Visual inspection robot for detecting inner surface defects of oil and gas pipeline

Abstract

The invention relates to a visual inspection robot for oil and gas pipeline inner surface defect detection, and belongs to the technical field of pipeline detection, the visual inspection robot comprises a power unit, an image acquisition unit and a rotary cleaning unit, the power unit is used for driving the robot to move in an oil and gas pipeline; the image acquisition unit comprises a camera and a processing module; the camera is arranged at the tail end of the image acquisition unit and used for shooting oil and gas pipeline inner surface images; the camera comprises a glass lens; the processing module is used for performing defect detection according to the shot inner surface image of the oil and gas pipeline; the power unit comprises a leather cup driving unit, and the visual inspection robot is driven to move in the pipeline through the pressure difference between the rear end and the front end of the leather cup driving unit; the rotary cleaning unit drives the glass lens to rotate, and stains on the outer side of the glass lens are thrown away.

Description

Technical Field

[0001] The present invention belongs to the technical field of pipeline detection, and particularly relates to a vision detection robot for detecting internal surface defects of oil and gas pipelines. Background Art

[0002] The internal detection of oil and gas pipelines is crucial for ensuring pipeline safety. The internal detection of oil and gas pipelines is of great importance for ensuring pipeline safety because during long-term use, oil and gas pipelines may suffer from a series of problems such as corrosion, cracks, blockages, etc. If these problems are not discovered and addressed in a timely manner, they may lead to serious safety hazards and even accidents.

[0003] Currently, the main method for oil and gas pipeline detection robots is image detection. By taking pictures of the internal pipeline images through a camera, and then judging whether there are defects on the pipeline surface based on the images.

[0004] When operating inside the pipeline, it is easily affected by internal pipeline stains. Especially in those environments that are wet, greasy or contain corrosive substances, stains are likely to adhere to the lens of the robot's camera. When these stains drip or accumulate on the lens, it will significantly reduce the light transmittance and shooting quality of the camera, resulting in blurred or distorted images. This decline in image quality directly affects the identification and analysis of pipeline defects, leading to inaccurate or missed detection results, greatly increasing potential risks. Summary of the Invention

[0005] To solve the above problems in the prior art, the present invention provides a vision detection robot for detecting internal surface defects of oil and gas pipelines.

[0006] To achieve the above object, the technical solution adopted by the present invention is: The present invention provides a vision detection robot for detecting internal surface defects of oil and gas pipelines, including: a power unit, an image acquisition unit, and a rotary cleaning unit; The power unit is used to drive the robot to move inside the oil and gas pipeline; The image acquisition unit includes: a camera and a processing module; The camera is arranged at the end of the image acquisition unit and is used to take pictures of the internal surface of the oil and gas pipeline; the camera includes a glass lens; The processing module is used to detect defects based on the pictures taken of the internal surface of the oil and gas pipeline; The power unit includes a leather cup drive unit; Through the pressure difference between the rear end and the front end of the leather cup drive unit, the vision detection robot is driven to move inside the pipeline; The rotary cleaning unit drives the glass lens to rotate and flings off the stains outside the glass lens.

[0007] The beneficial effects of the present invention are as follows: The present invention provides a vision inspection robot for detecting internal surface defects of oil and gas pipelines. The rotating cleaning unit can drive the glass lens to rotate, fling off the stains outside the glass lens, and prevent these stains from dripping or accumulating on the lens, resulting in blurred or distorted images, ensuring the image quality, improving the accuracy of detecting surface defects of oil and gas pipelines, and reducing potential risks. Description of the Drawings

[0008] Figure 1 It is a schematic structural diagram of a vision inspection robot for detecting internal surface defects of a single-section oil and gas pipeline provided by the present invention.

[0009] Figure 2 It is a schematic structural diagram of a vision inspection robot for detecting internal surface defects of a double-section oil and gas pipeline provided by the present invention.

[0010] Figure 3 It is a schematic external structural diagram of an image acquisition unit provided by the present invention.

[0011] Figure 4 It is a schematic internal structural diagram of an image acquisition unit provided by the present invention.

[0012] Figure 5 It is a schematic internal structural diagram of an image acquisition unit from another perspective provided by the present invention.

[0013] Figure 6 It is a schematic structural diagram of a rotating cleaning unit provided by the present invention.

[0014] Figure 7 It is a schematic structural diagram of another rotating cleaning unit provided by the present invention.

[0015] Figure 8 It is a schematic structural diagram of yet another rotating cleaning unit provided by the present invention.

[0016] Description of the Reference Numerals: 1, power unit; 2, image acquisition unit; 3, anti-collision head; 4, anti-collision head base; 5, odometer wheel unit; 51, odometer wheel; 52, odometer wheel base; 53, odometer wheel bracket; 6, rear-end connection cover; 7, cross universal joint; 8, front-end connection cover; 9, tail-end seal head; 10, rear-end seal head; 11, processing module; 111, computer; 112, power control board; 12, camera; 121, glass lens; 13, battery unit; 14, leather cup drive unit; 15, rotating cleaning unit; 151, rotating mechanism housing; 1511, air inlet hole; 1512, air outlet hole; 152, rotating fan blade; 153, DC motor; 154, rotating gear; 155, rotating pulley; 156, belt. Detailed Embodiments

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] The specific embodiments provided by the present invention are as follows:

[0019] Refer to the instruction manual Figures 1 to 8 An embodiment of the present invention provides a visual inspection robot for inner surface defect detection of oil and gas pipelines, comprising: a power unit 1, an image acquisition unit 2 and a rotating cleaning unit 15.

[0020] The power unit 1 is used to drive the robot to move inside the oil and gas pipeline.

[0021] Alternatively, the power unit 1 may use an electric motor to drive the robot to move through a transmission system such as gears, chains or belts.

[0022] Furthermore, the power unit 1 may have a built-in backup power supply to extend the battery life, thereby ensuring that the robot can operate stably for a long time in the pipeline.

[0023] The image acquisition unit 2 includes a camera 12 and a processing module 11 .

[0024] The camera 12 is disposed at the rear end of the image acquisition unit 2 and is used to capture images of the inner surface of the oil and gas pipeline. The camera 12 includes a glass lens 121 .

[0025] The processing module 11 is used to perform defect detection based on the captured image of the inner surface of the oil and gas pipeline. The defect detection algorithm specifically used by the processing module 11 will be described later.

[0026] The power unit 1 comprises a leather cup driving unit 14. The visual inspection robot is driven to move in the pipeline by the pressure difference between the rear end and the front end of the leather cup driving unit 14.

[0027] The rotating cleaning unit 15 drives the glass lens 121 to rotate, and removes the dirt on the outer side of the glass lens 121 .

[0028] The beneficial effects of the present invention are embodied in:

[0029] The present invention provides a visual inspection robot for detecting inner surface defects of oil and gas pipelines. The rotating cleaning unit can drive the glass lens to rotate, and remove the stains on the outer side of the glass lens to prevent these stains from dripping or accumulating on the lens to cause image blur or distortion, thereby ensuring image quality, improving the accuracy of oil and gas pipeline surface defect detection, and reducing potential risks.

[0030] In the actual application process, the vision inspection robot can adopt a multi-section structure. Taking the double-section structure as a possible implementation manner, the vision inspection robot adopts a double-section structure. The first section contains a power unit 1, and the second section contains an image acquisition unit 2, a leather cup driving unit 14, and a rotary cleaning unit 15. The first section and the second section are connected by a hinge. Specifically, the rear connection cover 6 of the first section is connected to the front connection cover 8 of the second section through a cross universal joint 7.

[0031] In the present invention, a multi-section structure is adopted. When the pipe diameter changes, the robot can adjust the angle and shape through two hinged segments, so as to pass through the pipe more smoothly and avoid jamming problems caused by narrow or curved pipes. This design not only improves the passing performance of the robot in complex pipes, but also enhances the overall stability and flexibility, ensuring that the robot can efficiently execute tasks in a changing pipe environment and perform high-quality image acquisition, thereby improving the accuracy and reliability of defect detection.

[0032] Furthermore, the external structure of the vision inspection robot is a pressure-bearing cylinder body, which is made of stainless steel material to withstand the high-pressure environment inside the pipe. Both ends of the cylinder body are open, and a flange is installed at the front end opening to fix the entire system structure and ensure stability during operation in the pipe.

[0033] Furthermore, sealed aluminum alloy sealing heads are respectively installed at both ends of the pressure-bearing cylinder body, that is, the tail-end sealing head 9 and the rear-end sealing head 10 are sealed by adding sealing rings between the sealing head and the cylinder body. In addition to being responsible for sealing, the tail-end sealing head is also used to install aviation plugs for connecting with external lines to achieve data transmission.

[0034] In a possible implementation manner, the rotary cleaning unit 15 includes: a rotary mechanism housing 151 and a rotary fan blade 152. An air inlet hole 1511 and an air outlet hole 1512 are provided on the rotary mechanism housing 151, and an air flow channel is provided between the air inlet hole 1511 and the air outlet hole 1512. A rotary fan blade 152 is arranged in the air flow channel. The rotary fan blade 152 is fixedly connected to the glass lens 121. The air flow at the rear end of the leather cup driving unit 14 enters the air flow channel through the air inlet hole 1511, drives the rotary fan blade 152 to rotate, and then drives the glass lens 121 to rotate, so as to throw off the stains outside the glass lens 121.

[0035] In the present invention, the airflow of the rotary cleaning unit 15 drives the rotary fan blades. The force of the airflow can efficiently and continuously drive the rotary fan blades, not only ensuring the stability of the cleaning effect, but also avoiding the influence on the image quality due to the accumulation of stains, thereby improving the image clarity and detection accuracy. This design requires no manual intervention and can automatically clean the lens during the operation of the robot, reducing the maintenance cost and improving the reliability of long-term operation.

[0036] In the actual application process, due to the unstable air pressure in the pipeline, there may be a phenomenon that the pressure difference at both ends of the rubber cup of the rear-end robot is large while the pressure difference at both ends of the rubber cup of the front-end robot is small. As a result, the rear-end robot pushes the front-end robot forward, and the cross universal joint coupling in the middle of the two works in a folded state, affecting the overall movement of the robot and the shooting effect of the rear-end robot. Therefore, a rotary cleaning mechanism with a rear-end motor version is designed. The sealing rubber cup of the rear-end robot is changed to a supporting rubber cup, and the rotary fan blade is redesigned as a rotary gear. The rotary gear is driven to rotate by the gear installed on the motor, thereby realizing the cleaning of the rear-end part of the rear-end robot.

[0037] In a possible implementation manner, the rotary cleaning unit 15 includes: a rotary mechanism housing 151, a DC motor 153, and a rotary gear 154. The DC motor 153 and the rotary gear 154 are arranged inside the rotary mechanism housing 151. The DC motor 153 is drivingly connected to the rotary gear 154. The glass lens 121 is fixedly connected to the rotary gear 154. The DC motor 153 drives the rotary gear 154 to rotate, and then drives the glass lens 121 to rotate, throwing off the stains on the outside of the glass lens 121.

[0038] In the present invention, the DC motor 153 drives the rotary gear 154, and then drives the glass lens 121 to rotate, thereby removing the stains on the outside of the glass lens 121. The DC motor 153 provides stable power, and the rotary gear 154 can efficiently transmit the rotational force, ensuring the continuous rotation of the glass lens 121 and removing the stains. This driving method has a simple and reliable structure and can automatically clean the glass lens 121 during the operation of the robot, avoiding the decline of image quality caused by the accumulation of stains, ensuring the image clarity and detection accuracy, and thus improving the effect of surface defect detection of the oil and gas pipeline and the long-term stable operation ability of the robot.

[0039] In a possible implementation, the rotary cleaning unit 15 includes: a rotary mechanism housing 151, a DC motor 153, and a rotary pulley 155. A DC motor 153, a rotary pulley 155, and a belt 156 are disposed inside the rotary mechanism housing 151. The DC motor 153 is drivingly connected to the rotary pulley 155 through the belt 156. The glass lens 121 is fixedly connected to the rotary pulley 155. The DC motor 153 drives the rotary pulley 155 to rotate through the belt 156, and further drives the glass lens 121 to rotate, so as to fling off the stains on the outer side of the glass lens 121.

[0040] In the present invention, belt drive has the advantages of simple structure, stable operation, low noise, strong buffering and vibration damping ability, etc., which can effectively reduce the vibration and wear caused by direct rigid connection and extend the system life. At the same time, this cleaning mechanism can continuously work during the operation of the device, ensuring that the lens is always clean and transparent, guaranteeing the image acquisition quality, and improving the accuracy and stability of oil and gas pipeline defect detection. In a possible implementation, the vision inspection robot further includes: a battery unit 13, a mileage wheel unit 5, and an LED supplementary lighting unit.

[0041] The battery unit 13 is used to supply power to the camera 12 and the processing module 11.

[0042] The mileage wheel unit 5 is used to record the movement mileage.

[0043] The LED supplementary lighting unit is used to supplement light to the inside of the oil and gas pipeline.

[0044] It should be noted that in the vision inspection robot, the design of the power system is a key part. The battery is first connected to the switch circuit, and the load end of the switch circuit is connected to the power supply part of the power circuit. The power circuit is responsible for providing the required voltages for the mileage wheel, the camera optocoupler input, the industrial camera, and the pocket computer.

[0045] Specifically, the mileage wheel and the camera optocoupler input obtain a 5V voltage, and the industrial camera and the pocket computer respectively obtain a 12V voltage. To prevent the pocket computer from shutting down due to the decrease in the battery voltage during the lighting of the LED lamp, a farad capacitor is connected in parallel across the battery to provide an instantaneous high current output and ensure the stability of the power supply when the device is working under high load.

[0046] In a possible implementation, it is characterized in that the glass lens 121 on the camera 12 is made of sapphire material.

[0047] Among them, the sapphire material can withstand a pressure of up to 10 Mpa, ensuring the protection of the internal system in a high-pressure environment and preventing damage to the camera 12, etc. The protection of the glass lens 121 made of sapphire material enables the camera 12 to work in a high-pressure environment of 10 Mpa.

[0048] In the present invention, the hardness and pressure resistance of the sapphire material effectively protect the camera lens and the internal system, preventing damage or breakage in harsh environments, and ensuring the reliability and long-term stability of image acquisition. It not only improves the durability of the device, but also extends the service life of the camera, ensuring continuous high-quality defect detection in high-pressure and harsh environments.

[0049] In a possible implementation manner, the processing module 11 includes: a computer 111, a power control board 112, and an LED control board. The power control board 112 is connected to the battery unit 13. The power control board 112 and the camera 12 are both connected to the computer 111. The camera 12, the odometer wheel unit 5, the LED control board, and the LED fill light unit are all connected to the power control board. The LED control board is used to implement the fill light function of the LED fill light unit. By configuring the LED fill light unit, clear images can be ensured to be captured in low-light environments.

[0050] In the present invention, when the light in the oil and gas pipeline is insufficient for the robot, the LED fill light unit can provide sufficient light sources to ensure that the camera 12 can capture clear and bright images. This design not only improves the quality of image acquisition, but also enhances the ability to perform detection tasks in complex and low-light environments, ensuring the accuracy and reliability of defect detection.

[0051] In a possible implementation manner, the odometer wheel unit 5 includes: an odometer wheel 51, an odometer wheel base 52, an odometer wheel bracket 53, a rotating shaft, a magnet, and a magnetic encoder. The odometer wheel bracket 53 is connected to the odometer wheel base 52. The odometer wheel 51 is arranged on the odometer wheel bracket 53 through the rotating shaft. A radially magnetized magnet is installed on the rotating shaft, and the magnet is located at the geometric center of the rotating shaft. When the vision inspection robot runs in the pipeline, the odometer wheel 51 contacts the inner wall of the pipeline and drives the rotating shaft connected by a key to rotate. When the rotating shaft rotates, the magnet rotates with the rotating shaft, and the magnetic encoder senses the position change of the magnet and generates a pulse signal, and triggers the camera 12 to capture an image through the pulse signal.

[0052] It should be noted that when the system runs in the pipeline, the odometer wheel contacts the inner wall of the pipeline and rotates therewith. The magnetic encoder can sense the mileage of the odometer wheel running and generate a pulse signal by cooperating with the radially magnetized magnet installed on the shaft of the odometer wheel. These pulse signals are then sent to the optocoupler input of the industrial camera. The optocoupler input outputs a photographing pulse signal according to the received number of pulses and a preset number of pulses.

[0053] Further, when the optocoupler input collects a preset number of pulse signals, that is, when the odometer wheel has run a fixed mileage, the camera 12 will send a photographing pulse signal to the power supply circuit. After receiving this pulse signal, the MOS transistor controls the battery to be directly connected to the LED light board, so that the LED supplementary lighting unit is lit. At the same time, the camera 12 starts image acquisition to ensure that clear images of the inner wall of the pipeline can be captured even in low-light environments.

[0054] In the present invention, by using the odometer wheel unit in combination with a magnetic encoder, the movement mileage of the visual inspection robot in the pipeline can be accurately recorded. The odometer wheel contacts the inner wall of the pipeline, drives the rotating shaft to rotate, the magnet rotates accordingly, and the magnetic encoder generates a pulse signal and triggers the camera to take a picture. This design can ensure that image acquisition is synchronized with the movement of the robot, enabling the camera to automatically take pictures after a predetermined movement distance, effectively improving the accuracy and timeliness of image acquisition. In addition, combining with the LED supplementary lighting unit ensures that clear images can be obtained even in low-light environments, improving the efficiency and reliability of the detection task and ensuring that defects on the inner wall of the pipeline can be accurately identified.

[0055] In a possible implementation manner, the visual inspection robot further includes: a collision avoidance head 3. The collision avoidance head is arranged at the front end of the power unit 1 and is used to prevent the visual inspection robot from colliding with the inner surface of the oil and gas pipeline and causing damage. The collision avoidance head 3 is installed on the collision avoidance head base 4.

[0056] In the present invention, the collision avoidance head can absorb impact force, protect the running stability of the robot in a complex or irregular pipeline environment, reduce the occurrence of equipment failures, extend the service life of the robot, and ensure the smooth progress of the detection task at the same time.

[0057] In a possible implementation manner, the visual inspection robot further includes: an eddy current detection unit. The eddy current detection unit is connected to the battery unit 13. In the forward direction of the visual inspection robot, the eddy current detection unit is in the front and the camera 12 is in the back. When the eddy current detection unit detects an abnormality on the wall surface of the oil and gas pipeline, it triggers the start signal of the camera 12, and takes a picture of the inner surface of the oil and gas pipeline through the camera 12.

[0058] In the present invention, when the eddy current detection unit discovers a problem, it can trigger the start signal of the camera and perform image acquisition in a timely manner. This design can ensure that the camera takes pictures at the position where potential defects are detected, thereby improving the accuracy and efficiency of defect detection. Identifying abnormalities in advance and accurately triggering image acquisition avoids the acquisition of invalid images, optimizes the detection process, and improves the detection ability of the robot in the pipeline. At the same time, it can ensure that the camera is only started when needed, avoiding invalid image acquisition, thereby saving energy consumption. By reducing unnecessary photographing and image processing, the robot can use the battery more efficiently, extend the running time, and improve the overall energy efficiency.

[0059] In a possible implementation, the vision detection robot further includes: a cleaning unit. In the forward direction of the vision detection robot, the cleaning unit is in the front and the camera 12 is in the rear. After the cleaning unit cleans the oil stain on the inner surface of the oil and gas pipeline, the inner surface image of the oil and gas pipeline is captured by the camera 12.

[0060] In the present invention, by configuring a cleaning unit in the vision detection robot and setting it at the front end, the oil stains and other dirt on the inner surface of the pipeline can be effectively cleaned, ensuring that the image captured by the camera reflects the true situation of the pipeline, thereby improving the accuracy of pipeline surface defect detection.

[0061] In a possible implementation, the processing module 11 is specifically configured to:

[0062] S1: Obtain the inner surface image of the oil and gas pipeline captured by the camera 12.

[0063] S2: Preprocess the inner surface image of the oil and gas pipeline.

[0064] Among them, the preprocessing includes: noise reduction processing and contrast enhancement.

[0065] Specifically, the interference of uneven light or environmental noise is removed by median filtering or adaptive filtering technology.

[0066] Furthermore, histogram equalization or gamma correction technology is used to improve the image brightness and contrast, highlighting the detailed features of the defect area.

[0067] S3: According to the preprocessed inner surface image of the oil and gas pipeline, defect detection of the inner surface of the oil and gas pipeline is performed through a convolutional neural network.

[0068] Specifically, the present invention can adopt an improved YOLOv5 network structure to achieve defect detection of the inner surface of the oil and gas pipeline.

[0069] The traditional YOLOv5 network structure mainly includes: input, backbone network, neck network, and head network. The inner surface image of the oil and gas pipeline is input. The backbone network is used to extract the image features of the inner surface image of the oil and gas pipeline. The neck network is used to fuse the extracted features. The head network is used to perform defect detection based on the fused features.

[0070] The backbone network mainly adopts the CSPDarknet53 structure. Based on the Darknet53 backbone network of YOLOv5 and drawing on the idea of CSPNet, a backbone structure with 5 CSP modules is designed. The convolutional kernel size in front of each CSP module is 3×3, and the stride is 2 for downsampling. The CSP module first divides the feature map of the basic layer into two parts and then merges them in a cross-stage manner, which can reduce the computational amount while ensuring the accuracy. Therefore, adopting the CSP network structure in the YOLOv5 backbone network has the advantages of enhancing the network's learning ability, reducing the computational bottleneck, and memory cost.

[0071] To further improve the feature extraction ability of the backbone network, a recursive gated convolution module is added to the backbone network. The recursive gated convolution module consists of a standard convolution, a linear mapping, and an element-wise multiplication. And layer normalization is added after the recursive gated convolution module. During the layer normalization process, the mean and variance of all channels are calculated and then normalized. The recursive gated convolution module adjusts the number of channels of the incoming feature map through two convolutional layers, and then divides the output feature of the depthwise separable convolution into multiple parts. Each part performs an element-wise multiplication operation with the previous part, and finally the output feature is obtained. Through the element-wise multiplication and recursive design, the interaction and fusion of high-order and low-order information of the feature map are realized, so that the information contained in the feature map is more abundant, the gradient diffusion phenomenon is reduced, and the feature extraction ability of the network is enhanced. The recursive operation is realized by continuously performing element-wise multiplication.

[0072] Furthermore, the recursive gated convolution module is specifically:

[0073]

[0074] Among them, x represents the input feature map, represents the linear projection mapping, which is used to mix channel information, and p 0 and q 0 represent two projection features segmented from the input feature map, f represents the depth convolution operation, represents the element-wise multiplication, and p 1 represents the interaction feature, represents the inverse linear projection mapping, which is used to remap the interaction feature back to the original channels of the input.

[0075] Furthermore, a recursive design can be introduced for multi-order mapping:

[0076]

[0077] Among them, p 0 and q 0 and …, q n-1Represents a set of projected features segmented from the input feature map.

[0078] Then perform gated convolution recursively:

[0079]

[0080] where p k+1 represents the (k + 1)-th order interaction feature, f k represents the k-th order depth convolution operation, q k represents the k-th order projected feature, g k represents a linear mapping used to match the k-th order feature dimension, and α represents a scaling factor.

[0081]

[0082] where, Identity represents the identity function, Linear represents the linear mapping function, C k-1 represents the number of feature channels of the (k - 1)-th order, C k represents the number of feature channels of the k-th order, C represents the number of feature channels of the input feature map, n represents the total order of the recursive gated convolution, and β represents the channel reduction factor, generally taken as 2.

[0083] In the present invention, by adding a recursive gated convolution module to the backbone network and combining layer normalization and multi-order recursive design, the feature extraction ability of the network can be effectively improved. The recursive gated convolution module realizes the interaction and fusion of high-order and low-order information of the feature map through element-wise multiplication, making the feature representation more abundant, while reducing the gradient diffusion phenomenon and enhancing the stability of the model. Layer normalization further standardizes the feature distribution and optimizes the network training process. In addition, through multi-order recursive design and gradual adjustment of the number of channels, the deep interaction of features and the efficient integration of information are realized. While ensuring that the computational cost is controllable, the accuracy of feature extraction and the model's representation ability for complex patterns are greatly improved.

[0084] In order to further improve the feature extraction ability of the backbone network for the inner surface defects of oil and gas pipelines and at the same time suppress the interference of useless features, the present invention introduces a coordinate attention mechanism, which takes into account the relationship between position information and channels, and can not only capture cross-channel information, but also capture direction-aware and position-aware information, so that the model can more accurately locate and identify the target area.

[0085] The coordinate attention mechanism decomposes global pooling into a pair of one-dimensional feature encoding operations. Given the input X, using pooling kernels of size (H, 1) and (1, W), each channel is encoded along the horizontal and vertical coordinates respectively, and the coordinate information embedding transformation is specifically:

[0086]

[0087] Among them, z h represents the horizontal direction perception feature map after horizontal pooling, represents the feature map of the c-th channel after horizontal pooling, x c represents the input feature map in the c-th channel, h represents the channel height, i represents the width direction index value, W represents the feature map width, z w represents the feature map after vertical pooling, represents the vertical direction perception feature map of the c-th channel after vertical pooling, w represents the channel width, j represents the height direction index value, H represents the feature map height.

[0088] Through the above two transformation operations, a pair of direction (X direction and Y direction) perception feature maps are obtained.

[0089] Input the X direction and Y direction perception feature maps into the shared 1×1 convolution transformation function to generate an intermediate feature map:

[0090]

[0091] Among them, f represents the intermediate feature map, δ represents the non-linear activation function, F 1 represents the 1×1 convolution transformation function, z h represents the horizontal direction perception feature map, z w represents the vertical direction perception feature map.

[0092] Decompose the intermediate feature map along the spatial dimension into two independent tensors to obtain two independent feature maps.

[0093] Use two 1×1 convolutions to convert the two independent feature maps into the same number of channels as the input respectively:

[0094]

[0095] Among them, g h represents the horizontal direction attention weight map, g w represents the vertical direction attention weight map, σ represents the Sigmoid activation function, F h represents the horizontal direction 1×1 convolution function, represents the horizontal direction independent feature map obtained by decomposition, F w represents the vertical direction 1×1 convolution function, represents the vertical direction independent feature map obtained by decomposition.

[0096] Process the input feature map according to the attention weight map:

[0097]

[0098] Among them, y c represents the output feature map after introducing coordinate attention in the c-th channel, represents the horizontal direction attention weight map in the c-th channel, represents the vertical direction attention weight map in the c-th channel.

[0099] In the present invention, by introducing a coordinate attention mechanism into the backbone network, the network's ability to extract features of inner surface defects of oil and gas pipelines is effectively improved, and at the same time, the interference of useless features is suppressed. The coordinate attention mechanism combines the relationship between position information and channels, and captures direction-aware and position-aware information by decomposing global pooling into one-dimensional feature encodings in the horizontal and vertical directions. Through the generation and fusion of horizontal and vertical direction attention weights, the network can more accurately locate and identify the target area, enhancing the accuracy and pertinence of feature expression, and enabling the model to have stronger robustness and performance in complex detection tasks.

[0100] The neck network adopts an FPN+PAN structure, where the FPN (Feature Pyramid Network) layer transmits strong semantic features from top to bottom, and the PAN (Path Aggregation Network) layer transmits strong localization features from bottom to top, aggregating features for different detection layers at different backbone layers to improve the feature extraction ability.

[0101] The head network can use the Softmax activation function for defect classification.

[0102] Optionally, the main defect categories include: deformation, dislocation, breakage, corrosion, oxide layer peeling, sediment, and penetration, etc.

[0103] In order to improve the fault detection accuracy of the convolutional neural network, it is necessary to optimize the network parameters of the convolutional neural network. The traditional method uses the gradient descent method for optimization. However, the optimization result of the gradient descent method usually depends on the selection of the initial parameters. Poor initial values may lead to low optimization efficiency or unsatisfactory results. And the gradient descent method is prone to falling into local optimal solutions in high-dimensional non-convex optimization problems and cannot find the global optimal solution. Therefore, the present invention adopts an improved heuristic algorithm (honey badger optimization algorithm) to optimize the network parameters of the convolutional neural network.

[0104] Among them, the Honey Badger Algorithm (HBA) is an intelligent optimization algorithm inspired by the foraging behavior of honey badgers. By simulating the olfactory tracking and digging strategies of honey badgers during the process of hunting for prey, it conducts global search and local optimization in the solution space. This algorithm combines a balance mechanism of exploration and exploitation, has a strong ability to jump out of local optima, and is suitable for solving complex non-linear optimization problems.

[0105] Specifically, a loss function for detecting inner surface defects of oil and gas pipelines can be constructed using a convolutional neural network. The loss function can adopt a bounding box loss function, a cross-entropy loss function, a confidence loss function, etc.

[0106] Furthermore, the fitness function of the honey badger optimization algorithm is constructed using the loss function.

[0107] Initialize honey badger individuals using Sine chaotic mapping. Each honey badger individual represents a set of feasible network parameters. Each honey badger individual consists of multiple dimensional components, and each component represents a network parameter:

[0108]

[0109] where, x i represents the initial position of the i-th honey badger individual, lb i represents the lower bound of the i-th feasible solution, ub i represents the upper bound of the i-th feasible solution, y i represents the chaotic number corresponding to the i-th honey badger individual, y i-1 represents the chaotic number corresponding to the (i - 1)-th honey badger individual, and μ represents the chaotic parameter, generally taking 0.99.

[0110] In the present invention, using the ergodicity and initial sensitivity of chaotic mapping helps to generate more uniformly distributed and wider coverage initial solutions in the solution space, thereby enhancing the diversity of the population and avoiding falling into local optimal solutions. At the same time, the Sine mapping is simple and efficient, and can improve the global search ability and convergence performance of the algorithm at a low computational cost, laying a good foundation for the subsequent optimization process.

[0111] Adopt an elite selection strategy to retain the first half of the honey badger individuals with the highest fitness values and discard the other half of the honey badger individuals to filter the population.

[0112] In the present invention, adopting an elite selection strategy to retain the first half of the honey badger individuals with the highest fitness values can effectively improve the overall quality of the population and accelerate the convergence speed of the algorithm. By weeding out the inferior and retaining the better individuals, it helps to inherit excellent characteristics and guide the search direction, reducing the interference of low-quality individuals to the optimization process, thereby improving the search efficiency and stability of the global optimal solution.

[0113] In the excavation stage, a random number r is introduced 1 , and it is selected in parallel between searching around the global optimal individual or the current individual to update the individual position:

[0114]

[0115] Wherein, represents the position of the i-th honey badger individual at the (t + 1)-th iteration, ω t represents the non-linear weight factor at the t-th iteration, represents the position of the i-th honey badger individual at the t-th iteration, x best represents the position where the global optimal individual is located, F represents the search direction control parameter, β represents the food acquisition ability of the honey badger individual, generally taking a fixed value of 6, I i represents the intensity factor of the i-th honey badger individual, α represents the density factor, r 1 、r 2 、r 3 and r 4 represent random numbers between 0 and 1.

[0116] In the present invention, introducing the random number r in the excavation stage 1 and performing parallel search between the global optimal individual and the current individual helps to achieve a dynamic balance between local development and global exploration. When , the individual mainly makes fine-tuning around itself to enhance the local search ability. When , the individual approaches the global optimal position to strengthen the global search ability. This mechanism can improve the convergence speed of the algorithm while effectively avoiding falling into local optimal solutions, enhancing the robustness and optimization efficiency of the algorithm. In addition, introducing non-linear perturbations using sine and cosine functions further increases the diversity of the search path and improves the adaptability to complex optimization problems.

[0117]

[0118] Wherein, t represents the current iteration number, and T represents the maximum iteration number.

[0119] In the present invention, adopting a non-linear weight factor can gradually reduce the weight during the iteration process, so that in the initial stage of the algorithm, the search process is more exploratory and can widely search the solution space. In the later stage, as the number of iterations increases, the weight gradually decreases, and the algorithm performs more local development to finely optimize the quality of the solution. This weight decay strategy can balance the global search and local optimization capabilities, avoid premature convergence, and ensure finding the global optimal solution or a solution closer to the global optimal solution.

[0120]

[0121] Among them, r 5 represents a random number between 0 and 1, and S represents the concentration intensity.

[0122] In the present invention, the idea of inverse distance is introduced into the intensity factor, so that individuals closer to the global optimal solution have a higher "perception" intensity, and thus participate in the search more actively. Combining the global optimal individual position, the current individual position and the position relationship of its neighbors, the perception intensity of the individual to the target is dynamically measured.

[0123]

[0124] Among them, C represents the density constant, and exp represents the exponential function with the natural constant as the base.

[0125] In the present invention, the density factor can make the density gradually decay with the number of iterations, so as to maintain a larger perturbation range at the initial stage of the algorithm and enhance the global exploration ability. As the iteration progresses, the perturbation gradually decreases, prompting the algorithm to focus more on local fine search in the later stage. Such an exponential decay mechanism helps to implement the strategy of "exploration first and exploitation later" in the optimization process, improve the overall convergence speed and optimization accuracy, and enhance the ability of the algorithm to jump out of the local optimum and approach the global optimum solution.

[0126]

[0127] Among them, r 6 represents a random number between 0 and 1.

[0128] In the honey collection stage, update the individual position:

[0129]

[0130] Among them, r 7 represents a random number between 0 and 1.

[0131] In the present invention, in the honey collection stage, the honey badger individuals can be appropriately adjusted according to the global optimal solution while maintaining exploration. The random factor r 7 provides more search diversity and prevents individuals from relying too much on the global optimal solution and resulting in premature convergence. Through the gradually decreasing density factor, the algorithm can gradually reduce the perturbation during the search process, enhance the local search ability, and finally improve the convergence accuracy and the ability to find the global optimal solution.

[0132] Update the fitness value of each honey badger individual and the global optimal individual.

[0133] Judge whether the current number of iterations has reached the maximum number of iterations. If so, output the set of network parameters represented by the honey badger individual with the highest current fitness. Otherwise, return to continue the iteration.

[0134] In the present invention, an improved honey badger optimization algorithm is used to optimize the network parameters of the convolutional neural network. Compared with the traditional gradient descent method, the honey badger optimization algorithm has stronger global search ability, can effectively avoid falling into local optimal solutions, and improve the reliability of the optimization results. It can significantly improve the optimization effect and application performance of the CNN, and provide more powerful model support for high-precision fault detection.

[0135] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principle and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A visual inspection robot for inner surface defect detection of oil and gas pipelines, characterized in that: include: A power unit (1), an image acquisition unit (2), and a rotation cleaning unit (15); The power unit (1) is used to drive the robot to move inside the oil and gas pipeline; The image acquisition unit (2) comprises: a camera (12) and a processing module (11); The camera (12) is arranged at the rear end of the image acquisition unit (2) and is used to capture images of the inner surface of the oil and gas pipeline; the camera (12) comprises a glass lens (121); The processing module (11) is used to perform defect detection based on the captured image of the inner surface of the oil and gas pipeline; The power unit (1) comprises a leather cup drive unit (14); The visual inspection robot is driven to move in the pipeline by a pressure difference between the rear end and the front end of the leather cup drive unit (14); The rotating cleaning unit (15) drives the glass lens (121) to rotate, thereby removing the dirt on the outside of the glass lens (121).

2. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1 is characterized in that: The rotary cleaning unit (15) comprises: a rotary mechanism housing (151) and rotary blades (152); An air inlet (1511) and an air outlet (1512) are provided on the rotating mechanism housing (151), and an air flow channel is provided between the air inlet (1511) and the air outlet (1512); The rotating fan blade (152) is arranged in the air flow channel; The rotating fan blade (152) is fixedly connected to the glass lens (121); The airflow at the rear end of the leather cup driving unit (14) enters the airflow channel through the air inlet hole (1511), driving the rotating fan blades (152) to rotate, thereby driving the glass lens (121) to rotate, thereby removing the stains on the outside of the glass lens (121).

3. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1 is characterized in that: The rotary cleaning unit (15) comprises: a rotary mechanism housing (151), a DC motor (153) and a rotary gear (154); The DC motor (153) and the rotating gear (154) are arranged in the rotating mechanism housing (151); The DC motor (153) is drivingly connected to the rotating gear (154); The glass lens (121) is fixedly connected to the rotating gear (154); The DC motor (153) drives the rotating gear (154) to rotate, thereby driving the glass lens (121) to rotate, thereby removing the dirt on the outside of the glass lens (121).

4. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1, characterized in that: The rotary cleaning unit (15) comprises: a rotary mechanism housing (151), a DC motor (153), a rotary pulley (155) and a belt (156); The DC motor (153), the rotating pulley (155) and the belt (156) are arranged in the rotating mechanism housing (151); The DC motor (153) is drivingly connected to the rotating pulley (155) via the belt (156); The glass lens (121) is fixedly connected to the rotating pulley (155); The DC motor (153) drives the rotating pulley (155) to rotate via the belt (156), thereby driving the glass lens (121) to rotate, thereby removing the stains on the outside of the glass lens (121).

5. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1, characterized in that: Also includes: A battery unit (13), a mileage wheel unit (5), and an LED fill light unit; The battery unit (13) is used to supply power to the camera (12) and the processing module (11); The mileage wheel unit (5) is used to record the movement mileage; The LED light-filling unit is used to fill light inside the oil and gas pipeline.

6. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1, characterized in that: The glass lens (121) on the camera (12) is made of sapphire material.

7. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1, characterized in that: Also includes: Anti-collision head (3); The anti-collision head (3) is arranged at the front end of the power unit (1) and is used to prevent the visual inspection robot from colliding with the inner surface of the oil and gas pipeline and causing damage.

8. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 5, characterized in that: The processing module (11) comprises: a computer (111), a power control board (112) and an LED control board; The power control board (112) is connected to the battery unit (13); the power control board (112) and the camera (12) are both connected to the computer (111); the camera (12), the mileage wheel unit (5), the LED control board and the LED fill light unit are all connected to the power control board; the LED control board is used to realize the fill light function of the LED fill light unit.

9. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 5, characterized in that: The mileage wheel unit (5) comprises: a mileage wheel (51), a mileage wheel base (52), a mileage wheel bracket (53), a rotating shaft, a magnet and a magnetic encoder; The mileage wheel bracket (53) is connected to the mileage wheel base (52), and the mileage wheel (51) is arranged on the mileage wheel bracket (53) via the rotating shaft; The magnet with radial magnetization is installed on the rotating shaft, and the magnet is located at the geometric center of the rotating shaft; When the visual inspection robot runs in the pipeline, the mileage wheel (51) contacts the inner wall of the pipeline, driving the rotating shaft connected by the key to rotate. When the rotating shaft rotates, the magnet rotates along with the rotating shaft, and the magnetic encoder senses the position change of the magnet and generates a pulse signal. The pulse signal triggers the camera (12) to capture an image.

10. The visual inspection robot for inner surface defect inspection of oil and gas pipelines according to claim 1, characterized in that: The processing module (11) is specifically used for: S1: Acquiring an image of the inner surface of the oil and gas pipeline taken by the camera (12); S2: preprocessing the inner surface image of the oil and gas pipeline; S3: Based on the preprocessed inner surface image of the oil and gas pipeline, defects on the inner surface of the oil and gas pipeline are detected through a convolutional neural network.

Images