Nonmetal electric energy metering box classification method and system
Through the combination of laser induced breakdown spectroscopy technology and classification model, the problems of low detection efficiency and low accuracy of the electrical energy metering box are solved, and the rapid and accurate identification of non-metallic electrical energy metering box is achieved, which improves detection efficiency and accuracy, and ensures that The reliability and safety of the power grid.
Patent Information
- Application Number
- CN202410857260.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2024-11-22
AI Technical Summary
The product quality of the electric energy metering box manufacturers is uneven, the traditional detection methods are inefficient and the accuracy is not high, and the effective technology for the grid inspection and outdoor survey of the electric energy metering box is difficult to ensure product quality.
Laser-induced breakdown spectroscopy technology is used to obtain the spectral data of the electrical energy metering box samples. After pre-processing and dimensionality reduction processing, the material is classified based on classification models such as support vector machines and self-organized mapping neural networks to achieve rapid non-metallic electrical energy metering box. Accurate identification.
It improves the accuracy and efficiency of detection, reduces power outage time, saves manpower and material resources, ensures the reliability and safety of power supply in the power grid, and reduces the economic losses of enterprises.
Smart Images

Figure CN119007879A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electric energy meter box detection, and in particular to a classification method and system for non-metallic electric energy meter boxes. Background Art
[0002] The electric energy meter box is the first layer of protection for metering equipment and key components such as smart electric energy meters, collection terminals and circuit breakers, and is also an important IoT sensing node that supports lean operation and maintenance of substations. At present, there are a large number of electric energy meter boxes in operation, and the operating conditions vary greatly. Their quality is directly related to the safe and reliable operation of the power grid, the high-quality and efficient electricity services, and the safety and stability of the economy and society. The two typical non-metallic materials used in non-metallic electric energy meter boxes are PC+ABS (flame retardant) (polycarbonate resin + acrylonitrile-butadiene-styrene resin) and SMC (glass reinforced plastic) (glass fiber reinforced unsaturated polyester molding compound), which are organic polymer compounds with similar elemental composition, good insulation performance, chemical corrosion resistance, long service life, and various installation forms and orientations. However, the quality of products produced by manufacturers of electric energy meter boxes on the market varies. Different manufacturers have different material selections, and the manufacturing process and quality of the same material are also different. This has caused the overall product quality to be unstable. There are even problems such as inferior materials for meter boxes (plastic materials are mostly recycled materials) and cutting corners. When actually connected to the network, it takes a lot of effort and cost to conduct inspections to ensure that the product quality meets the requirements; when inspecting whether the electric energy meter box needs to be replaced on site, the traditional manual inspection method has defects, and the results are greatly affected by the subjective judgment of the inspectors, and it is impossible to judge the material quality from a quantitative perspective. This not only makes it difficult to meet the requirements for enterprise asset life cycle management, but also brings considerable economic losses to the enterprise.
[0003] At present, the detection technology of electricity meter boxes is inefficient and has low accuracy. There is no detection technology for grid access detection and outdoor survey of electricity meter boxes, making it difficult to ensure product quality. Summary of the invention
[0004] The purpose of the present invention is to provide a non-metallic electric energy meter box classification method in order to address the problems that the product quality of current electric energy meter box manufacturers is uneven, the currently commonly used detection methods are inefficient and have low accuracy, and there is no detection technology for electric energy meter box access detection and outdoor survey, thereby realizing the rapid classification of non-metallic electric energy meter box materials.
[0005] The purpose of the present invention can be achieved by the following technical solutions:
[0006] A method for classifying non-metallic electric energy meter boxes comprises the following steps:
[0007] The spectral data of the electric energy meter box samples were obtained by using laser induced breakdown spectroscopy technology;
[0008] Preprocessing and dimension reduction processing are performed on the collected spectral data;
[0009] Based on the principal component spectral data after dimensionality reduction, the classification results of the electric energy meter box samples are obtained through the classification model.
[0010] Furthermore, the preprocessing includes discrete wavelet transform, baseline correction and noise reduction processing.
[0011] Furthermore, in the discrete wavelet transform, a heuristic wavelet threshold function method is used to perform threshold processing on the wavelet coefficients to screen the wavelet coefficients.
[0012] Furthermore, the dimension reduction process is implemented by principal component analysis.
[0013] Furthermore, the classification model is a model constructed based on a support vector machine, a self-organizing map neural network, naive Bayes, a random forest or a decision tree.
[0014] The present invention also provides a non-metallic electric energy meter box classification system, comprising:
[0015] Spectral data acquisition equipment, used to obtain spectral data of electric energy meter box samples using laser induced breakdown spectroscopy technology;
[0016] Spectral data processing equipment, used for preprocessing and dimensionality reduction of the collected spectral data;
[0017] The meter box classification device is used to obtain the category results of the electric energy meter box samples through the classification model based on the principal component spectrum data after dimensionality reduction processing.
[0018] Furthermore, the preprocessing includes discrete wavelet transform, baseline correction and noise reduction processing.
[0019] Furthermore, in the discrete wavelet transform, a heuristic wavelet threshold function method is used to perform threshold processing on the wavelet coefficients to screen the wavelet coefficients.
[0020] Furthermore, the dimension reduction process is implemented by principal component analysis.
[0021] Furthermore, the classification model is a model constructed based on a support vector machine, a self-organizing map neural network, naive Bayes, a random forest or a decision tree.
[0022] Compared with the prior art, the present invention has the following beneficial effects:
[0023] 1. The present invention adopts laser induced breakdown spectroscopy technology to classify non-metallic electric energy meter boxes, which causes almost no damage to the equipment and can quickly determine the material and state of the electric energy meter box, effectively improving the accuracy and efficiency of detection.
[0024] 2. The present invention adopts laser induced breakdown spectroscopy technology, which is convenient for network access detection and maintenance personnel to carry out repairs, reduces power outage time, saves manpower and material resources, ensures the safety of staff, and has important engineering significance and economic benefits for improving the reliability of power supply of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS
[0025] Figure 1 It is a principle block diagram of the present invention;
[0026] Figure 2 It is the structural diagram of LIBS detection system;
[0027] Figure 3 The linear SVM classification results of SMC with different formulations are shown in Figure 2.
[0028] Figure 4 The SOM neural network clustering results of SMC with different formulations are shown in Figure 4, where (4a) is the SOM neighbor weight distance and (4b) is the hit number;
[0029] Figure 5 The linear SVM classification results of PC+ABS with different formulations are shown below;
[0030] Figure 6 The SOM neural network clustering results of different formulations of PC+ABS are shown in Figure 4, where (4a) is the SOM neighbor weight distance and (4b) is the hit number. DETAILED DESCRIPTION
[0031] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.
[0032] Abbreviations involved:
[0033] Laser-induced breakdown spectroscopy, LIBS
[0034] Support Vector Machine: Support Vector Machine, SVM
[0035] Self-organized Map: Self-organized Map, SOM
[0036] Glass fiber reinforced plastic: SMC
[0037] Polycarbonate resin + acrylonitrile butadiene styrene resin: PC + ABS
[0038] Discrete Wavelet Transform: DWT
[0039] The invention provides a classification method for non-metallic electric energy meter boxes. The method utilizes laser-induced breakdown spectroscopy (LIBS) technology. A laser emits high-energy laser to reach the surface of a sample, and the sample is ablated and excited to form plasma. The atomic emission spectrum of the plasma is collected by a spectrometer to identify the elemental components of the sample, thereby performing material identification, classification, qualitative and quantitative analysis.
[0040] Laser induced breakdown spectroscopy is an atomic emission spectroscopy measurement technology with the advantages of fast measurement speed, online detection, high sensitivity, simultaneous analysis of multiple elements, and near non-destructiveness. Its principle is to focus the laser energy on the surface of the material through the optical path, ablate the surface of the material to vaporize the trace sample to form a high-temperature, high-electron-density plasma, and radiate a large number of atomic and ion linear spectra during the cooling process. The spectral signal is collected through the optical fiber, and then the elemental composition and state characteristics of the ablated material are analyzed based on the collected characteristic spectral information. Based on the above-mentioned laser induced breakdown spectroscopy technology, the present invention can quickly and accurately realize the online classification and identification of non-metallic electric energy metering boxes without causing any damage to the equipment.
[0041] Example 1
[0042] The present embodiment provides a method for classifying non-metallic electricity meter boxes, comprising the following steps: obtaining spectral data of electricity meter box samples using laser induced breakdown spectroscopy technology; preprocessing and dimensionality reduction processing the collected spectral data; and obtaining the category results of the electricity meter box samples through a classification model based on the principal component spectral data after dimensionality reduction processing.
[0043] (1) Construction of LIBS detection system
[0044] In this embodiment, the spectrum data of the electric energy meter box sample is collected by the LIBS detection system. Figure 2As shown, the LIBS detection system used in this embodiment includes a laser emission system, an optical path system, a delay controller 3, a spectrometer 4 and a computer 5, wherein the laser emission system includes a power supply unit 1 and a Nd:YAG laser 2, the optical path system includes an optical reflector 6, a focusing lens 7, etc., and the spectrometer is provided with a collection probe. The pulsed laser emitted by the laser bombards the electric energy meter box sample 8 after being focused by the optical lens, and a small part of the sample is heated, ablated, dissociated, excited, and ionized, and finally forms a plasma that is approximately electrically neutral. After the pulsed laser stops, the plasma gradually cools down and expands, and the atoms and ions in the excited state in the plasma transition from a high energy level to a low energy level, emitting photons of a specific frequency, forming a spectrum containing specific information, and the wavelength of the spectrum contains the element type information of the constituent substance, and the intensity of the spectrum contains the information of the content of the substance component. Under the control of the delay controller, the spectrometer collects the spectral information emitted by the plasma through the collection probe, and finally transmits the data to the computer for data analysis, and the qualitative and quantitative analysis of the sample to be tested can be achieved based on these data. By selecting appropriate laser energy, light collection angle and spectrometer delay time, a spectral signal with high signal-to-noise ratio and signal-to-background ratio can be obtained.
[0045] (2) Spectral data processing and element identification
[0046] In this embodiment, discrete wavelet transform (DWT) is used to process the spectral data, and baseline correction and noise reduction are completed at the same time. Specifically, the wavelet transform parameters selected are as follows: the wavelet base is db5, the wavelet decomposition layer number is 5 layers, and the heuristic wavelet threshold (Heursure) function method is selected to perform threshold processing on the wavelet coefficient. The wavelet coefficient is screened by the threshold segmentation method to achieve the removal of high-frequency noise. Finally, the wavelet coefficient after correction is used to perform wavelet reconstruction to obtain the processed spectral data. The ridge method based on continuous wavelet transform is selected to screen the spectral peaks. Combined with the range of light intensity detected by the spectrometer, the spectral peaks with relatively large spectral line intensity are selected in the case of clarifying the approximate element types of the sample. The information of the standard spectral lines of the atomic spectral database of the National Institute of Standards and Technology (National Institute of Standards and Technology, NIST) (including element types and wavelengths, spectral line transition probability, upper and lower energy level differences, etc.) is referred to. The resolution of the spectrometer (there is a drift of the center wavelength of the spectral line) is comprehensively considered, and the spectral peaks and corresponding intensities of different elements are finally determined.
[0047] (3) Classification of non-metallic electricity meter boxes
[0048] The spectral data after wavelet transform is processed by principal component analysis to reduce the dimension of the data, and the principal components and their contribution rate and cumulative contribution rate to the data variance are obtained, and the principal components are used as the input data of the classification model.
[0049] In this embodiment, reference Figure 1 As shown in the figure, support vector machine (SVM) and self-organized map (SOM) neural network were used to classify the samples: three SMC materials with different formulations were qualitatively classified, supervised classification was performed using linear SVM, and five-fold cross validation was performed, with an accuracy rate of 99.9%, as shown in the figure. Figure 3 As shown in Figure 2, using the unsupervised classification of the SOM neural network, the classification accuracy of the three samples exceeded 90%. Figure 4 As shown in the figure, three PC+ABS materials with different formulations were qualitatively classified, supervised classification was performed using linear SVM, and five-fold cross validation was performed, with an accuracy rate of 98.4%. Figure 5 As shown in Figure 2, using the unsupervised classification of the SOM neural network, the classification accuracy of the three samples is about 90%. Figure 6 As shown. Figure 3 , Figure 4 , Figure 5 , Figure 6 As shown, the classification accuracy of the classification model for the samples reaches more than 90%, indicating that the classification method of the present invention has a high classification accuracy for non-metallic electric energy meter boxes.
[0050] Example 2
[0051] This embodiment provides a non-metallic electric energy meter box classification system, including a spectral data acquisition device, a spectral data processing device and a meter box finalization classification device, wherein the spectral data acquisition device is used to obtain spectral data of electric energy meter box samples using laser induced breakdown spectroscopy technology. The specific structure of the spectral data acquisition device is referred to Figure 2 As shown; the spectral data processing equipment is used to preprocess and reduce the dimension of the collected spectral data; the meter box classification equipment is used to obtain the category results of the electric energy meter box samples through the classification model based on the principal component spectral data after the dimensionality reduction processing.
[0052] The rest is the same as in Example 1.
[0053] The preferred specific embodiments of the present invention are described in detail above. It should be understood that a person skilled in the art can make many modifications and changes based on the concept of the present invention without creative work. Therefore, any technical solution that can be obtained by a person skilled in the art through logical analysis, reasoning or limited experiments based on the concept of the present invention on the basis of the prior art should be within the scope of protection determined by the claims.
Claims
1. A classification method for non-metallic electric energy meter boxes, It is characterized in that The following steps are involved: The spectral data of the electric energy meter box samples were obtained by using laser induced breakdown spectroscopy technology; Preprocessing and dimension reduction processing are performed on the collected spectral data; Based on the principal component spectral data after dimensionality reduction, the classification results of the electric energy meter box samples are obtained through the classification model.
2. The non-metallic electric energy meter box classification method according to claim 1, It is characterized in that The preprocessing includes discrete wavelet transform, baseline correction and noise reduction.
3. The non-metallic electric energy meter box classification method according to claim 2, It is characterized in that In the discrete wavelet transform, a heuristic wavelet threshold function method is used to perform threshold processing on wavelet coefficients to screen the wavelet coefficients.
4. The non-metallic electric energy meter box classification method according to claim 1, It is characterized in that The dimensionality reduction process is achieved by principal component analysis.
5. The non-metallic electric energy meter box classification method according to claim 1, It is characterized in that The classification model is a model constructed based on support vector machine, self-organizing map neural network, naive Bayes, random forest or decision tree.
6. A classification system for non-metallic electric energy meter boxes, It is characterized in that include: Spectral data acquisition equipment, used to obtain spectral data of electric energy meter box samples using laser induced breakdown spectroscopy technology; Spectral data processing equipment, used for preprocessing and dimensionality reduction of the collected spectral data; The meter box classification device is used to obtain the category results of the electric energy meter box samples through the classification model based on the principal component spectrum data after dimensionality reduction processing.
7. The non-metallic electric energy meter box classification system according to claim 6, It is characterized in that The preprocessing includes discrete wavelet transform, baseline correction and noise reduction.
8. The non-metallic electric energy meter box classification system according to claim 7, It is characterized in that In the discrete wavelet transform, a heuristic wavelet threshold function method is used to perform threshold processing on wavelet coefficients to screen the wavelet coefficients.
9. The non-metallic electric energy meter box classification system according to claim 6, It is characterized in that The dimensionality reduction process is achieved by principal component analysis.
10. The non-metallic electric energy meter box classification system according to claim 6, It is characterized in that The classification model is a model constructed based on support vector machine, self-organizing map neural network, naive Bayes, random forest or decision tree.
