Anti-corrosion control system and method for pipeline operation robot

Abstract

The invention discloses an anti-corrosion control system and method for a pipeline operation robot, and relates to the technical field of robot control, and the system is composed of a plurality of functional modules, and comprises a multi-sensor fusion module which obtains a pipeline image in real time through a camera; meanwhile, a sensor is used for dynamically collecting internal environment parameters of the pipeline; the three-dimensional modeling and positioning module is used for constructing a pipeline three-dimensional model based on an SLAM algorithm and the pipeline image and positioning the position of the pipeline where the robot is located in real time; the intelligent decision-making module is used for identifying a pipeline corrosion area and position by utilizing YOLOv5 based on the pipeline image and performing corrosion analysis on the corrosion area, and the corrosion analysis comprises corrosion thickness and form contour; measuring the corrosion thickness of the corrosion area by utilizing electromagnetic eddy current detection, and extracting the shape contour of the corrosion area by utilizing deep learning model image recognition; and triggering a self-adaptive control strategy according to the internal environment parameters of the pipeline and the pipeline position of the robot.

Description

Technical Field

[0001] The present invention relates to the technical field of robot control, and specifically to an anti-corrosion control system and method for a pipeline operation robot. Background Art

[0002] A reservoir water conveyance tunnel consists of an inlet sluice, a tunnel section, and an outlet sluice chamber. After a long time of water flow action, the steel lining of the water conveyance tunnel is severely corroded and needs anti-corrosion treatment; the inlet of the tunnel is at the bottom of the reservoir. Due to this special terrain restriction, the entry and exit of construction workers and ventilation can only be carried out unilaterally through the outlet; a certain reservoir was built in the 1970s, and the water conveyance pipeline has been eroded by a high-humidity and multi-sediment environment for a long time, and the corrosion problem is serious; under such harsh working conditions, if the traditional manual anti-corrosion operation method is adopted, not only is the safety risk extremely high, the efficiency is also very low, and the construction quality fluctuates greatly and is difficult to guarantee; in order to complete the tunnel repair task safely, efficiently, and with quality assurance, after careful consideration on site, it was decided to abandon the traditional method, adopt an unmanned operation mode, and introduce a special anti-corrosion robot to carry out the operation.

[0003] The traditional manual operation mode has long faced multiple severe challenges, and its inherent defects in the operation environment and operation process have severely restricted the construction efficiency and quality assurance; construction workers need to complete the operation in a highly restricted physical space, and the operation area usually presents complex characteristics such as narrowness, high humidity, and high dust concentration; moreover, traditional manual operations rely on experience-based judgments, with a significant lack of standardization. Construction workers need to hold electric tools to process the inner wall of the pipeline section by section, and their operation speed is limited by human endurance and equipment power. The daily effective operation duration is usually less than 6 hours; at the quality control level, the uncontrollable factors of the traditional mode are more prominent, lacking real-time monitoring means, and relying entirely on visual inspection and post-event sampling inspection, resulting in a significant lag in defect discovery. Summary of the Invention

[0004] To achieve the above objectives, the present invention is realized through the following technical solutions: An anti-corrosion control system for a pipeline operation robot, comprising: A multi-sensor fusion module that obtains pipeline images in real time through a camera; at the same time, it uses sensors to dynamically collect pipeline internal environment parameters; A three-dimensional modeling and positioning module that constructs a three-dimensional pipeline model based on the SLAM algorithm and pipeline images, and real-time locates the position of the robot in the pipeline; The intelligent decision-making module, based on the pipeline image, uses YOLOv5 to identify the corrosion area and location of the pipeline, and conducts corrosion analysis on the corrosion area. The corrosion analysis includes corrosion thickness and morphological contour. It measures the corrosion thickness of the corrosion area using electromagnetic eddy current detection, and extracts the morphological contour of the corrosion area using deep learning model image recognition. According to the internal environment parameters of the pipeline and the position of the robot in the pipeline, it triggers an adaptive control strategy. The operation execution and interaction module, based on the location and corrosion information of the corrosion area, polishes and cleans the corrosion area, and at the same time conducts corrosion analysis and makes intelligent decisions on the corrosion area until the corrosion area has been completely cleared.

[0005] Furthermore, the process of obtaining the pipeline image is as follows: The camera includes a visible light camera and an infrared camera, which obtains the pipeline image in real time. It conducts online monitoring of the image quality of the camera, and evaluates the image clarity through the PSNR (Peak Signal-to-Noise Ratio). If the detected image is blurred, it automatically switches to the thermal imaging mode; otherwise, it preferentially uses the visible light camera.

[0006] Furthermore, the process of collecting the environment parameters is as follows: It dynamically collects the internal environment parameters of the pipeline through sensors. The environment parameters include temperature and humidity, harmful gas values, moving distance, and magnetic field strength values.

[0007] Furthermore, the specific process of constructing the 3D pipeline model is as follows: Based on RGB-D vision and combined with the pipeline image, it constructs a 3D model of the pipeline: S201: Set the initial position and direction, and load the prior map of the pipeline; S202: Construct a factor graph, capture the depth information of the scene, and generate local point cloud data; fuse the local point cloud data into a global point cloud model, and use Poisson reconstruction to generate a triangular mesh model; S203: Refine the 3D model and conduct environmental semantic annotation, perform surface smoothing on the generated triangular mesh model, use a deep learning model to classify the point cloud data, and construct a dynamic environment model with the optimized 3D model and real-time sensor data.

[0008] Furthermore, the process of real-time positioning of the robot's location in the pipeline is as follows: The distance between the robot and the fixed base station is obtained by using UWB ranging technology, and the angular velocity and acceleration data of the IMU inertial measurement unit are combined; the depth image and color image of the inner wall of the pipeline are collected in real time through the RGB-D camera, ORB and SIFT feature points are extracted, and feature matching is performed with the prior map to calculate the current pose, position and direction of the robot; the local point cloud data is aligned with the global point cloud model, and the cumulative drift caused by sensor noise is corrected by minimizing the pose error; when the robot passes through the mapped area, the cumulative error is corrected through feature matching and the map gap is closed.

[0009] Further, the process of identifying the corrosion area and location of the pipeline is as follows: Identify the corrosion area and location of the pipeline: S301: Use the improved YOLOv5 model to detect pitting corrosion by introducing the BiFPN feature pyramid and the CBAM attention mechanism; S302: Input the pipeline image into the YOLOv5 model to output the bounding box and category of the corrosion area; S303: Refine the contour of the corrosion area by combining semantic segmentation, generate a pixel-level corrosion morphology map, and classify the corrosion type.

[0010] Further, the morphological contour and the corrosion thickness include: The probe scans the corrosion area and records the signal changes on the impedance plan view; using the calibration data, the thickness-impedance relationship is fitted to calculate the actual corrosion thickness; Based on the deep learning image, use the semantic segmentation model to extract the pixel-level contour of the corrosion area, and combine the geometric analysis algorithm to calculate the morphological parameters.

[0011] Further, the process of the adaptive control strategy is as follows: Through the multi-dimensional scoring model, the corrosion depth weight, the morphological complexity and the environmental deterioration factor are output; using the dynamic risk matrix, combined with the corrosion location, the medium flow rate and the historical corrosion rate, a risk heat map is generated.

[0012] Further, the process of making an intelligent decision on the corrosion area is as follows: During the grinding process, the electromagnetic eddy current detection is continuously used to verify the corrosion depth, and at the same time, the images before and after grinding are compared in real time by YOLOv5. If the corrosion area is not completely removed, the processing time is extended until it is completely removed.

[0013] An anti-corrosion control method for a pipeline operation robot includes the following steps: Step 1: Obtain the pipeline image in real time through the camera; dynamically collect the internal environment parameters of the pipeline through the sensor; Step 2: Based on the SLAM algorithm and pipeline images, construct a 3D pipeline model and real-time locate the position of the robot in the pipeline. Step 3: Based on the pipeline images, use YOLOv5 to identify the corrosion area and location in the pipeline, and conduct corrosion analysis on the corrosion area. The corrosion analysis includes corrosion thickness and morphological contour. Use electromagnetic eddy current detection to measure the corrosion thickness of the corrosion area, and use deep learning model image recognition to extract the morphological contour of the corrosion area. Trigger the adaptive control strategy based on the internal environment parameters of the pipeline and the position of the robot in the pipeline. Step 4: Based on the position and corrosion information of the corrosion area, polish and clean the corrosion area, and at the same time conduct corrosion analysis to make intelligent decisions on the corrosion area until the corrosion area has been completely removed.

[0014] The anti-corrosion control system and method for a pipeline operation robot provided by the present invention have the following beneficial effects: (1) Through the unmanned operation mode, the robot can independently complete high-risk operations such as rust removal and spraying, avoiding personnel from entering narrow, high-humidity, and high-dust environments, significantly reducing safety risks. Combining technologies such as RGB-D vision, UWB positioning, and IMU inertial navigation, the robot can achieve real-time precise positioning with an accuracy of centimeter level, construct a 3D pipeline model, provide a high-precision spatial reference for operation path planning, and at the same time, based on YOLOv5 corrosion detection, electromagnetic eddy current thickness measurement, and multi-dimensional scoring model, dynamically adjust grinding parameters such as rotation speed and pressure processing strategies to ensure that the corrosion area is completely removed without damaging the healthy area.

[0015] (2) By dynamically collecting the internal environment parameters of the pipeline through sensors such as temperature and humidity, harmful gases, and magnetic field intensity, combined with the generation of corrosion heat maps, early warning of high-risk areas, correcting the cumulative pose error, closing the map gap, ensuring that the 3D model is consistent with the actual pipeline state, supporting long-term operation and maintenance data comparison, and through double verification of electromagnetic eddy current detection and image recognition, ensuring the processing quality of the corrosion area, reducing the rework rate. At the same time, the system adopts a modular design, and the power tractor, grinding vehicle, and spraying vehicle operate independently, which can quickly adapt to different pipe diameters and operation requirements, and improve the equipment reuse rate. Description of the Drawings

[0016] Figure 1 It is a schematic diagram of the system flow of the present invention; Figure 2 It is a schematic diagram of the overall method of the present invention. Detailed Embodiments

[0017] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0018] Embodiment 1 Please refer to Figure 1 , Embodiment 1 of this application provides an anti-corrosion control system for a pipeline operation robot, and this system includes: A multi-sensor fusion module that obtains pipeline images in real time through a camera; dynamically collects pipeline internal environment parameters through sensors; Composition of the anti-corrosion robot: The total length of the robot is 6.8 meters. The device consists of a power tractor, a grinding and rust-removing vehicle, a dust-absorbing and cleaning vehicle, a spraying and anti-corrosion vehicle, as well as a wireless remote control module and a control box; Obtaining pipeline images: Among them, the camera is located in the power tractor, and real-time image transmission is carried out through a wireless bridge. There is no delay within 500 meters. At the same time, the 10-inch display of the control box can display 4 video images simultaneously; the camera includes a visible light camera and an infrared camera, and obtains image data inside the pipeline in real time; Visible light camera: A 4-million-pixel high-definition camera; Infrared camera: Used to penetrate interferences such as smoke and oil stains; The camera continuously takes images of the inner wall of the pipeline at a fixed frame rate, and cooperates with the LED lamp to ensure clear images in low-light environments; uses an infrared or thermal imaging camera to penetrate oil stains; At the same time, online monitoring of the image quality of the camera is carried out, and the image clarity is evaluated through PSNR (Peak Signal-to-Noise Ratio) or SSIM (Structural Similarity Index); if blurred images or environmental interferences are detected, such as too high smoke concentration, it automatically switches to the thermal imaging mode; Dynamically collecting pipeline internal environment parameters: Temperature and humidity sensor: Through Sending an initialization command to configure the working mode of the sensor; for example, The default I²C address is 0x38, and it is necessary to send , a soft reset command to ensure that the sensor enters the normal working state; first send a read command, wait for the sensor to complete the measurement, and then read the original data returned by the sensor, usually a 16-bit temperature value and a 16-bit humidity value. Multiply the 16-bit original value by the calibration coefficient to obtain the temperature or humidity; according to the calibration parameters provided in the sensor manual, perform secondary correction on the data; send the final result through the wireless module; Harmful gas detection sensor: For , , etc., the response time < 1s. Send an initialization command to configure the gas detection range, such as the CO concentration range of 0–500 ppm, range of 0–5000 ppm; send a read command to trigger measurement, read the raw data returned by the sensor, and directly output the ppm value. For example, the concentration range of SGP30 is 400–8192 ppm. Output the gas concentration value and set the threshold alarm. For example, when the CO concentration > 50 ppm, trigger the alarm; Ultrasonic ranging array: Connect each ultrasonic module's Trig pin to the Trig pin. Use multi-channel GPIO to control the Trig pins of 8 channels, sharing one Echo pin or assigning an independent Echo pin for each channel; perform initialization configuration to set the GPIO pins as output and input, and set a timer to measure the high-level time of the Echo signal; send a high-level pulse to the Trig pin to start ultrasonic emission. Capture the high-level duration of the Echo pin through the timer, and through multi-channel polling: trigger the measurement of 8 channels in sequence, and record the distance data of the robot's movement within the channel; Through the collaborative work of multiple ultrasonic sensors, a non-contact measurement system for the distance, position, and shape of the target object is realized; calculate the distance based on the time difference between the ultrasonic emission and reception, and improve the measurement accuracy and environmental perception ability through multi-sensor data fusion; a single ultrasonic sensor emits high-frequency ultrasonic waves (usually 40 kHz) and receives the reflected wave, and calculates the distance based on the time difference of the sound wave's round-trip and the speed of sound; For example: The typical working process of the HC-SR04 module. Send a high level to the TRIG pin → emit 8 square waves of 40 kHz → the receiver detects the echo → the ECHO pin outputs the high-level duration; The ultrasonic ranging array can cover a wider detection range by arranging multiple ultrasonic sensors at different positions, and determine the three-dimensional position of the target through triangulation or time-difference analysis; industrial robots can perceive the surrounding environment in real time through the ultrasonic array and achieve autonomous obstacle avoidance in combination with path planning algorithms; Electromagnetic field strength detector: Use the interface to connect the detector. If it is an analog output, the voltage needs to be converted to a digital signal through an ADC; send a configuration command to set the measurement range and sampling rate, and record the zero-offset value in an electromagnetic interference-free environment, obtain the magnetic field strength value returned by the sensor, convert the ADC reading to a voltage value, and convert the digital value to The unit uses a moving average or a low-pass filter to reduce high-frequency noise, outputs the real-time magnetic field intensity value, and marks abnormal fluctuations; There may be toxic and harmful gases in a closed or semi-closed pipeline environment, such as , , etc. By monitoring the concentration of these gases in real time, the safety of the robot and the operator can be ensured, and the risks of poisoning or explosion can be avoided; and through an ultrasonic ranging array, the robot can be prevented from colliding with the internal structure of the pipeline or other obstacles during movement.

[0019] The three-dimensional modeling and positioning module constructs a three-dimensional model of the pipeline based on the SLAM algorithm and pipeline images, and real-time locates the position of the robot in the pipeline; Constructing a three-dimensional model of the pipeline: By obtaining the visual information and environmental parameters inside the pipeline in real time, and combining RGB-D visual SLAM and UWB positioning technologies, using the pipeline images collected by the camera and the dynamically collected environmental parameters, the precise positioning of the robot and the mapping of the surrounding environment are completed in real time. It can not only accurately determine the real-time position of the robot in the pipeline, but also construct a complete three-dimensional model of the pipeline based on the real-time collected data, providing high-precision spatial reference and environmental perception support for subsequent operations; For the initial position and orientation of the robot, use the known map information to help initialize, extract features from the sensor data and perform matching; S201: Set the initial position and orientation, load the prior map of the pipeline (usually a two-dimensional map), and extract feature points such as ORB and SIFT from the environmental image, and calculate the pose change of the robot through feature matching; ORB: Use the FAST algorithm to detect corner feature points in the image, and assign a rotation direction to each feature point through the gray centroid method, and generate a binary feature vector in combination with the BRIEF descriptor; SIFT: Detect scale-space extreme points through the Gaussian difference pyramid, and calculate the gradient direction histogram for each key point to achieve rotation invariance, and finally generate a 128-dimensional floating-point descriptor; Feature matching: Compare the feature vectors of two images using the Hamming distance, screen out the matching pairs with the highest similarity, and remove outliers, that is, false matches, through the Lowe ratio test or the RANSAC algorithm, and retain the feature point pairs with high confidence; Pose calculation: Based on the matched feature point pairs, estimate the motion parameters of the camera or the robot relative to the environment using the fundamental matrix; extract the rotation matrix and translation vector by decomposing the matrix, and combine the prior map or the initial pose to finally determine the three-dimensional pose change of the robot in the environment; S202: Construct a factor graph to minimize the pose error, fuse the local point cloud data into a global point cloud model, and generate a triangular mesh model using the Poisson reconstruction or Marching Cubes algorithm; Factor graph optimization: Use the local point cloud data collected by the robot at different positions as nodes in the factor graph. Calculate the relative pose constraints between adjacent nodes through feature matching to establish a graph structure. Optimize the factor graph using the non-linear least squares method to correct the cumulative pose error caused by slipping, vibration, or sensor noise, ensuring the spatial consistency of the point cloud data in the global coordinate system; Point cloud fusion: Based on the optimized pose information, align the local point cloud data to the global coordinate system through rigid transformation, and use voxel filtering or octree structure to remove duplicates and merge overlapping regions to generate a continuous global point cloud model; Mesh reconstruction: Poisson reconstruction: By solving the Poisson equation, convert the normal vector information of the point cloud into an implicit function field, and then generate a smooth and watertight triangular mesh model through isosurface extraction; By dividing the three-dimensional voxel space, interpolate the point cloud density distribution and extract the isosurface to generate a triangular mesh, which is suitable for processing regularly sampled or low-noise point cloud data; S203: Refine the three-dimensional model and perform environmental semantic annotation. Smooth the surface of the generated triangular mesh model, use a deep learning model to classify the point cloud data, and construct a dynamic environment model with the optimized three-dimensional model and real-time sensor data to reflect the corrosion state inside the pipeline that changes over time; Locate the position of the robot in the pipeline in real time: Use UWB ranging technology to obtain the accurate distance between the robot and the fixed base station, with an accuracy of up to centimeter level. Combine the angular velocity and acceleration data of the IMU inertial measurement unit, and fuse multi-source data through the extended Kalman filter or particle filter algorithm to eliminate the errors of a single sensor, such as the drift of the IMU and the multipath interference of the UWB; In a complex pipeline environment, the UWB provides a global positioning reference, and the IMU compensates for the short-term motion trajectory to achieve continuous and stable positioning; Real-time collect the depth image and color image of the inner wall of the pipeline through an RGB-D camera, extract feature points such as ORB and SIFT, perform feature matching with the prior map (the two-dimensional map in S201), and calculate the current pose, position, and direction of the robot; Align the local point cloud data with the global point cloud model, and correct the cumulative drift caused by sensor noise or motion uncertainty by minimizing the pose error; When the robot passes through the mapped area, identify repeated scenes through feature matching, such as loop detection, correct the cumulative error, and close the map gap; Example: The water conveyance tunnel of a certain reservoir is a steel pipe with a diameter of 1300 mm. Due to long-term sediment erosion, the inner wall is severely corroded. In the repair project, a robot is equipped with an RGB-D camera to collect depth images and color images of the inner wall of the pipeline in real time, and combined with a prior two-dimensional map for positioning and mapping; The RGB-D camera collects 60 frames of images per second with a depth accuracy of ±2 cm. ORB feature points and SIFT feature points are extracted from the color images. When the robot moves in the straight section of the pipeline, the threaded joint feature points on the pipe wall are quickly detected by ORB. In the elbow area, SIFT identifies irregular spots formed by corrosion as the basis for robust matching. The feature points of the current frame are matched with the prior two-dimensional map, and high-confidence matching pairs are screened using the Hamming distance. After removing outliers through the RANSAC algorithm, the PnP algorithm is used to calculate the rotation matrix and translation vector of the camera relative to the prior map to determine the current pose of the robot. If the cumulative pose error causes the robot to misjudge as deviating from the pipeline axis, the pose is corrected through the matched elbow feature points; The RGB-D camera generates local point clouds every 1 m of movement, covering 360° around the entire circumference of the pipeline. The ICP algorithm is used to align the local point cloud with the global point cloud model, and the Euclidean distance between the point clouds is minimized to correct sensor noise or motion drift. For example, in the outlet gate chamber section of the pipeline, the local point cloud of the robot is offset by 2 cm due to sliding and is restored to the designed axis position after ICP alignment. When the robot returns to the mapped area, the feature points of the current frame are quickly matched with the historical map feature library through the bag-of-words model to identify repeated scenes. Through factor graph optimization, constraints are established between the current pose and the prior pose to correct the cumulative drift and close the map gap. After closing, the global point cloud model automatically merges the repeated areas to generate a continuous and seamless three-dimensional model of the pipeline.

[0020] The intelligent decision-making module, based on pipeline images, uses YOLOv5 to identify the corrosion areas and locations of the pipeline, and conducts corrosion analysis on the corrosion areas. The corrosion analysis includes corrosion thickness and morphological contour. The electromagnetic eddy current detection is used to measure the corrosion thickness of the corrosion area, and the deep learning model image recognition is used to extract the morphological contour of the corrosion area. According to the internal environment parameters of the pipeline, combined with the position of the robot in the pipeline, an adaptive control strategy is triggered; Identify the corrosion areas and locations of the pipeline: S301: Use an improved YOLOv5 model to improve the detection accuracy of pitting corrosion by introducing the BiFPN feature pyramid and the CBAM attention mechanism. Among them, the training data set contains labeled images covering various corrosion types; S302: Input the pipeline image into the YOLOv5 model, and output the bounding boxes and categories of the corrosion areas, such as cracks, holes, and corrosion patches; S303: Refine the contour of the corroded area by combining semantic segmentation to generate a pixel-level corrosion morphology map; and perform corrosion type classification. Classify the detected areas according to 7 types of corrosion labels, such as fracture, surface corrosion, deformation, etc. Corrosion thickness: An alternating magnetic field is emitted by an eddy current probe to induce eddy currents on the metal surface. The material thickness is deduced based on the change in eddy current impedance. The probe scans the corroded area and records the signal changes on the impedance plan view. Using calibration data, such as reference samples with different thicknesses, fit the thickness-impedance relationship to calculate the actual corrosion thickness with an error ≤ 0.1 mm. Example: The eddy current probe emits a high-frequency alternating magnetic field through an electromagnetic coil. This magnetic field penetrates the surface of the metal material and induces eddy currents related to the material properties. The thickness and conductivity of the pipeline affect the flow path and intensity of the eddy currents, resulting in changes in the impedance of the probe coil. The change in impedance value is recorded by a high-precision sensor. The eddy current probe is moved to scan the corroded area point by point to generate an impedance plan view, such as a phase-amplitude diagram. The signal differences in different areas reflect the non-uniformity of the material thickness. Experiments are carried out using standard samples with known thicknesses, such as metal test blocks with different thicknesses, to measure their corresponding impedance values and establish a mathematical model of thickness and impedance, such as linear regression or polynomial fitting. Substitute the impedance data of the scanned corroded area into the calibration model to reverse-calculate the actual thickness values of each point, so that the detection system error is controlled within 0.1 mm to meet the high-precision detection requirements. Morphological contour: Based on deep learning images, use a semantic segmentation model to extract the pixel-level contour of the corroded area, and combine geometric analysis algorithms to calculate morphological parameters, such as corrosion volume. Map the bounding boxes of YOLOv5 and the thickness data of electromagnetic eddy currents to the same coordinate system to generate the thickness and morphological features of the three-dimensional corrosion model. Adaptive control strategy: Output the corrosion depth weight, morphological complexity, and environmental deterioration factor through a multi-dimensional scoring model. Corrosion depth weight: A thickness loss ≥ 30% of the wall thickness is a high risk, 10% - 30% is a medium risk, and < 10% is a low risk. Morphological complexity: Irregular corrosion, such as cracks and pits, needs to be processed first, and regular corrosion, such as uniform thinning, can be postponed. Environmental deterioration factor: High High concentration, high SRB bacteria content, or high-temperature and high-pressure environment requires increasing the processing priority. Use a dynamic risk matrix, combined with the corrosion location, medium flow rate, and historical corrosion rate, to generate a risk heat map. Table 1: Strategy classification and parameter mapping table: Dynamic adjustment logic: Force control based on corrosion thickness: Dynamically adjust the mechanical arm pressure through the PID algorithm. For example, when the thickness loss is 20%, the pressure linearly increases from 0.5 MPa to 1.5 MPa; Path optimization based on morphology: Adopt a "Z" - shaped trajectory to cover grooved corrosion and perform fixed - point multiple impacts on pitting corrosion; Environmental adaptive compensation: Reduce the grinding speed in high - temperature (>60°C) or high - humidity environments to avoid overheating damage. When the concentration > 100,000 mg / L, increase the spraying frequency of the corrosion inhibitor; The operation execution and interaction module, based on the location and corrosion information of the corrosion area, grinds and cleans the corrosion area, and at the same time conducts corrosion analysis to make intelligent decisions on the corrosion area until the corrosion area is completely cleared; During the grinding process, continuously use electromagnetic eddy current detection to verify the thickness recovery effect, and at the same time, use YOLOv5 to compare the images before and after grinding in real - time. If the corrosion area is not completely cleared, extend the processing time; Safety threshold protection: Maximum processing time limit: Single - time grinding does not exceed 30 seconds to avoid excessive wear of healthy areas; If abnormal pipeline vibration or sudden temperature rise is detected, immediately pause the operation; Judge the complete clearance of the corrosion area and conduct multi - source data verification; Use electromagnetic eddy current detection to confirm that the thickness of the corrosion area has recovered to more than 95% of the original value and stop grinding the corrosion area; Through image recognition, use YOLOv5 to detect no residual corrosion patches and stop grinding.

[0021] Example 2 Please refer to Figure 2 , based on Example 1, Embodiment 2 of this application also provides an anti - corrosion control method for a pipeline operation robot, including the following specific steps: Step 1: Obtain pipeline images in real - time through a camera; Dynamically collect pipeline internal environment parameters through sensors; Step 2: Based on the SLAM algorithm and pipeline images, construct a three - dimensional model of the pipeline and real - time locate the position of the robot in the pipeline; Step 3: Based on the pipeline images, use YOLOv5 to identify the corrosion area and location of the pipeline, and conduct corrosion analysis on the corrosion area. The corrosion analysis includes corrosion thickness and morphological contour; Use electromagnetic eddy current detection to measure the corrosion thickness of the corrosion area, use deep - learning model image recognition to extract the morphological contour of the corrosion area; According to the pipeline internal environment parameters, combined with the position of the robot in the pipeline, trigger the adaptive control strategy; Step 4: Based on the location and corrosion information of the corroded area, polish and clean the corroded area, and at the same time conduct corrosion analysis to make an intelligent decision on the corroded area until the corroded area has been completely removed.

[0022] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those of ordinary skill in the art will realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in hardware or software depends on the specific application and design constraints of the technical solution.

[0023] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0024] The above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present application, and all of them should be covered by the protection scope of the present application.

Claims

1. An anti-corrosion control system for a pipeline operation robot, characterized in that, The system includes: A multi-sensor fusion module that obtains pipeline images in real time through a camera; meanwhile, it uses sensors to dynamically collect the internal environment parameters of the pipeline; A 3D modeling and positioning module that constructs a 3D model of the pipeline based on the SLAM algorithm and pipeline images, and real-time locates the position of the robot in the pipeline; An intelligent decision-making module that, based on pipeline images, uses YOLOv5 to identify the corrosion areas and positions in the pipeline, and conducts corrosion analysis on the corrosion areas. The corrosion analysis includes corrosion thickness and morphological contour; it measures the corrosion thickness of the corrosion area using electromagnetic eddy current detection, and extracts the morphological contour of the corrosion area using deep learning model image recognition; according to the internal environment parameters of the pipeline and combined with the position of the robot in the pipeline, it triggers an adaptive control strategy; An operation execution and interaction module that, based on the position and corrosion information of the corrosion area, polishes and cleans the corrosion area, and at the same time conducts corrosion analysis and makes intelligent decisions on the corrosion area until the corrosion area has been completely removed.

2. The anti-corrosion control system for a pipeline operation robot according to claim 1, characterized in that, The process of obtaining the pipeline image is as follows: The camera includes a visible light camera and an infrared camera, which obtains pipeline images in real time; it conducts online monitoring of the image quality of the camera, and evaluates the image clarity through the PSNR peak signal-to-noise ratio; if a blurred image is detected, it automatically switches to the thermal imaging mode; otherwise, it preferentially uses the visible light camera.

3. The anti-corrosion control system for a pipeline operation robot according to claim 1, characterized in that, The process of collecting the environment parameters is as follows: Through sensors, it dynamically collects the internal environment parameters of the pipeline; the environment parameters include temperature and humidity, harmful gas values, moving distance, and magnetic field strength values.

4. The anti-corrosion control system for a pipeline operation robot according to claim 3, characterized in that, The specific process of constructing the 3D model of the pipeline is as follows: Based on RGB-D vision, combined with pipeline images, it constructs a 3D model of the pipeline: S201: Set the initial position and direction, and load the prior map of the pipeline; S202: Construct a factor graph, capture the depth information of the scene, and generate local point cloud data; fuse the local point cloud data into a global point cloud model, and use Poisson reconstruction to generate a triangular mesh model; S203: Refine the 3D model and conduct environmental semantic annotation, perform surface smoothing on the generated triangular mesh model, use a deep learning model to classify the point cloud data, and construct a dynamic environment model with the optimized 3D model and real-time sensor data.

5. The anti-corrosion control system for a pipeline operation robot according to claim 4, characterized in that, The process of real-time locating the position of the robot in the pipeline is as follows: It uses UWB ranging technology to obtain the distance between the robot and the fixed base station, combined with the angular velocity and acceleration data of the IMU inertial measurement unit; through the RGB-D camera, it real-time collects the depth image and color image of the inner wall of the pipeline, extracts ORB and SIFT feature points, and conducts feature matching with the prior map to calculate the current pose, position, and direction of the robot; aligns the local point cloud data with the global point cloud model, and corrects the cumulative drift caused by sensor noise by minimizing the pose error; when the robot passes through the mapped area, it corrects the cumulative error and closes the map gap through feature matching.

6. The anti-corrosion control system for a pipeline operation robot according to claim 1, characterized in that, The process of identifying the corrosion areas and positions in the pipeline is as follows: Identify the corrosion areas and positions in the pipeline: S301: Detect pitting corrosion using an improved YOLOv5 model by introducing the BiFPN feature pyramid and the CBAM attention mechanism. S302: Input the pipeline image into the YOLOv5 model to output the bounding boxes and classes of the corroded areas. S303: Refine the contour of the corroded area by combining semantic segmentation to generate a corrosion morphology map and classify the corrosion types.

7. The anti-corrosion control system for a pipeline operation robot according to claim 6, characterized in that, The morphological contour and the corrosion thickness include: The probe scans the corroded area and records the signal changes on the impedance plan view; using the calibration data, fit the thickness-impedance relationship to calculate the actual corrosion thickness. Based on the deep learning image, use the semantic segmentation model to extract the pixel-level contour of the corroded area, and combine the geometric analysis algorithm to calculate the morphological parameters.

8. The anti-corrosion control system for a pipeline operation robot according to claim 7, characterized in that, The process of the adaptive control strategy is as follows: Output the corrosion depth weight, morphological complexity, and environmental deterioration factor through a multi-dimensional scoring model; use the dynamic risk matrix, combine the corrosion location, medium flow rate, and historical corrosion rate to generate a risk heat map.

9. The anti-corrosion control system for a pipeline operation robot according to claim 1, characterized in that, The process of making an intelligent decision on the corroded area is as follows: Continuously use electromagnetic eddy current detection to verify the corrosion depth during the grinding process, and at the same time compare the images before and after grinding in real time through YOLOv5. If the corroded area is not completely cleared, extend the processing time until it is cleared.

10. A corrosion prevention control method for a pipeline operation robot, characterized in that, It includes the following steps: Step 1: Obtain the pipeline image in real time through the camera; dynamically collect the internal environment parameters of the pipeline through the sensor. Step 2: Based on the SLAM algorithm and the pipeline image, construct a 3D pipeline model and real-time locate the position of the robot in the pipeline. Step 3: Based on the pipeline image, use YOLOv5 to identify the corroded area and location of the pipeline, and conduct corrosion analysis on the corroded area. The corrosion analysis includes corrosion thickness and morphological contour; use electromagnetic eddy current detection to measure the corrosion thickness of the corroded area, and use deep learning model image recognition to extract the morphological contour of the corroded area; according to the internal environment parameters of the pipeline, combined with the position of the robot in the pipeline, trigger the adaptive control strategy. Step 4: Based on the position and corrosion information of the corroded area, grind and clean the corroded area, and at the same time conduct corrosion analysis to make an intelligent decision on the corroded area until the corroded area has been completely cleared.

Images