Lightweight power transmission line icing thickness intelligent evaluation method and system
Through the neural network model combining micrometeorological and ice-covered image data, intelligent evaluation of ice-covered thickness of transmission lines is achieved, solving the detection deviation and untimely problems in the existing technology, improving the accuracy and reliability of the evaluation, and reducing the Maintenance cost.
Patent Information
- Application Number
- CN202411585713.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-08
- Publication Date
- 2024-12-06
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the prior art, there are deviations and untimely conditions for ice-covering detection of transmission lines, and the sensing equipment is frequently maintained, so efficient and rapid measurements of full coverage cannot be achieved, resulting in missing ice-covering thickness detection and affecting the safety of the power system.
Using the intelligent evaluation method of ice-covered thickness of lightweight transmission lines, by obtaining preset distribution maps and micrometeorological data of the transmission line, combining ice-covered image data, training is used for the neural network model, and ice-covered evaluation model is established to achieve ice-covered thickness intelligent assessment and risk level division.
It improves the accuracy and reliability of ice-covered thickness evaluation, reduces the maintenance needs of sensing equipment, achieves fast and efficient evaluation with full coverage, reduces labor costs, and avoids the problem of untimely detection feedback.
Smart Images

Figure CN119091235A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field related to power grid safety technology, and in particular to a method and system for intelligently assessing ice thickness of lightweight transmission lines. Background Art
[0002] With the continuous advancement of economic and technological development and grid transformation, the coverage of transmission lines is expanding. Icing disasters on transmission lines pose a serious threat to the safe operation of lines, which may lead to serious accidents such as insulator breakage, line dancing, and tower collapse. It brings huge risks to the safe and reliable operation of the power system and huge losses to regional economic development. Therefore, it is very important to timely understand the icing situation of transmission lines and prevent and deal with icing disasters to ensure the safety and stability of power transmission.
[0003] In the prior art, most of them are implemented by sensor measurement. However, the use of sensor equipment requires long-term maintenance and inspection to determine whether the processor is working normally and stably. Once a problem occurs in the sensor and is not discovered in time, it will cause deviations in the feedback results of ice measurement, affecting the timeliness of de-icing. Even long-term failure to de-ice in time may cause cable safety problems. At the same time, when this type of monitoring method is used, its ice measurement is a point measurement method, and it is impossible to achieve efficient and rapid measurement with full coverage. This leads to the amplification of meteorological factors in local areas due to terrain and environment reasons, and the ice thickness is higher, but the point measurement method cannot fully cover the entire area, resulting in missed detection. Summary of the invention
[0004] The purpose of the present invention is to provide a method and system for intelligently assessing ice thickness of lightweight transmission lines to solve the problems raised in the above-mentioned background technology.
[0005] To achieve the above object, the present invention provides the following technical solutions: A lightweight transmission line ice thickness intelligent assessment method comprises the following steps: Obtain a preset distribution map of the transmission line, perform cloud-based meteorological search based on the preset distribution map, and obtain micro-meteorological data corresponding to the location of the transmission line; Performing a visual inspection of ice coverage based on a preset distribution map of the transmission line, acquiring ice coverage image data corresponding to the position of the transmission line, and matching the position of the transmission line with the micro-meteorological data; Preprocessing the ice image data to obtain ice thickness, and training a neural network model through the obtained array of micrometeorological data and the ice thickness to establish an ice assessment model, in which the micrometeorological data is the assessment input data and the ice thickness is the assessment output data; The ice thickness is divided based on the ice risk level of the cable, and the ice thickness division is mapped to the ice assessment model and acts on the assessment output data. The ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
[0006] As a further solution of the present invention: the ice image data is preprocessed to obtain the ice thickness, and the neural network model is trained by the obtained array of micrometeorological data and the ice thickness, and the step of establishing the ice assessment model specifically includes: Acquiring acquisition parameter information of the ice-covered image data, and collaboratively scaling the power transmission lines in the ice-covered image data based on the acquisition parameter information so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measuring the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; Inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; Inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; The meteorological four-dimensional feature vector and the image four-dimensional feature vector are input into a fully connected layer for training, and an ice coverage assessment model is established based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
[0007] As a further solution of the present invention: the micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
[0008] As a further solution of the present invention: the fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer, the input layer and the first hidden layer neurons are fully connected to each other, the first hidden layer and the second hidden layer neurons are fully connected to each other, and the fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
[0009] As a further solution of the present invention: the convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence.
[0010] The embodiment of the present invention aims to provide a lightweight transmission line ice thickness intelligent assessment system, comprising: An environmental synchronization module, used to obtain a preset distribution map of the transmission line, perform cloud-based meteorological retrieval based on the preset distribution map, and obtain micro-meteorological data corresponding to the location of the transmission line; A visual synchronization module, used to perform a visual inspection of ice coverage based on a preset distribution map of the transmission line, obtain ice coverage image data corresponding to the position of the transmission line, and match the position of the transmission line with the micro-meteorological data; A model training module is used to pre-process the ice-covered image data to obtain the ice-covered thickness, and train a neural network model through the obtained array of micro-meteorological data and the ice-covered thickness to establish an ice-covered evaluation model, in which the micro-meteorological data is the evaluation input data and the ice-covered thickness is the evaluation output data; The output optimization module is used to classify the ice thickness based on the ice risk level of the cable, and map the ice thickness classification to the ice assessment model and act on the assessment output data, wherein the ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
[0011] As a further solution of the present invention: the model training module includes: a data preprocessing unit, configured to obtain acquisition parameter information of the ice-covered image data, and collaboratively scale the power transmission lines in the ice-covered image data based on the acquisition parameter information, so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measure the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; A meteorological feature unit, used for inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; An image feature unit, used for inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; A fully connected training unit is used to input the meteorological four-dimensional feature vector and the image four-dimensional feature vector into a fully connected layer for training, and to establish an ice cover assessment model based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
[0012] As a further solution of the present invention: the micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
[0013] As a further solution of the present invention: the fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer, the input layer and the first hidden layer neurons are fully connected to each other, the first hidden layer and the second hidden layer neurons are fully connected to each other, and the fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
[0014] As a further solution of the present invention: the convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence.
[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: the neural network model is trained through the coordinated use of micrometeorological data and image data, so as to realize the acquisition of the icing assessment model, and then achieve the purpose of using the icing assessment model to perform intelligent icing assessment on transmission lines, and the result-oriented model training and correction method based on image data can easily reach the deviation coefficient judgment of the impact of non-uniform objective factors such as environmental distribution on the assessment, thereby improving the accuracy and reliability of icing thickness assessment. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 The flowchart is a method for intelligently assessing ice thickness on lightweight transmission lines.
[0017] Figure 2 A flowchart of the training steps in an intelligent assessment method for ice thickness on lightweight transmission lines.
[0018] Figure 3 This is a block diagram of a lightweight transmission line ice thickness intelligent assessment system. DETAILED DESCRIPTION
[0019] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0020] The specific implementation of the present invention is described in detail below in conjunction with specific embodiments.
[0021] like Figure 1 The method for intelligently assessing ice thickness of a lightweight transmission line provided by an embodiment of the present invention comprises the following steps: S10, obtaining a preset distribution map of the transmission line, performing cloud-based meteorological search based on the preset distribution map, and obtaining micro-meteorological data corresponding to the location of the transmission line; S20, performing a visual inspection of ice coverage based on a preset distribution map of the transmission line, acquiring ice coverage image data corresponding to the position of the transmission line, and matching the position of the transmission line with the micro-meteorological data; S30, preprocessing the ice-covered image data to obtain ice-covered thickness, and training a neural network model through the obtained array of micro-meteorological data and the ice-covered thickness to establish an ice-covered evaluation model, in which the micro-meteorological data is evaluation input data, and the ice-covered thickness is evaluation output data; S40, classifying the ice thickness based on the ice risk level of the cable, and mapping the ice thickness classification to the ice assessment model and acting on the assessment output data, wherein the ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
[0022] In this embodiment, a lightweight transmission line ice thickness intelligent assessment method is provided. The neural network model is trained by the coordinated use of micro-meteorological data and image data, so that the ice assessment model is obtained, and then the purpose of using the ice assessment model to perform fast and efficient intelligent assessment of transmission line ice is achieved. The result-oriented model training and correction method based on image data can easily achieve the deviation coefficient judgment of the impact of non-uniform objective factors such as environmental distribution on the assessment, thereby improving the accuracy and reliability of ice thickness assessment. In the prior art, the methods for detecting ice thickness of transmission lines include mechanical methods, optical methods, and electromagnetic methods. Methods, ultrasonic detection methods, etc., among which the mechanical method usually uses an ice detection rod or an ice sensor to directly measure the ice thickness; the optical method uses a lidar or a laser rangefinder to detect the ice thickness by passing through the ice layer at the speed of light, or obtains an ice thickness image and calculates the ice thickness based on an image processing algorithm; the electromagnetic and ultrasonic detection methods detect and judge the ice thickness by electromagnetic wave and ultrasonic wave penetration respectively; in these existing technical solutions, measurement is used to achieve this, but in the setting and use of the sensor, long-term maintenance and inspection are required to determine whether the processor is working properly It works very stably, and once the sensor has a problem and cannot be discovered in time, it will cause deviations in the feedback results of ice measurement, affecting the timeliness of de-icing. Even long-term failure to de-icing in time may cause cable safety problems. At the same time, when this type of monitoring method is used, its ice measurement belongs to the point selection measurement method, and it is impossible to achieve efficient and fast measurement with full coverage. This leads to the amplification of meteorological factors in local areas due to terrain and environment, and the ice thickness is higher. However, because the detection method cannot fully cover the whole area, it leads to missed detection, which eventually leads to the occurrence of safety accidents. The solution proposed in this embodiment is to first patrol The ice image data of various parts of the transmission cables are obtained through inspection, and the micro-meteorological data of each transmission cable is obtained from the Meteorological Bureau. Through such data training, the influence of environmental distribution in different sections of the cable will be reflected in the training model, and then the model can be used to predict and judge the cable icing situation in a certain section under the condition of known meteorological data. Compared with the fixed-point measurement method, the model-based evaluation method can more timely and efficiently cover the icing situation of various parts of the transmission line. At the same time, it eliminates the maintenance process of the sensor equipment, reduces labor costs, and can avoid the untimely feedback of icing detection caused by damage to the sensor equipment.
[0023] like Figure 2 As shown, as another preferred embodiment of the present invention, the ice image data is preprocessed to obtain the ice thickness, and the neural network model is trained by the obtained array of micrometeorological data and the ice thickness, and the steps of establishing the ice assessment model specifically include: S31, obtaining acquisition parameter information of the ice-covered image data, and collaboratively scaling the power transmission lines in the ice-covered image data based on the acquisition parameter information, so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measuring the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; S32, inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; S33, inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; S34, inputting the meteorological four-dimensional feature vector and the image four-dimensional feature vector into a fully connected layer for training, and establishing an ice cover assessment model based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
[0024] Furthermore, the micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
[0025] Furthermore, the fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer. The input layer and the first hidden layer neurons are fully connected to each other, and the first hidden layer and the second hidden layer neurons are fully connected to each other. The fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
[0026] Furthermore, the convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence.
[0027] In this embodiment, on the basis of the above-mentioned embodiment, the training of the ice assessment model and the selection of the relevant parameter model are further defined. Specifically, before the model training is performed according to the image data, the image data needs to be preprocessed, that is, the ice thickness data contained in the image data is extracted. The embodiment shows that the ice thickness includes multi-directional thickness distribution information, that is, the ice condition corresponds to a four-dimensional vector as a label, that is, no ice corresponds to a vector [1,0,0,0], The corresponding vector for light icing is [0,1,0,0], the corresponding vector for moderate icing is [0,0,1,0], and the corresponding vector for heavy icing is [0,0,0,1]. Training the network model can further achieve the training optimization effect of the model. At the same time, the evaluation method uses a low-computational fully-connected neural network to extract the features of micrometeorological data, and uses a 3-layer convolutional layer plus a pooling layer to extract the features of the image. Finally, a fully-connected layer is used to simultaneously use the features of two dimensions to evaluate the ice thickness. It has low computational complexity and can be used in low-cost, low-computing power, and low-power terminal devices. It can be distributed and does not require much cost and environmental construction. The low-power terminal can achieve long battery life and has higher stability, that is, environmental changes will not have much impact on its working environment and load support. ; Based on the above embodiments, the specific training steps can be summarized as follows: Step 1, using sensors deployed on transmission lines to obtain micro-meteorological data near transmission lines and using cameras deployed on transmission lines to obtain image data of ice-covered transmission lines; Step 2, processing the size of the ice-covered transmission line image into an image of uniform size, and using the ice thickness manually measured by a measuring tool as an image label; Step 3, inputting the micro-meteorological data into a fully connected neural network model 1, and outputting a meteorological four-dimensional feature vector V1; Step 4, inputting the processed image into a convolutional neural network model 2, and outputting a four-dimensional feature vector V2 of the image; Step 5, inputting the meteorological four-dimensional feature vector V1- and the image four-dimensional feature vector V2 into the fully connected layer, and the fully connected layer outputs the classification result of the ice thickness of the transmission line, and the ice thickness is divided into four categories: no ice, light ice, moderate ice, and heavy ice.
[0028] To summarize, the lightweight intelligent assessment method for ice thickness of transmission lines based on images and micrometeorological data used in this embodiment utilizes a combination of neural networks, images and micrometeorological data, makes full use of image and micrometeorological information, and improves the accuracy and reliability of ice thickness prediction; through training and feature extraction, the neural network model can effectively process a large amount of images and data to achieve rapid prediction and assessment; this intelligent, efficient and accurate method provides important support for the operation and management of power systems; compared with traditional methods, this method overcomes the problem that the model is too complex when using images and micrometeorological information at the same time, making it impossible to use it on low-cost, low-computing power terminals, and can achieve efficient and accurate assessment of the ice thickness of transmission lines; this is of great significance to improving the safety and stability of power systems, reducing the risk of accidents caused by icing, and further improving the reliability and power supply quality of power grids.
[0029] like Figure 3 As shown, the present invention also provides a lightweight transmission line ice thickness intelligent assessment system, which comprises: The environmental synchronization module 100 is used to obtain a preset distribution map of the transmission line, perform cloud-based meteorological retrieval based on the preset distribution map, and obtain micro-meteorological data corresponding to the location of the transmission line; A visual synchronization module 200 is used to perform a visual inspection of ice coverage based on a preset distribution map of the transmission line, obtain ice coverage image data corresponding to the position of the transmission line, and match the position of the transmission line with the micro-meteorological data; A model training module 300 is used to pre-process the ice image data to obtain ice thickness, and train a neural network model through the obtained array of micro-meteorological data and the ice thickness to establish an ice assessment model, in which the micro-meteorological data is the assessment input data and the ice thickness is the assessment output data; The output optimization module 400 is used to classify the ice thickness based on the ice risk level of the cable, and map the ice thickness classification to the ice assessment model and act on the assessment output data, wherein the ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
[0030] As another preferred embodiment of the present invention, the model training module includes: a data preprocessing unit, configured to obtain acquisition parameter information of the ice-covered image data, and collaboratively scale the power transmission lines in the ice-covered image data based on the acquisition parameter information, so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measure the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; A meteorological feature unit, used for inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; An image feature unit, used for inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; A fully connected training unit is used to input the meteorological four-dimensional feature vector and the image four-dimensional feature vector into a fully connected layer for training, and to establish an ice cover assessment model based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
[0031] As another preferred embodiment of the present invention, the micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
[0032] As another preferred embodiment of the present invention, the fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer. The input layer and the first hidden layer neurons are fully connected to each other, and the first hidden layer and the second hidden layer neurons are fully connected to each other. The fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
[0033] As another preferred embodiment of the present invention, the convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence. Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the program can be stored in a non-volatile computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
[0034] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the disclosure in the specification and examples. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the claims.
[0035] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.
Claims
1. A lightweight transmission line ice thickness intelligent assessment method, characterized in that: Contains steps: Obtain a preset distribution map of the transmission line, perform cloud-based meteorological search based on the preset distribution map, and obtain micro-meteorological data corresponding to the location of the transmission line; Performing a visual inspection of ice coverage based on a preset distribution map of the transmission line, acquiring ice coverage image data corresponding to the position of the transmission line, and matching the position of the transmission line with the micro-meteorological data; Preprocessing the ice image data to obtain ice thickness, and training a neural network model through the obtained array of micrometeorological data and the ice thickness to establish an ice assessment model, in which the micrometeorological data is the assessment input data and the ice thickness is the assessment output data; The ice thickness is divided based on the ice risk level of the cable, and the ice thickness division is mapped to the ice assessment model and acts on the assessment output data. The ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
2. According to claim 1, a lightweight transmission line ice thickness intelligent assessment method is characterized in that: The steps of preprocessing the ice image data to obtain ice thickness, and training a neural network model by using the obtained array of micrometeorological data and the ice thickness to establish an ice assessment model specifically include: Acquiring acquisition parameter information of the ice-covered image data, and collaboratively scaling the power transmission lines in the ice-covered image data based on the acquisition parameter information so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measuring the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; Inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; Inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; The meteorological four-dimensional feature vector and the image four-dimensional feature vector are input into a fully connected layer for training, and an ice coverage assessment model is established based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
3. The method for intelligently assessing ice thickness of lightweight transmission lines according to claim 2, characterized in that: The micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
4. The method for intelligently assessing ice thickness of lightweight transmission lines according to claim 3 is characterized in that: The fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer. The input layer and the first hidden layer neurons are fully connected to each other, and the first hidden layer and the second hidden layer neurons are fully connected to each other. The fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
5. The method for intelligently assessing ice thickness of lightweight transmission lines according to claim 4, characterized in that: The convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence.
6. A lightweight transmission line ice thickness intelligent assessment system, characterized in that: Include: An environmental synchronization module, used to obtain a preset distribution map of the transmission line, perform cloud-based meteorological retrieval based on the preset distribution map, and obtain micro-meteorological data corresponding to the location of the transmission line; A visual synchronization module, used to perform a visual inspection of ice coverage based on a preset distribution map of the transmission line, obtain ice coverage image data corresponding to the position of the transmission line, and match the position of the transmission line with the micro-meteorological data; A model training module is used to pre-process the ice-covered image data to obtain the ice-covered thickness, and train a neural network model through the obtained array of micro-meteorological data and the ice-covered thickness to establish an ice-covered evaluation model, in which the micro-meteorological data is the evaluation input data and the ice-covered thickness is the evaluation output data; The output optimization module is used to classify the ice thickness based on the ice risk level of the cable, and map the ice thickness classification to the ice assessment model and act on the assessment output data, wherein the ice thickness classification includes no ice, light ice, moderate ice and heavy ice.
7. The lightweight transmission line ice thickness intelligent assessment system according to claim 6 is characterized in that: The model training module includes: a data preprocessing unit, configured to obtain acquisition parameter information of the ice-covered image data, and collaboratively scale the power transmission lines in the ice-covered image data based on the acquisition parameter information, so that the scales of the power transmission lines in the plurality of ice-covered image data are unified, and then measure the ice thickness based on visual mapping, wherein the ice thickness includes multi-directional thickness distribution information; A meteorological feature unit, used for inputting the micro-meteorological data into a fully connected neural network model to obtain a meteorological four-dimensional feature vector; An image feature unit, used for inputting the ice thickness into a convolutional neural network model to obtain a four-dimensional feature vector of the image; A fully connected training unit is used to input the meteorological four-dimensional feature vector and the image four-dimensional feature vector into a fully connected layer for training, and to establish an ice cover assessment model based on a fully connected neural network model, a convolutional neural network model and a fully connected layer.
8. The lightweight transmission line ice thickness intelligent assessment system according to claim 7, characterized in that: The micro-meteorological data includes wind speed data, wind direction data, temperature and humidity data, air pressure data, and precipitation and snowfall data near the transmission line; The ice-covered image data includes an ice-covered image of a power transmission line.
9. The lightweight transmission line ice thickness intelligent assessment system according to claim 8, characterized in that: The fully connected neural network model includes an input layer, a first hidden layer and a second hidden layer. The input layer and the first hidden layer neurons are fully connected to each other, and the first hidden layer and the second hidden layer neurons are fully connected to each other. The fully connected neural network model determines the range of the number of hidden layer nodes according to the number of input nodes and output nodes, and finally determines the hidden nodes.
10. The lightweight transmission line ice thickness intelligent assessment system according to claim 9, characterized in that: The convolutional neural network model includes a convolutional layer, a pooling layer, a convolutional layer, a pooling layer, a convolutional layer, a pooling layer and a fully connected layer arranged in sequence.
Citation Information
Patent Citations
Power transmission line icing prediction method based on quantum particle swarm and wavelet nerve network
CN105139274A
Modeling method for power transmission line icing thickness prediction model based on PR-KELM
CN110929611A
Icing thickness detection method and device, electronic equipment and storage medium
CN117495954A
Lightweight power transmission line icing thickness intelligent evaluation method
CN117710741A
Predicting ice coating status on transmission lines
US20140067271A1
