An unmanned aerial vehicle urban road roadside facility inspection route autonomous planning method

Abstract

The present application relates to the field of unmanned aerial vehicle roadside facility inspection flight path planning, and specifically discloses a kind of unmanned aerial vehicle city road roadside facility inspection flight path autonomous planning method, comprising: first obtain sparse milepost of one side or both sides of city road and roadside facility GPS point sequence Pmile, generate high-density target tracking trajectory by smooth curve interpolation;Combined with the best shooting distance Dopt and 10 meters horizontal no-fly safety distance D NFZ Revision D opt , along the outside normal reverse generation parallel initial flight path;For minimum curvature smoothing, dynamic drift angle calculation and vertical safety distance overhead obstacle avoidance, generate optimized flight path and write into task file, embed milepost GPS coordinate and associate flight path point.The present application solves the problems of low inspection efficiency, difficult data positioning and rough attitude control, and is suitable for city expressway, mountain road and other scenes.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) roadside facility inspection route planning technology, specifically a method for autonomous planning of UAV roadside facility inspection routes in urban areas. Background Technology

[0002] In the field of drone-based urban roadside facility inspection route planning technology, the regular inspection of roadside guardrails, streetlights, cables and other facilities is a key link in ensuring the safe operation of urban roads, and efficient and accurate drone route planning directly determines the quality and efficiency of the inspection.

[0003] Existing drone inspection route planning methods mainly employ parallel routes along the centerline of urban roads or manually set waypoints, but these methods have significant limitations:

[0004] The inspection is inefficient and of inconsistent quality: Existing methods cannot ensure that the drone payload always maintains the best side-view angle and distance to the roadside vertical facilities, resulting in insufficient validity of the captured data. In addition, double-sided inspections require separate route planning, resulting in redundant round-trip flights and low overall efficiency.

[0005] Difficulty in data location and correlation: Traditional drone aerial photography data lacks direct binding with key nodes of urban roads, and the data on defects or facility status obtained during inspections cannot be accurately traced back to specific locations, causing great inconvenience to subsequent maintenance and repair work;

[0006] Attitude control lacks structured optimization: existing routes mostly control the attitude of UAVs, such as yaw angle, by setting fixed angles or simply along the route direction. They fail to make dynamic and fine adjustments based on the geometric characteristics of roadside facilities, which cannot meet the precise requirements of side-view tracking and shooting.

[0007] Poor adaptability to complex scenarios: In scenarios with dense obstacles such as overpasses on urban expressways and overhead power lines, traditional manual waypoint planning is prone to problems such as excessive track curvature and inaccurate obstacle avoidance, which not only affects flight stability but also poses safety hazards. Furthermore, it is difficult to quickly locate the damaged facilities during emergency inspections after urban disasters.

[0008] In summary, existing drone-based route planning technologies for inspecting roadside facilities in urban areas suffer from problems such as low efficiency, inaccurate data association, coarse attitude control, and insufficient adaptability to complex scenarios. There is an urgent need for an autonomous planning method that can integrate structured information about urban roads, dynamically optimize flight paths and attitudes, accurately avoid obstacles, and bind data with location information to improve inspection efficiency and accuracy. Summary of the Invention

[0009] To address the aforementioned problems in the prior art, this invention provides an autonomous route planning method for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas. This method solves problems such as low inspection efficiency, difficulty in associating data with mileage, coarse attitude control, and insufficient obstacle avoidance, thereby improving inspection efficiency, positioning accuracy, and flight safety.

[0010] To achieve the above objectives, this invention proposes an autonomous route planning method for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas, comprising:

[0011] S1. Obtain roadside structured input: Obtain the GPS point sequence Pmile of sparse mileage markers or roadside facilities on one or both sides of urban roads.

[0012] S2. Generate target tracking trajectory: Based on the GPS point sequence Pmile, a high-density roadside facility target tracking trajectory is generated using a smooth curve interpolation algorithm, which serves as the desired target tracking line for the UAV camera. ;

[0013] S3. Initial track generation under constraints: Based on the preset optimal shooting distance D opt And the horizontal no-fly distance D over urban roads NFZ Forced D opt Take max(D) opt D NFZ Based on the corrected D opt and The outer normal direction is used to generate the reverse calculation with respect to the direction of the outer normal. Parallel initial tracks ;

[0014] S4. Dynamic Attitude and Path Integration Optimization: Initial Track The minimum curvature smoothing algorithm is applied, dynamic yaw angle is calculated, and vertical obstacle avoidance is performed. The UAV attitude is calculated simultaneously to generate the final optimized trajectory A'.

[0015] S5. Data Association and Output: The final optimized flight track A' generated after optimization is written into the flight mission file, and the GPS coordinates of the mile markers are embedded in the mission file and associated with the nearest track point to achieve precise binding of flight and shooting data with urban road mileage information.

[0016] Preferably, in S4, the initial track is... The specific steps for performing minimum curvature smoothing algorithm processing, dynamic yaw angle calculation, and vertical obstacle avoidance processing include:

[0017] S41. Smoothing algorithm processing: [This part is incomplete and requires further context.] The waypoints on the surface are processed using a curvature-constrained smoothing algorithm to meet the minimum turning radius constraint of the UAV;

[0018] S42. Dynamic Yaw Angle Calculation: Dynamically calculates any waypoint on the track after processing by S41. Yaw angle Ensure that the main optical axis of the drone camera is precisely pointed to the target tracking trajectory. The corresponding target point G on j ;

[0019] S43. Vertical Safety Avoidance of Overhead Obstacles: When the flight path passes through an overhead obstacle O k When the horizontal projection area is reached, the flight altitude ZA' of the flight segment is forcibly raised to meet the vertical safety constraints:

[0020] ZA'>ZO+H safety ;

[0021] Where ZO is the elevation of the highest point of the obstacle, and H safety For vertical safety constraints, ZA' is the flight altitude of the segment.

[0022] 3. The method for autonomous planning of flight routes for UAV inspection of roadside facilities in urban areas according to claim 2, characterized in that, in S41, the curvature-constrained smoothing algorithm is a Dubins curve or a G curve. 2 Continuous cubic spline interpolation.

[0023] Preferably, in S42, any waypoint on the track processed by S41 Yaw angle The calculation formula is:

[0024] ;

[0025] In the formula, These are the horizontal coordinates of the target point. These are the horizontal coordinates of the waypoints. It is the camera's pitch correction setting.

[0026] Preferably, in S43, the vertical safety constraint H safety At a height of 20 meters, during the process of safely avoiding the overhead obstacle vertically, the UAV continuously maintains its yaw angle while flying at an altitude of ZA'. Precisely pointing to the target and tracking the trajectory .

[0027] Preferably, in S1, the specific method for obtaining roadside structured input is to collect or retrieve records from the urban road basic database through on-site GPS acquisition equipment. The obtained GPS point sequence Pmile must contain the three-dimensional coordinate information of longitude, latitude and altitude corresponding to each point.

[0028] Preferably, in S2, the smooth curve interpolation algorithm is cubic spline interpolation or B-spline interpolation.

[0029] Preferably, in S3, the horizontal no-fly safety distance D above the urban roads NFZ It is 10 meters.

[0030] Preferably, in S5, the specific steps for data association and output are as follows: after organizing all waypoint coordinates of the optimized track A' in flight order, write them into a standard format flight mission file; extract the GPS three-dimensional coordinates of each mile marker; calculate the straight-line distance between each coordinate and all waypoints on the optimized track A'; uniquely bind each mile marker to the waypoint with the smallest distance; and mark the bound mile marker identification information in the corresponding waypoint record of the flight mission file to complete the accurate association between the captured data and the urban road mileage information.

[0031] Therefore, this invention proposes an autonomous route planning method for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas, which has the following advantages:

[0032] (1) Achieve integrated optimization of flight path and attitude, rely on the core planning logic of side-looking tracking-smooth trajectory change, reduce redundant flight, adapt to various urban road inspection scenarios, dynamically adjust the yaw angle of the UAV, ensure the effectiveness and consistency of roadside facility photography, and significantly improve the efficiency of inspection operations.

[0033] (2) Construct a precise correlation mechanism between inspection data and urban road mileage information to solve the problem of traditional inspection data location backtracking; through multiple safety constraints and dynamic obstacle avoidance processing, ensure the safety and stability of UAV flight operations and improve the reliability of inspection operations.

[0034] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0035] Figure 1 This is a schematic diagram of the overall process of the present invention, which describes an autonomous route planning method for inspecting roadside facilities on urban roads using unmanned aerial vehicles (UAVs). Detailed Implementation

[0036] To make the technical solutions, advantages, and objectives of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below. 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 described embodiments of the present invention without creative effort are within the protection scope of this application.

[0037] Unless otherwise defined, the technical or scientific terms used in this invention shall have the ordinary meaning as understood by one of ordinary skill in the art to which this invention pertains.

[0038] like Figure 1 As shown, the present invention provides an autonomous route planning method for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas, comprising:

[0039] S1. Obtain roadside structured input: Obtain the GPS point sequence Pmile of sparse mileage markers or roadside facilities on one or both sides of urban roads.

[0040] The specific method for obtaining roadside structured input is to collect data through on-site GPS acquisition equipment or retrieve records from the urban road basic database. The obtained GPS point sequence Pmile must contain the three-dimensional coordinate information of longitude, latitude, and altitude for each point.

[0041] S2. Generate target tracking trajectory: Based on the GPS location sequence Pmile, a smooth curve interpolation algorithm is used to generate high-density roadside facility target tracking trajectories, which serve as the desired target tracking line for the UAV camera. The smooth curve interpolation algorithm is either cubic spline interpolation or B-spline interpolation.

[0042] S3. Initial track generation under constraints: Based on the preset optimal shooting distance D opt And the horizontal no-fly distance D over urban roads NFZ Forced D opt Take max(D) opt D NFZ Based on the corrected D opt and The outer normal direction is used to generate the reverse calculation with respect to the direction of the outer normal. Parallel initial tracks The safe distance for horizontal no-fly zones over urban roads is 10 meters.

[0043] S4. Dynamic Attitude and Path Integration Optimization: Initial Track The minimum curvature smoothing algorithm is applied, dynamic yaw angle is calculated, and vertical obstacle avoidance is performed. The UAV attitude is calculated simultaneously to generate the final optimized trajectory A'.

[0044] For the initial track The specific steps for performing minimum curvature smoothing algorithm processing, dynamic yaw angle calculation, and vertical obstacle avoidance processing include:

[0045] S41. Smoothing algorithm processing: [This part is incomplete and requires further context.] The waypoints on the surface are processed using a curvature-constrained smoothing algorithm to meet the minimum turning radius constraint of the UAV;

[0046] Curvature-constrained smoothing algorithms include Dubins curves or cubic spline interpolation for G2 continuity;

[0047] S42. Dynamic Yaw Angle Calculation: Dynamically calculates any waypoint on the track after processing by S41. Yaw angle Ensure that the main optical axis of the drone camera is precisely pointed to the target tracking trajectory. The corresponding target point G on j ;

[0048] Dynamic yaw angle The calculation formula is:

[0049] ;

[0050] In the formula, These are the horizontal coordinates of the target point. These are the horizontal coordinates of the waypoints. It is the camera's pitch correction setting.

[0051] S43. Vertical Safety Avoidance of Overhead Obstacles: When the flight path passes through an overhead obstacle O k When the horizontal projection area is reached, the flight altitude ZA' of the flight segment is forcibly raised to meet the vertical safety constraints:

[0052] ZA'>ZO+H safety ;

[0053] Where ZO is the elevation of the highest point of the obstacle, and H safety For vertical safety constraints, ZA' is the flight altitude of the segment.

[0054] The vertical safety constraint Hsafety is 20 meters. During the vertical safety avoidance process, the UAV maintains its yaw angle continuously while flying at an altitude ZA'. Precisely pointing to the target and tracking the trajectory .

[0055] S5. Data Association and Output: The final optimized flight track A' generated after optimization is written into the flight mission file, and the GPS coordinates of the mile markers are embedded in the mission file and associated with the nearest track point to achieve precise binding of flight and shooting data with urban road mileage information.

[0056] The specific steps for data association and output are as follows: After organizing all waypoint coordinates of optimized track A' in flight order, write them into a standard format flight mission file; extract the GPS three-dimensional coordinates of each mile marker; calculate the straight-line distance between each coordinate and all waypoints on optimized track A'; uniquely bind each mile marker to the waypoint with the smallest distance; and mark the bound mile marker identification information in the corresponding waypoint record of the flight mission file to complete the accurate association between the captured data and urban road mileage information.

[0057] This invention employs a data input module, a target modeling module, an initial trajectory calculation module, a dynamic attitude and path optimization module, and a data association and output module to implement an autonomous route planning method for unmanned aerial vehicles (UAVs) to inspect roadside facilities in urban areas.

[0058] Example: Scenario of inspecting facilities on both sides of an urban expressway.

[0059] This embodiment is applied to the inspection of facilities on both sides of a city ring expressway. The road section is 15 kilometers long and is equipped with continuous guardrails, intermittent streetlights and traffic signs on both sides. There are 3 overpasses and 8 overhead cable corridors in the road section. Traditional inspection requires planning routes on both sides separately, which has problems such as redundant round-trip flights, insufficient accuracy of obstacle avoidance for overpasses and cables, and inability to directly correlate the shooting data with the mileage markers. The present invention is used to carry out autonomous route planning and inspection operations.

[0060] Hardware equipment: Multi-rotor UAV (equipped with GPS positioning module, high-definition camera payload, and lidar obstacle avoidance sensor) and GPS data acquisition equipment (positioning accuracy ±0.1 meters) are selected.

[0061] Basic data: The urban road database of this expressway was retrieved to obtain the GPS point sequence of sparse kilometer markers on both sides (1 point per kilometer, a total of 30 points, including three-dimensional coordinates of longitude, latitude, and altitude, such as the coordinates of point AK1+000 being 116.35°E, 39.92°N, and 45 meters above sea level). At the same time, GPS information of key points of street light bases on both sides was collected as a supplement.

[0062] Preset parameters: Optimal shooting distance D opt The preset safe distance for horizontal no-fly zones over urban roads is 8 meters (D). NFZ The document specifies a height of 10 meters, with a vertical safety constraint H. safety The minimum turning radius of the UAV is 5 meters, and the camera pitch angle correction term is 20 meters. Set it to 0.05 rad.

[0063] S1: Obtain roadside structured input:

[0064] By combining database retrieval with on-site data collection, GPS point sequences (Pmiles) of sparse mile markers and street light bases on both sides of the expressway were obtained. The left Pmile sequence contains 15 mile marker points and 8 street light base points, while the right Pmile sequence contains 15 mile marker points and 7 street light base points. All points contain complete three-dimensional coordinate information of longitude, latitude, and altitude, forming a structured input dataset.

[0065] S2: Generate target tracking trajectory:

[0066] Based on the Pmile sequence, a cubic spline interpolation algorithm (a type of smooth curve interpolation algorithm) was used for high-density interpolation processing, with an interpolation interval of 0.5 meters, to generate target tracking trajectories for roadside facilities on both sides. and Both tracks follow the distribution of facilities on both sides, serving as the target tracking lines for the drone camera to ensure full coverage of guardrails, streetlights, and traffic signs.

[0067] S3: Initial trajectory generation under constraints:

[0068] Based on the preset parameters, the optimal shooting distance D opt =8 meters, horizontal no-fly distance D above city roads NFZ =10 meters, D is mandatory according to the rules. opt Take max(8,10) = 10 meters; based on the corrected D opt =10 meters, respectively along and Using the outer normal direction, a reverse calculation is performed to generate an initial left-side track parallel to the two target tracking trajectories. and the initial track on the right The two initial tracks maintain a fixed distance of 10 meters from the corresponding target tracking line.

[0069] S4: Integrated optimization of dynamic attitude and path:

[0070] Smooth trajectory change: The Dubins curve algorithm (a type of curvature-constrained smoothing algorithm) is used for... and The waypoints on the flight path are subjected to curvature constraint processing to ensure that the curvature of all flight segments meets the constraint of the minimum turning radius of 5 meters for the UAV, so that the flight path is smooth and without abrupt changes;

[0071] Dynamic yaw angle calculation: for the optimized trajectory and Each waypoint According to the formula Dynamic calculation of yaw angle , It is the horizontal coordinate of the target point on the target tracking trajectory. It optimizes the horizontal coordinates of the waypoints to ensure that the main optical axis of the drone camera always points precisely to the facilities on both sides;

[0072] Vertical safety avoidance: Using drone obstacle detection sensors, the horizontal projection areas of three overpasses and eight overhead cables are identified. When the flight path crosses this area, the corresponding flight segment's altitude ZA' is forcibly raised to ensure that ZA' > ZO + 20 meters (ZO is the altitude of the highest point of the obstacle), and the yaw angle is maintained continuously during the raising process. Pointing to target tracking trajectory It does not affect the shooting angle.

[0073] S5: Data Association and Output:

[0074] The optimized bilateral flight path and The data is integrated into a single continuous flight mission file using a single-sided priority + dual-sided switching logic. All GPS coordinates of the mile markers are embedded, and the straight-line distance between each mile marker and all waypoints on the track is calculated one by one. Each mile marker is uniquely bound to the waypoint with the smallest distance, and the mile marker is marked in the corresponding waypoint record in the mission file (e.g., AK3+120). This completes the precise association between the captured data and the urban road mileage information. The mission file is then imported into the UAV flight control system to perform inspection operations.

[0075] The flight path planned using the method of this invention enables integrated inspection of facilities on both sides of urban expressways, eliminating the need for separate planning of routes on both sides and reducing redundant flights. After smoothing, the flight path is stable, obstacle avoidance is precise, and there is no risk of collision. The drone's yaw angle is dynamically adjusted to ensure that guardrails, streetlights, and traffic signs on both sides are at the optimal side-viewing angle, resulting in clear and effective data. All inspection data is accurately linked to mileage information on mileage markers, allowing for direct tracing of facility locations via mileage markers, providing precise data support for maintenance work, and fully adapting to the complex inspection scenarios of urban expressways.

[0076] Therefore, this invention provides an autonomous flight path planning method for UAV inspection of roadside facilities in urban areas. Using sparse mile markers or GPS locations of roadside facilities as core input, a high-density target tracking trajectory is generated through smooth curve interpolation. An initial flight path is generated by combining the optimal shooting distance and horizontal no-fly constraints. Then, through minimum curvature smoothing, dynamic yaw angle adjustment, and vertical obstacle avoidance, the flight path and attitude are optimized in an integrated manner. Finally, the optimized flight path is precisely bound to the mile marker coordinates for output. This method completely solves the problems of low efficiency, difficult data positioning, coarse attitude control, and insufficient obstacle avoidance in complex scenarios associated with traditional inspection methods. It eliminates the need for repeated manual planning, is adaptable to various scenarios such as urban expressways and mountain roads, and achieves autonomous, precise, and safe inspection operations, providing efficient data support for roadside facility maintenance.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for autonomously planning flight routes for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas, characterized in that, include: S1. Obtain roadside structured input: Obtain the GPS point sequence Pmile of sparse mileage markers or roadside facilities on one or both sides of urban roads; The specific method for obtaining roadside structured input is to collect data through on-site GPS data collection equipment or retrieve records from the urban road basic database. The obtained GPS point sequence Pmile must contain the three-dimensional coordinate information of longitude, latitude and altitude corresponding to each point. S2. Generate target tracking trajectory: Based on the GPS location sequence Pmile, a high-density roadside facility target tracking trajectory is generated using a smooth curve interpolation algorithm, which serves as the desired target tracking trajectory for the UAV camera. ; S3. Initial track generation under constraints: Based on the preset optimal shooting distance D opt And the horizontal no-fly distance D over urban roads NFZ Forced D opt Take max(D) opt D NFZ Based on the corrected D opt and The outer normal direction is used to generate the reverse calculation with respect to the direction of the outer normal. Parallel initial tracks ; S4. Dynamic Attitude and Path Integration Optimization: Initial Track The minimum curvature smoothing algorithm is applied, dynamic yaw angle is calculated, and vertical obstacle avoidance is performed. The UAV attitude is calculated simultaneously to generate the final optimized trajectory A'. S5. Data Association and Output: The final optimized track A' generated after optimization is written into the flight mission file, and the GPS coordinates of the mile markers are embedded in the mission file and associated with the nearest track point to achieve precise binding of flight and shooting data with urban road mileage information; In S4, the initial track The specific steps for performing minimum curvature smoothing algorithm processing, dynamic yaw angle calculation, and vertical obstacle avoidance processing include: S41. Smoothing algorithm processing: [This part is incomplete and requires further context.] The waypoints on the surface are processed using a curvature-constrained smoothing algorithm to meet the minimum turning radius constraint of the UAV; S42. Dynamic Yaw Angle Calculation: Dynamically calculates any waypoint on the track after processing by S41. Yaw angle Ensure that the main optical axis of the drone camera is precisely pointed to the target tracking trajectory. The corresponding target point G on j Any waypoint on the track processed by S41 Yaw angle The calculation formula is: ； In the formula, These are the horizontal coordinates of the target point. These are the horizontal coordinates of the waypoints. It is the camera's pitch correction factor; S43. Vertical Safety Avoidance of Overhead Obstacles: When the flight path passes through an overhead obstacle O k When the horizontal projection area is reached, the flight altitude ZA' of the flight segment is forcibly raised to meet the vertical safety constraints: ZA'>ZO+H safety ; Where ZO is the elevation of the highest point of the obstacle, and H safety For vertical safety constraints, ZA' is the flight altitude for the segment; In S5, the specific steps for data association and output are as follows: After organizing all waypoint coordinates of the optimized track A' in flight order, write them into a standard format flight mission file; extract the GPS three-dimensional coordinates of each mile marker; calculate the straight-line distance between each coordinate and all waypoints on the optimized track A'; uniquely bind each mile marker to the waypoint with the smallest distance; and mark the bound mile marker identification information in the corresponding waypoint record of the flight mission file to complete the accurate association between the captured data and the urban road mileage information.

2. The method for autonomous planning of flight routes for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas according to claim 1, characterized in that, In S41, the curvature-constrained smoothing algorithm is a Dubins curve or a G curve. 2 Continuous cubic spline interpolation.

3. The method for autonomous planning of flight routes for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas according to claim 1, characterized in that, In S43, the vertical safety constraint H safety At a height of 20 meters, during the process of safely avoiding the overhead obstacle vertically, the UAV continuously maintains its yaw angle while flying at an altitude of ZA'. Precisely pointing to the target and tracking the trajectory .

4. The method for autonomous planning of flight routes for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas according to claim 1, characterized in that, In S2, the smooth curve interpolation algorithm is cubic spline interpolation or B-spline interpolation.

5. The method for autonomous planning of flight routes for unmanned aerial vehicle (UAV) inspection of roadside facilities in urban areas according to claim 1, characterized in that, In S3, the horizontal no-fly safety distance D above the city roads NFZ It is 10 meters.

Images