] Step S301, generating an image information data array of the insulator according to the infrared image information;

[0040] The step of generating the image information data array of the insulator according to the infrared image information comprises the following steps:

[0041] Step S401, extracting features from the infrared image information to obtain grayscale image information of the insulator;

[0042] Step S402: generating an image information data array of the insulator according to the grayscale image information.

[0043] Specifically, feature extraction can help us extract useful information from infrared images and convert image information into data arrays, which can facilitate data processing and analysis and improve processing efficiency.

[0044] Step S302, generating a spectrum data array of the insulator according to the spectrum information;

[0045] Step S303, using a deep learning model of a convolutional neural network, information fusion is performed on the image information data array and the spectral data array to obtain a contamination data array of the insulator, where the contamination data array is the contamination data of the insulator.

[0046] The process of fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator includes the following steps:

[0047] Step S501, using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array, to obtain first characteristic data and second characteristic data respectively;

[0048] Step S502, performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data;

[0049] Step S503, using a feature hierarchical method to process the first processed data and the second processed data respectively to obtain a first optimal feature subset and a second optimal feature subset;

[0050] Step S504: using two-dimensional correlation analysis to fuse the first optimal feature subset and the second optimal feature subset to obtain the contamination data array of the insulator.

[0051] Specifically, the continued projection algorithm and the uninformative variable removal algorithm can reduce the complexity of the model and improve the accuracy and robustness of the model prediction.

[0052] Specifically, by fusing infrared image information and spectral information, the accuracy of the data can be improved, thereby gaining a more comprehensive understanding of the contamination of insulators and more accurately classifying the contamination level based on the two characteristics of the image's physical information and the element's chemical composition.

[0053] Wherein, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes:

[0054] The infrared image information and the spectral information are preprocessed by using a preprocessing algorithm, wherein the preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scatter correction algorithm, and a multi-order derivative algorithm.

[0055] Specifically, combining multiple algorithms can achieve full collaboration between the two types of data, reduce model complexity, and improve the accuracy and robustness of model predictions.

[0056] In some instances, the data is preprocessed before fusion, and the preprocessing algorithms used include smoothing algorithms, variable standardization, multivariate scatter correction, multi-order derivatives, etc.

[0057] The continuous projection algorithm and the uninformative variable removal algorithm are used to extract the characteristic bands in the data array, and the extracted characteristic data are subjected to wavelet three-dimensional decomposition. The optimal characteristic subset is generated using the characteristic stratification method. In this process, the influence of three different strategies at the spectral level, characteristic level and decision level on the fusion effect is analyzed. The optimal characteristic subset is fused using two-dimensional correlation analysis. The fused data array is used as the input variable. The deep learning algorithms such as support vector machine, random forest algorithm and convolutional neural network can be combined to establish the fusion model of the learner using the optimal weighted fusion strategy, so as to achieve full collaboration between the two types of data, reduce the complexity of the model, and improve the accuracy and robustness of the model prediction.

[0058] According to the sample data of different pollution levels, four levels of pollution samples were selected. 100 infrared and LIBS spectral data of each pollution level sample were collected for modeling. A pollution level classification method based on convolutional neural network was established to classify the collected samples by pollution level. After the model was established, the model was tested and verified on actual samples.

[0059] In some instances, multi-input convolutional neural networks are used to attribute two types of parameters, image and element spectrum, through gradient weighted class activation mapping, and data fusion of dirty infrared images and atomic LIBS spectra is performed.

[0060] Different parameter variables pass through multiple convolutional layers of the neural network, and feature learning is completed after the pooling layer. They are fused multiple times at different feature layers. At the same time, T-SNE is used to perform nonlinear unsupervised dimensionality reduction on the data at each level of the neural network to realize the visualization of the multi-input network data fusion process; the attribution algorithm is used to count the importance of each variable in the LIBS spectrum and infrared image and mark the representative elements or molecules, and finally the model classification results are output; the corrected linear unit activation function is used to improve the nonlinear mapping ability of the CNN model.

[0061] Rectified Linear Unit (Relu) function:

[0062] Based on the conditional independence assumption, LIBS and infrared images under the same category of labels (labels are divided according to different staging stages) are randomly combined multiple times to achieve data set expansion and further improve the classification accuracy of the association discrimination model.

[0063] In order to facilitate modeling, the characteristic data of the target sample in the sample is represented by a 2×7 vector E: E i =[e i1 , e i2 , ..., e iL ], i = 1, 2, ... N, where i represents the sample number, e represents each feature quantity, and the values ​​of L and N are 8 and 96 respectively. The CNN model mainly consists of a shared module and a classification module. The shared module is responsible for feature extraction, including 1 batch normalization layer, 2 convolution layers and 2 maximum pooling layers. In the convolution layer, a convolution kernel with a dimension of 3×3 is used to perform convolution operations on each input data matrix and output a feature matrix. Each convolution layer is followed by a 2×2 maximum pooling layer to extract the maximum value in each pooling matrix and filter out unimportant information in the feature matrix. The shared module extracts the main features of the data to complete the classification task. The classification module mainly includes 1 stretching layer, 2 Dropout layers and 3 fully connected layers, and the last layer is a fully connected layer of size 8. For each input sample, the classification module will output a 1×8 vector. Each value in this vector represents the probability distribution of the input fusion feature at different pollution levels. The pollution level to which the input sample belongs can be judged by the maximum value in the vector.

[0064] Taking into account the model differentiation caused by different regions and different pollution types, the model is de-differentiated and universalized with the help of model transfer learning and standardization, further improving the performance of the fusion model.

[0065] Step S204, determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0066] Specifically, the two-dimensional image information and the element spectrum information are fused to form an automatic discrimination method for the pollution level by fusing multiple feature information, which saves manpower and time costs and improves the accuracy of the data.

[0067] Determining the contamination level of the insulator according to the contamination data of the insulator comprises the following steps:

[0068] Step S601, combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, wherein the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network;

[0069] Step S602, inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0070] Wherein, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further comprises the following steps:

[0071] Step S701, obtaining a target mapping relationship, where the target mapping relationship is a mapping relationship between the geographical range where the insulator is located and the operating parameters of the fusion classification model;

[0072] Step S702, determining target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship;

[0073] Step S703, based on the model transfer learning capability of the fusion classification model, the operating parameters of the fusion classification model are adjusted to the target operating parameters to obtain an updated classification model, and the updated classification model corresponds one-to-one to the target geographical range.

[0074] Specifically, considering the model differentiation caused by different regions and different pollution types, model de-differentiation and universality are achieved with the help of model transfer learning and standardization, further improving the performance of the fusion model.

[0075] Specifically, the optimal weighted fusion strategy is used to establish a fusion model of the learner to achieve full collaboration between the two types of data, reduce model complexity, and improve the accuracy and robustness of model predictions.

[0076] The above-mentioned detection method of the insulator contamination degree based on the fusion of infrared image and LIBS spectrum multi-source information of the present application first uses infrared imaging technology to obtain the infrared image information of the insulator at the current moment; then uses laser induced breakdown spectroscopy technology to obtain the spectrum information of the insulator, and the spectrum information at least includes the atomic spectrum information of the insulator at the current moment; then through the deep learning model of the convolutional neural network, the infrared image information and the spectrum information are extracted and fused to obtain the insulator contamination data, and the insulator contamination data includes the grayscale data of the insulator and the atomic data of the insulator; finally, the insulator contamination level is determined according to the insulator contamination data, wherein the insulator contamination degree is proportional to the insulator contamination level. The method can fuse two types of information, two-dimensional image information and element spectrum, and more accurately classify the contamination level from the two types of features of the physical information of the image and the chemical composition of the element. By superimposing and fusing the two-dimensional image and the contamination element spectrum, a multi-source data source is formed, and the contamination level is accurately judged with the help of multi-source information, which solves the problem that the method of detecting the insulator pollution level using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and the accuracy of the judgment result is low.

[0077] In order to enable those skilled in the art to more clearly understand the technical solution of the present application, the implementation process of the insulator contamination degree detection method of the present application will be described in detail below in conjunction with specific embodiments.

[0078] The embodiment of the present application also provides a device for detecting the contamination degree of an insulator. It should be noted that the device for detecting the contamination degree of an insulator in the embodiment of the present application can be used to execute the detection method for the contamination degree of an insulator provided in the embodiment of the present application. The device is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, the implementation in hardware, or a combination of software and hardware, is also possible and conceivable.

[0079] The following is an introduction to the insulator contamination detection device provided in an embodiment of the present application.

[0080] Figure 3 Schematic diagram of an insulator contamination detection device according to an embodiment of the present application. Figure 3As shown, the device includes: a first acquisition unit 10, a second acquisition unit 20, a fusion unit 30 and a determination unit 40, the first acquisition unit 10 is used to obtain infrared image information of the insulator at the current moment by using infrared imaging technology; the second acquisition unit 20 is used to obtain spectral information of the insulator by using laser induced breakdown spectroscopy technology, and the spectral information at least includes atomic spectral information of the insulator at the current moment; the fusion unit 30 is used to extract and fuse the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, and the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; the determination unit 40 is used to determine the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0081] The above-mentioned insulator contamination degree detection device based on infrared image and LIBS spectrum multi-source information fusion of the present application includes: a first acquisition unit, a second acquisition unit, a fusion unit and a determination unit, the first acquisition unit is used to use infrared imaging technology to obtain infrared image information of the insulator at the current moment; the second acquisition unit is used to use laser induced breakdown spectroscopy technology to obtain spectral information of the insulator, and the spectral information at least includes atomic spectral information of the insulator at the current moment; the fusion unit is used to extract and fuse the infrared image information and spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, and the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; the determination unit is used to determine the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator. The method can fuse two types of information, namely, two-dimensional image information and element spectrum information, and more accurately classify the pollution level from the two types of characteristics, namely, the physical information of the image and the chemical composition of the elements. By superimposing and fusing the two-dimensional image and the pollution element spectrum, a multi-source data source is formed, and accurate judgment of the pollution level is achieved with the help of multi-source information. This solves the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results.

[0082] In some examples, the fusion unit includes a first generation module, a second generation module and a first fusion module, wherein the first generation module is used to generate the image information data array of the insulator according to the infrared image information; the second generation module is used to generate the spectrum data array of the insulator according to the spectrum information; the first fusion module is used to fuse the image information data array and the spectrum data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, and the contamination data array is the contamination data of the insulator. By fusing the infrared image information and the spectrum information, the accuracy of the data can be improved, thereby more comprehensively understanding the contamination of the insulator, and more accurately classifying the contamination level from the two types of features of the physical information of the image and the chemical composition of the elements.

[0083] In some examples, the first generation module includes a first extraction module and a third generation module, the first extraction module is used to perform feature extraction on the infrared image information to obtain the grayscale image information of the insulator; the third generation module is used to generate the image information data array of the insulator according to the grayscale image information. Feature extraction can help us extract useful information from infrared images, convert image information into data arrays, and make data processing and analysis more convenient, thereby improving processing efficiency.

[0084] In some examples, the first fusion module includes a second extraction module, a first decomposition module, a first processing module and a second fusion module. The second extraction module is used to extract the characteristic bands in the image information data array and the characteristic bands in the spectral data array respectively by using a continuous projection algorithm and an uninformative variable removal algorithm to obtain first characteristic data and second characteristic data respectively; the first decomposition module is used to perform wavelet three-dimensional decomposition on the first characteristic data and the second characteristic data respectively to obtain first processed data and second processed data; the first processing module is used to process the first processed data and the second processed data respectively by using a characteristic stratification method to obtain a first optimal characteristic subset and a second optimal characteristic subset; the second fusion module is used to perform fusion processing on the first optimal characteristic subset and the second optimal characteristic subset by using a two-dimensional correlation analysis to obtain the contamination data array of the insulator. Through the continuous projection algorithm and the uninformative variable removal algorithm, the model complexity can be reduced and the accuracy and robustness of the model prediction can be improved.

[0085] In some examples, before the image information data array and the spectral data array are fused by the deep learning model of the convolutional neural network to obtain the pollution data array of the insulator, the device further includes a second processing module, which is used to preprocess the infrared image information and the spectral information using a preprocessing algorithm, and the preprocessing algorithm includes at least one of the following: smoothing algorithm, variable standardization algorithm, multivariate scattering correction algorithm, and multi-order derivative algorithm. Combining multiple algorithms can achieve full collaboration between the two types of data, reduce model complexity, and improve the accuracy and robustness of model prediction.

[0086] In some examples, the determination unit includes a first establishment module and a third processing module. The first establishment module is used to combine a deep learning algorithm and adopt an optimal weighted fusion strategy to establish a fusion classification model based on a learner. The deep learning algorithm includes at least one of the following: The third processing module is used for a support vector machine, a random forest algorithm, and a convolutional neural network; the pollution data array of the insulator is input into the fusion classification model, and the output of the fusion classification model is the pollution level of the insulator. The fusion model of the learner is established by adopting an optimal weighted fusion strategy to achieve full collaboration between the two types of data, reduce model complexity, and improve the accuracy and robustness of model prediction.

[0087] In some examples, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the device further includes a first acquisition module, a first determination module and a fourth processing module, wherein the first acquisition module is used to acquire a target mapping relationship, wherein the target mapping relationship is a mapping relationship between the geographical range where the insulator is located and the operating parameters of the fusion classification model; the first determination module is used to determine the target operating parameters of the fusion classification model according to the target geographical range where the insulator is located and the target mapping relationship; the fourth processing module is used to adjust the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model, and obtain an updated classification model, wherein the updated classification model corresponds to the target geographical range one by one. Considering the model differentiation caused by different regions and different pollution types, the model de-differentiation and universality are realized with the help of model transfer learning and standardization, and the performance of the fusion model is further improved.

[0088] The above-mentioned insulator contamination detection device includes a processor and a memory, and the above-mentioned first acquisition unit and the like are stored in the memory as program units, and the processor executes the above-mentioned program units stored in the memory to realize the corresponding functions. The above-mentioned modules are all located in the same processor; or, the above-mentioned modules are located in different processors in any combination.

[0089] The processor includes a kernel, which retrieves the corresponding program unit from the memory. One or more kernels can be set, and the kernel parameters can be adjusted to solve the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results.

[0090] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM, and the memory includes at least one memory chip.

[0091] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the insulator contamination degree detection method.

[0092] Specifically, the detection method of insulator contamination degree includes:

[0093] Step S201, using infrared imaging technology to obtain infrared image information of the insulator at the current moment;

[0094] Specifically, the wavelength of infrared imaging is an order of magnitude larger than that of visible light. Infrared thermal imaging generally works at a wavelength of 8 to 14 μm. Compared with visible light (0.4 to 0.8 μm wavelength), it has better smoke penetration ability and is not affected by strong light and glare. Compared with ultraviolet light sources, it is not easy to attenuate when propagating in the air. When interacting with pollution, it has a better penetration depth and can obtain more comprehensive pollution information.

[0095] Step S202, using laser induced breakdown spectroscopy to obtain spectrum information of the insulator, wherein the spectrum information at least includes atomic spectrum information of the insulator at the current moment;

[0096] Specifically, laser-induced breakdown spectroscopy technology is highly sensitive to the atomic spectrum information of insulators and can detect the presence of contaminants, which helps to promptly discover contamination and defects on the surface of insulators. The two-dimensional image is superimposed and fused with the spectrum of contaminant elements to form a multi-source data source, which can help to accurately judge the contamination level with the help of multi-source information.

[0097] Step S203, extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, where the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator;

[0098] Specifically, through the deep learning model of convolutional neural network (CNN), the two types of data sources, image information data array and spectral data array, are fused to form a new data array. Through large-scale data training and learning, the efficiency of data processing and analysis is improved, saving manpower and time costs.

[0099] Step S204, determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0100] Specifically, the two-dimensional image information and the element spectrum information are fused to form an automatic discrimination method for the pollution level by fusing multiple feature information, which saves manpower and time costs and improves the accuracy of the data.

[0101] Optionally, the infrared image information and the spectral information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, including: generating an image information data array of the insulator according to the infrared image information; generating a spectral data array of the insulator according to the spectral information; and fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, wherein the contamination data array is the insulator contamination data.

[0102] Optionally, generating an image information data array of the insulator based on the infrared image information includes: extracting features from the infrared image information to obtain grayscale image information of the insulator; and generating an image information data array of the insulator based on the grayscale image information.

[0103] Optionally, the image information data array and the spectral data array are fused through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, including: using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array to respectively obtain first characteristic data and second characteristic data; performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; using a feature hierarchical method to process the first processed data and the second processed data to obtain a first optimal characteristic subset and a second optimal characteristic subset; using two-dimensional correlation analysis to fuse the first optimal characteristic subset and the second optimal characteristic subset to obtain the contamination data array of the insulator.

[0104] Optionally, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes: performing data preprocessing on the infrared image information and the spectral information using a preprocessing algorithm, wherein the preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

[0105] Optionally, determining the pollution level of the insulator according to the pollution data of the insulator includes: combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0106] Optionally, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further includes: obtaining a target mapping relationship, which is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; determining the target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; and adjusting the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model to obtain an updated classification model, wherein the updated classification model corresponds one-to-one to the target geographical range.

[0107] An embodiment of the present invention provides a processor, and the processor is used to run a program, wherein the method for detecting the contamination degree of an insulator is executed when the program is run.

[0108] Specifically, the detection method of insulator contamination degree includes:

[0109] Step S201, using infrared imaging technology to obtain infrared image information of the insulator at the current moment;

[0110] Specifically, the wavelength of infrared imaging is an order of magnitude larger than that of visible light. Infrared thermal imaging generally works at a wavelength of 8 to 14 μm. Compared with visible light (0.4 to 0.8 μm wavelength), it has better smoke penetration ability and is not affected by strong light and glare. Compared with ultraviolet light sources, it is not easy to attenuate when propagating in the air. When interacting with pollution, it has a better penetration depth and can obtain more comprehensive pollution information.

[0111] Step S202, using laser induced breakdown spectroscopy to obtain spectrum information of the insulator, wherein the spectrum information at least includes atomic spectrum information of the insulator at the current moment;

[0112] Specifically, laser-induced breakdown spectroscopy technology is highly sensitive to the atomic spectrum information of insulators and can detect the presence of contaminants, which helps to promptly discover contamination and defects on the surface of insulators. The two-dimensional image is superimposed and fused with the spectrum of contaminant elements to form a multi-source data source, which can help to accurately judge the contamination level with the help of multi-source information.

[0113] Step S203, extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, where the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator;

[0114] Specifically, through the deep learning model of convolutional neural network (CNN), the two types of data sources, image information data array and spectral data array, are fused to form a new data array. Through large-scale data training and learning, the efficiency of data processing and analysis is improved, saving manpower and time costs.

[0115] Step S204, determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0116] Specifically, the two-dimensional image information and the element spectrum information are fused to form an automatic discrimination method for the pollution level by fusing multiple feature information, which saves manpower and time costs and improves the accuracy of the data.

[0117] Optionally, the infrared image information and the spectral information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, including: generating an image information data array of the insulator according to the infrared image information; generating a spectral data array of the insulator according to the spectral information; and fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, wherein the contamination data array is the insulator contamination data.

[0118] Optionally, generating an image information data array of the insulator based on the infrared image information includes: extracting features from the infrared image information to obtain grayscale image information of the insulator; and generating an image information data array of the insulator based on the grayscale image information.

[0119] Optionally, the image information data array and the spectral data array are fused through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, including: using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array to respectively obtain first characteristic data and second characteristic data; performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; using a feature hierarchical method to process the first processed data and the second processed data to obtain a first optimal characteristic subset and a second optimal characteristic subset; using two-dimensional correlation analysis to fuse the first optimal characteristic subset and the second optimal characteristic subset to obtain the contamination data array of the insulator.

[0120] Optionally, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes: performing data preprocessing on the infrared image information and the spectral information using a preprocessing algorithm, wherein the preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

[0121] Optionally, determining the pollution level of the insulator according to the pollution data of the insulator includes: combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0122] Optionally, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further includes: obtaining a target mapping relationship, which is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; determining the target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; and adjusting the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model to obtain an updated classification model, wherein the updated classification model corresponds one-to-one to the target geographical range.

[0123] An embodiment of the present invention provides a device, the device including a processor, a memory, and a program stored in the memory and executable on the processor, and when the processor executes the program, at least the following steps are implemented:

[0124] Step S201, using infrared imaging technology to obtain infrared image information of the insulator at the current moment;

[0125] Step S202, using laser induced breakdown spectroscopy to obtain spectrum information of the insulator, wherein the spectrum information at least includes atomic spectrum information of the insulator at the current moment;

[0126] Step S203, extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, where the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator;

[0127] Step S204, determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0128] The devices in this article can be servers, PCs, PADs, mobile phones, etc.

[0129] Optionally, the infrared image information and the spectral information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, including: generating an image information data array of the insulator according to the infrared image information; generating a spectral data array of the insulator according to the spectral information; and fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, wherein the contamination data array is the insulator contamination data.

[0130] Optionally, generating an image information data array of the insulator based on the infrared image information includes: extracting features from the infrared image information to obtain grayscale image information of the insulator; and generating an image information data array of the insulator based on the grayscale image information.

[0131] Optionally, the image information data array and the spectral data array are fused through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, including: using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array to respectively obtain first characteristic data and second characteristic data; performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; using a feature hierarchical method to process the first processed data and the second processed data to obtain a first optimal characteristic subset and a second optimal characteristic subset; using two-dimensional correlation analysis to fuse the first optimal characteristic subset and the second optimal characteristic subset to obtain the contamination data array of the insulator.

[0132] Optionally, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes: performing data preprocessing on the infrared image information and the spectral information using a preprocessing algorithm, wherein the preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

[0133] Optionally, determining the pollution level of the insulator according to the pollution data of the insulator includes: combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0134] Optionally, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further includes: obtaining a target mapping relationship, which is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; determining the target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; and adjusting the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model to obtain an updated classification model, wherein the updated classification model corresponds one-to-one to the target geographical range.

[0135] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing a program for initializing at least the following method steps:

[0136] Step S201, using infrared imaging technology to obtain infrared image information of the insulator at the current moment;

[0137] Step S202, using laser induced breakdown spectroscopy to obtain spectrum information of the insulator, wherein the spectrum information at least includes atomic spectrum information of the insulator at the current moment;

[0138] Step S203, extracting and fusing the infrared image information and the spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, where the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator;

[0139] Step S204, determining the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

[0140] Optionally, the infrared image information and the spectral information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, including: generating an image information data array of the insulator according to the infrared image information; generating a spectral data array of the insulator according to the spectral information; and fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain a contamination data array of the insulator, wherein the contamination data array is the insulator contamination data.

[0141] Optionally, generating an image information data array of the insulator based on the infrared image information includes: extracting features from the infrared image information to obtain grayscale image information of the insulator; and generating an image information data array of the insulator based on the grayscale image information.

[0142] Optionally, the image information data array and the spectral data array are fused through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, including: using a continuous projection algorithm and an uninformative variable removal algorithm to respectively extract characteristic bands in the image information data array and characteristic bands in the spectral data array to respectively obtain first characteristic data and second characteristic data; performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; using a feature hierarchical method to process the first processed data and the second processed data to obtain a first optimal characteristic subset and a second optimal characteristic subset; using two-dimensional correlation analysis to fuse the first optimal characteristic subset and the second optimal characteristic subset to obtain the contamination data array of the insulator.

[0143] Optionally, before fusing the image information data array and the spectral data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes: performing data preprocessing on the infrared image information and the spectral information using a preprocessing algorithm, wherein the preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

[0144] Optionally, determining the pollution level of the insulator according to the pollution data of the insulator includes: combining a deep learning algorithm and adopting an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; inputting the pollution data array of the insulator into the fusion classification model, and obtaining the output of the fusion classification model as the pollution level of the insulator.

[0145] Optionally, after the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further includes: obtaining a target mapping relationship, which is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; determining the target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; and adjusting the operating parameters of the fusion classification model to the target operating parameters based on the model transfer learning capability of the fusion classification model to obtain an updated classification model, wherein the updated classification model corresponds one-to-one to the target geographical range.

[0146] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0147] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0148] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes 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 generate 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 device that provides the functions specified in a block or multiple blocks.

[0149] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0150] 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 for the functions specified in one or more boxes.

[0151] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0152] The memory may include non-permanent memory in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0153] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0154] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0155] From the above description, it can be seen that the above embodiments of the present application achieve the following technical effects:

[0156] 1) The above-mentioned insulator contamination degree detection method based on infrared image and LIBS spectrum multi-source information fusion of the present application first uses infrared imaging technology to obtain the infrared image information of the insulator at the current moment; then uses laser induced breakdown spectroscopy technology to obtain the spectral information of the insulator, and the spectral information at least includes the atomic spectral information of the insulator at the current moment; then, through the deep learning model of the convolutional neural network, the infrared image information and the spectral information are extracted and fused to obtain the insulator contamination data, and the insulator contamination data includes the grayscale data of the insulator and the atomic data of the insulator; finally, the contamination level of the insulator is determined according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator. The method can fuse two types of information, namely, two-dimensional image information and element spectrum information, and more accurately classify the pollution level from the two types of characteristics, namely, the physical information of the image and the chemical composition of the elements. By superimposing and fusing the two-dimensional image and the pollution element spectrum, a multi-source data source is formed, and accurate judgment of the pollution level is achieved with the help of multi-source information. This solves the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results.

[0157] 2) The above-mentioned insulator contamination degree detection device based on infrared image and LIBS spectrum multi-source information fusion of the present application includes: a first acquisition unit, a second acquisition unit, a fusion unit and a determination unit, the first acquisition unit is used to use infrared imaging technology to obtain infrared image information of the insulator at the current moment; the second acquisition unit is used to use laser induced breakdown spectroscopy technology to obtain spectral information of the insulator, and the spectral information at least includes atomic spectral information of the insulator at the current moment; the fusion unit is used to extract and fuse the infrared image information and spectral information through a deep learning model of a convolutional neural network to obtain insulator contamination data, and the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; the determination unit is used to determine the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator. The method can fuse two types of information, namely, two-dimensional image information and element spectrum information, and more accurately classify the pollution level from the two types of characteristics, namely, the physical information of the image and the chemical composition of the elements. By superimposing and fusing the two-dimensional image and the pollution element spectrum, a multi-source data source is formed, and accurate judgment of the pollution level is achieved with the help of multi-source information. This solves the problem that the method of detecting the pollution level of insulators using infrared imaging technology in the prior art is greatly affected by weather, lacks comprehensive judgment, and has low accuracy of judgment results.

[0158] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for detecting insulator contamination based on multi-source information fusion of infrared image and LIBS spectrum, characterized in that: include: Use infrared imaging technology to obtain the infrared image information of the insulator at the current moment; Using laser induced breakdown spectroscopy technology to obtain spectrum information of the insulator, the spectrum information at least includes atomic spectrum information of the insulator at the current moment; The infrared image information and the spectrum information are extracted and fused through a deep learning model of a convolutional neural network to obtain insulator contamination data, wherein the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; The contamination level of the insulator is determined according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

2. The method according to claim 1, characterized in that Through the deep learning model of the convolutional neural network, the infrared image information and the spectrum information are extracted and fused to obtain insulator contamination data, including: Generate an image information data array of the insulator according to the infrared image information; generating a spectrum data array of the insulator according to the spectrum information; Through the deep learning model of the convolutional neural network, the image information data array and the spectral data array are fused to obtain the contamination data array of the insulator, and the contamination data array is the insulator contamination data.

3. The method according to claim 2, characterized in that Generating an image information data array of the insulator according to the infrared image information includes: Extracting features from the infrared image information to obtain grayscale image information of the insulator; An image information data array of the insulator is generated according to the grayscale image information.

4. The method according to claim 2, characterized in that: By using a deep learning model of a convolutional neural network, the image information data array and the spectrum data array are fused to obtain a contamination data array of the insulator, including: Using a continuous projection algorithm and an uninformative variable removal algorithm, respectively extracting characteristic bands in the image information data array and characteristic bands in the spectral data array, to obtain first characteristic data and second characteristic data, respectively; Performing three-dimensional wavelet decomposition on the first characteristic data and the second characteristic data to obtain first processed data and second processed data; Using a feature hierarchical method to process the first processed data and the second processed data respectively to obtain a first optimal feature subset and a second optimal feature subset; The first optimal feature subset and the second optimal feature subset are fused by using two-dimensional correlation analysis to obtain the contamination data array of the insulator.

5. The method according to claim 2, characterized in that: Before fusing the image information data array and the spectrum data array through a deep learning model of a convolutional neural network to obtain the contamination data array of the insulator, the method further includes: A preprocessing algorithm is used to perform data preprocessing on the infrared image information and the spectral information. The preprocessing algorithm includes at least one of the following: a smoothing algorithm, a variable normalization algorithm, a multivariate scattering correction algorithm, and a multi-order derivative algorithm.

6. The method according to claim 1, characterized in that Determining the contamination level of the insulator according to the contamination data of the insulator includes: Combined with a deep learning algorithm, and using an optimal weighted fusion strategy to establish a fusion classification model based on a learner, the deep learning algorithm includes at least one of the following: a support vector machine, a random forest algorithm, and a convolutional neural network; The pollution data array of the insulator is input into the fusion classification model, and the output of the fusion classification model is the pollution level of the insulator.

7. The method according to claim 6, characterized in that After the pollution data array of the insulator is input into the fusion classification model and the output of the fusion classification model is obtained as the pollution level of the insulator, the method further includes: Acquire a target mapping relationship, where the target mapping relationship is a mapping relationship between the geographical range of the insulator and the operating parameters of the fusion classification model; Determining target operating parameters of the fusion classification model according to the target geographical range of the insulator and the target mapping relationship; Based on the model transfer learning capability of the fusion classification model, the operating parameters of the fusion classification model are adjusted to the target operating parameters to obtain an updated classification model, and the updated classification model corresponds one-to-one to the target geographical range.

8. A device for detecting the degree of insulator contamination based on the fusion of infrared image and LIBS spectrum multi-source information, characterized in that: include: A first acquisition unit is used to acquire infrared image information of the insulator at the current moment by using infrared imaging technology; A second acquisition unit is used to acquire spectrum information of the insulator by using laser induced breakdown spectroscopy technology, wherein the spectrum information at least includes atomic spectrum information of the insulator at a current moment; A fusion unit, used for extracting and fusing the infrared image information and the spectrum information through a deep learning model of a convolutional neural network to obtain insulator contamination data, wherein the insulator contamination data includes grayscale data of the insulator and atomic data of the insulator; A determination unit is used to determine the contamination level of the insulator according to the contamination data of the insulator, wherein the contamination degree of the insulator is proportional to the contamination level of the insulator.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein when the program is executed, the device where the computer-readable storage medium is located is controlled to execute the insulator contamination degree detection method based on infrared image and LIBS spectrum multi-source information fusion as described in any one of claims 1 to 7.

10. An electronic device, characterized in that: include: One or more processors, a memory, and one or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include a method for executing the insulator contamination detection method based on multi-source information fusion of infrared images and LIBS spectra as described in any one of claims 1 to 7.

Citation Information

Patent Citations

  • Insulator equivalent salt deposit density detection method cooperating with hyperspectral and infrared technologies

    CN112414950A

  • Insulator pollution degree detection method based on multi-source spectrum sensing technology

    CN115047010A

  • Measurement method of fouling of insulator, measurement device and measurement program

    JP2017198593A

Cited By

  • Waste lithium battery recycling method based on multi-stage crushing and intelligent sorting

    CN119560669A

  • Mineral oxygen element analysis method based on multi-mode LIBS

    CN120084777A

  • Insulator pollution degree detection method and system based on multi-source spectrum sensing

    CN120495303A

  • A method and system for detecting the pollution level of insulators based on multi-source spectral sensing

    CN120495303B