An Adaptive Robotic Repair Method and System for Power Pipelines Based on a Defect Database

Abstract

This invention relates to the field of power pipeline maintenance technology, and particularly to an adaptive power pipeline robot repair method and system based on a defect database. The method includes: acquiring power pipeline defect data in real time through a data acquisition component, and extracting multi-dimensional defect feature parameters from the defect data; constructing an initial power pipeline defect model based on the multi-dimensional defect feature parameters and historical defect data, and performing similarity matching between real-time power pipeline defect data and the data in the initial power pipeline defect model to obtain an updated power pipeline defect model; after repair, a repair quality inspection sensor performs a comprehensive inspection of the repaired area; if the inspection results are unsatisfactory, a repair strategy adjustment algorithm is used to regenerate the optimal repair strategy based on real-time power pipeline defect data. This invention significantly reduces the need for manual intervention and improves repair efficiency through automated defect identification, strategy generation, and execution.

Description

Technical Field

[0001] This application relates to the field of power pipeline maintenance technology, and in particular to an adaptive power pipeline robot repair method and system based on a defect database. Background Technology

[0002] As a crucial component of power transmission systems, power pipelines operate in complex environments, frequently facing challenges such as mechanical damage, corrosion, and thermal expansion and contraction. This often leads to various defects including cracks, holes, and surface corrosion. Failure to repair these defects promptly can severely impact the safe operation of power pipelines and even cause widespread power outages.

[0003] Traditional power pipeline repair mainly relies on manual labor, and the repair process includes welding, filling, and spraying. However, manual repair has the following problems:

[0004] (1) Low efficiency: Manual operation is time-consuming and subject to environmental limitations, such as difficulty in operation under high temperature and high pressure conditions.

[0005] (2) Unstable quality: The repair effect depends on the operator's experience and skill level, which can easily lead to inconsistent repair quality.

[0006] (3) High risk: When repairing in-service power pipelines, operators face the threat of high voltage and harsh environment.

[0007] (4) Limitations: Traditional methods are difficult to deal with complex and diverse defects, especially those involving high-precision and high-intensity repair tasks.

[0008] In recent years, with the rapid development of robotics and artificial intelligence, robot-based power pipeline repair technology has begun to receive widespread attention. These robots typically possess automation and remote control capabilities, which can reduce the risks associated with human intervention and improve repair efficiency. However, existing robotic repair solutions still have the following technical shortcomings:

[0009] (1) Lack of intelligence: Many repair robots only have fixed repair strategies and cannot flexibly adjust the repair plan according to the specific characteristics of the defect.

[0010] (2) Lack of real-time optimization capability: It is difficult to optimize the repair parameters based on real-time feedback during the repair process, resulting in unstable repair effect.

[0011] (3) Lack of knowledge accumulation mechanism: Repair experience is not systematically managed, and repair data is not used for the optimization of subsequent tasks. Summary of the Invention

[0012] Purpose of the invention: The purpose of this application is to provide an adaptive power pipeline robot repair method based on a defect database, in order to solve the technical problems existing in the prior art, such as insufficient intelligence and precision in the repair process, poor adaptability of repair schemes, low repair efficiency, and unstable repair quality. The present invention also provides an adaptive power pipeline robot repair system based on a defect database.

[0013] Technical solution:

[0014] In view of the above problems, this application provides an adaptive electric pipeline robot repair method based on a defect database, the method comprising:

[0015] The system acquires power pipeline defect data in real time using a data acquisition component, and extracts multidimensional defect feature parameters from the defect data.

[0016] The initial model of the power pipeline defect is constructed based on the multidimensional defect feature parameters and historical defect data. The real-time power pipeline defect data is then matched with the data in the initial model to obtain the updated model of the power pipeline defect. The optimal repair strategy for the real-time power pipeline defect data is obtained based on the similarity matching value.

[0017] Based on the obtained optimal repair strategy, the movement path of the repair tool is planned to ensure that the repair operation covers all defect areas. The repair tool operates and repairs according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multi-dimensional feature parameters of the corresponding power pipeline defect.

[0018] After the repair is completed, the repair quality inspection sensor performs a comprehensive inspection of the repaired area. If the inspection results are not ideal, the repair strategy adjustment algorithm is used to regenerate the best repair strategy based on the real-time power pipeline defect data, and the power pipeline defect update model and repair trajectory model are updated.

[0019] Repeat the above operations to identify and repair more electrical pipeline defects and continuously optimize the repair strategy.

[0020] Furthermore, including:

[0021] The method of acquiring power pipeline defect data in real time through the acquisition component includes:

[0022] The acquisition components are positioned in a suitable location to accurately collect defect data from power pipelines. The acquisition components include a lidar sensor, a vision camera, and an ultrasonic sensor. The lidar sensor is used to scan the geometry of the pipeline surface to obtain high-precision three-dimensional point cloud data. The vision camera is used to capture high-definition images of the pipeline surface to assist in identifying the visual features of defects. The ultrasonic sensor is used to detect defects inside the pipeline to obtain the geometric shape, size, and material information of the defects in real time.

[0023] The collected data were filtered and denoised, and the multimodal data were fused to generate unified defect description data.

[0024] Furthermore, including:

[0025] The extraction of multidimensional defect feature parameters from the defect data includes:

[0026] Edge detection and morphological analysis are performed on the high-definition images captured by the vision camera to extract two-dimensional features of defects;

[0027] The three-dimensional point cloud data acquired by lidar is processed to construct a model, and the three-dimensional model of the corresponding defect is reconstructed, thereby calculating the volume and depth of the defect.

[0028] A multidimensional feature model is constructed based on the two-dimensional features, the three-dimensional model, and the distance and location information of the defects collected by the ultrasonic sensor. The geometric shape, geometric features, material features, and surface texture features of the defects are extracted from the multidimensional feature model.

[0029] Furthermore, including:

[0030] The power pipeline defect update model includes:

[0031] The initial model of the power pipeline defect is the storage of historical defect data in the database. Each record contains: defect type, geometric features, material properties, surface texture features, and the best repair strategy corresponding to the defect.

[0032] The real-time multidimensional defect feature parameters corresponding to the power pipeline defect data are updated into the initial power pipeline defect model;

[0033] The multidimensional defect feature parameters are matched with the feature vector formed by the features in each record. The repair strategy corresponding to the historical record with the highest similarity result is taken as the best repair strategy for the current defect.

[0034] Furthermore, including:

[0035] The historical defect data is stored in a database, including:

[0036] The database has a two-level index structure, including: a first-level index that uses a clustering algorithm to initially classify historical defect data according to defect type or material characteristics, with each category forming a sub-database; and a second-level index that constructs a multi-dimensional index structure within each sub-database. The multi-dimensional index structure performs fast nearest neighbor search by recursively dividing the feature space and reducing the dimensionality of high-dimensional features.

[0037] Furthermore, including:

[0038] The multidimensional defect feature parameters are similar to the feature vector formed by the features in each record, including:

[0039] The feature vector of each historical defect data record is standardized to obtain the first feature vector.

[0040] The second feature vector is obtained by standardizing and reducing the dimensionality of the feature vector formed by the multidimensional defect feature parameters.

[0041] Set the weights for the corresponding feature dimensions, and calculate the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector;

[0042] If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching.

[0043] Furthermore, including:

[0044] The step of setting weights for corresponding feature dimensions and calculating the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector includes:

[0045] The second feature vector is pre-classified according to the defect type, and the corresponding sub-database is selected;

[0046] In the multidimensional index structure of the selected sub-database, a nearest neighbor search is performed with the second feature vector as the query point to obtain the k most similar historical defect records.

[0047] Sort the k results according to weighted distance, and select the N results with the smallest distance;

[0048] Extract the corresponding set of repair strategies from the top N most similar defect records;

[0049] If the set of repair strategies is greater than 1, the best repair strategy is selected by voting or weighted average strategy.

[0050] Furthermore, including:

[0051] If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching, including:

[0052] After each repair is completed, the repair quality is evaluated. Assuming the best strategy corresponding to the historical defect record matched in this repair is used to repair the current defect, if the repair quality is high, it indicates that the feature weight allocation is reasonable; if the evaluation result is that the repair quality is low, it indicates that the weights of some feature dimensions in the similarity metric are set improperly, and the weights are updated through gradient descent.

[0053] A corresponding loss function is constructed based on the updated weights. The loss function includes the expected value of the repair quality. When the repair quality is lower than the expected value, the relevant features in the matched defect records are analyzed. If a certain feature has a high correlation with the repair success rate in the database, the weight of that feature is increased. If a certain feature shows a low correlation with the repair success rate in multiple repairs, its corresponding weight is decreased. This update process is performed online, and the weights are recalculated after each batch of repair tasks is completed.

[0054] Furthermore, including:

[0055] The method also includes:

[0056] If the detection results are satisfactory, the repair trajectory and the corresponding repair effect in the repair trajectory model will be updated to the power pipeline defect update model to form a historical repair case record.

[0057] The power pipeline defect update model is continuously optimized and updated using machine learning algorithms.

[0058] On the other hand, the present invention also provides an adaptive power pipeline robot repair system based on a defect database, the system comprising:

[0059] The acquisition module is used to acquire power pipeline defect data in real time through the acquisition components, and extract multi-dimensional defect feature parameters from the defect data.

[0060] The repair strategy generation module is used to construct an initial model of the power pipeline defect based on the multidimensional defect feature parameters and historical defect data, and to perform similarity matching between the real-time power pipeline defect data and the data in the initial model of the power pipeline defect to obtain the updated model of the power pipeline defect; and to obtain the optimal repair strategy for the real-time power pipeline defect data based on the similarity matching value.

[0061] The repair execution module is used to plan the movement path of the repair tool according to the obtained optimal repair strategy, ensuring that the repair operation covers all defect areas. The repair tool performs repair operations according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multi-dimensional feature parameters of the corresponding power pipeline defect.

[0062] The repair quality detection and feedback module is used to perform a comprehensive inspection of the repaired area by the repair quality detection sensor after the repair is completed. If the inspection result is not ideal, the best repair strategy is regenerated based on the real-time power pipeline defect data using the repair strategy adjustment algorithm, and the power pipeline defect update model and repair trajectory model are updated.

[0063] The iterative module is used to repeatedly execute the operations of the above modules, identify and repair more power pipeline defects, and continuously optimize the repair strategy.

[0064] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0065] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0066] This invention first fuses data from multiple sensors to obtain richer and more accurate defect feature parameters. Traditional techniques rely solely on visual or LiDAR data for two-dimensional or three-dimensional parameter extraction. This invention, however, combines multimodal data fusion with the advantages of different sensors. This not only improves the accuracy and reliability of defect identification but also provides high-quality input data for subsequent adaptive repair strategy generation. This organically integrates defect feature extraction with subsequent adaptive repair strategy generation and quality feedback optimization, forming a closed-loop system.

[0067] Specifically: First, the similarity matching of this invention is not limited to simple single parameters such as length and width, but is based on multi-dimensional, deep feature parameters obtained through multimodal data fusion. Second, the weights of each feature parameter are not fixed, but are automatically optimized through machine learning models, dynamically adjusting feature weights over long-term data accumulation to achieve more accurate similarity matching. Traditional techniques generally use fixed or empirically set similarity calculation methods, lacking this dynamic optimization and weight update mechanism. Third, this invention deeply couples similarity matching with continuous learning and optimization of the defect database. The similarity matching results are not used only in a single task, but are continuously updated with subsequent repair feedback data to update the database and matching algorithm. Finally, similarity matching assists in generating adaptive repair strategies, rather than simply classifying or ranking. This invention considers the correspondence between defects and repair solutions during the matching process; that is, similarity matching not only matches "similar defects" but also directly associates them with "the corresponding optimal repair solution and its parameters."

[0068] Furthermore, after the repair is completed, the present invention uses real-time detection of the repair effect (such as surface integrity and adhesion strength) and the feedback results to optimize the repair strategy, thereby ensuring the reliability of the repair quality.

[0069] Therefore, this invention supports the dynamic selection and switching of various repair tools and materials, and is applicable to various defect types such as cracks, pores, and corrosion. Through integrated modular design, the system has good scalability to adapt to more complex repair needs in the future. Furthermore, this invention, through intelligent technology and self-learning mechanisms, realizes the generation and optimization of dynamic repair strategies, significantly improving repair efficiency, quality, and adaptability.

[0070] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0071] To more clearly illustrate the technical solutions in this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0072] Figure 1 This is a flowchart illustrating the adaptive power pipeline robot repair method based on a defect database proposed in this application.

[0073] Figure 2 This is a schematic diagram of the adaptive power pipeline robot repair system based on a defect database, as described in this application. Detailed Implementation

[0074] The purpose of this invention is to provide an adaptive power pipeline robot repair method and system based on a defect database. Through intelligent technology and self-learning mechanism, it realizes the generation and optimization of dynamic repair strategies, significantly improving repair efficiency, quality and adaptability. Specific objectives include: (1) Enhancing dynamic adjustment capability: This invention collects defect features in real time through built-in multimodal sensors and dynamically generates repair strategies in combination with the defect database. It can flexibly adjust the repair method according to parameters such as defect shape, size and material, ensuring efficient and accurate repair process. (2) Improving intelligence level: This invention uses machine learning algorithms (such as neural networks and reinforcement learning) to perform in-depth analysis of defect data, optimize repair strategies, and enable the robot to have adaptive capabilities, achieving high-quality repair in complex defect scenarios. (3) Optimizing defect database utilization: This invention constructs a continuously updated defect database. The robot automatically records repair data during the repair process and feeds it back to the database for updates, realizing the effective accumulation of historical experience and continuous optimization of future repair strategies. (4) Achieving dynamic feedback of repair quality: After the repair is completed, this invention detects the repair effect (such as surface integrity and adhesion strength) in real time and uses the feedback results to optimize the repair strategy, ensuring the reliability of repair quality. (5) Improved adaptability and scalability: This invention supports the dynamic selection and switching of various repair tools and materials, and is applicable to various defect types such as cracks, holes, and corrosion. Through integrated modular design, the system has good scalability and can adapt to more complex repair needs in the future.

[0075] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0076] First, please refer to the appendix. Figure 1 This application proposes an adaptive robotic repair method for power pipelines based on a defect database. The aim is to achieve efficient, accurate, and intelligent power pipeline repair by intelligently identifying pipeline defects, dynamically adjusting repair strategies, and utilizing a continuously updated defect database. The technical solution includes the following main components:

[0077] (1) Defect Information Collection and Analysis

[0078] a) Defect Information Collection

[0079] The robot scans the pipe surface using built-in multimodal sensors (such as LiDAR, vision cameras, and ultrasonic sensors) to acquire information in real time about the geometry, size, and material of defects. Defect characteristics include, but are not limited to:

[0080] ① Crack: length, width, and depth;

[0081] ② Hole: diameter and depth;

[0082] ③ Corrosion: area, depth, and distribution range.

[0083] The raw data collected will be transmitted to the data processing module for further analysis.

[0084] b) Defect Feature Extraction and Classification

[0085] Based on data processing algorithms (such as edge detection, morphological analysis, and 3D reconstruction), the robot extracts key features of defects. Defects are then classified into different repair categories (such as cracks, holes, and corrosion) according to their geometry, size, and material type.

[0086] c) Matching the defect database

[0087] The defect information, after feature extraction, will be matched against a pre-established defect database. This database contains information on various types of defect features, repair strategies, repair materials, and process parameters. The database uses a similarity matching algorithm to select the repair solution most similar to the current defect, which will then serve as a preliminary repair suggestion.

[0088] (2) Adaptive generation of repair strategies

[0089] a) Selection of repair tools and materials

[0090] Based on factors such as defect type, shape, size, and material, the system will automatically select the most suitable repair tools (e.g., filling tools, polishing tools, curing equipment) and repair materials (e.g., hot melt coatings, polymer fillers). Different defect types require different repair tools and materials, for example:

[0091] For cracks, choose a suitable filler material;

[0092] For corrosion, choose corrosion-resistant coating materials.

[0093] b) Setting process parameters

[0094] The system automatically adjusts the process parameters of the repair process, including temperature, pressure, and speed, based on the defect characteristics and repair materials. Precise control of these process parameters ensures effective repair for different tasks. Examples include the application temperature of the hot-melt coating, the application pressure of the filler material, and the curing temperature.

[0095] c) Repair strategy optimization

[0096] Through machine learning algorithms (such as reinforcement learning or neural networks), the system can provide real-time feedback on the repair progress during multiple repair processes and automatically optimize the repair strategy based on the feedback results. The goal of optimization is to improve repair quality and reduce repair time.

[0097] (3) Execution and feedback mechanism of the repair process

[0098] a) Repair task execution

[0099] When performing repair tasks, the repair robot operates according to the selected repair strategy. Through a precise motion control system, the robot performs repair operations such as coating, filling, welding, and polishing.

[0100] b) Repair result feedback and real-time monitoring

[0101] After the repair is completed, the robot uses built-in sensors (such as force sensors and adhesion force detectors) to detect the repair effect. Specific feedback includes:

[0102] Surface integrity restoration: Inspect the repair area for flatness, absence of cracks or peeling.

[0103] Repair adhesion strength: Ensure that the repair material adheres firmly to the original pipe surface.

[0104] c) Adjustment of dynamic repair strategy

[0105] If the repair effect is not ideal, the robot will automatically adjust the repair strategy based on the detected repair quality problems (such as insufficient adhesion, uneven surface, etc.) and re-execute the repair task.

[0106] (4) Defect Database and Learning System

[0107] a) Database construction and maintenance

[0108] One of the key aspects of this invention is the establishment of a defect database containing various defect types and repair solutions. This database includes historical repair data, repair strategies, and repair results. It not only stores basic defect parameters but also the optimal repair solution and its process parameters for each defect.

[0109] b) Self-learning mechanism

[0110] During the repair process, the robot continuously feeds back data and results to the database. Through self-learning and data accumulation, the repair strategies in the database are continuously optimized, enabling the robot to improve future repair plans based on historical data.

[0111] c) Continuous optimization

[0112] As repair tasks progress, the defect database is gradually improved, and the robot's repair capabilities and intelligence level are continuously enhanced. The robot can not only quickly generate repair strategies based on current defects, but also extract experience from historical repair data to continuously improve adaptive repair strategies.

[0113] The above solution automatically selects appropriate repair tools (such as filling, coating, and curing devices) and materials based on the characteristics of the defect type, material, and location, and adaptively adjusts the repair process parameters (such as temperature, pressure, and speed) to meet the repair needs of various complex pipeline defects. Specifically, it includes the following steps:

[0114] Step S1: Acquire power pipeline defect data in real time through the acquisition component, and extract multidimensional defect feature parameters from the defect data.

[0115] This embodiment dynamically generates a repair strategy adapted to the current defect characteristics based on real-time acquired defect features (including defect shape, size, material, etc.), by comparing them with historical data and repair strategies stored in the defect database, using a similarity matching algorithm and a machine learning algorithm. This includes:

[0116] The acquisition components are positioned in a suitable location to accurately collect defect data from power pipelines. The acquisition components include a lidar sensor, a vision camera, and an ultrasonic sensor. The lidar sensor is used to scan the geometry of the pipeline surface to obtain high-precision three-dimensional point cloud data. The vision camera is used to capture high-definition images of the pipeline surface to assist in identifying the visual features of defects. The ultrasonic sensor is used to detect defects inside the pipeline to obtain the geometric shape, size, and material information of the defects in real time.

[0117] The collected data were filtered and denoised, and the multimodal data were fused to generate unified defect description data.

[0118] The step of extracting multidimensional defect feature parameters from the defect data includes:

[0119] Edge detection and morphological analysis are performed on the high-definition images captured by the vision camera to extract two-dimensional features of defects.

[0120] Specifically, in this embodiment, edge detection and morphological analysis are performed on the images acquired by the visual camera to extract two-dimensional features of defects such as cracks and holes. This embodiment can employ the Canny edge detection algorithm to extract the edges of cracks and holes in the image. For morphological analysis, this embodiment can utilize morphological operations (dilation, erosion) to enhance defect features, facilitating subsequent classification.

[0121] The three-dimensional point cloud data acquired by lidar is processed to construct a model, and the three-dimensional model of the corresponding defect is reconstructed, thereby calculating the volume and depth of the defect.

[0122] In this embodiment, the 3D reconstruction can be performed using PCL (Point Cloud Library) to reconstruct the 3D model of the point cloud data, generating a 3D model of the defect. The volume is calculated using a volume measurement algorithm to calculate the specific size parameters of the defect.

[0123] A multidimensional feature model is constructed based on the two-dimensional features, the three-dimensional model, and the distance and location information of the defects collected by the ultrasonic sensor. The geometric shape, geometric features, material features, and surface texture features of the defects are extracted from the multidimensional feature model.

[0124] Specifically, in this embodiment, a feature extraction algorithm can be used to combine image and point cloud data to extract multi-dimensional feature parameters such as the geometric shape, size (e.g., length, width, depth), and material of the defect.

[0125] This application utilizes multiple sensor fusion technologies, such as lidar, vision camera, and ultrasonic sensor, to detect and analyze pipeline defects in real time, extracting multi-dimensional feature parameters of the defects to provide accurate data support for the generation of subsequent adaptive repair strategies.

[0126] Step S2: Construct an initial model of the power pipeline defect based on the multidimensional defect feature parameters and historical defect data, and perform similarity matching between the real-time power pipeline defect data and the data in the initial model to obtain an updated model of the power pipeline defect; and obtain the optimal repair strategy for the real-time power pipeline defect data based on the similarity matching value. Furthermore, after the repair operation is completed, this application performs quality inspection on the repaired area, and returns the repair quality information to the defect database and learning system through a feedback mechanism to adjust and optimize the repair strategy, achieving closed-loop control of quality inspection—feedback—optimization.

[0127] Step S2 in this embodiment specifically includes:

[0128] The power pipeline defect update model includes:

[0129] The initial model of the power pipeline defect is the storage of historical defect data in the database. Each record contains: defect type, geometric features, material properties, surface texture features, and the best repair strategy corresponding to the defect.

[0130] The real-time multidimensional defect feature parameters corresponding to the power pipeline defect data are updated into the initial power pipeline defect model;

[0131] The multidimensional defect feature parameters are matched with the feature vector formed by the features in each record. The repair strategy corresponding to the historical record with the highest similarity result is taken as the best repair strategy for the current defect.

[0132] Furthermore, including:

[0133] The historical defect data is stored in a database, including:

[0134] The database has a two-level index structure, including: a first-level index that uses a clustering algorithm to initially classify historical defect data according to defect type or material characteristics, with each category forming a sub-database; and a second-level index that constructs a multi-dimensional index structure within each sub-database. The multi-dimensional index structure performs fast nearest neighbor search by recursively dividing the feature space and reducing the dimensionality of high-dimensional features.

[0135] The classification algorithm employs Support Vector Machine (SVM): it classifies defects (such as cracks, holes, corrosion, etc.) based on extracted feature parameters. The corresponding clustering algorithm subdivides complex defects into different categories, refining the repair strategy.

[0136] This embodiment includes:

[0137] The multidimensional defect feature parameters are similar to the feature vector formed by the features in each record, including:

[0138] The feature vector of each historical defect data record is standardized to obtain the first feature vector.

[0139] The second feature vector is obtained by standardizing and reducing the dimensionality of the feature vector formed by the multidimensional defect feature parameters.

[0140] Set the weights for the corresponding feature dimensions, and calculate the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector;

[0141] If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching.

[0142] The step of setting weights for corresponding feature dimensions and calculating the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector includes:

[0143] The second feature vector is pre-classified according to the defect type, and the corresponding sub-database is selected;

[0144] In the multidimensional index structure of the selected sub-database, a nearest neighbor search is performed with the second feature vector as the query point to obtain the k most similar historical defect records.

[0145] Sort the k results according to weighted distance, and select the N results with the smallest distance;

[0146] Extract the corresponding set of repair strategies from the top N most similar defect records;

[0147] If the set of repair strategies is greater than 1, the best repair strategy is selected by voting or weighted average strategy.

[0148] If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching, including:

[0149] After each repair is completed, the repair quality is evaluated. Assuming the best strategy corresponding to the historical defect record matched in this repair is used to repair the current defect, if the repair quality is high, it indicates that the feature weight allocation is reasonable; if the evaluation result is that the repair quality is low, it indicates that the weights of some feature dimensions in the similarity metric are set improperly, and the weights are updated through gradient descent.

[0150] A corresponding loss function is constructed based on the updated weights. The loss function includes the expected value of the repair quality. When the repair quality is lower than the expected value, the relevant features in the matched defect records are analyzed. If a certain feature has a high correlation with the repair success rate in the database, the weight of that feature is increased. If a certain feature shows a low correlation with the repair success rate in multiple repairs, its corresponding weight is decreased. This update process is performed online, and the weights are recalculated after each batch of repair tasks is completed.

[0151] Specifically, the similarity matching algorithm mentioned above includes: Euclidean distance calculation: calculating the Euclidean distance between the current defect features and the defect features in the database, and selecting the most similar repair scheme. Weighted matching: assigning different weights based on the important features of the defect (such as depth and material) to improve the accuracy of matching.

[0152] Step S3: Based on the obtained optimal repair strategy, plan the movement path of the repair tool to ensure that the repair operation covers all defect areas. The repair tool operates and repairs according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multi-dimensional feature parameters of the corresponding power pipeline defect.

[0153] Step S3 in this embodiment specifically includes:

[0154] Filling Tool: Employs an automatic filling device that heats the filler material (such as a hot-melt coating) via a heating element and applies it evenly to the defect area through a nozzle. Coating Equipment: An integrated spraying system allows for precise control of coating thickness and uniformity, suitable for coating repair of corroded areas. Curing Device: Uses an infrared curing lamp to quickly cure the applied repair material, ensuring repair quality.

[0155] Motion Control System: Servo Motor: A high-precision servo motor is used to drive the repair tool, achieving precise position and speed control. Motion Control Software: Based on the ROS platform, motion control algorithms are developed to ensure the repair tool moves accurately along the planned path.

[0156] Step S4: After the repair is completed, the repair quality detection sensor performs a comprehensive inspection of the repaired area. If the inspection result is not ideal, the repair strategy adjustment algorithm is used to regenerate the best repair strategy based on the real-time power pipeline defect data, and the power pipeline defect update model and repair trajectory model are updated.

[0157] If the detection results are satisfactory, the repair trajectory and the corresponding repair effect in the repair trajectory model will be updated to the power pipeline defect update model to form a historical repair case record.

[0158] The power pipeline defect update model is continuously optimized and updated using machine learning algorithms.

[0159] In this embodiment, the machine learning algorithm includes: Neural Network: Constructing a multilayer perceptron (MLP) model, taking defect features as input, and outputting optimized repair strategy parameters. Reinforcement Learning: Optimizing the repair strategy through a reward mechanism, enabling the robot to learn and improve itself in continuous repair tasks.

[0160] Therefore, the specific working principle and operation steps of step S4 in this embodiment include:

[0161] Quality Inspection: After the repair is completed, the repair quality inspection sensor conducts a comprehensive inspection of the repaired area to evaluate the surface integrity, adhesion strength and material properties.

[0162] Feedback Analysis: The feedback control unit analyzes the detection data to determine whether the repair quality meets the standards.

[0163] Strategy Adjustment: If the detection results are not ideal, the repair strategy adjustment algorithm regenerates the repair strategy based on the feedback information, adjusts the repair parameters (such as temperature and pressure), and instructs the repair execution module to perform the repair operation again.

[0164] Step S5: Repeat the above operations to identify and repair more electrical conduit defects and continuously optimize the repair strategy.

[0165] Specifically, the operation process of this method includes:

[0166] (1) Startup and initialization

[0167] After the robot starts up, it performs a self-check to ensure that all sensors and actuators are working properly. It then initializes the defect database and learning system, and loads the latest repair strategies.

[0168] (2) Defect Information Collection

[0169] The robot moves inside or on the surface of the pipe, scanning the pipe with the sensor system of the defect information acquisition module to obtain defect data in real time.

[0170] (3) Defect identification and feature extraction

[0171] The collected raw data is transmitted to the defect identification and feature extraction module, which extracts the geometric shape, size, material and other feature parameters of the defects and classifies them.

[0172] (4) Repair strategy generation

[0173] Based on the extracted defect features, the repair strategy generation module searches for the most suitable repair solution in the defect database using a similarity matching algorithm. Machine learning algorithms are then used to optimize the repair strategy, generating an adaptive repair solution.

[0174] (5) Repair execution

[0175] The repair execution module selects appropriate repair tools and materials based on the generated repair strategy, plans the repair path, and executes the repair operation.

[0176] (6) Repair quality inspection

[0177] After the repair is completed, the repair quality inspection and feedback module conducts a comprehensive inspection of the repaired area to evaluate the repair effect.

[0178] (7) Feedback and Optimization

[0179] The detection results are fed back to the defect database and learning system, and repair data is recorded. The learning and optimization module analyzes the feedback information, optimizes the repair strategy, and improves the effectiveness of future repair tasks.

[0180] (8) Execute in a loop

[0181] The robot continues to scan the pipeline as needed, identifying and repairing more defects. The entire process forms a closed loop, continuously optimizing the repair strategy.

[0182] On the other hand, this invention also provides an adaptive power pipeline robot repair system based on a defect database, wherein the modules communicate and transmit data through an internal data bus to ensure coordinated operation of the system. The system architecture is as follows: Figure 2 The system includes:

[0183] The acquisition module is used to acquire power pipeline defect data in real time through acquisition components, and to extract multi-dimensional defect feature parameters from the defect data.

[0184] Specifically, the acquisition module includes a defect information acquisition module and a defect identification and feature extraction module. The defect information acquisition module is used to acquire pipeline defect data through sensors and transmit the data to the defect identification and feature extraction module through an internal data bus.

[0185] The defect identification and feature extraction module is used to process and analyze data, extract defect feature parameters, and transmit these feature information to the repair strategy generation module through the internal data bus.

[0186] The defect information acquisition module in this embodiment is implemented through the following:

[0187] Components of the defect information acquisition module:

[0188] (1) LiDAR sensor: used to scan the geometry of the pipe surface and obtain high-precision three-dimensional point cloud data.

[0189] (2) Vision camera: used to capture high-definition images of the pipe surface to help identify the visual features of defects.

[0190] (3) Ultrasonic sensor: used to detect defects inside the pipe, such as internal corrosion and voids.

[0191] (4) Data processing unit: preprocesses the collected raw data, including filtering, noise reduction and data fusion.

[0192] Working principle and operating procedures:

[0193] (1) Data acquisition: The sensor system (LiDAR, vision camera, ultrasonic sensor) carried by the robot scans the surface and interior of the pipe simultaneously to obtain the geometric shape, size and material information of the defects in real time.

[0194] (2) Data preprocessing: The data processing unit filters and denoises the raw data and fuses the multimodal data to generate unified defect description data.

[0195] Components of the defect identification and feature extraction module:

[0196] (1) Image processing submodule: responsible for processing images acquired by the vision camera, such as edge detection and morphological analysis, and extracting two-dimensional features of defects.

[0197] (2) Point cloud processing submodule: responsible for processing the three-dimensional point cloud data collected by lidar, performing three-dimensional reconstruction and defect volume calculation.

[0198] (3) Feature extraction algorithms: including shape recognition, size measurement, material analysis and other algorithms, to extract key feature parameters of defects.

[0199] Working principle and operating procedures:

[0200] (1) Image analysis: The image processing submodule performs edge detection and morphological analysis on the images acquired by the vision camera, and extracts two-dimensional features of defects such as cracks and holes.

[0201] (2) Point cloud analysis: The point cloud processing submodule processes the three-dimensional point cloud data collected by the lidar, reconstructs the three-dimensional model of the defect, and calculates the volume and depth of the defect.

[0202] (3) Feature extraction: The feature extraction algorithm combines image and point cloud data to extract multi-dimensional feature parameters such as the geometric shape, size (e.g., length, width, depth), and material of the defect.

[0203] This invention not only improves the accuracy and reliability of defect identification through data fusion algorithms and feature extraction, but also provides high-quality input data for the subsequent generation of adaptive repair strategies. Furthermore, it organically combines defect feature extraction with subsequent adaptive repair strategy generation and quality feedback optimization to form a closed-loop system.

[0204] The repair strategy generation module is used to construct an initial model of the power pipeline defect based on the multi-dimensional defect feature parameters and historical defect data, and to perform similarity matching between the real-time power pipeline defect data and the data in the initial model to obtain an updated model of the power pipeline defect; and to obtain the optimal repair strategy for the real-time power pipeline defect data based on the similarity matching value. Specifically, it obtains historical repair experience and strategies by accessing the defect database and learning system, matches and optimizes the current defect features, generates a corresponding repair plan, and sends it to the repair execution module via the internal data bus.

[0205] The repair strategy generation module in this embodiment specifically includes the following parts:

[0206] (1) Defect database: Stores the characteristic parameters, best repair strategies, repair materials and process parameters of various defects.

[0207] (2) Similarity matching algorithm: used to match the current defect with known defects in the database to determine the most suitable repair strategy.

[0208] 1. Hierarchical database structure:

[0209] The defect database stores a large amount of historical defect data. Each record includes the defect type, geometric features (length, width, depth), material properties (conductivity, stress distribution parameters), surface texture features (edge ​​intensity extracted based on image processing, local binary mode LBP features), and the best repair strategy for the defect (including the type of repair material, heating temperature, filling pressure, curing time, etc.).

[0210] To improve retrieval efficiency and matching accuracy, this invention employs a two-level index structure for the database:

[0211] (a) First-level clustering index: First, clustering algorithms such as K-Means are used to group existing defect samples according to defect type or material characteristics. For example, crack type defects, corrosion type defects, and hole type defects are initially classified, and each category forms a sub-database.

[0212] (b) Two-dimensional spatial index structure: A multi-dimensional index structure (such as KD-Tree) is built within each sub-database. KD-Tree performs fast nearest neighbor search after recursively partitioning the feature space and reducing the dimensionality of high-dimensional features.

[0213] 2. Feature standardization and dimensionality reduction:

[0214] The feature vector F_db(i) of each defect record is standardized:

[0215]

[0216] Where, μ j and σ j These are the mean and standard deviation of feature dimension j, respectively, obtained statistically from historical data. Standardization eliminates the influence between features of different dimensions and orders of magnitude.

[0217] Furthermore, when the feature dimensionality is too high, PCA (Principal Component Analysis) can be used to reduce the dimensionality of the features, selecting the top few principal components with a cumulative variance contribution rate of over 90% to construct the KD-Tree index. This dimensionality reduction process can significantly improve retrieval speed.

[0218] 3. Similarity Measurement and Algorithm Steps

[0219] (1) Definition of similarity distance:

[0220] Let the current defect feature vector be denoted as F_current, which, after undergoing the same standardization and dimensionality reduction, becomes F_current'. This is recorded in the database as F_db(i)'. Define the weighted Euclidean distance:

[0221]

[0222] Where m is the number of features after dimensionality reduction, w j These are the weights for feature dimension j. Initial values ​​for these weights can be obtained from expert experience or by training a linear regression or neural network on a dedicated training set.

[0223] (2) Adaptive weight update and feedback mechanism:

[0224] Whenever a repair is completed, the present invention evaluates the repair quality Q (determined based on comprehensive indicators such as adhesion, surface smoothness, electrical performance scoring, etc.). Suppose the historical defect record matched in this repair is F_db(k), and its corresponding strategy S_k is used to repair the current defect. If the repair quality Q is high, it indicates that the feature weight allocation is reasonable; if Q is low, it means that the weight settings of some feature dimensions in the similarity measurement are inappropriate. The present invention updates w through gradient descent j as follows:

[0225]

[0226] where the loss function L can be defined as:

[0227]

[0228] When Q < Q_ideal (lower than the expected value), analyze the relevant features in the matched defect records. If a certain type of feature (such as the material conductivity feature) has a high correlation with the repair success rate in the database, moderately increase the weight w of this dimension of the feature j ; if a certain feature (such as the texture feature) shows unimportance or misleadingness in multiple repairs, reduce its weight.

[0229] This update process is carried out online. After each batch of repair tasks (such as N repairs) is completed, recalculate the weights and continuously optimize the accuracy of similarity matching.

[0230] 4. Retrieval and Matching Process

[0231] (1) Retrieval process:

[0232] (a) Input: The multi-modal fusion feature F_current of the current defect data.

[0233] (b) Preprocessing: Standardize F_current and multiply it by the PCA matrix to achieve dimensionality reduction, obtaining F_current'.

[0234] (c) Primary classification: Pre-classify F_current' according to the defect type and select the corresponding sub-database (such as the crack sub-database).

[0235] (d) Secondary index search: In the KD-Tree of the selected sub-database, perform a nearest neighbor search (k-NN query) with F_current' as the query point to obtain k most similar historical defect records F_db(i)'.

[0236] (e) Similarity ranking: Sort these k results according to the weighted distance d(F_current', F_db(i)') and select the top N results with the smallest distance.

[0237] (2) Strategy selection and comprehensive decision-making:

[0238] Extract the corresponding repair strategy set S = {S_1, S_2, ..., S_N} from the N most similar defect records.

[0239] If N>1, a voting or weighted average strategy can be used to select the final repair solution. For example, each strategy can be weighted according to its historical repair success rate and average repair quality score, and the strategy S_best with the highest expected repair effect can be selected.

[0240] 5. Feedback and Updates:

[0241] After the repair is completed, record the repair effect Q and the matching error and feature weight change trend corresponding to the strategy S_best. This information is then sent back to the database for use in the next round of weight updates and strategy optimization, gradually improving the performance of the matching algorithm.

[0242] This embodiment features a special design and improvement in the selection of feature parameters and the method for calculating the matching degree.

[0243] (a) Multidimensional feature input: The similarity matching of this invention is not only based on simple single parameters such as length and width, but also on multidimensional and deep feature parameters (such as the shape complexity of defects, material properties, corrosion depth distribution, stress concentration degree, etc.) obtained through multimodal data fusion.

[0244] (b) Adaptive Weight Allocation: In this invention, the weights of each feature parameter are not fixed, but are automatically optimized through a machine learning model. The feature weights are dynamically adjusted over long-term data accumulation to achieve more accurate similarity matching. Traditional techniques generally use fixed or empirically set similarity calculation methods, lacking this dynamic optimization and weight update mechanism.

[0245] The closed-loop combination of similarity matching and learning systems in this invention specifically includes:

[0246] This invention deeply couples similarity matching with continuous learning and optimization of a defect database. The similarity matching results are not used in a single task, but are continuously updated with subsequent repair feedback data to improve both the database and the matching algorithm.

[0247] (a) Continuously optimized matching model: Traditional technologies use fixed matching rules and algorithms in similarity matching. This invention, however, uses machine learning and reinforcement learning to continuously adjust the matching algorithm based on historical repair results, making the matching results more relevant and instructive for actual repair effects.

[0248] (b) Dynamic evolution of the database: Unlike existing technologies that rely on static feature templates, the defect database of this invention is dynamically updated. The matching algorithm continuously improves its accuracy and practicality with the constantly enriched training data, so that similarity matching is no longer a simple invocation of existing methods, but a core link that runs through the entire system's self-learning cycle.

[0249] The linkage between this invention and specific repair strategies and process parameters specifically includes:

[0250] The ultimate goal of similarity matching is to assist in generating adaptive repair strategies, rather than simply classifying or ranking them. This invention considers the correspondence between defects and repair solutions during the matching process; that is, similarity matching not only matches "similar defects" but also directly associates them with "the corresponding optimal repair solution and its parameters."

[0251] (a) Personalized matching indicators: The matching process incorporates considerations such as subsequent repair process parameters, material selection, and tool compatibility, enabling the similarity matching results to be directly used for repair strategy generation. Traditional similarity matching is mostly limited to the comparison of feature similarity. This invention incorporates effectiveness indicators such as the applicability of executable strategies and historical repair success rates into the matching calculation, which goes far beyond the level of traditional pure feature matching.

[0252] (b) Dynamic evolution of the database: Unlike existing technologies that rely on static feature templates, the defect database of this invention is dynamically updated. The matching algorithm continuously improves its accuracy and practicality with the constantly enriched training data, so that similarity matching is no longer a simple invocation of existing methods, but a core link that runs through the entire system's self-learning cycle.

[0253] The repair execution module is used to plan the movement path of the repair tool according to the obtained optimal repair strategy, ensuring that the repair operation covers all defect areas. The repair tool performs repair operations according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multi-dimensional feature parameters of the corresponding power pipeline defects.

[0254] The repair execution module in this embodiment specifically includes the following:

[0255] (1) Repair tool system: including filling tools, coating equipment, curing device, etc., which are operated according to the repair strategy.

[0256] (2) Motion control system: controls the motion path and operation accuracy of the repair tool.

[0257] (3) Energy supply system: Provides the energy required for repair tools and control systems.

[0258] Working principle and operating procedures:

[0259] (1) Tool configuration: Select the appropriate repair tools according to the repair strategy (e.g., filling tools are used for crack repair, and coating equipment is used for corrosion repair).

[0260] (2) Path planning: The motion control system plans the motion path of the repair tool according to the geometry and location of the defect, ensuring that the repair operation covers all defect areas.

[0261] (3) Perform repair: The repair tool operates according to the planned path, applies repair materials (such as hot melt coatings, polymer fillers) and performs curing treatment to complete the repair of defects.

[0262] The repair quality detection and feedback module is used to perform a comprehensive inspection of the repaired area by the repair quality detection sensor after the repair is completed. If the inspection result is not ideal, the module regenerates the best repair strategy based on the real-time power pipeline defect data using the repair strategy adjustment algorithm, and updates the power pipeline defect update model and repair trajectory model.

[0263] Specifically, the repair quality detection and feedback module in this embodiment includes the following:

[0264] (1) Repair quality detection sensors: including force sensors, adhesion force detectors, surface smoothness detectors, etc., are used to detect the repair effect.

[0265] (2) Feedback control unit: Receives data from the detection sensors, evaluates the repair quality and generates feedback information.

[0266] (3) Repair strategy adjustment algorithm: Adjust the repair strategy based on feedback information and optimize subsequent repair operations.

[0267] Working principle and operating procedures:

[0268] (1) Quality inspection: After the repair is completed, the repair quality inspection sensor will conduct a comprehensive inspection of the repaired area to evaluate the surface integrity, adhesion strength and material properties of the repaired area.

[0269] (2) Feedback analysis: The feedback control unit analyzes the detection data to determine whether the repair quality meets the standards.

[0270] (3) Strategy adjustment: If the detection results are not ideal, the repair strategy adjustment algorithm regenerates the repair strategy based on the feedback information, adjusts the repair parameters (such as temperature and pressure), and instructs the repair execution module to perform the repair operation again.

[0271] Specifically, the technical implementation of the repair quality inspection and feedback module includes:

[0272] Repair quality inspection sensors: ① Force sensor: Measures the applied force of the repair material to ensure the mechanical properties during filling and coating. ② Adhesion force detector: Detects the adhesion strength between the repair material and the pipe surface to prevent material peeling. ③ Surface smoothness detector: Detects the surface smoothness of the repair area through laser scanning or tactile sensors to ensure that the repaired surface is free of obvious defects.

[0273] Feedback Control Unit: ① Data Analysis Module: Analyzes data from the detection sensors to assess whether the repair quality meets the standards. ② Decision Module: Based on the analysis results, decides whether the repair strategy needs to be adjusted and the repair task needs to be re-executed.

[0274] Repair strategy adjustment algorithm: ① Adaptive adjustment: Dynamically adjust repair parameters (such as temperature, pressure, and speed) based on feedback results to optimize the repair process. ② Re-repair instruction: If the detection results are unsatisfactory, generate a new repair instruction to guide the repair execution module to perform the repair operation again.

[0275] The defect database and learning module are used to store and manage historical data on various defect types and remediation strategies. Machine learning algorithms are used to continuously optimize and update the remediation strategies, providing better decision support for future remediation tasks. This invention constructs and uses a continuously updated defect database to record and analyze remediation results. Through the continuous accumulation of historical data and feedback information, it performs self-learning and optimization, enabling remediation strategies to continuously improve as remediation experience grows.

[0276] Specifically, the defect database and learning system in this embodiment consist of the following components:

[0277] (1) Database management system: used to store and manage defect data and repair strategies.

[0278] (2) Data acquisition module: Collects detailed data for each repair operation, including defect characteristics, repair strategy, repair effect, etc.

[0279] (3) Learning and Optimization Module: Utilize machine learning algorithms to analyze and optimize data in the database, thereby improving the intelligence level of the repair strategy.

[0280] Working principle and operating procedures:

[0281] (1) Data recording: After each repair operation, the repair quality detection and feedback module will feed back the repair data (defect characteristics, repair strategy, repair effect) to the data acquisition module.

[0282] (2) Data storage: The database management system stores the collected data in the defect database to form a record of historical repair cases.

[0283] (3) Data Analysis and Learning: The learning and optimization module analyzes the stored data, identifies the effectiveness of the repair strategies, and uses machine learning algorithms to continuously optimize and update the repair strategy database.

[0284] (4) Continuous optimization: Through continuous learning and optimization, the repair strategies in the defect database are gradually improved, enhancing the system's adaptability and intelligence.

[0285] In this embodiment, the technical implementation of the defect database and learning module includes:

[0286] Database Management System: ① Data Storage: Use a relational database (such as MySQL) to store defect data, remediation strategies, and remediation results. ② Data Indexing: Establish an efficient indexing mechanism to accelerate defect matching and strategy retrieval.

[0287] Data Acquisition Module: ① Data Recording: Automatically records detailed data for each repair operation, including defect characteristics, repair strategies, and repair effects. ② Data Synchronization: Synchronizes data to a cloud database via wireless communication (such as Wi-Fi, 4G / 5G) for centralized data management and backup.

[0288] Learning and Optimization Module: ① Data Analysis: Utilizing big data analytics, conduct in-depth analysis of stored data to identify the strengths and weaknesses of remediation strategies. ② Model Training: Based on the collected remediation data, train machine learning models to improve the intelligence level of remediation strategies. ③ Strategy Update: Update the optimized remediation strategies to the defect database for reference in subsequent remediation tasks.

[0289] To facilitate understanding of the above solution, the following example is provided:

[0290] Crack Repair

[0291] (1) Startup and initialization

[0292] After the robot starts up, it performs a self-check to ensure that the LiDAR, vision camera, and ultrasonic sensors are working properly. The defect database and learning system load the latest repair strategies.

[0293] (2) Defect Information Collection

[0294] The robot moves along the surface of the pipe and scans with lidar and vision cameras to identify a linear crack that is 20 millimeters long and 2 millimeters wide.

[0295] (3) Defect identification and feature extraction

[0296] The image processing submodule extracts the crack edges using the Canny edge detection algorithm. The point cloud processing submodule reconstructs the 3D model of the crack and calculates the crack depth to be 3 mm.

[0297] (4) Repair strategy generation

[0298] A similarity matching algorithm identifies similar thermal stress crack repair cases in the defect database. A machine learning algorithm optimizes the repair strategy, determining the use of a high-strength hot-melt filler material, setting the filling temperature at 200℃, and the filling path as a gradual expansion from the crack center outwards.

[0299] (5) Repair execution

[0300] The repair execution module selects a filling tool and applies hot-melt filler material evenly through a nozzle. The motion control system moves the repair tool along the crack path, applying pressure to ensure uniform material filling.

[0301] (6) Repair quality inspection and feedback

[0302] The repair quality inspection sensor detected insufficient adhesion in the filled area, revealing some areas with inadequate adhesion. The feedback control unit analyzed the results and instructed an adjustment to the repair strategy, increasing the filling temperature to 220°C and re-performing the repair operation.

[0303] (7) Strategy Optimization

[0304] After the repair is completed, the repair quality inspection sensor confirms that the repair effect meets the standards. The data acquisition module records the repair data, and the learning and optimization module updates the repair strategy in the defect database.

[0305] Corrosion Repair

[0306] (1) Startup and initialization

[0307] After the robot is started, it completes a self-check and loads the defect database and learning system.

[0308] (2) Defect Information Collection

[0309] The robot used ultrasonic sensors to scan and detect a circular corrosion area with a diameter of 50 millimeters, made of carbon steel.

[0310] (3) Defect identification and feature extraction

[0311] The image processing submodule extracts the edges of the eroded region through morphological analysis. The point cloud processing submodule calculates the area of ​​the eroded region to be 200 square millimeters and the depth to be 5 millimeters.

[0312] (4) Repair strategy generation

[0313] Similarity matching algorithms found similar carbon steel corrosion repair cases in the defect database. Machine learning algorithms optimized the repair strategy, determining to use a layered application of a corrosion-resistant coating, with the first layer being 0.5 mm thick and the second layer 0.8 mm thick, and the curing temperature set at 180℃.

[0314] (5) Repair execution

[0315] The repair module selects a coating device and uses a spray system to evenly apply the first layer of corrosion-resistant coating. The curing device then activates an infrared curing lamp to cure the first layer of coating. The above steps are repeated to apply the second layer of coating and perform the curing process.

[0316] (6) Repair quality inspection and feedback

[0317] A surface smoothness tester measures the coating thickness and uniformity to confirm that it meets the set standards. The data acquisition module records the repair data, while the learning and optimization module updates the repair strategy and optimizes the coating thickness and curing parameters for subsequent similar repair tasks.

[0318] In summary, the adaptive power pipeline robot repair method and system based on a defect database provided in this application have the following technical effects:

[0319] (1) Significantly improved intelligence and adaptive capabilities

[0320] The drawbacks of existing technologies: Current power pipeline repair robots typically employ fixed repair strategies and cannot dynamically adjust the repair methods based on the specific characteristics of different defects (such as shape, size, and material). This results in unstable repair effects when faced with complex and diverse defects, making it difficult to meet the demands for high-quality repairs.

[0321] Advantages of this invention: This invention, through an adaptive repair strategy generation method based on defect features, combined with multimodal sensor data and machine learning algorithms, can automatically identify the specific characteristics of defects and dynamically generate the optimal repair strategy. In this way, the robot can flexibly handle various complex defects, ensuring the accuracy and consistency of each repair operation.

[0322] (2) Efficient data utilization and continuous optimization

[0323] Limitations of existing technologies: While existing repair robots may be equipped with defect databases, these databases are often static and lack real-time updates and self-learning capabilities. The selection of repair strategies relies on pre-defined rules and cannot be optimized or improved based on actual repair data.

[0324] Advantages of this invention: This invention introduces a defect database and a learning system. By continuously recording the data and results of each repair operation, machine learning algorithms are used to analyze and optimize the database. This allows the repair strategy to be continuously optimized based on historical data and actual results, improving the system's self-learning ability and intelligence level, and significantly improving repair efficiency and quality.

[0325] (3) Multimodal sensor fusion enhances defect recognition accuracy

[0326] Disadvantages of existing technologies: Existing repair robots typically rely on a single type of sensor (such as a camera or lidar) for defect detection, resulting in insufficient identification of defect features and affecting the accuracy of repair strategies.

[0327] Advantages of this invention: This invention employs multimodal sensor fusion technology, including LiDAR, a vision camera, and an ultrasonic sensor, to comprehensively collect information on the geometry, size, and material of defects. By fusing data from different sensors, the accuracy of defect identification and feature extraction is improved, providing reliable data support for the subsequent generation of adaptive repair strategies.

[0328] (4) Real-time repair quality detection and closed-loop feedback control

[0329] The shortcomings of existing technologies: After completing the repair operation, existing repair robots lack an effective quality inspection and feedback mechanism, which makes it impossible to detect and correct the deficiencies in the repair in a timely manner, resulting in inconsistent repair quality and difficulty in achieving high-standard repair results.

[0330] Advantages of this invention: This invention designs a repair quality detection and feedback module. After repair is completed, the repair effect is comprehensively evaluated using multiple sensors, including force sensors, adhesion force detectors, and surface smoothness detectors. The detection results are transmitted to the defect database and learning system via a feedback control unit, enabling real-time adjustment and optimization of the repair strategy. This closed-loop control mechanism ensures the quality of each repair operation, significantly improving the reliability and consistency of the repair process.

[0331] (5) Highly modular design and strong system scalability

[0332] Disadvantages of existing technologies: Existing repair robots often lack modularity in their design, making it difficult to flexibly configure and expand them according to different repair needs, thus limiting their application scope and adaptability.

[0333] Advantages of this invention: This invention adopts a modular system architecture, allowing each functional module (such as the defect information acquisition module and the repair execution module) to be independently upgraded and replaced, facilitating system expansion and maintenance. Simultaneously, the system supports dynamic selection and switching of various repair tools and materials, enhancing the robot's adaptability to diverse defects and expanding its application scenarios and service life.

[0334] (6) Improve repair efficiency and reduce operating costs

[0335] Disadvantages of existing technologies: Existing repair methods rely on manual operation or require frequent manual intervention, resulting in low repair efficiency and high labor costs. Furthermore, inconsistent repair quality may necessitate multiple repair operations, further increasing operating costs.

[0336] Advantages of this invention: This invention significantly reduces the need for manual intervention and improves repair efficiency through automated defect identification, strategy generation, and execution. Simultaneously, continuously optimized repair strategies reduce the probability of repair failures and rework, lowering operating costs. Overall, this invention can complete more repair tasks in a shorter time, improving the economy and sustainability of power pipeline maintenance.

[0337] (7) Wide applicability and good environmental adaptability

[0338] Disadvantages of existing technology: Existing repair robots have poor adaptability when facing power pipes of different types and materials, and are difficult to work stably under various environmental conditions.

[0339] Advantages of this invention: Through an adaptive repair strategy and multimodal sensors, this invention can adapt to different types and materials of power pipelines, exhibiting excellent environmental adaptability. Furthermore, the robot design considers special environmental requirements such as high voltage and corrosion resistance, enabling it to operate stably under various complex working conditions and expanding its application range.

[0340] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0341] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. An adaptive electric pipeline robot repair method based on a defect database, characterized in that, The method includes: Real-time power pipeline defect data is acquired by the acquisition component, and multi-dimensional defect feature parameters are extracted from the defect data. An initial model of power pipeline defects is constructed based on the multidimensional defect feature parameters and historical defect data. The real-time power pipeline defect data is then matched with the data in the initial model to obtain an updated model of power pipeline defects. The optimal repair strategy for the real-time power pipeline defect data is then obtained based on the similarity matching value. Based on the obtained optimal repair strategy, the movement path of the repair tool is planned to ensure that the repair operation covers all defect areas. The repair tool operates and repairs according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multidimensional defect feature parameters of the corresponding power pipeline defect data. After the repair is completed, the repair quality inspection sensor performs a comprehensive inspection of the repaired area. If the inspection results are not ideal, the repair strategy adjustment algorithm is used to regenerate the best repair strategy based on the real-time power pipeline defect data, and the power pipeline defect update model and repair trajectory model are updated. Repeat the above operations to identify and repair more electrical pipeline defects and continuously optimize the repair strategy.

2. The adaptive power pipeline robot repair method based on a defect database as described in claim 1, characterized in that, The method of acquiring power pipeline defect data in real time through the acquisition component includes: The acquisition components are positioned in a suitable location to accurately collect defect data from power pipelines. The acquisition components include a lidar sensor, a vision camera, and an ultrasonic sensor. The lidar sensor is used to scan the geometry of the pipeline surface to obtain high-precision three-dimensional point cloud data. The vision camera is used to capture high-definition images of the pipeline surface to assist in identifying the visual features of defects. The ultrasonic sensor is used to detect defects inside the pipeline to obtain the geometric shape, size, and material information of the defects in real time. The collected data were filtered and denoised, and the multimodal data were fused to generate unified defect description data.

3. The adaptive power pipeline robot repair method based on a defect database as described in claim 2, characterized in that, The extraction of multidimensional defect feature parameters from the defect data includes: Edge detection and morphological analysis are performed on the high-definition images captured by the vision camera to extract two-dimensional features of defects; The three-dimensional point cloud data acquired by lidar is processed to construct a model, and the three-dimensional model of the corresponding defect is reconstructed, thereby calculating the volume and depth of the defect. A multidimensional feature model is constructed based on the two-dimensional features, the three-dimensional model, and the distance and location information of the defect collected by the ultrasonic sensor. The geometric shape, geometric features, material features, and surface texture features of the defect are extracted from the multidimensional feature model.

4. The adaptive power pipeline robot repair method based on a defect database as described in claim 1, characterized in that, The power pipeline defect update model includes: The initial model of the power pipeline defect is the storage of historical defect data in the database. Each record contains: defect type, geometric features, material properties, surface texture features, and the best repair strategy corresponding to the defect. The multidimensional defect feature parameters corresponding to the real-time power pipeline defect data are updated in the initial power pipeline defect model; The multidimensional defect feature parameters are matched with the feature vector formed by the features in each record. The repair strategy corresponding to the historical defect data with the highest similarity result is taken as the best repair strategy for the current defect.

5. The adaptive power pipeline robot repair method based on a defect database as described in claim 4, characterized in that, The historical defect data is stored in a database, including: The database has a two-level index structure, including: a first-level index that uses a clustering algorithm to initially classify historical defect data according to defect type or material characteristics, with each category forming a sub-database; and a second-level index that constructs a multi-dimensional index structure within each sub-database. The multi-dimensional index structure performs fast nearest neighbor search by recursively dividing the feature space and reducing the dimensionality of high-dimensional features.

6. The adaptive power pipeline robot repair method based on a defect database as described in claim 4, characterized in that, The multidimensional defect feature parameters are similar to the feature vector formed by the features in each record, including: The feature vector of each historical defect data record is standardized to obtain the first feature vector. The second feature vector is obtained by standardizing and reducing the dimensionality of the feature vector formed by the multidimensional defect feature parameters. Set the weights for the corresponding feature dimensions, and calculate the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector; If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching.

7. The adaptive power pipeline robot repair method based on a defect database as described in claim 6, characterized in that, The step of setting weights for corresponding feature dimensions and calculating the corresponding weighted Euclidean distance based on the first feature vector and the second feature vector includes: The second feature vector is pre-classified according to the defect type, and the corresponding sub-database is selected; In the multidimensional index structure of the selected sub-database, a nearest neighbor search is performed using the second feature vector as the query point to obtain... k The most similar historical defect record; Based on the weighted Euclidean distance, this k Sort the most similar historical defect records and select the N results with the smallest distance; Extract the corresponding set of repair strategies from the top N most similar defect records; If the set of repair strategies is greater than 1, the best repair strategy is selected by voting or weighted average strategy.

8. The adaptive power pipeline robot repair method based on a defect database as described in claim 6, characterized in that, If the repair quality is low, the weights are updated using gradient descent to continuously optimize the accuracy of similarity matching, including: After each repair is completed, the repair quality is evaluated. Assuming that the best repair strategy corresponding to the historical defect data record matched in this repair is used to repair the current defect, if the repair quality is high, it indicates that the feature weight allocation is reasonable; if the evaluation result is that the repair quality is low, it indicates that the weights of some feature dimensions in the similarity measure are set improperly, and the weights are updated through gradient descent.

9. The adaptive power pipeline robot repair method based on a defect database as described in claim 1, characterized in that, The method also includes: If the detection results are satisfactory, the repair trajectory and the corresponding repair effect in the repair trajectory model will be updated to the power pipeline defect update model to form a historical repair case record. The power pipeline defect update model is continuously optimized and updated using machine learning algorithms.

10. An adaptive power pipeline robot repair system based on a defect database, characterized in that, The system includes: The acquisition module is used to acquire real-time power pipeline defect data through the acquisition components, and to extract multi-dimensional defect feature parameters from the defect data. The repair strategy generation module is used to construct an initial model of power pipeline defects based on the multidimensional defect feature parameters and historical defect data, and to perform similarity matching between the real-time power pipeline defect data and the data in the initial model of power pipeline defects to obtain an updated model of power pipeline defects; and to obtain the optimal repair strategy for the real-time power pipeline defect data based on the similarity matching value. The repair execution module is used to plan the movement path of the repair tool according to the obtained optimal repair strategy, ensuring that the repair operation covers all defect areas. The repair tool performs repair operations according to the planned path, and a repair trajectory model is constructed based on the obtained movement path and the multi-dimensional defect feature parameters of the corresponding power pipeline defect data. The repair quality detection and feedback module is used to perform a comprehensive inspection of the repaired area by the repair quality detection sensor after the repair is completed. If the inspection result is not ideal, the best repair strategy is regenerated based on the real-time power pipeline defect data using the repair strategy adjustment algorithm, and the power pipeline defect update model and repair trajectory model are updated. The iterative module is used to repeatedly perform the above operations, identify and repair more power pipeline defects, and continuously optimize the repair strategy.

Images