A Pipeline Robot Path Planning and Detection Method Based on Multi-Data Fusion

Abstract

The present invention relates to the technical field of path planning, and specifically provides a path planning and detection method for pipeline robots based on multi-data fusion, including the following steps: collecting data from vision, ultrasound, magnetic flux leakage, and lidar, aligning the timestamps of the data and correcting the spatial coordinates, performing data fusion and verifying the consistency of the data to generate an environmental data set. In the present invention, by constructing a three-dimensional pipeline model, it not only provides a visual representation of the pipeline, but also enhances the understanding of the internal structure of the pipeline, enabling the maintenance team to formulate repair strategies more effectively. The access priority and optimized access path determined through cost-benefit analysis allow the robot to perform inspections in the most efficient order, reducing the detection time and operating costs, ensuring immediate fault diagnosis and rapid response, significantly improving the accuracy of detection and the operating efficiency of the robot, significantly enhancing the efficiency and accuracy of pipeline detection, reducing manual intervention, and lowering the maintenance costs.

Description

Technical Field

[0001] The present invention relates to the technical field of path planning, and particularly to a path planning and detection method for a pipeline robot based on multi-data fusion. Background Art

[0002] The technical field of path planning involves the cross-field of computer science, robotics, and automation engineering, mainly focusing on how to enable a robot or an automated system to find the optimal path for completely detecting an entire pipe network in a constrained environment. Integrating various technologies such as algorithm design, machine learning, and sensor data processing to handle and solve navigation problems in a dynamic environment. The application of path planning is very extensive, covering various scenarios such as industrial automation, autonomous driving vehicles, and unmanned aerial vehicle navigation. The core goal is to improve the efficiency and safety of the path, reduce energy consumption and time costs.

[0003] Among them, the path planning and detection method for a pipeline robot based on multi-data fusion refers to the technology of using integrated multiple sensors and data sources to optimize the path selection and environmental detection efficiency of the pipeline robot. Its main use is in pipeline maintenance and inspection, where an intelligent robot system automatically detects problems such as pipeline cracks, corrosion, and sediment accumulation. By fusing data from different sensors, such as visual, acoustic, magnetic, and pressure data, the robot can more accurately analyze the pipeline state, providing scientific data support and decision-making basis for operation and maintenance. This method can significantly improve the efficiency and accuracy of pipeline detection, reduce manual intervention, and lower maintenance costs.

[0004] Existing path planning technologies rely on single or a small number of data sources when dealing with navigation problems in a dynamic environment, resulting in limitations in environmental perception and increased errors. The lack of efficient data fusion processing logic and precise spatial correction technology limits the navigation efficiency and accuracy of the robot in a complex environment. The path planning of the existing technology fails to make full use of real-time data and lacks a dynamic adjustment mechanism, resulting in the robot not being sensitive enough to respond to sudden environmental changes. For example, when an unexpected pipeline blockage or damage occurs, the robot cannot adjust the path in time, delaying the problem handling time, affecting the efficiency and timeliness of pipeline detection and maintenance, and increasing the risk and cost of operation and maintenance. Summary of the Invention

[0005] The purpose of the present invention is to solve the disadvantages existing in the prior art, and to propose a path planning and detection method for a pipeline robot based on multi-data fusion.

[0006] To achieve the above purpose, the present invention adopts the following technical scheme: A path planning and detection method for a pipeline robot based on multi-data fusion, including the following steps:

[0007] S1: Collect data from vision, ultrasound, magnetic flux leakage, and lidar, perform timestamp alignment and spatial coordinate correction on the data, conduct data fusion, and verify the consistency of the data to generate an environmental dataset;

[0008] S2: Based on the environmental dataset, extract structural information and identify abnormal signals, mark the defects and abnormal areas of the pipeline, analyze the internal environment of the pipeline, draw the structural condition of the pipeline, and generate a 3D pipeline model;

[0009] S3: According to the 3D pipeline model, compare the data of detection requirements and costs between pipeline nodes, determine the access priority of nodes through cost-benefit analysis, and generate an optimized access path;

[0010] S4: Use the optimized access path to guide the actions of the robot, adjust the speed and steering parameters of the pipeline robot, detect and verify the predetermined path trajectory, and generate an optimized path planning diagram;

[0011] S5: When the robot moves along the optimized path planning diagram, continuously collect pipeline environment and robot status data, analyze abnormal data in real time, detect data points deviating from the normal state, and generate real-time monitoring results;

[0012] S6: According to the feedback of the real-time monitoring results, adjust the path and speed of the robot, detect the pipeline, evaluate the quality and integrity of the dataset, and generate a path evaluation result.

[0013] As a further solution of the present invention, the environmental dataset includes sensor data with timestamp alignment, measurement points after spatial coordinate correction, and multi-sensor fusion data. The 3D pipeline model includes the spatial layout of the pipeline, connection nodes, and defect locations. The optimized access path includes the economic evaluation, access priority, and cost calculation of pipeline detection nodes. The optimized path planning diagram includes the adjusted speed and steering parameters of the robot, predetermined trajectory, and key points on the path. The real-time monitoring results include records of environmental changes, real-time data of the robot status, and data analysis deviating from the normal state. The path evaluation result includes the detection of new defects, the effectiveness of the path, and the evaluation of dataset integrity.

[0014] As a further solution of the present invention, the steps of collecting data from vision, ultrasound, magnetic flux leakage, and lidar, performing timestamp alignment and spatial coordinate correction on the data, conducting data fusion, and verifying the consistency of the data to generate an environmental dataset are specifically as follows:

[0015] S101: Collect vision, ultrasound, magnetic flux leakage, and lidar data, synchronize the time frequency of the sensors during data collection, and add a timestamp synchronized with the clock to each data point to generate a time-synchronized dataset;

[0016] S102: Using the time synchronization data set, coordinate calibration is performed on visual, ultrasonic, magnetic flux leakage, and lidar data using a unified spatial reference framework. By adjusting the sensor position parameters and calibration angles, the consistency of the data in space is checked to generate a spatial correction data set.

[0017] S103: Based on the spatial correction data set, the corrected data is fused, the consistency between visual, ultrasonic, magnetic flux leakage, and lidar data is compared, and the abnormal areas in the data are marked through differential point analysis to construct an environment data set.

[0018] As a further solution of the present invention, the steps of extracting structural information and identifying abnormal signals based on the environment data set, identifying the defects and abnormal areas of the pipeline, analyzing the internal environment of the pipeline, and drawing the structural condition of the pipeline to generate a three-dimensional pipeline model are specifically as follows:

[0019] S201: According to the environment data set, the structural information of the pipeline is extracted, the abnormal signals in the data are identified, the pipeline defects and abnormal areas are located through the analysis of signal strength and continuity to generate a defect identification result. According to the defect size confidence of the differential detection method, the Bayesian method is used to obtain the comprehensive confidence of the defect size at each defect position.

[0020] S202: Using the defect identification result, combined with the structural information of the pipeline, the internal state of the pipeline is generated, the structure diagram of the pipeline is drawn, the physical structure and internal characteristics of the pipeline are analyzed, and a three-dimensional pipeline model is generated.

[0021] As a further solution of the present invention, according to the three-dimensional pipeline model, the data of the detection requirements and costs between pipeline nodes are compared, and the access priority of the nodes is determined through cost-benefit analysis to generate the steps of an optimized access path are specifically as follows:

[0022] S301: Based on the three-dimensional pipeline model, by comparing the data differences between different nodes, the access priority of the nodes is determined to generate a node priority analysis result.

[0023] S302: Using the node priority analysis result, perform a cost minimization path analysis, refer to the optimal allocation of time and resources, calculate the sequence of the lowest cost access path, and obtain a cost-based access order analysis result.

[0024] S303: According to the cost-based access order analysis result, referring to the real-time operating conditions and dynamic resource allocation, determine the access path for pipeline inspection and maintenance to generate an optimized access path.

[0025] As a further solution of the present invention, the steps of using the optimized access path to guide the movement of the robot, adjusting the speed and steering parameters of the pipeline robot, detecting and verifying the predetermined path trajectory, and generating an optimized path planning diagram are specifically as follows:

[0026] S401: Based on the optimized access path, adjust the speed and steering parameters of the pipeline robot, make adjustments according to the requirements and obstacle information of the target pipeline area, verify that the robot moves along the predetermined trajectory, and generate the adjustment result of the robot action parameters;

[0027] S402: Through the adjustment result of the robot action parameters, test the forward path of the robot, verify the adaptability of speed and steering in the pipeline environment, verify the consistency between the robot behavior and the predetermined path, and generate the path verification result;

[0028] S403: Utilize the path verification result, integrate the motion data of the robot and the path planning requirements, draw the real-time operation path diagram of the robot, and generate an optimized path planning diagram.

[0029] As a further solution of the present invention, when the robot moves along the optimized path planning diagram, the steps of continuously collecting pipeline environment and robot status data, analyzing abnormal data in real time, detecting data points deviating from the normal state, and generating real-time monitoring results are specifically as follows:

[0030] S501: Based on the optimized path planning diagram, estimate the real-time position of the robot and monitor the environmental changes, and generate the environmental change recognition result;

[0031] S502: Adopt the environmental change recognition result, conduct abnormal analysis on the collected information, identify the data points deviating from the preset parameters, determine the potential risks and abnormal areas through comparative analysis, and generate the abnormal analysis result;

[0032] S503: According to the abnormal analysis result, analyze the detected abnormal data points, adjust the monitoring strategy and response measures, verify the safety and efficiency of the robot operation, and generate the real-time monitoring result.

[0033] As a further solution of the present invention, according to the feedback of the real-time monitoring result, the steps of adjusting the path and speed of the robot, detecting the pipeline, evaluating the quality and integrity of the data set, and generating the path evaluation result are as follows:

[0034] S601: Based on the real-time monitoring result, dynamically optimize the path and speed of the robot, avoid risks and optimize efficiency, and generate the path adjustment result;

[0035] S602: Use the path adjustment result to verify data consistency and coverage, analyze the quality and integrity of the collected data set, check the reliability and validity of the data, and generate a data quality assessment result;

[0036] S603: Through the data quality assessment result, with reference to the real-time effect of path adjustment and the integrity of the data set, evaluate the path planning and operation effect of the robot, and generate a path assessment result.

[0037] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0038] In the present invention, by integrating vision, ultrasonic, magnetic flux leakage, and lidar data, and optimizing the processing logic of timestamp alignment and spatial coordinate correction for the data, the accuracy and consistency of environmental data are improved. The environmental data set improves the recognition rate of pipeline defects and abnormal areas, which is crucial in pipeline detection. The three-dimensional pipeline model not only provides a visual representation of the pipeline but also enhances the understanding of the internal structure of the pipeline, enabling the maintenance team to formulate repair strategies more effectively. The access priority and optimized access path determined through cost-benefit analysis allow the robot to perform inspections in the most efficient order, reducing inspection time and operating costs. The generation of real-time monitoring results and the feedback mechanism ensure instant fault diagnosis and rapid response, significantly improving the accuracy of detection and the operating efficiency of the robot, significantly enhancing the efficiency and accuracy of pipeline detection, reducing manual intervention, and lowering maintenance costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the main steps of the present invention;

[0040] Figure 2 It is a detailed schematic diagram of S1 of the present invention;

[0041] Figure 3 It is a detailed schematic diagram of S2 of the present invention;

[0042] Figure 4 It is a detailed schematic diagram of S3 of the present invention;

[0043] Figure 5 It is a detailed schematic diagram of S4 of the present invention;

[0044] Figure 6 It is a detailed schematic diagram of S5 of the present invention;

[0045] Figure 7 It is a detailed schematic diagram of S6 of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0046] To make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0047] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation to the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more unless otherwise specifically defined.

[0048] Please refer to Figure 1 , the present invention provides a technical solution, a pipeline robot path planning and detection method based on multi-data fusion, including the following steps:

[0049] S1: Collect data from vision, ultrasound, magnetic flux leakage and lidar, perform timestamp alignment and spatial coordinate correction on the data, perform data fusion and check the consistency of the data to generate an environmental data set;

[0050] S2: Analyze the environmental data set, extract structural information and identify abnormal signals, identify defects and abnormal areas of the pipeline, analyze the internal environment of the pipeline, draw the geometry and structure of the pipeline, and generate a three-dimensional pipeline model;

[0051] S3: According to the three-dimensional pipeline model, compare the detection requirements and cost data between pipeline nodes, determine the access priority of nodes through cost-benefit analysis, and generate an optimized access path according to the priority ranking;

[0052] S4: Use the optimized access path to guide the action of the robot, adjust the speed and steering parameters of the pipeline robot, detect and verify the predetermined path trajectory, monitor the approach of key nodes to avoid collisions, and generate an optimized path planning diagram;

[0053] S5: When the robot moves along the optimized path planning diagram, continuously collect pipeline environment and robot status data, perform real-time anomaly analysis on the collected data, detect data points deviating from the normal state, and feedback anomaly information to generate real-time monitoring results;

[0054] S6: According to the feedback of the real-time monitoring results, adjust the path and speed of the robot, detect the pipeline, evaluate the quality and integrity of the data set, and generate a path evaluation result.

[0055] The environmental data set includes sensor data with timestamp alignment, measurement points after spatial coordinate correction, and data from multi-sensor fusion. The three-dimensional pipeline model includes the spatial layout of the pipeline, connection nodes, and defect locations. The optimized access path includes economic evaluation of pipeline inspection nodes, access priority, and cost calculation. The optimized path planning diagram includes the robot's adjusted speed and steering parameters, the predetermined trajectory, and key points in the path. The real-time monitoring results include records of environmental changes, real-time data on the robot's status, and data analysis of deviations from the normal state. The path evaluation results include the detection of new defects, the effectiveness of the path, and the evaluation of the integrity of the data set.

[0056] See also Figure 2 , collect data from vision, ultrasound, magnetic flux leakage and lidar, align the data timestamps and correct the spatial coordinates, perform data fusion and check the consistency of the data. The specific steps to generate the environmental data set are:

[0057] S101: Collect visual, ultrasonic, magnetic flux leakage and lidar data, synchronously set the time frequency of the sensor during data collection, add a timestamp synchronized with the clock to each data point, and generate a time-synchronized data set. The execution process is as follows;

[0058] In the process of synchronously setting the time frequency of sensors, it is necessary to first ensure that the time frequency parameters of each sensor - vision, ultrasound, magnetic flux leakage and lidar - are accurately defined and configured, including the sampling frequency and response time of each sensor, and the synchronization mechanism with the system master clock. Ensuring the time consistency of data acquisition is the key. Through precise time marking, each data point is added with a timestamp synchronized with the system clock. The timestamp not only records the specific moment when the data point is collected, but also identifies the time sequence of the data points. It is the basis for constructing a time-synchronized data set. It is also necessary to consider the time offset and drift correction between sensors to ensure the continuous synchronization of each sensor data during long-term operation. Through strict control and adjustment of parameters, time errors can be reduced in analysis and subsequent processing, the accuracy and efficiency of data processing can be improved, and a time-synchronized data set can be generated.

[0059] S102: Using the time-synchronized data set and a unified spatial reference frame, coordinate calibration is performed on the visual, ultrasonic, magnetic flux leakage and lidar data. By adjusting the sensor position parameters and calibration angles, the consistency of the data in space is checked and the execution process of generating a spatial correction data set is as follows;

[0060] By adjusting the sensor position parameters and calibration angle, the consistency of the data in space is checked, according to the formula Calculate the spatial consistency between sensors. In the formula, x1, y1, and z1 represent the coordinates of a sensor before calibration, and x2, y2, and z2 represent the coordinates of the same sensor after calibration. d sync represents the spatial distance between the coordinates before and after calibration.

[0061] Detailed explanation of the formula and the derivation process of the formula calculation: There is a vision sensor and a lidar sensor. The initial position of the vision sensor is (0, 0, 0), and it moves to (1, 1, 1) after calibration. The initial position of the lidar is (2, 2, 2), and its position after calibration is (3, 3, 3). Use the above formula to calculate the spatial distance between the two sensors before and after calibration, and the distance of the vision sensor will be obtained The distance of the lidar The consistency of the data in space can be verified through the calculation method.

[0062] S103: Based on the spatial correction data set, fuse the corrected data, compare the consistency between vision, ultrasonic, magnetic flux leakage, and lidar data, mark the abnormal areas in the data through differential point analysis, and the execution process of constructing the environmental data set is as follows;

[0063] During the process of fusing vision, ultrasonic, magnetic flux leakage, and lidar data based on the spatial correction data set, it is necessary to deeply compare the data of each sensor to determine the corresponding relationship in space and time. Through advanced data processing technologies, such as multi-sensor data fusion algorithms, high-precision alignment of vision data with ultrasonic and lidar data can be achieved. The algorithm involves complex geometric transformations and time series analysis to ensure that the data obtained from different sensors can be correctly represented in a unified coordinate system. Mark the abnormal areas in the data through differential point analysis. The key lies in using the algorithm to perform statistical analysis on the data set, identify the data points that deviate from the normal mode, mark the abnormal areas, and provide accurate basic data for subsequent environmental modeling and analysis to construct the environmental data set.

[0064] Please refer to Figure 3 , based on the environmental data set, extract structural information and identify abnormal signals, identify the defects and abnormal areas of the pipeline, analyze the internal environment of the pipeline, draw the structural situation of the pipeline, and the steps to generate a three-dimensional pipeline model are specifically as follows:

[0065] S201: According to the environmental data set, extract the structural information of the pipeline, identify the abnormal signals in the data, locate the pipeline defects and abnormal areas through the analysis of signal strength and continuity, generate the defect recognition results, and according to the defect size confidence of the differential detection method, use the Bayesian method to obtain the comprehensive confidence of the defect size at each defect position. The execution process is as follows;

[0066] By analyzing the environmental dataset of the pipeline, extracting structural information, and identifying abnormal signals, this process relies on precise data processing and analysis techniques to collect and organize data from the environmental dataset, ensuring data integrity and accuracy. Signal processing techniques are used to filter out noise in the data to improve the accuracy of signal identification. Data analysis methods such as Fourier transform or wavelet transform are utilized to analyze the frequency and intensity of the signals to determine the abnormal patterns of the signals. Through these analyses, the defects and abnormal areas existing in the pipeline are identified. A series of precise data analyses and processing provide the basic data and preconditions for the subsequent calculation of the confidence level of defect sizes, and the comprehensive confidence level of the defect sizes at each defect location is obtained.

[0067] S202: According to the defect identification results, combined with the structural information of the pipeline, generate the internal state of the pipeline, draw the structure diagram of the pipeline, analyze the physical structure and internal characteristics of the pipeline, and the execution process of generating the 3D pipeline model is as follows;

[0068] Use the Bayesian method to calculate the comprehensive confidence level of defect sizes. Assume that the uncertainty of the measured defects remains unchanged, and the probability distribution of the defect sizes evaluated after the previous inspection is a normal distribution N(1,1 2 ), the time from the previous inspection to the pipeline being put into use is 1 year, and the corrosion rate is obtained by dividing the mean value of the defect sizes by the time, with a corrosion rate of 1 mm / year. Conduct another inspection in the second year, set the variance of the defect sizes to remain unchanged, and the prior probability distribution of the defect sizes is π(θ) ∼ N(2,1 2 ). The defect size obtained by one detection method is 1.8, and the variance of this detection method is 0.9 2 , and the likelihood function of this detection method According to Bayes' formula The posterior distribution probability of the mean value of the defect sizes is π1(θ|x) ∼ N(1.89, 0.67 2 ). The probability distribution of the defect size obtained by the second detection method is 2.1 mm, and the variance of this detection method is 1.2 2 , then the posterior distribution of the mean value of the defect sizes obtained by this method is π2(θ|x) ∼ N(2.04, 0.77 2 ). Using the Monte Carlo method, sample the distribution of corrosion defects from multiple detection methods, and use the maximum likelihood estimation to fit the sampling distribution of the mean value to a normal distribution. For the results obtained from the above two detection methods, after sampling and fitting, the new defect size distribution is N(1.96, 0.72 2 ) mm, and the updated corrosion growth rate is 0.98 year / mm. The above is only an example using two detection methods, and other additional detection methods can also be processed according to the same steps.

[0069] Please refer toFigure 4 , according to the three-dimensional pipeline model, compare the detection requirements and cost data between pipeline nodes, and determine the access priority of nodes through cost-benefit analysis. The specific steps for generating the optimized access path are as follows:

[0070] S301: Based on the three-dimensional pipeline model, by comparing the data differences between different nodes, determine the access priority of nodes, and the execution process for generating the node priority analysis result is as follows;

[0071] Based on the three-dimensional pipeline model, by comparing the data differences between different nodes, use three-dimensional coordinates to represent the position of each node, evaluate the observation information between each pair of nodes, calculate the difference between the two, including using the data difference method to determine the observation position between nodes, and by calculating the total difference of all node pairs, sort each node to determine the access priority of each node. This not only helps to clarify that the node is a key node in the network, but also provides basic data for subsequent path optimization analysis. This method enables the screening of high-priority nodes from batch data to optimize resource allocation and time management, and generate the node priority analysis result.

[0072] S302: Adopt the node priority analysis result to conduct a cost minimization path analysis, and calculate the sequence of the lowest-cost access path with reference to the optimal allocation of time and resources. The execution process for obtaining the cost-based access order analysis result is as follows;

[0073] According to the detection requirements between pipeline nodes, select the path with the lowest cost, and with reference to the optimal allocation of time and resources, calculate the total cost C of the lowest-cost path according to the formula The total cost C of the lowest-cost path min . In the formula, w(v b , v b+1 ) represents the weight of the edge from node v b to v b + 1, and n b represents the number of nodes in the path. Explanation of the formula and the derivation process of the formula calculation: Consider a specific network topology with four nodes forming a path: v1 → v2 → v3 → v4, and the weights of each edge are 3, 2, and 5 respectively. According to the lowest-cost path analysis, the total cost from v1 to v4 needs to be calculated. Using the formula, the calculation process is as follows: The cost from v1 to v2 is 3. The cost from v2 to v3 is 2. The cost from v3 to v4 is 5. Substitute the values into the formula: C min = w(v1, v2) + w(v2, v3) + w(v3, v4) = 3 + 2 + 5 = 10. The total cost of the lowest-cost path is 10, indicating that under the optimal allocation of time and resources, the path provides the most cost-effective access sequence.

[0074] S303: According to the analysis result of the cost-based access order, with reference to the real-time operating conditions and dynamic resource allocation, determine the access path for pipeline inspection and maintenance, and the execution process for generating the optimized access path is as follows;

[0075] According to the analysis result of the cost-based access order, as well as the actual requirements of real-time operating conditions and dynamic resource allocation, determine the optimized access path for pipeline inspection and maintenance. The process involves comprehensively considering the access cost, urgency, and operation obstacles of each node, and calculating the most economical and efficient path through an algorithm model. It involves path optimization algorithms, such as variants of the shortest path algorithm or the traveling salesman problem algorithm in graph theory. The algorithm considers cost minimization and also the feasibility of the path, ensuring that the optimal inspection and maintenance operations are achieved within the limited resources and time, providing a maintenance team with the most reasonable inspection path dynamically generated based on the latest data, and obtaining the optimized access path.

[0076] Please refer to Figure 5 , and use the optimized access path to guide the actions of the robot. The steps for adjusting the speed and steering parameters of the pipeline robot, detecting and verifying the predetermined path trajectory, and generating the optimized path planning diagram are specifically as follows:

[0077] S401: Based on the optimized access path, adjust the speed and steering parameters of the pipeline robot, make adjustments according to the requirements and obstacle information of the target pipeline area, and verify that the robot moves along the predetermined trajectory. The execution process for generating the adjustment result of the robot's action parameters is as follows;

[0078] Based on the optimized access path, by adjusting the speed and steering parameters of the robot, in response to the changes in the pipeline environment and obstacle information, the operations include obtaining data on the specific requirements and obstacle types of the pipeline area, analyzing and processing these data to adjust the speed and steering of the robot, ensuring that the robot can effectively avoid obstacles and improve the movement efficiency. This adjustment process involves multiple parameters, such as the current speed of the robot, the target speed, the current steering angle, the target steering angle, as well as the distance and type of obstacles. Through real-time data analysis, adjust the parameters to adapt to the environmental changes and ensure that the robot moves along the predetermined trajectory, generating the adjustment result of the robot's action parameters.

[0079] S402: Through the adjustment result of the robot's action parameters, test the forward path of the robot, verify the adaptability of the speed and steering in the pipeline environment, and verify the consistency between the robot's behavior and the predetermined path. The execution process for generating the path verification result is as follows;

[0080] Based on the results of robot action parameter adjustment, test the forward path of the robot, mainly verifying the adaptability of the robot's speed and steering in the pipeline environment. This includes adjusting the speed and steering parameters of the robot under different types of pipeline conditions, such as diameter changes, bending degrees, and obstacles inside the pipeline, to test the response ability to complex environments. During the experiment, collect the real-time motion data of the robot through sensors and monitoring systems. By comparing the differences between the predetermined path and the actual running trajectory, evaluate the consistency of the robot's behavior, ensure that the robot can accurately navigate in the complex pipeline system. The test results will be used to further optimize the robot's action strategy, improve the efficiency and safety in actual operation, and generate path verification results.

[0081] S403: Using the path verification results, integrate the motion data of the robot and the path planning requirements, and draw the real-time operation path map of the robot. The execution process of generating the optimized path planning map is as follows;

[0082] Using the path verification results, integrate the motion data of the robot and the path planning requirements, and draw the real-time operation path map of the robot. The process involves advanced data processing and visualization technologies. By analyzing data such as the speed, steering, and position of the robot during the test, combined with the specific layout and operation requirements of the pipeline, it not only provides detailed guidance for the robot's operation, but also shows the predetermined and actual paths of the robot through graphics, helping engineers understand the action efficiency and accuracy of the robot in the pipeline, as well as the adjustments required in future operations, ensuring that the robot's operation is more in line with the actual application scenarios and requirements, and generating an optimized path planning map.

[0083] Please refer to Figure 6 , when the robot moves along the optimized path planning map, continuously collect pipeline environment and robot status data, and analyze abnormal data in real time to detect data points deviating from the normal state. The specific steps for generating real-time monitoring results are as follows:

[0084] S501: Based on the optimized path planning map, estimate the real-time position of the robot and monitor the environmental changes, and the execution process of generating the environmental change recognition results is as follows;

[0085] Based on the optimized path planning diagram, the operations performed include monitoring the real-time position of the robot and continuously monitoring changes in the environment. The execution process involves the precise determination of the robot's position and the real-time update of environmental parameters, including changes in environmental factors such as temperature, humidity, and light intensity. These monitoring data are collected through sensors and compared with a preset environmental model to identify any significant deviations or anomalies. The accuracy of environmental monitoring depends on the sensitivity of the sensors and the accuracy of the data processing algorithm. By combining real-time position data with environmental parameters, the changing trends of environmental conditions can be immediately displayed, providing a basis for subsequent robot navigation and path adjustment, and generating environmental change recognition results.

[0086] S502: Using the environmental change recognition results, perform anomaly analysis on the collected information, identify data points that deviate from the preset parameters, and determine potential risks and abnormal areas through comparative analysis. The execution process of generating the anomaly analysis results is as follows;

[0087] Using the environmental change recognition results, perform anomaly analysis on the collected information. The execution process includes the collection and analysis of data points, especially those that deviate from the preset parameters. These data points indicate potential risks and abnormal areas. The analysis methods involve comparative analysis and risk assessment. Statistical methods such as standard deviation and variance analysis are used during the analysis process to identify the degree of data deviation. The determination of abnormal areas is based on set thresholds, which are preset according to historical data and risk management strategies. Through these analyses, identify the risk factors existing near the robot's running path, list all identified abnormal areas and potential risks, provide reference and warning for the safe operation of the robot, and generate anomaly analysis results.

[0088] S503: According to the anomaly analysis results, analyze the detected abnormal data points, adjust the monitoring strategy and response measures, and verify the safety and efficiency of the robot operation. The execution process of generating the real-time monitoring results is as follows;

[0089] According to the anomaly analysis results, analyze the detected abnormal data points. During the process of adjusting the monitoring strategy and response measures, carefully analyze each data point identified as abnormal, determine the specific impact on the robot operation, and adjust the corresponding monitoring strategy for each type of anomaly, including more frequent data collection, adding additional sensor monitoring, or modifying the robot's behavior pattern to address the identified risks. For example, if the detected battery level is below the safety threshold, adjust the robot's route to shorten the task time or arrange for temporary charging to ensure the continuity of the operation and the safety of the robot. After implementing the adjustment, conduct on-site tests again to verify the effectiveness of the adjustment measures and the safety of the robot operation. Through real-time monitoring and feedback, the reliability of the robot system and the task execution efficiency can be significantly improved, generating real-time monitoring results.

[0090] Please refer to Figure 7 , according to the feedback of real-time monitoring results, adjust the path and speed of the robot, detect the pipeline, evaluate the quality and integrity of the dataset, and the steps to generate the path evaluation result are specifically as follows:

[0091] S601: Based on the real-time monitoring results, dynamically optimize the path and speed of the robot, avoid risks and optimize efficiency, and the execution process to generate the path adjustment result is as follows;

[0092] During the process of dynamically optimizing the path and speed of the robot based on the real-time monitoring results, it is necessary to collect the state data of the environment and the robot itself in real time, including the current position of the robot, the position and distance of the obstacles ahead, the current traveling speed, and the expected target path. Use the path planning algorithm to calculate the best traveling route, and adjust the speed in a timely manner to adapt to environmental changes. In this way, potential risks can be effectively avoided and the traveling efficiency can be optimized. If the sensor detects an obstacle ahead, the path will be recalculated to bypass the obstacle, and the speed will be adjusted according to the size and distance of the obstacle to ensure that the robot minimizes the collision risk while maintaining high efficiency. The adjustment of the path and speed will also consider the energy consumption of the robot and the time requirements of the predetermined task to ensure that the task is completed without failure due to excessive energy consumption. Through this dynamic optimization, the robot can maintain the best performance state in a complex and changeable environment.

[0093] S602: Adopt the path adjustment result to verify the data consistency and coverage, analyze the quality and integrity of the collected dataset, check the reliability and validity of the data, and the execution process to generate the data quality evaluation result is as follows;

[0094] Adopt the path adjustment result to verify the data consistency and coverage. Collect data according to the adjusted path, and conduct statistical analysis on the dataset, including data integrity, consistency, and coverage. It involves comparing and verifying information collected from multiple data sources, using statistical tests to evaluate the quality of the dataset. For example, by calculating indicators such as data missing rate, outlier detection, and data duplication rate, remove invalid or duplicate data to ensure the reliability and validity of the dataset, which helps to ensure the efficiency and accuracy of the dataset in practical applications, reflect the application potential and limitations of the dataset in real scenarios, and generate the data quality evaluation result.

[0095] S603: Through the data quality evaluation result, refer to the real-time effect of the path adjustment and the integrity of the dataset to evaluate the path planning and operation effect of the robot, and the execution process to generate the path evaluation result is as follows;

[0096] In the process of evaluating the path planning and operation effect of the robot based on the data quality assessment results, the real-time effect of path adjustment and the integrity of the data set are considered. The actual execution effect of the optimized path planning diagram is analyzed in detail, including the extensiveness of the path coverage, the uniformity of data collection, and the timeliness of task completion. Through the evaluation of dimensions, it can be determined whether the optimized path planning diagram effectively improves the operation efficiency and data quality. For example, if the new optimized path planning diagram can reduce the moving distance of the robot while increasing the detection frequency of key areas, it is considered to improve the efficiency and effect. Through repeated testing and adjustment, the path planning diagram is continuously optimized to ensure better adaptation to the changes in the pipeline environment and the needs of robot operation in future tasks, and path evaluation results are generated.

[0097] The above are only the preferred embodiments of the present invention and do not limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A pipeline robot path planning and detection method based on multi-data fusion, characterized in that, It includes the following steps: Collect data from vision, ultrasound, magnetic flux leakage, and lidar, perform timestamp alignment and spatial coordinate correction on the data, conduct data fusion and verify the consistency of the data, and generate an environmental dataset; Based on the environmental dataset, extract structural information and identify abnormal signals, mark the defects and abnormal areas of the pipeline, analyze the internal environment of the pipeline, draw the structural condition of the pipeline, and generate a three-dimensional pipeline model; According to the three-dimensional pipeline model, compare the data of detection requirements and costs between pipeline nodes, determine the access priority of nodes through cost-benefit analysis, and generate an optimized access path; Use the optimized access path to guide the actions of the robot, adjust the speed and steering parameters of the pipeline robot, detect and verify the predetermined path trajectory, and generate an optimized path planning diagram; When the robot moves along the optimized path planning diagram, continuously collect pipeline environment and robot status data, analyze abnormal data in real time, detect data points deviating from the normal state, and generate real-time monitoring results; According to the feedback of the real-time monitoring results, adjust the path and speed of the robot, detect the pipeline, evaluate the quality and integrity of the dataset, and generate a path evaluation result; Based on the three-dimensional pipeline model, determine the access priority of nodes by comparing the data differences between different nodes, and generate a node priority analysis result; Adopt the node priority analysis result, conduct a cost minimization path analysis, calculate the sequence of the lowest-cost access path with reference to the optimal allocation of time and resources, and obtain a cost-based access order analysis result; According to the cost-based access order analysis result, determine the access path for pipeline inspection and repair with reference to real-time operating conditions and dynamic resource allocation, and generate an optimized access path.

2. The method for pipeline robot path planning and detection based on multi-data fusion according to claim 1, characterized in that The environmental dataset includes sensor data with timestamp alignment, measurement points after spatial coordinate correction, and multi-sensor fusion data. The three-dimensional pipeline model includes the spatial layout of the pipeline, connection nodes, and defect locations. The optimized access path includes the economic evaluation of pipeline detection nodes, access priority, and cost calculation. The optimized path planning diagram includes the adjusted speed and steering parameters of the robot, predetermined trajectory, and key points on the path. The real-time monitoring results include records of environmental changes, real-time data of the robot status, and analysis of data deviating from the normal state. The path evaluation result includes newly detected defects, path effectiveness, and evaluation of dataset integrity.

3. The pipeline robot path planning and detection method based on multi-data fusion according to claim 1, wherein The steps of collecting data from vision, ultrasound, magnetic flux leakage, and lidar, performing timestamp alignment and spatial coordinate correction on the data, conducting data fusion and verifying the consistency of the data, and generating an environmental dataset are specifically as follows: Collect vision, ultrasound, magnetic flux leakage, and lidar data, synchronize the time frequency of the sensors during data collection, add a timestamp synchronized with the clock to each data point, and generate a time-synchronized dataset; Adopt the time-synchronized dataset, use a unified spatial reference framework to calibrate the coordinates of vision, ultrasound, magnetic flux leakage, and lidar data, check the spatial consistency of the data by adjusting the sensor position parameters and calibration angles, and generate a spatially corrected dataset; Based on the spatial calibration dataset, fuse the calibrated data, compare the consistency among visual, ultrasonic, magnetic flux leakage, and lidar data, mark the abnormal areas in the data through differential point analysis, and construct an environmental dataset.

4. The pipeline robot path planning and detection method based on multi-data fusion according to claim 1, characterized in that, Based on the environmental dataset, the steps of extracting structural information, identifying abnormal signals, identifying defects and abnormal areas of the pipeline, analyzing the internal environment of the pipeline, and drawing the structural condition of the pipeline to generate a three-dimensional pipeline model are specifically as follows: According to the environmental dataset, extract the structural information of the pipeline, identify the abnormal signals in the data, locate the pipeline defects and abnormal areas through the analysis of signal strength and continuity, generate defect recognition results, and adopt the Bayesian method according to the defect size confidence of the differential detection method to obtain the comprehensive confidence of the defect size at each defect position; According to the defect recognition results, combine the structural information of the pipeline to generate the internal state of the pipeline, draw the structure diagram of the pipeline, analyze the physical structure and internal characteristics of the pipeline, and generate a three-dimensional pipeline model.

5. The pipeline robot path planning and detection method based on multi-data fusion according to claim 1, characterized in that, The steps of using the optimized access path to guide the robot's actions, adjusting the speed and steering parameters of the pipeline robot, detecting and verifying the predetermined path trajectory, and generating an optimized path planning diagram are specifically as follows: Based on the optimized access path, adjust the speed and steering parameters of the pipeline robot, adjust according to the requirements and obstacle information of the target pipeline area, verify that the robot moves along the predetermined trajectory, and generate the robot action parameter adjustment result; Through the robot action parameter adjustment result, test the forward path of the robot, verify the adaptability of speed and steering in the pipeline environment, verify the consistency between the robot behavior and the predetermined path, and generate a path verification result; Utilize the path verification result to integrate the motion data of the robot and the path planning requirements, draw the real-time operation path diagram of the robot, and generate an optimized path planning diagram.

6. The pipeline robot path planning and detection method based on multi-data fusion according to claim 1, characterized in that, When the robot moves along the optimized path planning diagram, the steps of continuously collecting pipeline environment and robot state data, analyzing abnormal data in real time, and detecting data points deviating from the normal state to generate real-time monitoring results are specifically as follows: Based on the optimized path planning diagram, estimate the real-time position of the robot and monitor the environmental changes to generate an environmental change recognition result; Adopt the environmental change recognition result to conduct abnormal analysis on the collected information, identify the data points deviating from the preset parameters, determine the potential risks and abnormal areas through comparative analysis, and generate an abnormal analysis result; According to the abnormal analysis result, analyze the detected abnormal data points, adjust the monitoring strategy and response measures, verify the safety and efficiency of the robot operation, and generate real-time monitoring results.

7. The pipeline robot path planning and detection method based on multi-data fusion according to claim 1, characterized in that, According to the feedback of the real-time monitoring result, adjust the path and speed of the robot, detect the pipeline, and evaluate the quality and integrity of the dataset to generate a path evaluation result. The steps are specifically as follows: Based on the real-time monitoring result, dynamically optimize the path and speed of the robot, avoid risks and optimize efficiency, and generate a path adjustment result; Using the path adjustment result, verify data consistency and coverage, analyze the quality and integrity of the collected data set, check the reliability and validity of the data, and generate a data quality assessment result; Based on the data quality assessment result, evaluate the path planning and operation effect of the robot with reference to the real-time effect of path adjustment and the integrity of the data set, and generate a path assessment result.

Images