Multi-dimensional parameter calibration method for infrared temperature measurement of composite insulator

Abstract

The invention discloses a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator, and relates to the technical field of infrared temperature measurement, and the method comprises the steps: obtaining the infrared temperature measurement results of an infrared thermodetector and an infrared thermal imager for the composite insulator at different distances and visual angles; making an infrared temperature measurement data set based on the infrared temperature measurement result, and dividing the infrared temperature measurement data set into a training set and a verification set; training the infrared temperature measurement data set based on a limited learning machine to obtain an ELM model; and correcting the infrared temperature measurement result by using the ELM model to obtain a corrected result of the infrared temperature measurement result. The invention aims to eliminate the distance and visual angle influence in the temperature measurement process, improve the accuracy of infrared temperature measurement, not only can significantly improve the measurement precision, but also provides reliable technical support for insulator inspection of a power transmission line, and has important application value and wide market prospect.

Description

Technical Field

[0001] The present application relates to the field of infrared temperature measurement technology, and in particular to a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator. Background Art

[0002] For a long time, composite insulators of transmission lines have been exposed to complex outdoor environments and are susceptible to a variety of electrical, mechanical and environmental factors, making them prone to aging and deterioration. This aging process may cause partial discharge and heat accumulation inside the sheath, which in turn may lead to further aging of the core rod and reduce the mechanical strength of the insulator. Under extreme working conditions, these problems may even lead to serious faults such as string breakage or line drop, posing a major threat to the safe and stable operation of the power system.

[0003] In order to improve the efficiency and accuracy of inspections, drones equipped with infrared imagers have been widely used in transmission line inspections in recent years. However, during the shooting process, the accuracy of infrared temperature measurement is often significantly affected due to changes in measurement distance and viewing angle as well as interference from the outdoor environment, which makes it difficult to accurately obtain the true temperature of composite insulators, thus affecting the judgment and evaluation of insulator faults.

[0004] In recent years, researchers at home and abroad have conducted a lot of research on the accuracy correction of infrared temperature measurement, with the aim of improving the accuracy and reliability of measurement. The methods used include advanced infrared imaging technology, optimized image processing algorithms, and improved temperature compensation models. Although some progress has been made, there are still some shortcomings. The main problems include: the accuracy of temperature measurement is significantly affected by environmental factors, especially the changes in measurement angle and distance have caused great interference to the accuracy of temperature measurement; the existing calibration methods are usually complex and have limited applicability, and it is difficult to maintain consistent measurement results under different environmental conditions; in addition, real-time and automation levels still need to be further improved. These problems limit the widespread application of infrared temperature measurement technology in composite insulator detection, so more in-depth research and optimization are urgently needed.

[0005] Therefore, in order to meet actual needs, a composite insulator infrared temperature measurement distance viewing angle compensation calibration technology is provided. Summary of the invention

[0006] In view of the defects existing in the prior art, the purpose of this application is to provide a multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators, aiming to eliminate the influence of distance and viewing angle in the temperature measurement process and improve the accuracy of infrared temperature measurement. It can not only significantly improve the measurement accuracy, but also provide reliable technical support for insulator inspection of transmission lines. It has important application value and broad market prospects.

[0007] In order to achieve the above objectives, the technical solution adopted by this application is:

[0008] The present application provides a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator, the method comprising the following steps:

[0009] Obtain infrared temperature measurement results of composite insulators using infrared thermometers and infrared thermal imagers at different distances and viewing angles;

[0010] Based on the infrared temperature measurement results, an infrared temperature measurement data set is prepared and divided into a training set and a validation set;

[0011] The infrared temperature measurement data set is trained based on an ELM model.

[0012] The infrared temperature measurement result is corrected by using the ELM model to obtain a corrected result of the infrared temperature measurement result.

[0013] On the basis of the above technical solution, the method further comprises the following steps:

[0014] Based on the actual temperature result obtained by actual measurement of the composite insulator, a comparison is made with the correction result of the infrared temperature measurement result to determine whether the ELM model is qualified.

[0015] On the basis of the above technical solution, obtaining infrared temperature measurement results of the composite insulator at different distances and viewing angles by an infrared thermometer and an infrared thermal imager includes the following steps:

[0016] Suspension composite insulators;

[0017] Three temperature probes are set on the surface of the composite insulator, and another temperature probe is set to monitor the ambient temperature;

[0018] A voltage is applied to the high voltage end at the bottom of the composite insulator, and the temperature of the composite insulator is measured.

[0019] On the basis of the above technical solution, the infrared temperature measurement data set is trained based on a limited learning machine to obtain an ELM model, including the following steps:

[0020] Determine the number of layers, number of hidden neurons, and activation function of the ELM neural network;

[0021] Randomly generate the hidden layer weights and bias vectors of the ELM neural network, set the initial position and velocity of each particle, and randomly generate the weight matrix and bias vector of the hidden layer nodes;

[0022] The infrared temperature measurement value, viewing angle and distance in the infrared temperature measurement data set are passed into the ELM hidden layer, and the hidden layer output matrix and the weight of the output layer are calculated;

[0023] According to the hidden layer output matrix and the weight of the output layer, the prediction output is performed. After obtaining the result, the error between the prediction result and the actual target value is calculated to construct the loss function;

[0024] Based on the loss function, error correction is performed to obtain the ELM model.

[0025] On the basis of the above technical solution, the activation function is:

[0026] a l =σ(W l a l―1 +b l );in,

[0027] a l is the activation value, W l is the weight matrix, a l―1 is the activation value of the previous layer, b l is the bias vector.

[0028] Based on the above technical solution, the hidden layer output matrix is:

[0029] H=σ(W1X+b1); where

[0030] W1 is the generated hidden layer weight, b1 is the bias vector corresponding to the generated hidden layer, X is the input data, and σ is the activation function.

[0031] Based on the above technical solution, the loss function is:

[0032] in,

[0033] y i is the actual value, y′ i is the predicted value, and n is the number of samples.

[0034] Compared with the prior art, the advantages of this application are:

[0035] This application aims to eliminate the influence of distance and viewing angle in the temperature measurement process and improve the accuracy of infrared temperature measurement. It can not only significantly improve the measurement accuracy, but also provide reliable technical support for insulator inspection of transmission lines. It has important application value and broad market prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0037] Figure 1 A flow chart of the principle steps of a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0038] Figure 2 It is a wiring schematic diagram of an infrared temperature measurement test of a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0039] Figure 3 The infrared thermometer shooting posture diagram corresponding to the multi-dimensional parameter calibration method of infrared temperature measurement of composite insulators in the embodiment of the present application;

[0040] Figure 4 A schematic diagram of the temperature measurement results of an infrared thermometer corresponding to the multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0041] Figure 5 A schematic diagram of the temperature measurement results of an infrared thermal imager corresponding to the multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0042] Figure 6 A model iteration error curve diagram corresponding to the multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0043] Figure 7 A comparison diagram of correction effects corresponding to the multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to an embodiment of the present application;

[0044] Figure 8 It is a schematic diagram of the correction result of the traditional fitting model of the infrared thermometer corresponding to the multi-dimensional parameter calibration method of the composite insulator infrared temperature measurement according to the embodiment of the present application;

[0045] Fig. 9 A schematic diagram of a correction result of an infrared thermal imager machine learning model corresponding to a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0046] Fig.10 A schematic diagram of a correction result of a traditional fitting model of an infrared thermal imager corresponding to a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator according to an embodiment of the present application;

[0047] Fig.11 This is a schematic diagram of the correction results of the infrared thermal imager machine learning model corresponding to the multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators in an embodiment of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, 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 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 making creative work are within the scope of protection of this application.

[0049] The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0050] The embodiment of the present application provides a multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators, which aims to eliminate the influence of distance and viewing angle in the temperature measurement process and improve the accuracy of infrared temperature measurement. It can not only significantly improve the measurement accuracy, but also provide reliable technical support for insulator inspection of transmission lines. It has important application value and broad market prospects.

[0051] In order to achieve the above technical effects, the overall idea of ​​this application is as follows:

[0052] A multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator, the method comprising the following steps:

[0053] S1. Obtain infrared temperature measurement results of composite insulators at different distances and viewing angles using an infrared thermometer and an infrared thermal imager;

[0054] S2. Based on the infrared temperature measurement results, an infrared temperature measurement data set is prepared and divided into a training set and a validation set;

[0055] S3, training the infrared temperature measurement data set based on an ELM model;

[0056] S4. Correct the infrared temperature measurement result by using the ELM model to obtain a corrected infrared temperature measurement result.

[0057] The embodiments of the present application are further described in detail below in conjunction with the accompanying drawings.

[0058] See also Figures 1 to 11 As shown, the embodiment of the present application provides a multi-dimensional parameter calibration method for infrared temperature measurement of a composite insulator, the method comprising the following steps:

[0059] S1. Obtain infrared temperature measurement results of composite insulators at different distances and viewing angles using an infrared thermometer and an infrared thermal imager;

[0060] S2. Based on the infrared temperature measurement results, an infrared temperature measurement data set is prepared and divided into a training set and a validation set;

[0061] S3, training the infrared temperature measurement data set based on an ELM model;

[0062] S4. Correct the infrared temperature measurement result by using the ELM model to obtain a corrected infrared temperature measurement result.

[0063] In the embodiments of the present application, the aim is to eliminate the influence of distance and viewing angle in the temperature measurement process and improve the accuracy of infrared temperature measurement. It can not only significantly improve the measurement accuracy, but also provide reliable technical support for insulator inspection of transmission lines, which has important application value and broad market prospects.

[0064] Further, infrared temperature measurement results of the composite insulator at different distances and viewing angles are obtained by using an infrared thermometer and an infrared thermal imager;

[0065] Based on the infrared temperature measurement results, an infrared temperature measurement data set is prepared and divided into a training set and a validation set;

[0066] The infrared temperature measurement data set is trained based on an ELM model.

[0067] The infrared temperature measurement result is corrected by using the ELM model to obtain a corrected result of the infrared temperature measurement result.

[0068] Furthermore, the method further comprises the following steps:

[0069] Based on the actual temperature result obtained by actual measurement of the composite insulator, a comparison is made with the correction result of the infrared temperature measurement result to determine whether the ELM model is qualified.

[0070] Further, obtaining infrared temperature measurement results of the composite insulator at different distances and viewing angles using an infrared thermometer and an infrared thermal imager includes the following steps:

[0071] Suspension composite insulators;

[0072] Three temperature probes are set on the surface of the composite insulator, and another temperature probe is set to monitor the ambient temperature;

[0073] A voltage is applied to the high voltage end at the bottom of the composite insulator, and the temperature of the composite insulator is measured.

[0074] Further, the infrared temperature measurement data set is trained based on an ELM model, and the ELM model is obtained, including the following steps:

[0075] Determine the number of layers, number of hidden neurons, and activation function of the ELM neural network;

[0076] Randomly generate the hidden layer weights and bias vectors of the ELM neural network, set the initial position and velocity of each particle, and randomly generate the weight matrix and bias vector of the hidden layer nodes;

[0077] The infrared temperature measurement value, viewing angle and distance in the infrared temperature measurement data set are passed into the ELM hidden layer, and the hidden layer output matrix and the weight of the output layer are calculated;

[0078] According to the hidden layer output matrix and the weight of the output layer, the prediction output is performed. After obtaining the result, the error between the prediction result and the actual target value is calculated to construct the loss function;

[0079] Based on the loss function, error correction is performed to obtain the ELM model.

[0080] Furthermore, the activation function is:

[0081] a l =σ(W l a l―1 +b l );in,

[0082] a l is the activation value, W l is the weight matrix, a l―1 is the activation value of the previous layer, b l is the bias vector.

[0083] Furthermore, the hidden layer output matrix is:

[0084] H=σ(W1X+b1); where

[0085] W1 is the generated hidden layer weight, b1 is the bias vector corresponding to the generated hidden layer, X is the input data, and σ is the activation function.

[0086] Furthermore, the loss function is:

[0087] in,

[0088] y i is the actual value, y′ i is the predicted value, and n is the number of samples.

[0089] In order to solve the problem that the temperature measurement accuracy of composite insulators decreases due to changes in distance, viewing angle and environmental factors during infrared temperature measurement, an embodiment of the present application proposes an infrared temperature measurement compensation and calibration method based on the extreme learning machine algorithm (ELM). By introducing the ELM model, the infrared temperature measurement data can be accurately compensated and calibrated under different measurement distances and viewing angles. This method not only simplifies the complex calibration process, but also significantly improves the accuracy and reliability of temperature measurement. In actual drone infrared inspections, the method of the present invention can effectively reduce the interference of environmental factors on the temperature measurement results, realize accurate monitoring and fault judgment of the state of composite insulators, thereby improving the safety and operation efficiency of transmission lines. The technical solution of the embodiment of the present application includes the following steps:

[0090] Step 1: Build an infrared temperature measurement experimental platform for composite insulators to obtain infrared temperature measurement results of composite insulators at different distances and viewing angles using an infrared thermometer and an infrared thermal imager;

[0091] In order to obtain infrared temperature measurement data at different distances and angles, the present invention built an infrared temperature measurement experimental platform for composite insulators, conducted experiments on 110kV composite insulators, and studied the relationship between the measured values ​​and true values ​​of currently commonly used infrared temperature measurement devices, infrared laser thermometers and infrared thermal imagers, at different measurement distances and viewing angles.

[0092] The composite insulator infrared temperature measurement experimental platform includes a booster device, a suspension device, a 110kV composite insulator, a fiber optic temperature meter, an infrared thermal imager, and an infrared temperature meter;

[0093] The boosting device mainly includes a high-voltage power supply system and a high-voltage measurement system; the high-voltage power supply system includes a voltage regulator, a boosting transformer (150kV, 50kVA), and a protective resistor (10kΩ, 10kW); the high-voltage measurement system includes a voltmeter (1:1000); a suspension device is used to suspend a 110kV composite insulator, and the ground wire of the pressurizing equipment is connected to the top suspension insulator of the suspension vehicle. The model of the infrared thermal imager is FLIR E8, and the model of the infrared thermometer is FLUKE 572, which are used to measure the infrared temperature of the composite insulator. The optical fiber thermometer is used to measure the actual temperature of the 110kV composite insulator;

[0094] The 110kV composite insulator is connected to the primary side of the step-up transformer through a protective resistor, the secondary side of the step-up transformer is connected to the voltage regulator, and a voltmeter is connected in parallel to the secondary side of the step-up transformer;

[0095] The infrared temperature measurement experiment of composite insulators is carried out in the laboratory, so the wind speed is set to zero. The time for infrared temperature measurement is short, so the temperature and humidity are regarded as fixed values ​​and have no effect on the temperature measurement results.

[0096] The experimental steps are as follows:

[0097] 1. Suspension insulator:

[0098] According to the standard "GB / T 4585-2004 Artificial pollution test for high-voltage insulators for AC systems", hang the insulator for the test. Select an insulator and hang it vertically on the hanging device in the test room. Fix the top of the insulator to ensure stability. Measure the vertical distance from the insulator to the ground to ensure that any part of the insulator is at least 0.7 meters away from the ground;

[0099] 2. Install the temperature measuring device optical fiber thermometer:

[0100] After the insulator is fixed, install the optical fiber thermometer outside the safety fence of the test area. Adhere the three temperature probes (CH1, CH2, CH3) of the thermometer to the surface of the insulator through insulating double-sided tape, and place probe CH4 in the air to monitor the ambient temperature;

[0101] 3. Apply voltage to the high voltage end at the bottom of the insulator:

[0102] The live wire of the step-up transformer is connected to the bottom of the insulator through a copper wire, and the ground wire is connected to the top of the insulator to simulate actual operating conditions. The voltage is set to 1.1 times the operating voltage, ensuring that the safety guardrail and ground rod are in the correct position. The pressurization process uses AC power, and the voltage is set to 1.1 times 110 / √3kV, which is about 70kV, and the pressurization time is 2h;

[0103] 4. Temperature measurement of insulators:

[0104] After pressurization, the temperature of the insulator is measured using a fiber optic thermometer, and the infrared temperature of the insulator surface is recorded using an infrared thermometer and an infrared thermal imager. The measurement distance and viewing angle are adjusted according to the requirements of the drone infrared inspection. The measurement distance is set to 2 meters, 3 meters, 4 meters and 5 meters respectively, and the shooting viewing angle range is set to -30° to 30°, with a step of 10°. Three measurements are taken at each measurement point, and the average value is taken to ensure data stability. The temperature data of the fiber optic thermometer, infrared thermometer and infrared thermal imager are recorded.

[0105] 5. Replace the insulator and retest:

[0106] After the test record is completed, reduce the voltage through the voltage regulator and turn off the power supply, hang the ground rod, remove the insulator that has completed the test and replace it with a new test insulator. Repeat the above steps to complete the test of the remaining insulators in turn.

[0107] Step 2: Create an infrared temperature measurement dataset and divide the infrared temperature measurement dataset into a training set and a validation set.

[0108] The infrared temperature measurement data was sorted out, a data set was established, and it was divided into a training set and a validation set in a ratio of 8:2.

[0109] Step 3: The data set was trained using an ELM. ELM is a single hidden layer feedforward neural network that has the characteristics of fast training and good generalization ability. During the training process, ELM randomly generates the weights and biases of the hidden layer nodes and directly solves the output layer weights, thus avoiding the back-propagation training process of traditional neural networks.

[0110] Step 3.1, determine the number of layers, the number of hidden layer neurons and the activation function of the ELM network. The formula of the activation function is:

[0111] a l =σ(W l a l―1 +b l ) (1)

[0112] Among them, a l is the activation value, W l is the weight matrix, a l―1 is the activation value of the previous layer, b l is the bias vector.

[0113] Step 3.2: Randomly generate the hidden layer weight W1 and bias b1 of the ELM neural network, set the initial position and velocity of each particle, and then randomly generate the weight matrix and bias of the hidden layer nodes:

[0114]

[0115] b1~U(―1,1)(3)

[0116] In the formula is the activation function, N(0,1) represents the standard normal distribution, and U(-1,1) represents the uniform distribution.

[0117] Step 3.3, pass the input data infrared temperature measurement value, viewing angle and distance X into the ELM hidden layer, and calculate the hidden layer output matrix H:

[0118] H=σ(W1X+b1) (4)

[0119] Among them, W1 is the generated hidden layer weight, b1 is the bias vector corresponding to the generated hidden layer, X is the input data, and σ is the activation function.

[0120] ELM directly calculates the weight W2 of the output layer through the least squares method. The optimization formula of the output layer weight is:

[0121] W2=H + Y (5)

[0122] Among them, H + is the hidden layer output matrix H, Y is the target output, and W2 is the weight matrix of the output layer.

[0123] Step 3.4: According to the calculated W2 and H, make prediction output:

[0124] Y ―1 =HW2 (6)

[0125] Among them, Y ―1 is the prediction result.

[0126] After obtaining the results, calculate the error between the predicted results and the actual target value. The loss function is:

[0127]

[0128] y i is the actual value, y′ i is the predicted value, and n is the number of samples.

[0129] After calculation, the final ELM neural network model is obtained, including the hidden layer weight W1, bias b1 and output layer weight W2.

[0130] Step 4: Correct the infrared temperature measurement result through the ELM machine learning model, and record the temperature value to obtain the corrected result of infrared temperature measurement.

[0131] Step 5: Input the infrared temperature measurement data to be predicted, verify the ELM model correction results, and compare them with other models.

[0132] The details are as follows:

[0133] After the training is completed, the temperature measurement test is carried out again, the infrared temperature measurement results are corrected through the correction models of different methods, and the accurate temperature values ​​are recorded to obtain the corrected results of infrared temperature measurement.

[0134] The commonly used statistical regression models, including linear, nonlinear and exponential function fitting, and commonly used machine learning models, such as extreme learning machine (ELM), radial basis function neural network (RBF) and partial least squares regression (PLS), are used. At the same time, deep learning models, such as genetic algorithm-optimized BP neural network (GA-BP) and standard BP neural network, are combined to correct the infrared temperature measurement data respectively; the infrared temperature measurement data after statistical correction are compared and analyzed to compare the correction accuracy of the PSO-BP infrared temperature measurement correction model proposed by the present invention and other models. .

[0135] Compared with traditional methods, the present invention has significant advantages in intelligent calibration and defect detection accuracy. By accurately correcting and compensating infrared temperature measurement data through ELM, the deviation of temperature measurement results caused by environmental factors such as shooting position, distance, and viewing angle can be corrected in real time, thereby improving the quality of infrared temperature measurement images and helping inspection personnel to accurately determine the temperature rise and defect level of composite insulators. Through efficient analysis of infrared temperature measurement data and real-time defect detection, the present invention can identify major safety hazards in advance, ensure the safe operation of transmission lines, reduce the risk of line power outages caused by insulator failures, and thus improve the reliability and safety of the power grid system. This technology is not only suitable for inspection needs of different voltage levels such as 110kV-220kV, 500kV, ±500kV and ±800kV, but also has broad application potential and can be extended to other fields such as construction, agriculture, and environmental monitoring.

[0136] Infrared temperature measurement of composite insulators is a key means of evaluating their operating status. However, in actual operation, the shooting distance and viewing angle of the infrared lens will have a significant impact on the temperature detection accuracy of the composite insulator. In order to effectively reduce the interference of these external factors on the temperature measurement results, the technical solution of the embodiment of the present application proposes a multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators based on the ELM algorithm, which is used for infrared temperature correction of composite insulators. In actual implementation, the situation is as follows:

[0137] Firstly, a high voltage temperature rise test was conducted on FXBW4-110 / 70 110kV composite insulators in two normal and two contaminated conditions. Then, the surface temperature of the composite insulators was collected using three types of instruments: infrared thermal imager, infrared thermometer and optical fiber temperature measurement. The infrared temperature measurement value, distance and viewing angle were used as the input layer parameters of the ELM model, and the optical fiber temperature measurement value was used as the output layer parameter. The ELM algorithm was used for training. Then, traditional statistical regression methods such as linear, nonlinear and exponential, and machine learning models such as RBF and PLS were used to correct the infrared temperature of the composite insulator. The results show that the ELM neural network model is superior to other statistical regression methods and machine learning models in correcting the infrared temperature measurement error. The average errors of infrared thermometer and thermal imager were reduced from 17.52% and 9.96% to 1.92% and 1.44% respectively.

[0138] In specific operation, the WT-YH-500 optical fiber temperature measurement device is used to measure the temperature of the composite insulator surface. Its temperature measurement accuracy can reach ±0.5°C. The temperature data obtained by the optical fiber temperature measurement device can be used as an accurate value to evaluate the accuracy of other temperature measurement methods and as a data set input for the subsequent infrared temperature measurement correction method. The present invention selects two types of infrared temperature measuring instruments: one is an infrared laser thermometer, which realizes point temperature measurement through laser and does not have an image display function; the other is an infrared thermal imager, which can generate infrared images and display the temperature distribution of the object being measured. These two types of infrared temperature measurement results will be corrected later.

[0139] Four 110kV composite insulators of model FXBW4-110 / 70 were selected, all equipped with grading rings. One of them was used to simulate the operating state of a normal insulator, and the other three were used to simulate the operating state under dirty conditions. According to the standard process, the surfaces of the three insulators were cleaned, and then two insulator models with surface dirt were artificially manufactured using the solid coating method. The specific operation was to mix 40g of loess, 1000g of tap water and an appropriate amount of NaCl, and the dirty liquid was evenly covered on the surface of the two cleaned and dried insulators by pouring, and then placed in a natural environment to dry for 24 hours.

[0140] Through the above test steps, infrared temperature measurement data of common composite insulator types at different distances and viewing angles were obtained using infrared thermometers and infrared thermal imagers. The specific results are shown in the attached figure of the specification. Figure 3 , Figure 4 and Figure 5As shown. From the temperature measurement data, it can be seen that as the distance between the thermometer and the insulator increases, the temperature difference between the measured temperature and the true temperature generally increases. At the same time, the change in the pitch angle also significantly affects the temperature measurement accuracy. The larger the pitch angle, the greater the temperature measurement error. The maximum temperature difference of the infrared thermometer is 1.6°C, while the maximum temperature difference of the infrared thermal imager is 1.1°C. It can be determined from the basic principles of infrared temperature measurement and test results that distance and viewing angle are the key factors affecting the accuracy of infrared temperature measurement. Therefore, in order to ensure the accuracy and effectiveness of the infrared temperature measurement value, the corresponding temperature compensation method must be used to correct the influence of these factors on the infrared temperature measurement results.

[0141] The ELM algorithm proposed in the embodiment of the present application is used to train and correct the infrared temperature measurement results. The specific overall step flow chart is as follows: Figure 1 As shown in the figure, first, the preliminary infrared temperature measurement data of the target composite insulator is obtained through infrared experiments, and these data are collected and sorted to form the original data set. Subsequently, the data is input into the ELM model for correction and training to improve the temperature measurement accuracy. After the correction is completed, the results are evaluated. If the effect is not good, the correction is performed again; if the effect is ideal, the verified ELM model is exported. The model is used to perform temperature compensation on the newly collected infrared temperature measurement data, correct the temperature measurement errors caused by distance and viewing angle, and generate the corrected temperature results. To ensure the accuracy of the results, a verification experiment is carried out to compare the actual temperature value, infrared image and temperature measurement data.

[0142] In order to verify the effectiveness of the proposed temperature measurement correction algorithm, the present invention will re-conduct a verification test and use the currently commonly used temperature measurement correction algorithms including statistical regression correction methods of linear, nonlinear and exponential fitting corrections, and machine learning correction methods of PBF and PLS to compare with the ELM algorithm of the present invention.

[0143] The commonly used statistical regression models currently include linear, nonlinear, exponential and logarithmic function fitting. However, since the input viewing angle value has negative values, logarithmic and power function fitting is not considered in the present invention. The embodiment of the present application uses linear, nonlinear and exponential function fitting methods to perform fitting analysis on the infrared temperature measurement data set. The fitting results are shown in Table 1. Where, is the corrected temperature value of the infrared thermal imager, d is the distance, θ is the angle, T m1 The temperature value measured by the infrared thermal imager. is the corrected temperature value of the infrared thermometer, T m2 It is the temperature value measured by infrared thermometer.

[0144] Table 1 Fitting correction results

[0145]

[0146] The machine learning model training and verification operating system is Windows 10, the CPU is Intel i7-12700K, the GPU is NVIDIA GeForce RTX 4060, and the compilation software is Matlab R2022a. The specific correction steps are as follows:

[0147] 1) Dataset acquisition and preprocessing: The infrared temperature measurement data collected in the experiment are preprocessed, including data cleaning and missing value processing. Then, the data set is divided into a training set and a validation set in a ratio of 8.5:1.5.

[0148] 2) Training parameter setting: Determine the parameters of the machine learning model, such as the number of hidden layer nodes, maximum number of iterations, and error threshold.

[0149] 3) Model training: Use different algorithms to train the data set and export the trained model.

[0150] 4) Training effect evaluation: Obtain training evaluation indicators, such as root mean square error and accuracy, to evaluate the training effect.

[0151] 5) Re-conduct the temperature measurement test: Correct the infrared temperature measurement results through correction models of different methods, and record the accurate temperature value.

[0152] The specific correction data are shown in Table 2:

[0153] Table 2 Comparison of correction results

[0154]

[0155] When correcting the temperature measurement results of infrared thermometers and infrared thermal imagers, there are significant differences in the effects of using different algorithms. Before correction, the average error of infrared thermometers was 17.52%, and the average error of infrared thermal imagers was 9.96%. After correction by statistical regression and machine learning, the temperature measurement error was improved, but the correction effects of different algorithms were different. Figure 8 and Fig. 9 It can be seen that statistical regression methods such as linear, nonlinear and exponential have a certain correction effect on infrared temperature measurement, but the overall effect is limited, especially when dealing with complex data, the generalization ability is weak.

[0156] As shown in the attached figure of the specification Fig.10 and Fig.11It shows that the RBF model and PLS model have better correction effects at certain points, but the overall performance is not stable enough, with average errors of 4.7%, 3.26% and 7.38%, 3.16% respectively. In contrast, the ELM model performs more stably at most test points and has the best correction effect. After correction, the average errors of the infrared thermometer and infrared thermal imager dropped to 1.92% and 1.44% respectively. In summary, the ELM algorithm performs well in infrared temperature measurement correction and is suitable for high-precision correction applications. The accuracy of the temperature measurement results after correction by the ELM algorithm is significantly improved, which verifies the effectiveness of the correction algorithm proposed in the present invention in infrared temperature measurement correction.

[0157] In the description of the present application, it should be noted that the terms "upper", "lower", etc. indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present application. Unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be a connection between the two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present application can be understood according to the specific circumstances.

[0158] It should be noted that, in this application, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article 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, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0159] The above is only the specific implementation mode of the present application, so that those skilled in the art can understand or implement the present application. It will be apparent to those skilled in the art that various modifications to these embodiments are possible, and the general principles defined in the embodiments of the present application can be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to these embodiments shown in the embodiments of the present application, but will conform to the widest scope consistent with the principles and novel features applied for by the embodiments of the present application.

Claims

1. A multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators, characterized in that: The method comprises the following steps: Obtain infrared temperature measurement results of composite insulators using infrared thermometers and infrared thermal imagers at different distances and viewing angles; Based on the infrared temperature measurement results, an infrared temperature measurement data set is prepared and divided into a training set and a validation set; The infrared temperature measurement data set is trained based on an ELM model. The infrared temperature measurement result is corrected by using the ELM model to obtain a corrected result of the infrared temperature measurement result.

2. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 1, characterized in that: The method further comprises the following steps: Based on the actual temperature result obtained by actual measurement of the composite insulator, a comparison is made with the correction result of the infrared temperature measurement result to determine whether the ELM model is qualified.

3. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 1, characterized in that: Obtaining infrared temperature measurement results of composite insulators at different distances and viewing angles using an infrared thermometer and an infrared thermal imager includes the following steps: Suspension composite insulators; Three temperature probes are set on the surface of the composite insulator, and another temperature probe is set to monitor the ambient temperature; A voltage is applied to the high voltage end at the bottom of the composite insulator, and the temperature of the composite insulator is measured.

4. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 1, characterized in that: The infrared temperature measurement data set is trained based on an ELM model, and the ELM model is obtained, including the following steps: Determine the number of layers, number of hidden neurons, and activation function of the ELM neural network; Randomly generate the hidden layer weights and bias vectors of the ELM neural network, set the initial position and velocity of each particle, and randomly generate the weight matrix and bias vector of the hidden layer nodes; The infrared temperature measurement value, viewing angle and distance in the infrared temperature measurement data set are passed into the ELM hidden layer, and the hidden layer output matrix and the weight of the output layer are calculated; According to the hidden layer output matrix and the weight of the output layer, the prediction output is performed. After obtaining the result, the error between the prediction result and the actual target value is calculated to construct the loss function; Based on the loss function, error correction is performed to obtain the ELM model.

5. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 4, characterized in that: The activation function is: a l =σ(W l a l―1 +b l ); among them, a l is the activation value, W l is the weight matrix, a l―1 is the activation value of the previous layer, b l is the bias vector.

6. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 5, characterized in that: The hidden layer output matrix is: H=σ(W1X+b1); where W1 is the generated hidden layer weight, b1 is the bias vector corresponding to the generated hidden layer, X is the input data, and σ is the activation function.

7. The multi-dimensional parameter calibration method for infrared temperature measurement of composite insulators according to claim 6, characterized in that: The loss function is: in, y i is the actual value, y′ i is the predicted value, and n is the number of samples.

Images