Mountainous road unmanned aerial vehicle inspection track optimization method

Abstract

The application discloses a mountainous road unmanned aerial vehicle (UAV) inspection track optimization method, and belongs to the technical field of UAV track planning.The application solves the problems that the existing track path planning method does not realize sufficient sensing of a dynamic environment, does not consider the balance between mountainous risk and inspection efficiency, and has poor dynamic adaptability.The application firstly constructs a space risk field quantitative model based on a mountainous road inspection environment including a terrain and a wind field, and optimizes an inspection sequence of each inspection point;the space risk field quantitative model and the inspection sequence can be combined to realize joint optimization of risk and efficiency.Then, a group intelligence optimization algorithm designed by the application is used to establish a track optimization model, and an optimal track is obtained through the established track optimization model;finally, a 3-order B-spline curve is used to smooth the obtained optimal track, and finally, a road UAV inspection path suitable for a complex dynamic environment of a mountainous area is obtained.The method of the application can be applied to mountainous road UAV inspection track planning.

Description

Technical Field

[0001] This invention belongs to the field of UAV trajectory planning technology, specifically relating to a method for optimizing the trajectory of UAV inspections on mountain roads. Background Technology

[0002] Mountain roads face numerous challenges in daily inspection due to their complex terrain and variable weather. Traditional manual inspection methods are inefficient and pose safety hazards, making drone-based inspections of mountain roads a growing trend. However, existing drone flight path planning methods for mountain road inspections have the following shortcomings:

[0003] (1) Simple environmental modeling: Mountainous environments are generally regarded as a static set of obstacles, ignoring dynamic risk factors such as terrain undulations and wind field changes, resulting in insufficient safety of flight path;

[0004] (2) Single optimization objective: Generally, the shortest path is the core objective, without comprehensively considering the balance between mountain risks and inspection efficiency, which makes it difficult to meet the actual needs of mountain road inspection;

[0005] (3) Poor dynamic adaptability: When faced with sudden weather changes or temporary addition of inspection points, the route replanning response is delayed, affecting the continuity of inspection tasks.

[0006] Therefore, in order to address the problems of existing flight path planning methods failing to fully perceive the dynamic environment, failing to consider the balance between mountain risks and inspection efficiency, and having poor dynamic adaptability, there is an urgent need for a UAV flight path planning method that can accurately perceive mountain environmental risks, perform multi-objective collaborative optimization, and possess dynamic adaptability, so as to improve the safety, efficiency, and dynamic adaptability of mountain highway inspections. Summary of the Invention

[0007] The purpose of this invention is to address the problems of existing flight path planning methods failing to fully perceive the dynamic environment, failing to consider the balance between mountainous risks and inspection efficiency, and having poor dynamic adaptability. Therefore, this invention proposes a flight path optimization method for UAV inspections of mountainous highways, which achieves safe, efficient, and adaptive mountainous highway inspection path planning.

[0008] The technical solution adopted by this invention to solve the above-mentioned technical problems is: a method for optimizing the flight path of unmanned aerial vehicle (UAV) inspections on mountain roads, the method specifically including the following steps:

[0009] Step 1: Based on the planar coordinates of all points to be inspected, determine the inspection sequence list for each point by the UAV.

[0010] Step 2: Initialize the inspection point numbers in the list. ;

[0011] Step 3: Plan from the list of items... The first inspection point to the first The route to each inspection point;

[0012] The specific process of step three is as follows:

[0013] Step 3.1 Set the total number of iteration rounds to 1. The initial value for initializing the association information between adjacent path points is... ;

[0014] Determine the drone from the first The first inspection point to the first All the path points that the inspection point is allowed to pass through, will be the first The first inspection point to the first All allowed path points between inspection points and the first inspection point The set of points consisting of inspection points is denoted as . ;

[0015] Step 3.2: Initialize the number of iteration rounds ;

[0016] Step 3: Initialize the selected path point count ;

[0017] Steps three and four: Calculate the drone's performance in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of the nth path point is used to select the point corresponding to the highest probability calculated. 1 path point;

[0018] The selected number The path point and the selected path point The path point and the selected path point The path points between the path points are from the set Remove from the middle to obtain the updated set. ;

[0019] Step 35: Determine if the drone has been selected. One inspection point;

[0020] If the drone has already selected the first The number of inspection points indicates that all path points selected during the k-th iteration form the drone's path in the k-th iteration. The first inspection point to the first Inspection flight paths between inspection points Continue with step three six.

[0021] If the drone did not select the first Each inspection point then ordered Return to steps three and four;

[0022] Step 36: Calculate the inspection flight path Corresponding total cost of the trajectory And update the inspection tracks based on the total track cost corresponding to each inspection track in the window. The associated information for each road segment; and the number of iteration rounds are determined. Is it equal to ;

[0023] like Then let Based on the updated association information, return to step three.

[0024] like Then proceed to step three seven;

[0025] Step 37: Select the path with the minimum total path cost from the K paths, and use the selected path as the UAV's path from the Kth path. The first inspection point to the first The inspection flight track of each inspection point;

[0026] Step 38: Determine if the inspection point number has reached its maximum.

[0027] If the inspection point number reaches the maximum, proceed to step five;

[0028] If the inspection point number has not reached the maximum, proceed to step four;

[0029] Step 4, Order Return to step three;

[0030] Step 5: Smooth the flight path of the UAV from the starting point to each inspection point in the inspection sequence list to obtain the smoothed inspection flight path.

[0031] Furthermore, the specific process of step one is as follows:

[0032] Step 11: Denote the set of all points to be inspected as follows: Calculate the starting point and set of the drones respectively. The Euclidean distance of each point to be inspected is used to determine the first point to be inspected, and the point with the smallest Euclidean distance is selected as the first inspection point.

[0033] And the first inspection point was set up at the assembly point. Delete, and get the updated set of points to be inspected. ;

[0034] Steps 1 and 2: Initialize the number of iterations ;

[0035] Step 1 and Step 3: Calculate the first three parts separately. Inspection points and assembly The Euclidean distance of each point to be inspected is used to determine the point to be inspected that has the smallest Euclidean distance. One inspection point;

[0036] And the first Each inspection point from the assembly Delete, and get the updated set of points to be inspected. ;

[0037] Step 14, Order Return to steps one and three until the set of inspection points is empty.

[0038] Furthermore, in steps three and four, the calculation of the drone in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of each path point; specifically:

[0039]

[0040] In the formula, Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point , For the first In the round of iteration, starting from the first... From the path point to the first Local trajectory cost generated by each waypoint; For the first In the round of iteration, starting from the first... From the path point to the first The probability of path selection for each path point; Indicates the first After -1 iterations, the... From the path point to the first Information relating to each path point; Representing path points For set The point in the middle; Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point; Influence factors on the correlation information between path points; This refers to the heuristic information influencing factors between path points.

[0041] Furthermore, the aforementioned from the first From the path point to the first Local trajectory cost generated by each waypoint The calculation method is as follows:

[0042]

[0043] in, Indicates from the first From the path point to the first The risk and cost of each path point Indicates from the first From the path point to the first The time cost of each path point The weighting coefficient for risk cost. The weighting coefficient for time cost. Indicates the first Spatial risk field at each path point Indicates the first Spatial risk field at each path point Indicates the first Spatial coordinates of each path point Indicates the first Spatial coordinates of each path point This indicates the inspection speed of the drone.

[0044] Furthermore, the first Spatial risk field at each path point The calculation method is as follows:

[0045]

[0046] in, Indicates the first Terrain risks at each waypoint Indicates the first Wind field risk at each path point express The weighting coefficients, express The weighting coefficients.

[0047] Furthermore, the first Terrain risks at each waypoint The calculation method is as follows:

[0048] Step 1, calculate the first... Elevation risk at each path point :

[0049]

[0050] In the formula: For drones in the Flight altitude at each waypoint; For the first Ground elevation at the projected coordinates of each path point on the ground; For safety height threshold; The base of the natural logarithm; This represents the elevation risk attenuation coefficient.

[0051] Step 2, calculate the first Slope risk at each path point :

[0052]

[0053] In the formula: This refers to the slope risk coefficient. The norm operator for vectors;

[0054] Step 3, calculate the first... Risk of terrain undulation at each waypoint :

[0055]

[0056] In the formula: Indicates the first Taking the projection of each path point on the ground as the center of a circle, with... The number of elevation sampling points within a local area with a radius of ; For the local area of ​​the first Ground elevation of each sampling point; The safety factor is the terrain relief. The maximum elevation difference threshold;

[0057] Step 4: Based on elevation risk Slope risk and terrain relief risk Calculate terrain risk :

[0058]

[0059] In the formula: This represents the weighting coefficient corresponding to elevation risk. This represents the weighting coefficient corresponding to slope risk. This represents the weighting coefficient corresponding to the risk of terrain relief.

[0060] Furthermore, the first Wind field risk at each path point The calculation method is as follows:

[0061] Step (1) Calculate the first Wind speed risk at each path point :

[0062]

[0063] In the formula: The safe wind speed threshold; For shape parameters; For the first Wind speed at each path point;

[0064] Step (2), calculate the first Wind shear risk at each path point :

[0065]

[0066] In the formula: The weighting coefficient for space wind shear risk; The weighting coefficient for time-varying wind shear risk; for In a Cartesian coordinate system, the X-axis component... for In a spatial rectangular coordinate system, the Y-axis component, for The Z-axis component in a spatial rectangular coordinate system; The Frobenius norm operator for matrices;

[0067] Step (3), calculate the first Wind field risk at each path point :

[0068]

[0069] In the formula: This is the weighting coefficient for wind speed risk; The weighting coefficient for wind shear risk.

[0070] Furthermore, the inspection track Corresponding total cost of the trajectory The calculation method is as follows:

[0071]

[0072] in, Indicates the inspection track The corresponding risk and cost, Indicates the inspection track The corresponding time cost, Indicates the flight path The total number of path points on the path.

[0073] Furthermore, the inspection track is updated based on the total track cost corresponding to each inspection track within the window. The associated information for each road segment on the road; the specific process is as follows:

[0074]

[0075] In the formula: Q is the correlation information attenuation coefficient; Q is the correlation information constant. For the set of tracks within the window, when hour, ,when hour, , The preset window length; For a high-quality subset of tracks, i.e. from The M paths with the minimum total path cost are selected; g is... A flight path in the middle; The total trajectory cost of trajectory g; The weight of the high-quality trajectory g; For path segment indication functions, when track g contains the first segment... From the path point to the first When there are path segments between path points, The value is 1, otherwise The value of is 0;

[0076]

[0077] in, For high-quality track subsets The minimum total cost of all paths in the middle.

[0078] Furthermore, the specific process of step five is as follows:

[0079] Step 51: Divide the entire inspection track before smoothing into curve segments. Specifically, take the 0th to 3rd path point on the entire inspection track as the 0th curve segment, take the 1st to 4th path point as the 1st curve segment, and so on, taking the (W-4)th to (W-1)th path point as the (W-4)th curve segment.

[0080] Step 52, regarding spline parameters ,definition The range of values ​​is ,in , and define , ;

[0081] For spline parameters For any given value, first determine The curve segment number corresponding to the point , It is an integer, and Then calculate the B-spline basis functions. :

[0082]

[0083] Among them, intermediate variables , , ;

[0084] Calculate the spline parameters again Corresponding spatial coordinates :

[0085]

[0086] in, This represents the first point on the entire inspection track before smoothing. Spatial coordinates of path points ;

[0087] Step 53: The trajectory formed by the spatial coordinates corresponding to each spline parameter value constitutes the smoothed inspection track.

[0088] The beneficial effects of this invention are:

[0089] This invention first constructs a spatial risk field quantification model based on multi-source risk factors in mountainous highway inspection environments, including terrain and wind fields, and optimizes the inspection sequence of each inspection point. Combining the spatial risk field quantification model and the inspection sequence enables joint optimization of risk and efficiency, achieving a balance between risk and inspection efficiency in mountainous areas. Finally, a self-designed swarm intelligence optimization algorithm is used to establish a trajectory optimization model. The optimal trajectory is obtained through this model and smoothed using a 3rd-order B-spline curve, ultimately yielding a highway UAV inspection path adapted to the complex and dynamic environment of mountainous areas. Furthermore, when sudden weather changes or temporary additions of inspection points occur, trajectory planning can be re-planned at any time, improving the dynamic adaptability of the method.

[0090] The method for optimizing the flight path of UAV inspections on mountain roads constructed in this invention can solve the problems of insufficient dynamic risk perception, failure to consider the balance between mountain risks and inspection efficiency, and poor dynamic adaptability of existing path planning methods due to the complex and changeable mountain road environment. It can optimize the flight path planning scheme for mountain road inspections, which is of great significance for reducing UAV inspection accidents and improving the safety and efficiency of mountain road inspections. Attached Figure Description

[0091] Figure 1 This is a flowchart of a method for optimizing the flight path of a drone for inspecting mountain roads according to the present invention;

[0092] I represents the total number of inspection points in the inspection sequence list. Detailed Implementation

[0093] Specific implementation method one: Combining Figure 1 This embodiment describes a method for optimizing the flight path of a drone for inspecting mountain roads. The method specifically includes the following steps:

[0094] Step 1: Based on the planar coordinates of all points to be inspected, determine the inspection sequence list for each point by the UAV.

[0095] Step 2: Initialize the inspection point numbers in the list. ;

[0096] Step 3: Plan from the list of items... The first inspection point to the first The route to each inspection point;

[0097] It should be noted that when At that time, the plan was from the drone's starting point to the... The path to each inspection point, when At that time, the plan was for drones to start from the first The first inspection point to the first The route to each inspection point;

[0098] The specific process of step three is as follows:

[0099] Step 3.1 Set the total number of iteration rounds to 1. The initial value for initializing the association information between adjacent path points is... ( Indicates the adjacent first The path point and the first The initial value of the association information between path points is set in this invention. The value of is 1, when At that time, it represents the initial value of the association information between the inspection point and the first path point.

[0100] Determine the drone from the first The first inspection point to the first All the path points that the inspection point is allowed to pass through, will be the first The first inspection point to the first All allowed path points between inspection points and the first inspection point The set of points consisting of inspection points is denoted as . ;

[0101] Step 3.2: Initialize the number of iteration rounds ;

[0102] Step 3: Initialize the selected path point count The present invention sets that when planning the path between any two adjacent inspection points, the selected path points are numbered starting from 1, and the path point numbered 0 is the first inspection point among the adjacent inspection points.

[0103] Steps three and four: Calculate the drone's performance in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of the nth path point is used to select the point corresponding to the highest probability calculated. 1 path point;

[0104] The selected number The path point and the selected path point The path point and the selected path point Path points between path points from the set The process involves removing (since we determined all allowed pathpoints before planning began, and the principle is to avoid backtracking, after selecting each pathpoint, we need to remove all allowed pathpoints preceding that selected pathpoint from the candidate set) to obtain the updated set. ;

[0105] Step 35: Determine if the drone has been selected. One inspection point;

[0106] If the drone has already selected the first The number of inspection points indicates that all path points selected during the k-th iteration form the drone's path in the k-th iteration. The first inspection point to the first Inspection flight paths between inspection points Continue with step three six.

[0107] If the drone did not select the first Each inspection point then ordered Return to steps three and four;

[0108] Step 36: Calculate the inspection flight path Corresponding total cost of the trajectory And update the inspection tracks based on the total track cost corresponding to each inspection track in the window. The association information corresponding to each road segment on the track (each two adjacent waypoints on the track correspond to one road segment); and the number of iteration rounds is determined. Is it equal to ;

[0109] like Then let Based on the updated association information, return to step three.

[0110] like Then proceed to step three seven;

[0111] Step 37: Select the path with the minimum total path cost from the K paths, and use the selected path as the UAV's path from the Kth path. The first inspection point to the first The inspection flight path of each inspection point;

[0112] Step 38: Determine if the inspection point number has reached its maximum.

[0113] If the inspection point number reaches the maximum, proceed to step five;

[0114] If the inspection point number has not reached the maximum, proceed to step four;

[0115] Step 4, Order Return to step three;

[0116] Step 5: Smooth the flight path of the UAV from the starting point to each inspection point in the inspection sequence list to obtain the smoothed inspection flight path.

[0117] The total number of allowed path points between adjacent inspection points is determined in the following way:

[0118] First, a 3D terrain model (DEM) is constructed using an elevation map. Then, information on obstacles such as trees and rocks in the mountainous area is supplemented by LiDAR scanning or visual image recognition to build a complete 3D spatial scene. Next, a minimum flight altitude constraint and a safe obstacle buffer distance based on the fuselage size are set. The space is discretized into 3D candidate nodes that meet the altitude requirements. The Dijkstra path planning algorithm is used to check the connections between nodes one by one, eliminate collision risks, and retain feasible connections to form candidate path groups. Finally, nodes on all candidate paths are extracted to form a set of feasible path points, and interpolation and encryption processing is performed to obtain all feasible path points between the UAV from the (i-1)th inspection point to the ith inspection point.

[0119] Specific Implementation Method Two: This implementation method is a further limitation of Specific Implementation Method One. The specific process of step one is as follows:

[0120] Step 11: Denote the set of all points to be inspected as follows: Calculate the starting point and set of the drones respectively. The Euclidean distance of each point to be inspected is used to determine the first point to be inspected, and the point with the smallest Euclidean distance is selected as the first inspection point.

[0121] And the first inspection point was set up at the assembly point. Delete, and get the updated set of points to be inspected. ;

[0122] Steps 1 and 2: Initialize the number of iterations ;

[0123] Step 1 and Step 3: Calculate the first three parts separately. Inspection points and assembly The Euclidean distance of each point to be inspected is used to determine the point to be inspected that has the smallest Euclidean distance. One inspection point;

[0124] And the first Each inspection point from the assembly Delete, and get the updated set of points to be inspected. ;

[0125] Step 14, Order Return to steps one and three until the set of inspection points is empty.

[0126] The other steps and parameters are the same as in Specific Implementation Method 1.

[0127] When the set of inspection points is empty, the inspection order of all inspection points is determined. If the drone needs to return to the starting point, add the starting point to the end of the inspection point sequence. This will give you a list of inspection points ordered by Euclidean distance with no backtracking. Subsequent paths between adjacent inspection points can then be planned according to this inspection order.

[0128] Specific Implementation Method Three: This implementation method further defines Specific Implementation Method Two. In steps three and four, the calculation of the UAV in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of each path point; specifically:

[0129]

[0130] In the formula, Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point , For the first In the round of iteration, starting from the first... From the path point to the first Local trajectory cost generated by each waypoint; For the first In the round of iteration, starting from the first... From the path point to the first The probability of path selection for each path point; Indicates the first After -1 iterations, the th From the path point to the first Information relating to each path point; Representing path points For set The point in the middle; Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point; The influence factor of the correlation information between path points is set to 1 in this invention; The value of the heuristic information influence factor between path points is 2 in this invention.

[0131] The other steps and parameters are the same as in Specific Implementation Method 2.

[0132] Specific Implementation Method Four: This implementation method is a further limitation of Specific Implementation Method Three, wherein the description from the first... From the path point to the first Local trajectory cost generated by each waypoint The calculation method is as follows:

[0133]

[0134] in, Indicates from the first From the path point to the first The risk and cost of each path point Indicates from the first From the path point to the first The time cost of each path point The weighting coefficient for risk cost. The weighting coefficient for time cost. Indicates the first Spatial risk field at each path point Indicates the first Spatial risk field at each path point Indicates the first Spatial coordinates of each path point Indicates the first Spatial coordinates of each path point This indicates the inspection speed of the drone.

[0135] The other steps and parameters are the same as in Specific Implementation Method 3.

[0136] In this invention, The value range is 0.7 to 0.9. The value ranges from 0.1 to 0.3, and satisfies the following conditions: Furthermore, it should be noted that when hour, It indicates the first Spatial risk fields at each inspection point.

[0137] Specific Implementation Method Five: This implementation method is a further limitation of Specific Implementation Method Four, wherein the first... Spatial risk field at each path point The calculation method is as follows:

[0138]

[0139] in, Indicates the first Terrain risks at each waypoint Indicates the first Wind field risk at each path point express The weighting coefficients, express The weight coefficients, and satisfying .

[0140] The other steps and parameters are the same as in Specific Implementation Method Four.

[0141] Specific Implementation Method Six: This implementation method is a further limitation of Specific Implementation Method Five, wherein the first... Terrain risks at each waypoint The calculation method is as follows:

[0142] Step 1, calculate the first... Elevation risk at each path point :

[0143]

[0144] In the formula: For drones in the The flight altitude at each waypoint can be obtained using airborne GNSS equipment, barometric altimeter, and in conjunction with LiDAR or visual sensors. For the first The ground elevation at the projected coordinates of each path point on the ground can be obtained by matching high-precision DEM data with real-time supplementary measurement data. The safety height threshold ranges from 5m to 20m. The base of the natural logarithm; This represents the elevation risk attenuation coefficient, with a value range of 0.1m. -1 ~0.5m -1 .

[0145] Step 2, calculate the first Slope risk at each path point :

[0146]

[0147] In the formula: The slope risk coefficient ranges from 0.02 to 0.08. The norm operator for vectors is calculated by first finding the sum of squares of all elements of the vector, and then taking the square root of the sum.

[0148] Step 3, calculate the first... Risk of terrain undulation at each waypoint :

[0149]

[0150] In the formula: Indicates the first Taking the projection of each path point on the ground as the center of a circle, with... The number of elevation sampling points within a local area of ​​radius ( The value can be set based on experience. For the local area of ​​the first Ground elevation of each sampling point; The safety factor for terrain relief ranges from 0.6 to 0.8. This is the maximum elevation difference threshold that the drone can withstand, with a value ranging from 35m to 40m.

[0151] Step 4: Based on elevation risk Slope risk and terrain relief risk Calculate terrain risk :

[0152]

[0153] In the formula: This represents the weighting coefficient corresponding to elevation risk. This represents the weighting coefficient corresponding to slope risk. This represents the weighting coefficient corresponding to the risk of terrain relief, and satisfies... .

[0154] The other steps and parameters are the same as in Specific Implementation Method 5.

[0155] Specific Implementation Method Seven: This implementation method is a further limitation of Specific Implementation Method Six, wherein the first... Wind field risk at each path point The calculation method is as follows:

[0156] Step (1) Calculate the first Wind speed risk at each path point :

[0157]

[0158] In the formula: The safe wind speed threshold ranges from 8 m / s to 14 m / s. This is a shape parameter, with a value range of 0.3 s / m to 0.8 s / m; For the first Wind speed at each path point;

[0159] Step (2), calculate the first Wind shear risk at each path point :

[0160]

[0161] In the formula: This is a weighting coefficient for space wind shear risk. The value range is 0.6 to 0.9; The weighting coefficient for time-varying wind shear risk. The value range is 0.1 to 0.4; and it satisfies... ; for In a Cartesian coordinate system, the X-axis component... for In a spatial rectangular coordinate system, the Y-axis component, for In a Cartesian coordinate system, the Z-axis component... , , This can be obtained from meteorological data or real-time sensors; The Frobenius norm operator is used to calculate the Frobenius norm of a matrix. The Frobenius norm is calculated by first finding the sum of the squares of all elements in the matrix, and then taking the square root of the sum.

[0162] Step (3), calculate the first Wind field risk at each path point :

[0163]

[0164] In the formula: This is the weighting coefficient for wind speed risk, with a value ranging from 0.4 to 0.7. The weighting coefficient for wind shear risk ranges from 0.3 to 0.6 and satisfies the following conditions: .

[0165] The other steps and parameters are the same as in Specific Implementation Method Six.

[0166] Specific Implementation Method Eight: This implementation method is a further limitation of Specific Implementation Method Seven, wherein the inspection track... Corresponding total cost of the trajectory The calculation method is as follows:

[0167]

[0168] in, Indicates the inspection track The corresponding risk and cost, Indicates the inspection track The corresponding time cost, Indicates the flight path The total number of path points on the path.

[0169] The other steps and parameters are the same as in Specific Implementation Method Seven.

[0170] Specific Implementation Method Nine: This implementation method further defines Specific Implementation Method Eight, wherein the inspection track is updated based on the total track cost corresponding to each inspection track within the window. The associated information for each road segment on the road; the specific process is as follows:

[0171]

[0172] In the formula: is the correlation information attenuation coefficient (recommended value: 0.2); Q is the correlation information constant (recommended value: 400). For the set of tracks within the window, when hour, ,when hour, , For the preset window length, The recommended value is 5; For a high-quality subset of tracks, i.e. from The M tracks with the lowest total track cost are selected (M is the pre-set number of high-quality tracks, recommended to be 3; when k < M, it is considered...). g is A flight path in the middle; The total trajectory cost of trajectory g; The weight of the high-quality trajectory g; For path segment indication functions, when track g contains the first segment... From the path point to the first When there are path segments between path points, The value is 1, otherwise The value of is 0;

[0173]

[0174] in, For high-quality track subsets The minimum total cost of all paths in the middle.

[0175] The other steps and parameters are the same as in Specific Implementation Method 8.

[0176] Specific Implementation Method Ten: This implementation method is a further limitation of Specific Implementation Method Nine. The specific process of step five is as follows:

[0177] Step 51: Divide the entire inspection track before smoothing into curve segments. Specifically, take the 0th to 3rd path point on the entire inspection track as the 0th curve segment, take the 1st to 4th path point as the 1st curve segment, and so on, taking the (W-4)th to (W-1)th path point as the (W-4)th curve segment.

[0178] Where W represents the total number of discrete path points on the entire inspection track. It should be noted that the W path points include the UAV's starting point and all points in the inspection sequence list.

[0179] Step 52, regarding spline parameters ,definition The range of values ​​is ,in , and define , ;

[0180] Will The range of values ​​is divided into The intervals are defined as follows: [0,1) is the 0th interval, [1,2) is the 1st interval, and so on, with [W-4, W-3] being the (W-4)th interval. The number of intervals then equals the number of curve segments, which can be determined based on the spline parameters. The interval in which it is located determines the curve segment, i.e., the spline parameters. Within which interval range, spline parameters It corresponds to which curve segment;

[0181] For spline parameters For any given value, first determine The curve segment number corresponding to the point , It is an integer, and Then calculate the B-spline basis functions. :

[0182]

[0183] Among them, intermediate variables , , ;

[0184] Calculate the spline parameters again Corresponding spatial coordinates :

[0185]

[0186] in, This represents the first point on the entire inspection track before smoothing. Spatial coordinates of path points ;

[0187] Step 53: The trajectory formed by the spatial coordinates corresponding to each spline parameter value constitutes the smoothed inspection track.

[0188] The other steps and parameters are the same as in Specific Implementation Method Nine.

[0189] It should be noted that this implementation can select a point in the interval [0,1) at intervals of 0.1, i.e., 0, 0.1, 0.2 up to 0.9; and select a point in the interval [1,2) at intervals of 0.1, i.e., 1, 1.1, 1.2 up to 1.9, and so on, until the spline parameters are traversed. The various intervals.

[0190] The effects of the present invention are verified using the following embodiments:

[0191] A 15km long, high-altitude mountain road section with continuous sharp bends, steep slopes, and frequent canyon winds was selected as the experimental environment. An industrial-grade multi-rotor UAV equipped with GNSS, LiDAR, and meteorological sensors was used, employing commonly used... The algorithm is used as a comparison scheme. The algorithm uses a static terrain obstacle map, and the optimization objective is to find the shortest path. Comparative experimental results are shown in Table 1. The algorithm had 3 safety violations and an average risk value of 0.72. The proposed solution had 0 safety violations and a reduced average risk value of 0.28. The average distance between the flight path and the risk zone increased from 8.2m to 15.6m, an improvement of 90.2%. This effectively solved the problem of insufficient dynamic risk perception in traditional methods and verified that the proposed solution can significantly reduce UAV inspection accidents and improve safety.

[0192] Table 1 Comparison of Experimental Results

[0193]

[0194] The above examples of the present invention are merely illustrative of the computational model and process of the present invention, and are not intended to limit the implementation of the present invention. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is impossible to exhaustively list all possible implementations here. Any obvious variations or modifications derived from the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for optimizing the flight path of unmanned aerial vehicle (UAV) inspections on mountainous highways, characterized in that, The method specifically includes the following steps: Step 1: Based on the planar coordinates of all points to be inspected, determine the inspection sequence list for each point by the UAV; Step 2: Initialize the inspection point numbers in the list. ; Step 3: Plan from the list of items... The first inspection point to the first The route to each inspection point; The specific process of step three is as follows: Step 3.1 Set the total number of iterations to 1. The initial value for initializing the association information between adjacent path points is... ; Determine the drone from the first The first inspection point to the first All the path points that the inspection point is allowed to pass through, will be the first The first inspection point to the first All allowed path points between inspection points and the first inspection point The set of points consisting of inspection points is denoted as . ; Step 3.2: Initialize the number of iteration rounds ; Step 3: Initialize the selected path point count ; Steps three and four: Calculate the drone's performance in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of the nth path point is used to select the point corresponding to the highest probability calculated. 1 path point; The selected number The path point and the selected path point The path point and the selected path point Path points between path points from the set Remove from the middle to obtain the updated set. ; Step 35: Determine if the drone has been selected. One inspection point; If the drone has already selected the first The number of inspection points indicates that all path points selected during the k-th iteration form the drone's path in the k-th iteration. The first inspection point to the first Inspection flight paths between inspection points Continue with step three six. If the drone did not select the first Each inspection point then ordered Return to steps three and four; Step 36: Calculate the inspection flight path Corresponding total cost of the trajectory And update the inspection tracks based on the total track cost corresponding to each inspection track in the window. The associated information for each road segment on the track, and the total cost of the trajectory includes the cost of terrain risk. No. Terrain risks at each waypoint The calculation method is as follows: Step 1, calculate the first... Elevation risk at each path point : In the formula: For drones in the Flight altitude at each waypoint; For the first Ground elevation at the projected coordinates of each path point on the ground; For safety height threshold; The base of the natural logarithm; This represents the elevation risk attenuation coefficient; Step 2, calculate the first Slope risk at each path point : In the formula: This refers to the slope risk coefficient. The norm operator for vectors; Step 3, calculate the first... Risk of terrain undulation at each waypoint : In the formula: Indicates the first Taking the projection of each path point on the ground as the center of a circle, with... The number of elevation sampling points within a local area with a radius of 1; For the local area of ​​the first Ground elevation of each sampling point; The safety factor is the terrain relief. The maximum elevation difference threshold; Step 4: Based on elevation risk Slope risk and terrain relief risk Calculate terrain risk : In the formula: This represents the weighting coefficient corresponding to elevation risk. This represents the weighting coefficient corresponding to slope risk. This represents the weighting coefficient corresponding to the risk of terrain relief. And determine the number of iteration rounds. Is it equal to ; like Then let Based on the updated association information, return to step three. like Then proceed to step three seven; Step 37: Select the path with the minimum total path cost from the K paths, and use the selected path as the UAV's path from the Kth path. The first inspection point to the first The inspection flight path of each inspection point; Step 38: Determine if the inspection point number has reached its maximum. If the inspection point number reaches the maximum, proceed to step five; If the inspection point number has not reached the maximum, proceed to step four; Step 4, Order Return to step three; Step 5: Smooth the flight path of the UAV from the starting point to each inspection point in the inspection sequence list to obtain the smoothed inspection flight path.

2. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 1, characterized in that, The specific process of step one is as follows: Step 11: Denote the set of all points to be inspected as follows: Calculate the starting point and set of the drones respectively. The Euclidean distance of each point to be inspected is used to determine the first point to be inspected, and the point with the smallest Euclidean distance is selected as the first inspection point. And the first inspection point was set up at the assembly point. Delete, and get the updated set of points to be inspected. ; Steps 1 and 2: Initialize the number of iterations ; Step 1 and Step 3: Calculate the first three parts separately. Inspection points and assembly The Euclidean distance of each point to be inspected is used to determine the point to be inspected that has the smallest Euclidean distance. One inspection point; And the first Each inspection point from the assembly Delete, and get the updated set of points to be inspected. ; Step 14, Order Return to steps one and three until the set of inspection points is empty.

3. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 2, characterized in that, In steps three and four, the calculation of the drone in the selected... When there are path points, the set will be... Each point in the selection is the first point. The probability of each path point; specifically: In the formula, Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point , For the first In the round of iteration, starting from the first... From the path point to the first Local trajectory cost generated by each waypoint; For the first In the round of iteration, starting from the first... From the path point to the first The probability of path selection for each path point; Indicates the first After -1 iterations, the th From the path point to the first Information relating to each path point; Representing path points For set The point in the middle; Indicates the first In the first iteration, the... From the path point to the first Heuristic information for each path point; Influence factors on the correlation information between path points; The heuristic information influencing factor between path points.

4. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 3, characterized in that, The from the first From the path point to the first Local trajectory cost generated by each waypoint The calculation method is as follows: in, Indicates from the first From the path point to the first The risk and cost of each path point Indicates from the first From the path point to the first The time cost of each path point The weighting coefficient for risk cost. The weighting coefficient for time cost. Indicates the first Spatial risk field at each path point Indicates the first Spatial risk field at each path point Indicates the first Spatial coordinates of each path point Indicates the first Spatial coordinates of each path point This indicates the inspection speed of the drone.

5. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 4, characterized in that, The first Spatial risk field at each path point The calculation method is as follows: in, Indicates the first Terrain risks at each waypoint Indicates the first Wind field risk at each path point express The weighting coefficients, express The weighting coefficients.

6. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 5, characterized in that, The first Wind field risk at each path point The calculation method is as follows: Step (1) Calculate the first Wind speed risk at each path point : In the formula: The safe wind speed threshold; For shape parameters; For the first Wind speed at each path point; Step (2), calculate the first Wind shear risk at each path point : In the formula: The weighting coefficient for space wind shear risk; The weighting coefficient for time-varying wind shear risk; for In a Cartesian coordinate system, the X-axis component... for In a spatial rectangular coordinate system, the Y-axis component, for The Z-axis component in a spatial rectangular coordinate system; The Frobenius norm operator for matrices; Step (3), calculate the first Wind field risk at each path point : In the formula: This is the weighting coefficient for wind speed risk; The weighting coefficient for wind shear risk.

7. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 6, characterized in that, The inspection flight path Corresponding total cost of the trajectory The calculation method is as follows: in, Indicates the inspection track The corresponding risk and cost, Indicates the inspection track The corresponding time cost, Indicates the flight path The total number of path points on the path.

8. The method for optimizing the flight path of a UAV inspection of mountain roads according to claim 7, characterized in that, The inspection track is updated based on the total track cost corresponding to each inspection track within the window. The associated information for each road segment on the road; the specific process is as follows: In the formula: Q is the correlation information attenuation coefficient; Q is the correlation information constant. For the set of tracks within the window, when hour, ,when hour, , The preset window length; For a high-quality subset of tracks, i.e. from The M paths with the minimum total path cost are selected; g is... A flight path in the middle; The total trajectory cost of trajectory g; The weight of the high-quality trajectory g; For path segment indication functions, when track g contains the first segment... From the path point to the first When there are path segments between path points, The value is 1, otherwise The value of is 0; in, For high-quality track subsets The minimum total cost of all paths in the middle.

9. A method for optimizing the flight path of a UAV inspection of mountain roads according to claim 8, characterized in that, The specific process of step five is as follows: Step 51: Divide the entire inspection track before smoothing into curve segments. Specifically, take the 0th to 3rd path point on the entire inspection track as the 0th curve segment, take the 1st to 4th path point as the 1st curve segment, and so on, taking the (W-4)th to (W-1)th path point as the (W-4)th curve segment. Step 52, regarding spline parameters ,definition The range of values ​​is ,in , and define , ; For spline parameters For any given value, first determine The curve segment number corresponding to the point , It is an integer, and Then calculate the B-spline basis functions. : Among them, intermediate variables , , ; Calculate the spline parameters again Corresponding spatial coordinates : in, This represents the first point on the entire inspection track before smoothing. Spatial coordinates of path points ; Step 53: The trajectory formed by the spatial coordinates corresponding to each spline parameter value constitutes the smoothed inspection track.

Images