Method for planning detection path of multifunctional magnetic flux leakage detection robot

Abstract

The invention discloses a multifunctional magnetic flux leakage detection robot detection path planning method, and belongs to the technical field of robot control, and the method comprises the steps: obtaining the body structure characteristics of a target pressure vessel, and fitting the structure of the pressure vessel; according to the structure of the pressure vessel, excavating the internal structure change of the target pressure vessel, and setting an initial planning path; fitting a corrosion thinning distribution model by taking corrosion thinning indexes of the same kind of pressure vessels as correlation characteristics based on the work log of the target pressure vessel; in the corrosion thinning distribution model, marking an initial planning path, performing path extension iteration, and identifying a potential corrosion risk area; and updating the initial planning path, and issuing the initial planning path to the magnetic flux leakage detection robot for detection path navigation. The technical problems of low detection efficiency and insufficient risk identification caused by difficulty in accurately covering the corrosion area of the pressure vessel by a magnetic flux leakage detection robot in the prior art are solved.

Description

Technical Field

[0001] The invention relates to the technical field of robot control, and in particular to a detection path planning method for a multifunctional magnetic flux leakage detection robot. Background Art

[0002] Pressure vessels are widely used to store and transport various liquids and gases. However, these pressure vessels are prone to corrosion, cracks or wall thinning during long-term use due to chemical reactions, temperature changes and other factors, which seriously affect their safety and service life. In order to avoid safety accidents caused by defects such as corrosion, detecting the structural integrity of pressure vessels, especially corrosion and thinning, has become an important part of safety maintenance work in the industrial field. Magnetic flux leakage detection technology has become one of the mainstream technologies in pressure vessel detection due to its high sensitivity, non-contact detection method and high detection accuracy. Using magnetic flux leakage detection robots to achieve efficient and accurate detection path planning is of great significance to improving the coverage, accuracy and work efficiency of detection.

[0003] At present, magnetic flux leakage detection technology has become one of the effective methods for detecting corrosion of pressure vessels due to its high sensitivity and real-time performance. However, in complex container structures, the corrosion areas are unevenly distributed, and it is difficult for the detection path to accurately cover all risk points. Existing magnetic flux leakage detection robots lack intelligent path planning functions, making it difficult to fully identify and focus on high-risk corrosion areas, resulting in low detection efficiency and insufficient detection accuracy. For this reason, a multifunctional magnetic flux leakage detection robot detection path planning method is urgently needed to solve the technical problem that the magnetic flux leakage detection robot in the existing technology is difficult to accurately cover the corrosion area of ​​the pressure vessel, resulting in low detection efficiency and insufficient risk identification. Summary of the invention

[0004] The present application provides a detection path planning method for a multifunctional magnetic flux leakage detection robot, aiming to solve the technical problem in the prior art that it is difficult for magnetic flux leakage detection robots to accurately cover the corrosion area of ​​a pressure vessel, resulting in low detection efficiency and insufficient risk identification.

[0005] In view of the above problems, the present application provides a detection path planning method for a multifunctional magnetic flux leakage detection robot, the method comprising obtaining the main structural features of a target pressure vessel and fitting the pressure vessel structure, wherein the main structural features include vessel shape, vessel size, and vessel wall thickness; according to the pressure vessel structure, mining the internal structural changes of the target pressure vessel and setting an initial planning path; based on the work log of the target pressure vessel, taking the corrosion thinning indicators of similar pressure vessels as associated features, fitting a corrosion thinning distribution model, wherein the similar pressure vessels and the target pressure vessel belong to the same model, and the work log records multiple work task records with medium corrosion characteristics; in the corrosion thinning distribution model, marking the initial planning path and performing path extension iterations to identify potential corrosion risk areas; based on the potential corrosion risk areas, updating the initial planning path and sending it to the magnetic flux leakage detection robot for detection path navigation.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0007] The adoption of a technical solution based on the structural characteristics of the pressure vessel body, the corrosion thinning distribution model, and the iterative optimization of path extension solves the technical problem in the prior art that the leakage magnetic detection robot is difficult to accurately cover the corrosion area of ​​the pressure vessel, resulting in low detection efficiency and insufficient risk identification. By combining the shape, size, wall thickness, welding information, etc. of the container, marking the corrosion risk area and performing local and global path optimization, it is possible to fully cover the potential corrosion area and improve the accuracy of corrosion detection. In addition, based on real-time magnetic permeability information and optimized paths, the adaptability and stability of the leakage magnetic detection robot in complex structures are further improved, significantly improving the detection efficiency and accuracy.

[0008] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flow chart of a detection path planning method for a multifunctional magnetic flux leakage detection robot is provided for an embodiment of the present application.

[0010] Figure 2 A schematic flow chart of identifying potential corrosion risk areas in a detection path planning method for a multifunctional magnetic flux leakage detection robot is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0011] The overall idea of ​​the technical solution provided by this application is as follows:

[0012] The embodiment of the present application provides a method for planning the inspection path of a multifunctional magnetic flux leakage inspection robot. By acquiring the main structural features of the target pressure vessel (such as shape, size, wall thickness, etc.) and fitting the container structure, the path planning is performed in combination with the corrosion thinning distribution model. By mining the internal structural changes of the target container, the initial planning path is set, and the potential corrosion risk area is identified based on the corrosion thinning index to optimize the path. Finally, the path is updated and sent to the magnetic flux leakage inspection robot for accurate corrosion inspection to ensure the improvement of inspection efficiency and accuracy.

[0013] After introducing the basic principles of the present application, various non-limiting implementation methods of the present application will be specifically described below in conjunction with the drawings in the specification.

[0014] Examples, such as Figure 1 As shown, the embodiment of the present application provides a detection path planning method for a multifunctional magnetic flux leakage detection robot, the method comprising:

[0015] Step S100: acquiring the main structural features of the target pressure vessel and fitting the pressure vessel structure, wherein the main structural features include the vessel shape, the vessel size, and the vessel wall thickness.

[0016] Specifically, the target pressure vessel refers to the specific pressure vessel that needs to be inspected, such as a tank or pipeline used to store high-pressure gas in a chemical plant. The basic physical properties of the target container include shape (such as cylindrical, spherical, etc.), size (such as diameter, height) and wall thickness (thickness of the material). These features will affect the structural strength of the container and the inspection path planning. Fitting the pressure vessel structure refers to building a three-dimensional model of the container based on the acquired main structural features, so as to intuitively display the shape and internal structure of the container, and to be used for subsequent inspection path planning.

[0017] First, the shape and size information of the target pressure vessel is obtained through laser scanning, 3D modeling software (such as SolidWorks, AutoCAD) or image processing tools. For example, laser scanning can quickly capture the shape characteristics and size data of the container. Then, an ultrasonic thickness gauge is used to obtain the wall thickness data of the container, which is combined with the shape and size information to form the complete body structure characteristics of the container.

[0018] After the data acquisition is completed, the modeling tool is used to fit the three-dimensional structure of the container. The modeling process uses finite element analysis software (such as ANSYS) to adjust the model to make it conform to the actual structural characteristics of the pressure vessel. For example, when testing a cylindrical storage tank, the surface model of the tank body and its wall thickness distribution are constructed to lay the foundation for subsequent corrosion path planning.

[0019] By fitting the three-dimensional model in this step, the detection system can more intuitively understand the actual structure of the container and help the detection robot better plan the path.

[0020] Step S200: According to the pressure vessel structure, internal structural changes of the target pressure vessel are mined and an initial planning path is set.

[0021] Specifically, internal structural changes refer to the complex structural features inside the container, such as support columns, partitions, and weld locations. The initial planning path is the initial inspection path set based on the structural information and internal features of the pressure vessel, so that the inspection robot can traverse key parts along the optimal path.

[0022] First, the pressure vessel structure is analyzed in detail using 3D modeling tools (such as SolidWorks and AutoCAD) to identify and extract internal structural changes, such as the vessel's baffles, support frames, and welds. These internal structures will interfere with the detection signal, so they must be taken into account in path planning. For example, in a cylindrical container, if there are multiple layers of baffles inside, the location and arrangement of these baffles will affect the detection path.

[0023] Next, use a path planning algorithm (such as the A* algorithm or the Dijkstra algorithm) to set the initial planning path on the model. First, set the key parts inside the container (such as welds or thin-walled areas) as priority inspection areas, and then adjust the path according to the complexity of the container. For containers with multiple partitions, a partitioned inspection strategy can be adopted to set different path subsets within the partition to ensure that the robot can evenly cover each area.

[0024] This step analyzes the internal structural changes of the pressure vessel and sets the initial planning path. The inspection robot can cover all key areas more accurately and avoid missed inspections or repeated inspections due to complex structures.

[0025] Step S300: Based on the work log of the target pressure vessel, the corrosion thinning index of the similar pressure vessel is used as the correlation feature to fit the corrosion thinning distribution model, and the similar pressure vessel and the target pressure vessel are of the same model, and the work log records a plurality of work task records with medium corrosion characteristics.

[0026] Specifically, the work log refers to the historical data that records the working conditions of the target pressure vessel, including information such as pressure, temperature, working hours, maintenance records, etc. These data help identify the corrosion conditions that may occur in pressure vessels during long-term use. Similar pressure vessels refer to other containers with the same model, material and use environment as the target pressure vessel. The corrosion conditions of these similar containers can be used as a reference for the corrosion assessment of the target vessel. The corrosion thinning index is used to describe the degree of thinning of the pressure vessel material due to corrosion. It is usually expressed as the change in wall thickness or the thinning rate, reflecting the changes in the corrosion resistance of the material. The corrosion thinning distribution model is a mathematical model based on historical data (such as work logs and corrosion data of similar containers) to predict the corrosion thinning distribution in various areas of the pressure vessel. This model can help locate high corrosion risk areas.

[0027] First, collect and organize the work logs of the target pressure vessel, which come from the sensors of the monitoring system or manual records. The content of the work log usually includes the long-term pressure, temperature, operation frequency and maintenance of the vessel. These variables will affect the corrosion rate of the vessel.

[0028] Next, based on these work log data, correlation analysis is performed with the corrosion thinning data of pressure vessels of the same model. The corrosion thinning indicators of similar containers (such as wall thickness change data) are input into the model as features to form a reference data set. For example, by comparing the corrosion conditions of gas tanks of the same model, it is found that the corrosion in high-temperature areas or areas that are in contact with liquids all year round is more serious. Analyzing the location and degree of corrosion of these similar containers can help predict the corrosion pattern of the target container.

[0029] Subsequently, the corrosion thinning distribution model is fitted by a machine learning algorithm (such as regression analysis, Bayesian model or neural network). Specifically, the work log data of the target pressure vessel is collected, including information such as pressure, temperature, operation time and maintenance records. At the same time, the corrosion data of similar pressure vessels are collected as a reference. Next, these data are preprocessed, abnormal data are removed and normalized to ensure data consistency. Then, the preprocessed data is input into a machine learning model (such as a multivariate regression model, a Bayesian model or a neural network) for training, and the relationship between the working parameters and the degree of corrosion is used to fit the corrosion thinning trend. During the training process, the model parameters can be adjusted by cross-validation or other evaluation methods to improve the fitting accuracy. Finally, the trained model is applied to the target container to generate a corrosion thinning distribution map to predict the corrosion risk of each area and help the inspection robot plan an efficient inspection path.

[0030] The corrosion thinning distribution model established in this step can effectively predict the corrosion accumulation area of ​​the target container. Taking an oil storage tank as an example, if the model shows that the corrosion thinning rate in the high-pressure area is faster, this area can be marked as a key inspection area in advance. In this way, the inspection robot can give priority to covering these high-corrosion risk areas, reducing the time for comprehensive scanning and improving inspection efficiency. Ultimately, the model not only improves the accuracy of corrosion prediction, but also optimizes resource allocation, helping to extend the safe service life of the container.

[0031] Step S400: In the corrosion thinning distribution model, the initial planning path is marked and path extension iterations are performed to identify potential corrosion risk areas.

[0032] Specifically, the initial planning path refers to the starting inspection route set for the inspection robot during the inspection process to ensure that the key parts of the container are covered, which is the basis for subsequent optimization of the path. Path extension iteration refers to a path optimization method that gradually adjusts the inspection path to higher risk areas based on the current inspection results, so as to gradually cover potential corrosion areas during the inspection process. Potential corrosion risk areas refer to areas that are predicted to have a higher possibility of corrosion and need to be inspected in detail through the corrosion thinning distribution model.

[0033] First, the initial planning path is marked based on the corrosion thinning distribution model, that is, the inspection path of the inspection robot is pre-set on the area with higher corrosion risk. The corrosion thinning distribution model can be established by using predictive analysis tools (such as data analysis software or machine learning algorithms). For example, the corrosion distribution map generated by the previous model can set the initial planning path through the area with higher wall thinning rate, such as the weld in the container or the part that is exposed to high temperature medium all year round.

[0034] While the inspection robot is inspecting along the initial path, the system will collect and analyze the inspection data in real time, and use the path extension iteration algorithm to gradually optimize the path. Common methods include dynamic programming algorithms or feedback-based optimization algorithms. The logic of path extension iteration is: whenever new corrosion data or wall thickness change data is detected, the system will extend the inspection path to the newly identified corrosion risk area. For example, if an unexpected significant thinning of the wall thickness is detected near the weld, the path iteration algorithm will automatically generate a new path to expand the inspection range around the area in order to obtain more accurate corrosion distribution data.

[0035] For example, when inspecting a large oil storage tank, the corrosion thinning distribution model will mark the bottom area of ​​the tank wall as the initial path, because this is the area with long-term contact with the liquid and a higher risk of corrosion. During the robot inspection process, if the path extension iterative algorithm identifies abnormal thinning of the wall thickness in the extended part of this area, the system will automatically generate a new inspection point, extend the path in that direction, and continue to inspect the potential corrosion risk area.

[0036] Through this step, the path can be adjusted in real time during the inspection process to ensure that the inspection robot covers all high-risk areas in a timely manner.

[0037] Step S500: Based on the potential corrosion risk area, the initial planned path is updated and sent to the magnetic flux leakage detection robot for detection path navigation.

[0038] Specifically, path navigation refers to the process of the magnetic flux leakage detection robot moving along the planned path on the container surface, which involves path tracking and real-time adjustment to achieve efficient and accurate detection.

[0039] By comparing the data of potential corrosion risk areas, the initially generated initial planning path will be updated to ensure that the robot pays attention to the corrosion risk areas. First, the potential corrosion risk areas will be clearly identified, such as heat maps or area markers generated by data analysis tools (such as Python or MATLAB), and high-risk areas will be distinguished by color. In this way, the distribution of corrosion risk areas in pressure vessels can be intuitively identified.

[0040] Then, the initial planned path is updated and adjusted according to the marked corrosion risk areas. In order to optimize the path, common path planning algorithms (such as A* algorithm and Dijkstra algorithm) will be applied to generate new inspection routes by setting the corrosion risk areas as "priority inspection points" and allocating more inspection resources. For example, in the inspection of an oil storage tank, it is found that the corrosion risk at the bottom and corners is higher, then the initial path will be adjusted so that the robot passes through these areas more times and stays longer, thereby achieving focused inspection of risk areas.

[0041] Once the path update is completed, the final planned inspection path will be sent to the magnetic flux leakage inspection robot. After receiving the path instruction, the robot uses it as the basis for path navigation to ensure that it gives priority to covering the corrosion risk area during the inspection process. The path navigation system will combine the sensors on the robot (such as position sensors and magnetic induction sensors) to adjust the path in real time to ensure that the robot can accurately follow the updated path during movement.

[0042] This step makes the detection path more targeted by updating the path based on the corrosion risk area, thereby improving the detection accuracy and coverage of the corrosion area.

[0043] Furthermore, the main body structural features include container shape, container size, and container wall thickness, and the method includes: collecting welding processing information of the target pressure vessel; based on the welding processing information, highlighting the weld of the target pressure vessel, and determining the internal support structure in combination with the container shape and container size in the main body structural features; determining the internal partition layout through the internal support structure in combination with the container shape and container wall thickness in the main body structural features; based on the internal support structure and the internal partition layout, identifying internal structural changes of the target pressure vessel.

[0044] Specifically, welding processing information refers to detailed data about the welding parts of the pressure vessel during the manufacturing process, including the location, shape, length and direction of the weld. This information is of great reference value for subsequent inspections, because welds are often stress concentration points and corrosion-prone areas of pressure vessels. Weld highlighting refers to specially marking the weld locations on the structural diagram or model of the pressure vessel so that subsequent path planning can focus on these areas. Internal support structure refers to the support components inside the pressure vessel, such as support columns or frames, which are used to enhance the structural strength of the vessel. Internal support structures are usually key areas for inspection paths to bypass or be specially planned. Partition layout refers to the location and arrangement of partitions inside the container. Partitions are used to separate internal spaces to prevent fluid impact or increase stability. Its layout affects inspection path planning because the presence of partitions blocks the normal movement of the inspection robot.

[0045] First, collect welding process information through the manufacturer's production archives or data acquisition equipment (such as laser scanners or ultrasonic scanning equipment). This includes the specific location and length of the weld. For example, if there are multiple longitudinal welds and transverse welds on the container, this information will be recorded in the structural model. Next, use 3D modeling software (such as SolidWorks or AutoCAD) to highlight the welds in the model so that they are clearly visible in the model, so that subsequent path planning can focus on these areas.

[0046] Then, the internal support structure of the pressure vessel is determined in combination with the structural features of the main body (such as shape and size). For example, if the container is cylindrical and has multiple layers of annular support frames inside, the positions of these support frames will be marked in the model to help the robot avoid the support structure and ensure stable adsorption when planning the path. On this basis, the partition layout is further determined in combination with the shape and wall thickness data of the container. This can be achieved by analyzing the container structure diagram or three-dimensional model. For example, for a high-pressure storage tank, by determining its wall thickness and the distance between the support frames, the exact position of the internal partition can be identified in the model to guide the robot to reasonably avoid and plan during the inspection process.

[0047] This step makes inspection path planning more accurate by marking internal structural changes such as welds, support structures, and bulkhead layouts in the model. The inspection robot can prioritize weld areas, which are susceptible to corrosion and have weaker structures, while avoiding structural obstructions that affect inspection efficiency by properly detouring around support frames and bulkheads.

[0048] Furthermore, the internal partition layout is determined through the internal support structure in combination with the container shape and container wall thickness in the main body structural features. The method also includes: connecting the leakage magnetic detection robot to determine the adsorption force data and driving torque range required for safe adsorption; introducing a first control parameter combination corresponding to the walking mechanism of the leakage magnetic detection robot and a second control parameter combination corresponding to the driving module; based on the first control parameter combination corresponding to the walking mechanism and the second control parameter combination corresponding to the driving module, combined with the internal partition layout, path planning constraints are performed with the adsorption force data and driving torque range required for safe adsorption.

[0049] Specifically, the adsorption force data refers to the force required for the detection robot to adhere to the surface of the pressure vessel to ensure that it is firmly attached to the surface of the container during the detection process, especially when the container is in a vertical or bent position. The driving torque range refers to the torque required by the robot when performing the detection task to ensure that it can flexibly walk, bypass or cross the internal structure of the container. This range is usually determined by the design of the robot's walking and driving system. The walking mechanism refers to the robot's mobile device, including wheeled, tracked or magnetic mechanisms, which are responsible for driving the robot to move on the surface of the container. The first control parameter combination refers to the control parameters of the walking mechanism, such as moving speed, steering angle and acceleration, which determine the stability and accuracy of the robot's walking. The second control parameter combination of the drive module refers to the control parameter combination of the drive device, such as drive power, motor torque, etc., to ensure that the robot can move safely and effectively on the surface or internal structure of the container.

[0050] First, connect the magnetic flux leakage detection robot to the control system. The system needs to determine the adsorption force data required for the robot to adsorb at different positions and angles. This can be measured by sensors (such as force sensors or magnetometers). Taking a vertical steel pressure vessel as an example, the robot needs to generate sufficient magnetic adsorption to prevent it from sliding due to gravity. At the same time, the system will detect and record the adsorption force at different surface angles to ensure the stability of the adsorption force during path planning.

[0051] Next, the robot's walking mechanism control parameter combination (first control parameter combination) and drive module control parameter combination (second control parameter combination) are introduced. For example, for a magnetic robot, the control parameters of the walking mechanism may include travel speed and steering angle, and the control parameters of the drive module may include motor power and torque. In order to ensure the stability of the detection robot in different structural areas, these parameter combinations need to be adjusted according to the internal partition layout. For example, when the robot approaches the support frame or partition inside the container, the speed of the walking mechanism can be reduced, and the driving torque will be appropriately reduced to avoid collision or slippage.

[0052] Finally, these parameter combinations are combined with the adsorption force data and the driving torque range to form path planning constraints. Specifically, the path planning algorithm will constrain the detection path based on these data to ensure that the robot can both stably adsorb on the surface of the pressure vessel and flexibly move in different structural areas. For example, in the partition area of ​​the container, the robot will adjust the adsorption force and reduce the driving power to successfully complete the path.

[0053] Through this step, the inspection robot can achieve stable and safe movement in the complex internal structure of the container while ensuring the integrity and accuracy of the path.

[0054] Furthermore, a longitudinally arranged dual-probe array layout is set to receive magnetic permeability information of the target pressure vessel surface; based on the magnetic permeability information, it is determined whether it is below a magnetic permeability threshold; if it is below the magnetic permeability threshold, the corrosion degree analysis is performed against the initial planned path to determine the corrosion accumulation area; based on the corrosion accumulation area, the initial planned path is locally optimized.

[0055] Specifically, the dual-probe array layout refers to two sensor probes installed on the inspection robot, and they are arranged vertically. This layout allows the probes to simultaneously obtain inspection data from different locations, improving the density and accuracy of data collection. Magnetic permeability information indicates the ability of pressure vessel materials to conduct magnetic fields. Changes in magnetic permeability can reflect the corrosion condition inside the material. When it is below a certain threshold, it usually means that the material has corroded or thinned. The magnetic permeability threshold is used to determine the baseline value of corrosion. When the detected magnetic permeability is lower than this threshold, it indicates that there is corrosion or thinning in the area. Corrosion accumulation area Areas where corrosion is more severe and concentrated on the surface or inside of the container are high-risk areas that are prioritized for magnetic flux leakage inspection robots.

[0056] First, a longitudinally arranged dual-probe array is set up on the inspection robot, and a magnetic flux leakage sensor probe is used to receive magnetic permeability information on the surface of the pressure vessel. Since the two probes work simultaneously, they can compare and verify data in real time during the inspection process, reducing the error caused by a single probe. This dual-probe array is suitable for large pressure vessels, such as gas and oil tanks, and can ensure that the inspection path covers more areas and enhance the comprehensiveness of corrosion inspection.

[0057] Next, the magnetic permeability information obtained by the probe is compared with the preset magnetic permeability threshold. When the probe detects that the magnetic permeability of certain areas is lower than the threshold value, it often means that there is a thinning or corrosion problem in the material in that area. The system will mark these low magnetic permeability areas as corrosion accumulation areas. Taking the gas tank inspection as an example, the probe finds that the magnetic permeability of an area is lower than the set threshold, indicating that the area is severely corroded. The system will then mark this area and make it a key inspection point.

[0058] Then, based on the identified corrosion accumulation areas, the initial planned path is locally optimized. Through the path optimization algorithm (such as ant colony algorithm or genetic algorithm), the detection path will be extended to the corrosion accumulation area, so that the robot can increase the number of inspections or perform more intensive scanning in these high-risk areas to ensure the accuracy of the detection of the corrosion area. For example, if the corrosion area is in a corner of the container, the system will adjust the path to cover the area to avoid omissions.

[0059] Through this step, the inspection robot can more accurately identify and focus on inspecting areas of corrosion accumulation, achieving efficient allocation of inspection resources.

[0060] Further, such as Figure 2 As shown, in the corrosion thinning distribution model, the initial planned path is marked and the path extension iteration is performed to identify potential corrosion risk areas. The method includes: in the pressure vessel structure, the initial planned path is globally optimized in comparison with the potential corrosion risk area, and a first detection point sequence is set; in the pressure vessel structure, the initial planned path is locally optimized in comparison with the corrosion accumulation area, and a second detection point sequence is set; and according to the first detection point sequence and the second detection point sequence, the next round of path extension iteration is adjusted.

[0061] Specifically, global optimization refers to the overall optimization of the inspection path on the entire pressure vessel structure to ensure that all corrosion risk areas are covered. The path set by global optimization can systematically cover high corrosion risk areas to avoid missing key parts. Local optimization refers to local adjustments to the inspection path for specific areas in the container (such as corrosion accumulation areas) to ensure that these high-risk areas are inspected in detail. The first inspection point sequence refers to the main inspection points set during the global optimization process, which are used to cover the corrosion risk areas and ensure the overall structural safety of the container. The second inspection point sequence refers to the inspection points set during the local optimization process, which are used to specifically cover the corrosion accumulation areas in order to accurately detect these key parts. Path extension iteration refers to a method of gradually adjusting the path based on feedback from the inspection data, so that the inspection path can be dynamically extended to new high corrosion risk areas.

[0062] First, the initial planning path is marked based on the corrosion thinning distribution model, and the path extension iteration method is used to identify potential corrosion risk areas. Then, according to the distribution characteristics of corrosion risks, the initial path is globally optimized against these risk areas, and the first detection point sequence is set. During the global optimization process, algorithmic tools (such as genetic algorithms, particle swarm optimization, etc.) are used to ensure that the detection path covers all high corrosion risk areas. For example, when inspecting the tank structure, if the model marks the bottom edge and the connector area as risk areas, global optimization will include these locations in the first detection point sequence to ensure full coverage.

[0063] Subsequently, the initial path is locally optimized according to the specific distribution of the corrosion accumulation area, and the second detection point sequence is set. Local optimization uses a more sophisticated path planning algorithm, such as a local refinement search algorithm, to focus the path on the areas where corrosion is most concentrated, such as local welds or corners inside the container. In this case, the second detection point sequence will carefully cover these small areas, thereby improving the detection accuracy. Specifically, when the detection robot detects the weld area, local optimization can set dense detection points to obtain more detailed corrosion data.

[0064] After obtaining the first and second detection point sequences, the system adjusts the path extension iteration according to the two sequences, so that each iteration can dynamically adjust the path based on the feedback data. This dynamic path adjustment method can effectively ensure that the robot detects new corrosion risk areas. For example, if a new corrosion point is found in a certain inspection, the path extension iteration will add the vicinity of the point to the inspection range to ensure integrity.

[0065] This step combines global optimization with local optimization. The detection path not only covers the overall high-corrosion risk area, but also conducts in-depth detection of the local corrosion accumulation area, thus ensuring the comprehensiveness and accuracy of the detection. This not only improves the detection efficiency, but also enables corrosion problems to be discovered at an early stage, extending the safe service life of the pressure vessel.

[0066] Furthermore, according to the first detection point sequence and the second detection point sequence, the next round of path extension iteration is adjusted, and the method includes: based on the first detection point sequence, under the condition of hierarchical random extraction, setting multiple first diversion detection extension spaces; based on the second detection point sequence, under the condition of hierarchical random extraction, setting multiple second diversion detection extension spaces; randomly combining the multiple first diversion detection extension spaces and the multiple second diversion detection extension spaces, and performing path extension iteration on the initial planned path.

[0067] Specifically, the hierarchical random extraction condition refers to dividing the detection points into predetermined levels and randomly selecting some detection points in each level, so as to take into account both the detection coverage and detection efficiency. The first shunt detection extension space refers to the detection path extension space formed by the hierarchical random extraction based on the first detection point sequence. It contains important detection points in the global scope and is used to cover all potential corrosion areas. The second shunt detection extension space refers to the detection path extension space formed by the hierarchical random extraction based on the second detection point sequence. It contains detection points in the local corrosion accumulation area, which is used to accurately cover specific high corrosion risk areas. Random combination refers to arranging the first and second shunt detection extension spaces in a random combination to generate a more flexible detection path. This combination increases the diversity and coverage accuracy of the detection path.

[0068] Further adjustments are made based on the first detection point sequence and the second detection point sequence obtained in the previous round of path extension iteration to form multiple diversion detection extension spaces. First, the first detection point sequence is divided into multiple subsets according to the graded random extraction conditions to form the first diversion detection extension space. For example, assuming that the first detection point sequence has 9 detection points, it can be divided into three levels (such as 1-3, 4-6, 7-9), and a detection point is randomly extracted from each level, and finally a first diversion detection extension path of 3 detection points is formed. This processing method ensures that the path covers important detection points of different levels and improves the comprehensiveness of the path.

[0069] Next, the second detection point sequence is randomly selected in the same graded manner to form a second shunt detection extension space. For example, if the second detection point sequence has 6 points, it can be divided into two groups (such as 1-3 and 4-6), and one point is randomly selected in each group to form an extension path of 2 detection points. This shunt path is mainly concentrated in the local high corrosion risk area to ensure detailed coverage.

[0070] Subsequently, a new round of extension paths is constructed by randomly combining the first and second diversion detection extension spaces. The random combination makes the detection path coverage in the global and local ranges intersect with each other, thereby generating more flexible detection path planning. This random combination method uses a path coverage algorithm (such as the Monte Carlo method) to randomly generate paths, ensuring coverage while increasing the randomness of detection.

[0071] This step makes the detection path extension flexible and comprehensive through random extraction of grades and random combination of diversion space, which not only ensures the complete inspection of the global risk area, but also improves the detection accuracy of the local corrosion accumulation area. In this way, the detection path can balance the detection efficiency and coverage at multiple detection points, improving the detection accuracy and efficiency of the magnetic flux leakage detection robot.

[0072] Furthermore, the multiple first shunt detection extension spaces and the multiple second shunt detection extension spaces are randomly combined, and the initial planned path is iterated and path extension is performed. The method includes: performing an overlap analysis based on the potential corrosion risk area and the driving torque range, and calculating a first overlap; performing an overlap analysis based on the corrosion accumulation area and the driving torque range, and calculating a second overlap; and setting a newly added planned path through the first overlap and the second overlap.

[0073] Specifically, overlap analysis refers to the degree of path coverage evaluated in detection path planning, that is, the degree of overlap of adjacent detection points in the detection path. It is divided into "first overlap" and "second overlap", which correspond to the coverage density of global potential corrosion risk areas and local corrosion accumulation areas respectively. Specifically, the first overlap refers to the degree of path overlap for potential corrosion risk areas, which is mainly used to measure the coverage depth of global detection to ensure that the path repetition in these areas is moderate. The second overlap refers to the degree of path overlap for corrosion accumulation areas, which is concentrated in high corrosion risk areas to ensure that the density of local detection paths is high enough to improve detection accuracy.

[0074] Steps implementation process and methods and tools used:

[0075] In this step, the detection path is further optimized through overlap analysis. First, in the potential corrosion risk area, the first overlap analysis is performed based on the path repetition coverage in the area. The implementation method of the overlap analysis includes calculating the coverage overlap between path points to identify the repeatability of the path in the potential corrosion area. For example, the heat map of the detection path is analyzed by image processing software, and the paths with more repeated coverage are marked as high overlap areas, thereby providing a reference for adjusting the path density. The driving torque range will be used as a constraint to ensure that the overlap adjustment of the path will not affect the normal movement and stability of the robot.

[0076] Next, a second overlap analysis is performed in the corrosion accumulation area to further improve the path coverage density in the local area. For example, based on the detection results of the corrosion accumulation area, a hierarchical analysis tool (such as MATLAB's heat map function or simulation model) is used to calculate the path density and adjust the robot path to form a high-density detection path in the corrosion-prone area. The second overlap analysis makes the path distribution in the corrosion accumulation area more accurate.

[0077] Finally, by comprehensively analyzing the calculation results of the first and second overlaps, the system generates a new planning path. This new path is reasonably distributed in the potential corrosion and corrosion accumulation areas to ensure comprehensive and detailed detection coverage. The planning path optimization algorithm (such as the graph-based path optimization algorithm) combines the first and second overlap results and finally outputs an efficient and reasonable detection path to ensure that the path has an appropriate detection density in each area.

[0078] This step effectively improves the coverage and efficiency of the detection path by analyzing the overlap between potential corrosion risk areas and corrosion accumulation areas and setting new planning paths.

[0079] Furthermore, before sending it to the magnetic flux leakage detection robot for detection path navigation, the method also includes: obtaining the current coordinate position of the magnetic flux leakage detection robot; setting a detection path network based on the initial planned path and the newly added planned path, and the detection path network completely covers all detection points in the first detection point sequence and the second detection point sequence; marking the current coordinate position in the detection path network as the starting point of the detection path navigation, and outputting the optimal planned path.

[0080] Specifically, the current coordinate position refers to the real-time position coordinates of the inspection robot on the surface of the pressure vessel, which is usually obtained through sensors to ensure that the robot can be accurately positioned during the inspection process. The inspection path network refers to a network path system covering all inspection points, ensuring that each inspection point on the path is covered to achieve complete inspection. The optimal planned path refers to the most time-saving and coherent route covering all inspection points after integrating the initial path, the newly added path, and the inspection point sequence.

[0081] First, the current position of the magnetic flux leakage detection robot needs to be obtained, usually in real time through a GPS positioning system or an inertial navigation sensor. For example, the position sensor on the robot will continuously feedback the current coordinate information to help the system locate its actual position.

[0082] Next, the system will create a comprehensive inspection path network based on the initial planned path and the newly added planned path. The purpose of this path network design is to ensure that the robot can fully cover all areas that need to be inspected during movement, including the earliest determined inspection point sequence (the first inspection point sequence and the second inspection point sequence). To create this path network, commonly used path network generation tools such as MATLAB's path optimization toolbox and network graph generation algorithms (such as graph traversal algorithms) can help effectively integrate all inspection points into a path system. For example, if there are different high-risk corrosion points on the inner wall of a gas tank, these tools can generate a mesh path connecting the corrosion points to ensure that the robot can cover these points one by one.

[0083] Once the path network is created, the system will mark the current coordinate position of the inspection robot in the network as the starting point for navigation. This means that the robot will start from this position and cover all inspection points along the generated optimal planning path. To determine this optimal planning path, algorithms such as Dijkstra or A* algorithms will perform path optimization calculations, comprehensively consider the path length, inspection order, and number of turns, and output a shortest, highest coverage coherent path. For example, in a large chemical pressure vessel, through optimal path planning, the robot can move along the inner wall to each inspection point in turn, avoiding repeated paths and ensuring that each corrosion point is inspected.

[0084] This step enables the inspection robot to cover all key areas more efficiently through navigation planning of the optimal path, saving inspection time and improving inspection accuracy.

[0085] In summary, the detection path planning method of the multifunctional magnetic flux leakage detection robot provided in the embodiment of the present application has the following technical effects:

[0086] 1. By obtaining the main structural features of the target pressure vessel and fitting the pressure vessel structure, its internal structural changes are determined and the initial planning path is generated. This process makes the detection path more accurate and targeted, effectively covers key parts, improves the detection efficiency of the detection robot and the scientific nature of path planning, thereby improving the accuracy of the overall detection.

[0087] 2. By setting up a dual-probe array to obtain magnetic permeability information, the degree of corrosion is monitored in real time, and the path is optimized for low magnetic permeability areas. This local optimization can effectively deal with concentrated corrosion areas, improve the detection accuracy and efficiency of the detection robot, and make corrosion detection more targeted.

[0088] 3. By calculating the overlap between corrosion risk and corrosion accumulation areas and setting new planning paths, we can achieve focused detection of areas with concentrated corrosion distribution. This method improves the robot's detection density in high corrosion risk areas, facilitates more detailed defect detection, and enhances detection coverage.

[0089] Any step of the method described above can be stored as a computer instruction or program in an unlimited computer memory, and can be called and recognized by an unlimited computer processor to implement any method in the embodiments of the present application, without any unnecessary restrictions.

[0090] Furthermore, the first or second mentioned above not only represents an order relationship, but also represents a specific concept, and / or refers to the selection of multiple elements individually or in whole. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application intends to include these modifications and variations.

Claims

1. A detection path planning method for a multifunctional magnetic flux leakage detection robot, characterized in that: The method comprises: Acquire the main structural features of the target pressure vessel and fit the pressure vessel structure, wherein the main structural features include vessel shape, vessel size, and vessel wall thickness; According to the pressure vessel structure, mining the internal structure changes of the target pressure vessel and setting an initial planning path; Based on the work log of the target pressure vessel, the corrosion thinning index of the similar pressure vessel is used as the correlation feature to fit the corrosion thinning distribution model, wherein the similar pressure vessel and the target pressure vessel are of the same model, and the work log records a plurality of work task records with medium corrosion characteristics; In the corrosion thinning distribution model, marking the initial planning path and performing path extension iterations to identify potential corrosion risk areas; Based on the potential corrosion risk area, the initial planned path is updated and sent to the magnetic flux leakage detection robot for detection path navigation.

2. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 1, characterized in that: The body structural features include container shape, container size, and container wall thickness, and the method includes: Collect welding process information of target pressure vessel; Based on the welding processing information, the weld of the target pressure vessel is highlighted and marked, and the internal support structure is determined in combination with the vessel shape and the vessel size in the body structure characteristics; Determine the internal partition layout through the internal support structure and the container shape and container wall thickness in the main body structural features; Based on the internal support structure and the internal partition layout, internal structural changes of the target pressure vessel are identified.

3. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 2, characterized in that: Determining the internal partition layout by combining the internal support structure with the container shape and container wall thickness in the body structure features, the method further includes: Connecting the magnetic flux leakage detection robot to determine the adsorption force data and driving torque range required for safe adsorption; Introducing a first control parameter combination corresponding to the walking mechanism of the magnetic flux leakage detection robot and a second control parameter combination corresponding to the driving module; Based on the first control parameter combination corresponding to the walking mechanism and the second control parameter combination corresponding to the driving module, combined with the internal partition layout, the path planning constraints are performed using the adsorption force data and driving torque range required for safe adsorption.

4. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 3, characterized in that: Setting a longitudinally arranged dual-probe array layout to receive magnetic permeability information on the surface of the target pressure vessel; Based on the magnetic permeability information, determining whether it is below a magnetic permeability threshold; If it is lower than the magnetic permeability threshold, then analyzing the corrosion degree by comparing the initial planning path to determine the corrosion accumulation area; Based on the corrosion accumulation area, the initial planned path is locally optimized.

5. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 4, characterized in that: In the corrosion thinning distribution model, the initial planning path is marked and path extension iteration is performed to identify potential corrosion risk areas. The method includes: In the pressure vessel structure, globally optimizing the initial planned path in comparison with the potential corrosion risk area, and setting a first detection point sequence; In the pressure vessel structure, locally optimizing the initial planned path in accordance with the corrosion accumulation area, and setting a second detection point sequence; The next round of path extension iteration is adjusted according to the first detection point sequence and the second detection point sequence.

6. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 5, characterized in that: According to the first detection point sequence and the second detection point sequence, adjusting the next round of path extension iteration, the method includes: Based on the first detection point sequence, a plurality of first diversion detection extension spaces are set under a hierarchical random sampling condition; Based on the second detection point sequence, a plurality of second diversion detection extension spaces are set under a hierarchical random sampling condition; The plurality of first diversion detection extension spaces and the plurality of second diversion detection extension spaces are randomly combined, and the initial planned path is iterated and path extension is performed.

7. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 6, characterized in that: The plurality of first flow diversion detection extension spaces and the plurality of second flow diversion detection extension spaces are randomly combined, and the initial planned path is iterated and path extension is performed, wherein the method comprises: Performing overlap analysis based on the potential corrosion risk area and the driving torque range to calculate a first overlap; Performing overlap analysis based on the corrosion accumulation area and the driving torque range to calculate a second overlap; A newly added planned path is set according to the first overlap degree and the second overlap degree.

8. The detection path planning method of the multifunctional magnetic flux leakage detection robot according to claim 7, characterized in that: Before sending the detection path navigation to the magnetic flux leakage detection robot, the method further includes: Obtaining the current coordinate position of the magnetic flux leakage detection robot; Based on the initial planned path and the newly added planned path, a detection path network is set, wherein the detection path network completely covers all detection points in the first detection point sequence and the second detection point sequence; The current coordinate position is marked in the detection path network as the starting point of the detection path navigation, and the optimal planned path is output.

Images