Method for detecting damage of power supply cable by using machine learning
Through machine learning, the method of detecting power supply cable damage is used to use feature extraction and screening technology to form differential features for training prediction models, solving the problem of low detection efficiency of existing cables and achieving the effect of efficiently identifying the type of cable damage.
Patent Information
- Application Number
- CN202411663081.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-20
- Publication Date
- 2025-02-14
AI Technical Summary
The existing cable detection based on machine vision technology has the problems of complex data processing and low detection efficiency.
Machine learning is used to detect power supply cable damage, including collecting cable damage pictures from the database, performing feature extraction and screening, and forming differential features, which are used to train prediction models, and then quickly predict cable damage types.
By extracting the differential features of the damage picture, the feature processing is simplified, the efficiency of cable damage detection is improved, and the damage type can be quickly and accurately identified, reducing training difficulty and improving accuracy.
Smart Images
Figure CN119445255A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of detection and relates to damage detection. Background Art
[0002] At this stage, with the development of 5G technology and the promotion of equipment cloud interconnection, the demand for instant information transmission and processing is becoming stronger and stronger. At present, the inspection of power supply cables is still mainly manpower, with the help of image and video monitoring systems to transmit a large amount of multimedia data to the dispatching department. Operation and maintenance personnel still need to observe images frame by frame and use experience to evaluate, which is labor-intensive and inefficient.
[0003] The core of existing machine vision intelligent recognition technology is mainly deep learning algorithm. Since deep learning has been well applied in image processing fields such as face recognition and signal extraction, more and more scholars have tried to apply it to other fields. Existing cable detection based on machine vision technology still has problems such as complex data processing and low detection efficiency. Summary of the invention
[0004] The purpose of the present invention is to solve the problems of complex data processing and low detection efficiency in existing cable detection based on machine vision technology, and proposes a method for detecting power supply cable damage using machine learning.
[0005] A method for detecting power cable damage using machine learning, the method comprising the following contents:
[0006] Step 1: Collect various types of cable damage pictures and corresponding damage types from the database to form a data set;
[0007] Step 2, feature extraction: extract features of each type of cable damage image to obtain multiple features of each type of cable damage image;
[0008] Step 3: Feature screening: retain the difference features among the multiple features of each cable damage image, remove the features without difference, and take the difference features of each cable damage image and the corresponding damage type as a sample;
[0009] Step 4: Use the samples to train the prediction model to obtain a trained prediction model;
[0010] Step 5: Collect cable damage pictures, perform feature extraction and feature screening on the pictures in turn, obtain the screened features, input them into the trained prediction model, and obtain the damage type.
[0011] Preferably, in step 4, the prediction model includes an elastic network regression model, a support vector regression model, a random forest regression model or an extreme gradient boosting model.
[0012] Preferably, in step 1, the various types of cable damage images include breakage, scratches and aging.
[0013] Preferably, between step 3 and step 4, the difference features are screened, and the features finally screened and the corresponding damage types are used as a sample.
[0014] Preferably, the difference features are screened, and the specific process is:
[0015] Calculate the difference value of every two features, eliminate the two features corresponding to the smallest difference value, and use the remaining features as the final screened features.
[0016] The beneficial effects of the present invention are:
[0017] The present invention extracts the features of the damage image, removes the features without differences, and leaves the difference features, that is, the features of the damaged parts, and further analyzes the degree of difference of the damaged parts, that is, the depth of the damage, so that it is easier to distinguish various damage types and efficiently identify each damage type. When the model is trained, the cable damage type can be quickly predicted. The present invention has a simple feature processing method and a high damage prediction efficiency.
[0018] The present invention greatly reduces the difficulty of training and improves the accuracy of training. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 Flowchart of a method for detecting power cable damage using machine learning. DETAILED DESCRIPTION
[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0021] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.
[0022] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.
[0023] Example:
[0024] Combination Figure 1 This embodiment describes a method for detecting power cable damage using machine learning, the method comprising the following contents:
[0025] Step 1: Collect various types of cable damage pictures and corresponding damage types from the database to form a data set;
[0026] Step 2, feature extraction: extract features of each type of cable damage image to obtain multiple features of each type of cable damage image;
[0027] Step 3: Feature screening: retain the difference features among the multiple features of each cable damage image, remove the features without difference, and take the difference features of each cable damage image and the corresponding damage type as a sample;
[0028] Step 4: Use the samples to train the prediction model to obtain a trained prediction model;
[0029] Step 5: Collect cable damage pictures, perform feature extraction and feature screening on the pictures in turn, obtain the screened features, input them into the trained prediction model, and obtain the damage type.
[0030] Specifically, feature screening is performed to select prominent features that are more able to determine the type of damage, and reduce the features to facilitate prediction model processing and efficient identification.
[0031] The following are the types of prediction models:
[0032] In step 4, the prediction model includes an elastic net regression model, a support vector regression model, a random forest regression model, or an extreme gradient boosting model.
[0033] The following are the types of damage:
[0034] In step 1, images of various types of cable damage including breakage, scratches, and aging are taken.
[0035] The following is the further processing process:
[0036] Between step 3 and step 4, the difference features are screened, and the final screened features and corresponding damage types are taken as a sample.
[0037] The following further defines the screening process:
[0038] The specific process of screening the difference features is as follows:
[0039] Calculate the difference value of every two features, eliminate the two features corresponding to the smallest difference value, and use the remaining features as the final screened features.
[0040] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that features described in conjunction with individual embodiments may be used in other described embodiments.
Claims
1. A method for detecting power supply cable damage using machine learning, characterized in that: The method comprises the following: Step 1: Collect various types of cable damage pictures and corresponding damage types from the database to form a data set; Step 2, feature extraction: extract features of each type of cable damage image to obtain multiple features of each type of cable damage image; Step 3: Feature screening: retain the difference features among the multiple features of each cable damage image, remove the features without difference, and take the difference features of each cable damage image and the corresponding damage type as a sample; Step 4: Use the samples to train the prediction model to obtain a trained prediction model; Step 5: Collect cable damage pictures, perform feature extraction and feature screening on the pictures in turn, obtain the screened features, input them into the trained prediction model, and obtain the damage type.
2. The method for detecting power cable damage using machine learning according to claim 1, characterized in that: In step 4, the prediction model includes an elastic net regression model, a support vector regression model, a random forest regression model, or an extreme gradient boosting model.
3. The method for detecting power cable damage using machine learning according to claim 1, characterized in that: In step 1, images of various types of cable damage including breakage, scratches, and aging are taken.
4. The method for detecting power cable damage using machine learning according to claim 1, characterized in that: Between step 3 and step 4, the difference features are screened, and the final screened features and corresponding damage types are taken as a sample.
5. The method for detecting power cable damage using machine learning according to claim 4, characterized in that: The specific process of screening the difference features is as follows: Calculate the difference value of every two features, eliminate the two features corresponding to the smallest difference value, and use the remaining features as the final screened features.