A method and system for automatic inspection of offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration
By using a multi-airport collaborative drone system, data on offshore wind turbine blades is collected for comprehensive evaluation and spatial clustering. Drone formations are constructed, shortest paths are planned, and no-fly zone constraints are considered. This solves the problem of the single inspection method for offshore wind turbine blades and achieves efficient and accurate inspection task execution and resource optimization.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 国电投南通新能源有限公司
- Filing Date
- 2026-03-26
- Publication Date
- 2026-06-30
AI Technical Summary
The existing methods for inspecting offshore wind turbine blades are too simplistic and make it difficult to achieve reasonable coordinated scheduling and path planning of multiple drones. This results in untimely and inefficient inspection responses. Furthermore, the existing assessment methods lack modeling of the coupling relationship between the environment and blade operating conditions, making the assessment results susceptible to environmental disturbances and resulting in insufficient accuracy in anomaly identification.
By collecting status perception data of offshore wind turbine blades, an evaluation index such as power, rotational speed, vibration and wind speed is constructed. Combined with the wind direction-blade angle coupling correction model, the wind speed influence value is calculated, and the health value is fused to determine the abnormal state. Spatial clustering is used to form inspection groups. After quantifying the task load, the intelligent decision-making drone formation is made to plan the shortest path and combine the no-fly zone constraint to plan the flight path.
It significantly improves the accuracy of state perception and environmental adaptability, realizes closed-loop automation of anomaly detection, improves inspection efficiency and the accuracy of resource scheduling, reduces the invalid flight distance and transfer time of UAVs, and reduces energy consumption.
Smart Images
Figure CN122308455A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wind turbine blade inspection technology, specifically to an automatic inspection method and system for offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration. Background Technology
[0002] As the core component for wind energy capture and conversion, offshore wind turbine blades are exposed to a complex and ever-changing marine environment. Their structural health directly affects the operational safety and power generation efficiency of the wind turbine. Therefore, conducting efficient, accurate, and routine inspections of offshore wind turbine blades has become a key issue in ensuring the safe and stable operation of wind farms and reducing the total life-cycle maintenance costs. Currently, offshore wind turbine blade inspections mainly employ methods such as manual tower climbing, shipborne close-range observation, and single-unit drone inspections. Among these, manual inspections are significantly limited by sea conditions, weather conditions, and personnel safety, resulting in high operational risks, long cycles, and high costs. Shipborne inspections have limited mobility and coverage efficiency, making them unsuitable for the high-frequency inspection needs of large-scale wind turbine clusters. Drone inspections are primarily conducted as single-unit operations, periodic executions, or manual triggering, with task scheduling relying on human experience and lacking real-time perception and intelligent analysis of wind turbine operating status, making it difficult to achieve truly on-demand inspections and precise deployment. Meanwhile, in terms of blade health status assessment, existing technologies often rely on a single operating parameter or simple threshold for judgment, lacking modeling of the coupling relationship between the environment and blade operating conditions. This makes the assessment results susceptible to environmental disturbances, and the accuracy and foresight of anomaly identification are insufficient, making it difficult to provide reliable decision support for UAV inspection. To address the aforementioned shortcomings, a technical solution is provided. Summary of the Invention
[0003] To address the aforementioned shortcomings of existing technologies, this invention provides an automatic inspection method and system for offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration. This method effectively solves the problems of limited and singular inspection methods for offshore wind turbine blades in existing technologies, which make it difficult to rationally coordinate and schedule multiple UAVs and plan their paths, resulting in untimely inspection response and low inspection efficiency.
[0004] To achieve the above objectives, the present invention can be implemented through the following technical solutions: This invention provides an automatic inspection method for offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration, comprising the following steps: The state perception dataset of offshore wind turbine blades is analyzed to obtain the power assessment index, speed assessment index, vibration intensity assessment index, and wind speed influence value. The health value is then calculated and determined based on the health value to determine whether the state is abnormal. Spatial clustering analysis was performed on the offshore wind turbine blades identified as being in an abnormal state. Based on the geographical location information of the offshore wind turbine blades, inspection groups were constructed, and drone formations were created for these inspection groups. The method for constructing the drone formations was as follows: The scale of the inspection team's tasks is quantitatively assessed to obtain the theoretical total inspection time. Based on the theoretical total inspection time, the maximum sustainable flight time of a single drone, and the safety time margin, the required number of drones is determined. At the same time, the optimal drones are selected based on the comprehensive adaptability value to build a drone formation. Based on the shortest path principle and in conjunction with no-fly zone constraints, flight path planning is performed for drone formations.
[0005] Furthermore, the method for solving the power evaluation index and the speed evaluation index is as follows: The output power of each monitoring time point is extracted from the state-aware dataset, and the maximum and minimum output power are removed. The output power of the remaining monitoring time points is then averaged to obtain the average output power, which is used as the power evaluation index of the offshore wind turbine blades during the current monitoring period. The rotational speed at each monitoring time point is extracted from the state-aware dataset. The difference between the rotational speed at each monitoring time point and the preset reference rotational speed is calculated to obtain the rotational speed deviation value at each monitoring time point. The rotational speed deviation value is then averaged to obtain the average rotational speed deviation value, which is used as the rotational speed evaluation index of the offshore wind turbine blades during the current monitoring period.
[0006] Furthermore, the method for solving the vibration intensity evaluation index is as follows: The vibration intensity of each monitoring time point is extracted from the state-aware dataset. The vibration intensity of each monitoring time point is compared and analyzed with the preset reference vibration intensity. If the vibration intensity of a certain monitoring time point is greater than the preset reference vibration intensity, the monitoring time point is marked as an abnormal time point. The number of abnormal time points and their proportion in all monitoring time points are counted to obtain the abnormal vibration intensity proportion value, which is used as the vibration intensity assessment index of offshore wind turbine blades in the current monitoring period.
[0007] Furthermore, the method for solving the wind speed influence value is as follows: Extract the wind speed and wind direction angle of the environment within the preset monitoring range from the state-aware dataset, and obtain the rotation angle of the corresponding offshore wind turbine blades. Based on the angle relationship between the wind direction and the blade rotation direction, calculate the deviation value between the wind direction angle and the offshore wind turbine blade rotation angle. A correction factor is constructed based on the deviation between the wind direction angle and the rotation angle of the offshore wind turbine blades; The wind speed in the environment is then coupled with the correction factor to obtain the wind speed influence value.
[0008] Furthermore, the method for solving the health value is as follows: The health value is obtained by weighted and fused calculation of the power evaluation index, speed evaluation index, vibration intensity evaluation index and wind speed influence value of offshore wind turbine blades.
[0009] Furthermore, the method for constructing the inspection group is as follows: Based on the geographical location information of offshore wind turbine blades, abnormal offshore wind turbine blades whose spatial distance from each other is less than a preset distance threshold are automatically grouped into the same inspection group, thus forming a spatially concentrated inspection group.
[0010] Furthermore, the method for quantitatively evaluating the scale of the inspection team's tasks is as follows: Suppose the inspection group contains N offshore wind turbine blades in abnormal condition. Based on historical inspection data, determine the average operation time required for a single offshore wind turbine blade to complete one standard inspection. And the average ferry flight time between adjacent offshore wind turbine blades within the group is Therefore, the theoretical total inspection time for the inspection team is considered. To solve this problem, we use the following calculation formula: The theoretical total inspection time for the inspection group is obtained. .
[0011] Furthermore, the method of selecting the optimal drone based on the comprehensive adaptability value to construct the drone formation is as follows: Key performance parameters of each UAV are collected, including the remaining battery power, the estimated flight time to reach the first inspection task point, and the payload capacity. The key performance parameters are weighted and calculated to obtain the comprehensive adaptability value of each UAV. The drones are sorted from high to low based on their overall adaptability scores, and drones with the highest overall adaptability scores and sufficient numbers are selected to form a drone formation.
[0012] Furthermore, the method for flight path planning of the drone formation is as follows: The takeoff point of each drone, the wind turbine blade inspection target points assigned to the drone, and the preset return point after the mission are all mapped to the same flight coordinate system to form an operational space network that includes all inspection and flight key nodes. All areas within the operational airspace that need to be avoided are defined as no-fly zones, and in the flight coordinate system, no-fly zones are modeled as impassable areas with spatial boundaries; Based on this, for each drone, a path search algorithm is used to calculate the shortest feasible flight path from the takeoff point to the first inspection target point under the condition of satisfying the no-fly zone constraint; Then, based on the principle of spatial proximity, the next target point closest to the current point is selected from the target points that the drone has not yet inspected, and the path search algorithm is called again to plan the shortest feasible transfer path to the target point; By iterating in the above manner, the inspection target points assigned to the same drone are sequentially connected to form a flight path.
[0013] Furthermore, an automated inspection system for offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration includes: The operation status monitoring module is used to analyze the status perception dataset of offshore wind turbine blades to obtain power assessment index, speed assessment index, vibration intensity assessment index and wind speed influence value, and to perform comprehensive calculation to obtain health value. Based on the health value, it is determined whether it is an abnormal state. The drone matching module is used to perform spatial clustering analysis on offshore wind turbine blades identified as being in an abnormal state. Based on the geographical location information of the offshore wind turbine blades, it constructs inspection groups and then creates drone formations for these groups. The method for constructing the drone formations is as follows: The scale of the inspection team's tasks is quantitatively assessed to obtain the theoretical total inspection time. Based on the theoretical total inspection time, the maximum sustainable flight time of a single drone, and the safety margin, the required number of drones is determined. Simultaneously, the optimal drones are selected based on the comprehensive suitability value to construct a drone swarm. The path planning module is used to plan the flight paths of drone formations based on the shortest path principle and in conjunction with no-fly zone constraints.
[0014] The technical solution provided by this invention has the following advantages compared with the known prior art: 1. This invention collects data such as power, rotational speed, vibration, wind speed, and wind direction, and obtains individual evaluation indices through robustness processing. Then, it introduces a coupled correction model of wind direction and blade angle to calculate the wind speed influence value that reflects the actual aerodynamic load. Finally, it normalizes each index and weights and fuses it with a reference benchmark to output a unified quantitative health value. Based on this health value, the blade status can be automatically determined to be normal or abnormal, thereby significantly improving the accuracy of status perception, environmental adaptability, and early warning capability. This provides a core decision-making basis for shifting from periodic inspections to status-driven intelligent operation and maintenance, and realizes closed-loop automation of anomaly detection and response.
[0015] 2. This invention integrates nearby abnormal blades into inspection groups through spatial clustering, intelligently decides between single-machine or formation mode after quantitatively assessing the task load, accurately calculates the required number of drones, and finally selects the optimal drones to form a formation based on multi-dimensional performance evaluation. This achieves efficient aggregation of inspection tasks, precise scheduling and adaptive optimization of resources, and significantly improves overall inspection efficiency.
[0016] 3. This invention digitally models the physical space involved in the inspection operation and each inspection task node, mapping the UAV takeoff point, inspection target point, and return point to the same spatial coordinate system. Based on this, flight safety rules such as no-fly zones are introduced to impose clear safety constraints on the operational airspace. Subsequently, a path optimization algorithm is used to sequentially connect and jointly solve the above spatial nodes, thereby effectively reducing the invalid flight distance and transfer time of the UAV during the inspection process, improving the overall execution efficiency of a single inspection task, and reducing the energy consumption of the UAV. Attached Figure Description
[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0018] Figure 1 This is a flowchart illustrating the overall process of the present invention.
[0019] Figure 2 This is an overall module block diagram of the present invention. Detailed Implementation
[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0021] like Figure 1 As shown, an automatic inspection method for offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration includes the following steps: A1. During the current monitoring period, condition-aware data is collected from the offshore wind turbine blades to form a condition-aware dataset. This dataset includes the output power, rotational speed, and vibration intensity of the offshore wind turbine blades during operation, as well as the wind speed and direction angle within the preset monitoring range. Based on this, a comprehensive analysis of the condition-aware dataset is performed to construct and calculate the health value of the offshore wind turbine blades. The specific calculation process is as follows: The output power at each monitoring time point is extracted from the state-aware dataset, and the maximum and minimum output power are removed. The output power of the remaining monitoring time points is then averaged to obtain the average output power, which serves as the power assessment index for offshore wind turbine blades during the current monitoring period. ; Rotational speeds at each monitoring time point are extracted from the state-aware dataset. The difference between the rotational speed at each monitoring time point and the preset reference rotational speed is calculated to obtain the rotational speed deviation value for each monitoring time point. Then, the rotational speed deviation values are averaged to obtain the average rotational speed deviation value, which serves as the rotational speed evaluation index for offshore wind turbine blades during the current monitoring period. ; Vibration intensity at each monitoring time point is extracted from the state-aware dataset. The vibration intensity at each monitoring time point is compared with a preset reference vibration intensity. If the vibration intensity at a certain monitoring time point is greater than the preset reference vibration intensity, that monitoring time point is marked as an abnormal time point. The number of abnormal time points and their proportion among all monitoring time points are counted to obtain the abnormal vibration intensity proportion value, which serves as the vibration intensity assessment index for offshore wind turbine blades during the current monitoring period. ; The wind speed and wind direction angle of the environment within the preset monitoring range are extracted from the state-aware dataset. The preset monitoring range can be specifically set by those skilled in the art according to the actual situation. The rotation angle of the corresponding offshore wind turbine blades is obtained. Based on the angular relationship between the wind direction and the blade rotation direction, the deviation value between the wind direction angle and the rotation angle of the offshore wind turbine blades is calculated. The calculation formula is as follows: ,in, For wind direction angle, For rotation angle, This is the deviation value; The deviation between the wind direction angle and the rotation angle of the offshore wind turbine blades is used as a correction factor for wind speed directionality. The correction factor can be expressed as: ; Based on this, the directionality of the ambient wind speed is corrected, and the ambient wind speed is coupled with the correction factor in the calculation to obtain the wind speed influence value. The calculation formula is as follows: ,in, For the wind speed in the environment; Power assessment index of offshore wind turbine blades during the current monitoring period Speed evaluation index Vibration intensity assessment index and wind speed impact value After normalizing each indicator, a weighted fusion calculation is performed to obtain the health value of the offshore wind turbine blades. The calculation formula is as follows: ,in, , and These represent the preset reference power evaluation index, reference speed evaluation index, and reference vibration intensity evaluation index, respectively. , and These are respectively represented as the allowable power assessment index difference, speed assessment index difference, and vibration intensity assessment index difference. , and It is represented by the set weight coefficient, and satisfies ; The operating status of offshore wind turbine blades is determined based on their health values. Specifically, the health value of the offshore wind turbine blade is compared with a preset health comparison range. If the health value of the offshore wind turbine blade is outside the preset health comparison range, the offshore wind turbine blade is determined to be in an abnormal state. Conversely, if the health value of the offshore wind turbine blade is within the preset health comparison range, the offshore wind turbine blade is determined to be in a normal state. In one specific embodiment, the present invention collects data such as power, rotational speed, vibration, wind speed, and wind direction, obtains individual evaluation indices through robustness processing, and then introduces a coupled correction model of wind direction and blade angle to calculate the wind speed influence value reflecting the actual aerodynamic load. Then, each index is normalized and weighted and fused with a reference benchmark to output a unified quantitative health value. Based on this health value, the blade status can be automatically determined to be normal or abnormal. This method effectively integrates aerodynamic principles and operational data, significantly improving the accuracy of status perception, environmental adaptability, and early warning capability. It provides a core decision-making basis for shifting from periodic inspections to status-driven intelligent operation and maintenance, and realizes closed-loop automation of anomaly detection and response.
[0022] A2. Spatial clustering analysis is performed on offshore wind turbine blades identified as abnormal. Specifically, a clustering algorithm (such as the DBSCAN algorithm based on Euclidean distance) is used. Based on the geographical location information of the offshore wind turbine blades, abnormal offshore wind turbine blades whose spatial distance from each other is less than a preset distance threshold are automatically grouped into the same inspection group, thus forming a spatially concentrated inspection group. Each inspection group corresponds to a batch of spatially highly concentrated abnormal offshore wind turbine blades. The UAV formation can complete the inspection operation sequentially in a predetermined order during a single voyage, thereby effectively reducing the transfer flight distance and time of UAVs between different inspection targets and improving the overall inspection efficiency. The process of constructing the UAV formation is as follows: For a single inspection group, the scale of its inspection task is quantitatively assessed. Assuming the inspection group contains N offshore wind turbine blades in abnormal condition, based on historical inspection data, the average operation time required for a single offshore wind turbine blade to complete one standard inspection is determined as follows: The average operation time includes the drone's approach to the blade, data acquisition and withdrawal process, as well as the average transfer flight time between adjacent offshore wind turbine blades within the group. Therefore, the theoretical total inspection time for the inspection team is considered. To solve this problem, we use the following calculation formula: The theoretical total inspection time for the inspection group is obtained. ; Obtain the maximum sustainable flight time of a single drone. and safety time margin According to the formula: To obtain the required number of drones G, where, The preset collaborative compensation coefficient is used to characterize the additional time overhead caused by communication, avoidance and synchronization control during multi-machine collaboration. It should be noted that in actual scheduling, the number of drones must be an integer. Therefore, the result G is processed by the principle of rounding up, that is, the final formation size is the smallest integer not less than G. This principle ensures that the allocated drone resources are theoretically sufficient to cover the total task load and coordination overhead, providing the necessary resource redundancy and reliability guarantee for the successful execution of the task. After determining the required number of drones, the drones in the available drone resource pool are screened. Specifically, key performance parameters of each drone are collected from the available drone resource pool. These key performance parameters include the drone's remaining battery power, the estimated flight time to reach the first inspection task point, and the payload capacity. The available drone resource pool refers to the intelligent airports for offshore drones distributed around the offshore wind turbine blades. Each intelligent airport has functions such as automatic storage, charging or battery swapping, data communication, and meteorological environment monitoring for drones, which are used to ensure the rapid deployment and stable operation of drones. Subsequently, based on the collected key performance parameters, the comprehensive suitability value for each UAV is calculated according to a preset comprehensive evaluation method, wherein the comprehensive evaluation method is as follows: Let the remaining battery power of the i-th drone be... The estimated time to arrive at the first inspection task point is The payload capacity is Then its overall fitness value The calculation formula is: ,in, This represents the maximum remaining battery power of drones in the available drone resource pool, used for normalizing the remaining battery power. This represents the shortest estimated time for all drones to reach the first mission point, used to demonstrate time efficiency advantages. This represents the maximum available payload capacity of drones in the available drone resource pool, used for payload capacity normalization. , and All are weighting coefficients, and satisfy the following conditions: ; The drones are sorted from high to low based on their overall adaptability scores. Then, drones with the highest overall adaptability scores and whose numbers meet the aforementioned requirements are selected to form a drone formation for performing the inspection tasks of the current inspection team. In one specific embodiment, the present invention integrates adjacent abnormal leaves into inspection groups through spatial clustering, makes intelligent decisions on single-machine or formation mode after quantitatively evaluating the task load, accurately calculates the required number of drones, and finally selects the optimal drones to form a formation based on multi-dimensional performance evaluation. This achieves efficient aggregation of inspection tasks, precise scheduling and adaptive optimization of resources, and significantly improves the overall inspection efficiency.
[0023] A3. Based on the shortest path principle and in conjunction with no-fly zone constraints, collaborative flight path planning is performed for drone formations. The specific implementation process is as follows: Based on the geographical location information of all offshore wind turbine blades in the inspection group and the location information of the airport (take-off point) where each UAV in the formation is stationed, a unified inspection operation airspace model is constructed. Specifically, the take-off point of each UAV, the inspection target points of each wind turbine blade assigned to the UAV, and the preset return point after the mission are completed are uniformly mapped to the same flight coordinate system (such as a local three-dimensional coordinate system based on UTM projection), thereby forming an operation space network that includes all inspection and flight key nodes. Secondly, all areas that need to be avoided within the operational airspace (such as offshore oil and gas platforms, busy waterways, ecological protection zones, and temporary controlled airspace) are defined as no-fly zones. In the flight coordinate system, the no-fly zones are modeled as impassable areas with spatial boundaries to form flight safety constraints, so as to ensure that the generated UAV flight paths do not pass through the no-fly zones in subsequent path planning. Based on the aforementioned airspace model and no-fly zone constraints, with the optimization objective of minimizing the total flight distance and flight time of UAVs, the execution order and corresponding flight paths of the UAV formation's inspection tasks are planned. The specific planning method is as follows: For each UAV, a path search algorithm (such as the A* algorithm) is used to calculate the shortest feasible flight path from its take-off point to the first inspection target point under the condition of satisfying the no-fly zone constraints. Then, after completing the inspection of the current target point, according to the principle of spatial proximity, the next inspection target point closest to the current point is selected from the target points that the UAV has not yet inspected, and the path search algorithm is called again to plan the shortest feasible transfer path to the target point. By iterating in the above way, the inspection target points assigned to the same drone are sequentially connected until all inspection tasks of the drone are completed, and finally a flight path is formed from the take-off point, visiting all inspection target points in sequence and finally arriving at the return point. In one specific embodiment, the present invention digitally models the physical space involved in the inspection operation and each inspection task node, mapping the UAV takeoff point, inspection target point, and return point to the same spatial coordinate system. On this basis, flight safety rules such as no-fly zones are introduced to impose clear safety constraints on the operational airspace. Subsequently, a path optimization algorithm is used to sequentially connect and jointly solve the above spatial nodes. Thus, under the premise of strictly avoiding all no-fly zones and ensuring flight safety, the invalid flight distance and transfer time of the UAV during the inspection process are effectively reduced. This not only significantly improves the overall execution efficiency of a single inspection task but also reduces the energy consumption of the UAV.
[0024] like Figure 2 As shown, an automatic inspection system for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) includes an operation status monitoring module, a UAV matching module, and a path planning module. The operation status monitoring module is used to collect status perception data of offshore wind turbine blades during the current monitoring period and form a status perception dataset. The status perception dataset includes the output power, rotational speed and vibration intensity of the offshore wind turbine blades during operation, as well as the wind speed and wind direction angle of the environment within the preset monitoring range. Based on this, the status perception dataset is comprehensively analyzed to construct and calculate the health value of the offshore wind turbine blades. The module determines whether the condition is abnormal based on the comparison relationship between the health value and the preset health comparison interval. The process of constructing and calculating the health value of offshore wind turbine blades is as follows: The output power of each monitoring time point is extracted from the state-aware dataset, and the maximum and minimum output power are removed. The output power of the remaining monitoring time points is then averaged to obtain the average output power, which is used as the power evaluation index of the offshore wind turbine blades during the current monitoring period. The rotational speed at each monitoring time point is extracted from the state-aware dataset. The difference between the rotational speed at each monitoring time point and the preset reference rotational speed is calculated to obtain the rotational speed deviation value at each monitoring time point. The rotational speed deviation value is then averaged to obtain the average rotational speed deviation value, which is used as the rotational speed evaluation index of the offshore wind turbine blades in the current monitoring period. The vibration intensity of each monitoring time point is extracted from the state-aware dataset. The vibration intensity of each monitoring time point is compared and analyzed with the preset reference vibration intensity. If the vibration intensity of a monitoring time point is greater than the preset reference vibration intensity, the monitoring time point is marked as an abnormal time point. The number of abnormal time points and their proportion in all monitoring time points are counted to obtain the abnormal vibration intensity proportion value, which is used as the vibration intensity assessment index of offshore wind turbine blades in the current monitoring period. Extract the wind speed and wind direction angle of the environment within the preset monitoring range from the state-aware dataset, and obtain the rotation angle of the corresponding offshore wind turbine blades. Based on the angle relationship between the wind direction and the blade rotation direction, calculate the deviation value between the wind direction angle and the offshore wind turbine blade rotation angle. A correction factor is constructed based on the deviation between the wind direction angle and the rotation angle of the offshore wind turbine blades; Then, the environmental wind speed and the correction factor are coupled and calculated to obtain the wind speed influence value; The health value is obtained by weighted and fused calculation of the power evaluation index, speed evaluation index, vibration intensity evaluation index and wind speed influence value of offshore wind turbine blades.
[0025] The UAV matching module is used to perform spatial clustering analysis on offshore wind turbine blades that are determined to be in an abnormal state. Specifically, it uses a clustering algorithm to automatically group abnormal offshore wind turbine blades whose spatial distance from each other is less than a preset distance threshold into the same inspection group based on the geographical location information of the offshore wind turbine blades. This forms a spatially concentrated inspection group, and each inspection group corresponds to a batch of abnormal offshore wind turbine blades that are highly concentrated in space. The UAV formation can complete the inspection operation in a predetermined order during a single voyage. The process of building a drone swarm is as follows: Suppose the inspection group contains N offshore wind turbine blades in abnormal condition. Based on historical inspection data, determine the average operation time required for a single offshore wind turbine blade to complete one standard inspection. And the average ferry flight time between adjacent offshore wind turbine blades within the group is Therefore, the theoretical total inspection time for the inspection team is considered. To solve this problem, we use the following calculation formula: The theoretical total inspection time for the inspection group is obtained. ; Obtain the maximum sustainable flight time of a single drone. and safety time margin According to the formula: To obtain the required number of drones G, where, This is the preset collaborative compensation coefficient; Key performance parameters of each UAV are collected, including the remaining battery power, the estimated flight time to reach the first inspection task point, and the payload capacity. The key performance parameters are weighted and calculated to obtain the comprehensive adaptability value of each UAV. The drones are sorted from high to low based on their overall adaptability scores, and drones with the highest overall adaptability scores and sufficient numbers are selected to form a drone formation.
[0026] The path planning module is used to plan collaborative flight paths for drone formations based on the shortest path principle and in conjunction with no-fly zone constraints. Its specific implementation process is as follows: Based on the geographical location information of all offshore wind turbine blades in the inspection group and the location information of the airports where each UAV in the formation is stationed, a unified inspection operation airspace model is constructed. Specifically, the take-off point of each UAV, the inspection target points of each wind turbine blade assigned to the UAV, and the preset return point after the mission are completed are uniformly mapped to the same flight coordinate system (such as a local three-dimensional coordinate system based on UTM projection), thereby forming an operation space network that includes all inspection and flight key nodes. Secondly, all areas that need to be avoided within the operational airspace (such as offshore oil and gas platforms, busy waterways, ecological protection zones, and temporary controlled airspace) are defined as no-fly zones. In the flight coordinate system, the no-fly zones are modeled as impassable areas with spatial boundaries to form flight safety constraints, so as to ensure that the generated UAV flight paths do not pass through the no-fly zones in subsequent path planning. Based on the aforementioned airspace model and no-fly zone constraints, with the optimization objective of minimizing the total flight distance and flight time of UAVs, the execution order and corresponding flight paths of the UAV formation's inspection tasks are planned. The specific planning method is as follows: For each UAV, a path search algorithm (such as the A* algorithm) is used to calculate the shortest feasible flight path from its take-off point to the first inspection target point under the condition of satisfying the no-fly zone constraints. Then, after completing the inspection of the current target point, according to the principle of spatial proximity, the next inspection target point closest to the current point is selected from the target points that the UAV has not yet inspected, and the path search algorithm is called again to plan the shortest feasible transfer path to the target point. By iterating in the above manner, the inspection target points assigned to the same UAV are sequentially connected until all inspection tasks of the UAV are completed, ultimately forming a flight path that starts from the take-off point, visits all inspection target points in sequence, and finally arrives at the return point.
[0027] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims
1. A method for automatic inspection of offshore wind turbine blades using unmanned aerial vehicles (UAVs) based on multi-airport collaboration, characterized in that, Includes the following steps: The state perception dataset of offshore wind turbine blades is analyzed to obtain the power assessment index, speed assessment index, vibration intensity assessment index, and wind speed influence value. The health value is then calculated and determined based on the health value to determine whether the state is abnormal. Spatial clustering analysis was performed on the offshore wind turbine blades identified as being in an abnormal state. Based on the geographical location information of the offshore wind turbine blades, inspection groups were constructed, and drone formations were created for these inspection groups. The method for constructing the drone formations was as follows: The scale of the inspection team's tasks is quantitatively assessed to obtain the theoretical total inspection time. Based on the theoretical total inspection time, the maximum sustainable flight time of a single drone, and the safety time margin, the required number of drones is determined. At the same time, the optimal drones are selected based on the comprehensive adaptability value to build a drone formation. Based on the shortest path principle and in conjunction with no-fly zone constraints, flight path planning is performed for drone formations.
2. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for solving the power evaluation index and the speed evaluation index is as follows: The output power of each monitoring time point is extracted from the state-aware dataset, and the maximum and minimum output power are removed. The output power of the remaining monitoring time points is then averaged to obtain the average output power, which is used as the power evaluation index of the offshore wind turbine blades during the current monitoring period. The rotational speed at each monitoring time point is extracted from the state-aware dataset. The difference between the rotational speed at each monitoring time point and the preset reference rotational speed is calculated to obtain the rotational speed deviation value at each monitoring time point. The rotational speed deviation value is then averaged to obtain the average rotational speed deviation value, which is used as the rotational speed evaluation index of the offshore wind turbine blades during the current monitoring period.
3. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for solving the vibration intensity evaluation index is as follows: The vibration intensity of each monitoring time point is extracted from the state-aware dataset. The vibration intensity of each monitoring time point is compared and analyzed with the preset reference vibration intensity. If the vibration intensity of a certain monitoring time point is greater than the preset reference vibration intensity, the monitoring time point is marked as an abnormal time point. The number of abnormal time points and their proportion in all monitoring time points are counted to obtain the abnormal vibration intensity proportion value, which is used as the vibration intensity assessment index of offshore wind turbine blades in the current monitoring period.
4. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for solving the wind speed influence value is as follows: Extract the wind speed and wind direction angle of the environment within the preset monitoring range from the state-aware dataset, and obtain the rotation angle of the corresponding offshore wind turbine blades. Based on the angle relationship between the wind direction and the blade rotation direction, calculate the deviation value between the wind direction angle and the offshore wind turbine blade rotation angle. A correction factor is constructed based on the deviation between the wind direction angle and the rotation angle of the offshore wind turbine blades; The wind speed in the environment is then coupled with the correction factor to obtain the wind speed influence value.
5. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for calculating the health value is as follows: The health value is obtained by weighted and fused calculation of the power evaluation index, speed evaluation index, vibration intensity evaluation index and wind speed influence value of offshore wind turbine blades.
6. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for constructing the inspection group is as follows: Based on the geographical location information of offshore wind turbine blades, abnormal offshore wind turbine blades whose spatial distance from each other is less than a preset distance threshold are automatically grouped into the same inspection group, thus forming a spatially concentrated inspection group.
7. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for quantitatively assessing the task scale of the inspection team is as follows: Suppose the inspection group contains N offshore wind turbine blades in abnormal condition. Based on historical inspection data, determine the average operation time required for a single offshore wind turbine blade to complete one standard inspection. And the average ferry flight time between adjacent offshore wind turbine blades within the group is Therefore, the theoretical total inspection time for the inspection team is considered. To solve this problem, we use the following calculation formula: The theoretical total inspection time for the inspection group is obtained. .
8. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for selecting the optimal drone based on the comprehensive adaptability value to construct a drone formation is as follows: Key performance parameters of each UAV are collected, including the remaining battery power, the estimated flight time to reach the first inspection task point, and the payload capacity. The key performance parameters are weighted and calculated to obtain the comprehensive adaptability value of each UAV. The drones are sorted from high to low based on their overall adaptability scores, and drones with the highest overall adaptability scores and sufficient numbers are selected to form a drone formation.
9. The automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs) according to claim 1, characterized in that, The method for flight path planning of drone formations is as follows: The takeoff point of each drone, the wind turbine blade inspection target points assigned to the drone, and the preset return point after the mission are all mapped to the same flight coordinate system to form an operational space network that includes all inspection and flight key nodes. All areas within the operational airspace that need to be avoided are defined as no-fly zones, and in the flight coordinate system, no-fly zones are modeled as impassable areas with spatial boundaries; Based on this, for each drone, a path search algorithm is used to calculate the shortest feasible flight path from the takeoff point to the first inspection target point under the condition of satisfying the no-fly zone constraint; Then, based on the principle of spatial proximity, the next target point closest to the current point is selected from the target points that the drone has not yet inspected, and the path search algorithm is called again to plan the shortest feasible transfer path to the target point; By iterating in the above manner, the inspection target points assigned to the same drone are sequentially connected to form a flight path.
10. An automatic inspection system for offshore wind turbine blades based on multi-airport collaboration using unmanned aerial vehicles (UAVs), applied to the automatic inspection method for offshore wind turbine blades based on multi-airport collaboration using UAVs as described in claim 1, characterized in that... include: The operation status monitoring module is used to analyze the status perception dataset of offshore wind turbine blades to obtain power assessment index, speed assessment index, vibration intensity assessment index and wind speed influence value, and to perform comprehensive calculation to obtain health value. Based on the health value, it is determined whether it is an abnormal state. The drone matching module is used to perform spatial clustering analysis on offshore wind turbine blades identified as being in an abnormal state. Based on the geographical location information of the offshore wind turbine blades, it constructs inspection groups and then creates drone formations for these groups. The method for constructing the drone formations is as follows: The scale of the inspection team's tasks is quantitatively assessed to obtain the theoretical total inspection time. Based on the theoretical total inspection time, the maximum sustainable flight time of a single drone, and the safety margin, the required number of drones is determined. Simultaneously, the optimal drones are selected based on the comprehensive suitability value to construct a drone swarm. The path planning module is used to plan the flight paths of drone formations based on the shortest path principle and in conjunction with no-fly zone constraints.