Real-time flight route replanning method for grid-based low-altitude patrol

Abstract

The present application relates to unmanned aerial vehicle low-altitude patrol and flight path planning technical field, specifically to the flight route real-time replanning method for grid low-altitude patrol, in the present application, the blocked grid connected block is determined according to the grid traffic state and the original flight route, then the cut-out boundary and the cut-in boundary are formed according to the entering and leaving relationship of the original flight route and the blocked grid connected block, and the minimum connected bypass grid band is obtained by layer-by-layer expansion;Then, the reachable grid sub-band and the bypass grid flight path sequence are formed by combining the replanning trigger heading and the minimum turning radius propagation into the heading range and out of the heading range, so as to replace the blocked flight section with the bypass grid flight path sequence, so that the possibility of the grid conversion unable to continuously connect out and re-connect into the original flight route due to the heading and turning constraints is reduced while bypassing the impassable grid, and the real-time replanning flight route is formed.

Description

Technical Field

[0001] This invention relates to the field of low-altitude inspection and trajectory planning technology for unmanned aerial vehicles, specifically a real-time flight path replanning method for gridded low-altitude inspection. Background Technology

[0002] Low-altitude air patrols typically require aircraft to continuously patrol a target area along a pre-planned flight path. To facilitate the description of flyable areas, restricted areas, and route traversal locations within the patrol airspace, the airspace can be divided into several adjacent grids, and the accessibility status of each grid can be maintained based on changes in the airspace environment. The aircraft sequentially traverses the corresponding grids along the planned route to perform the patrol task. If the accessibility status of a grid changes during flight, affecting the subsequent portion of the original flight path, the unexecuted flight path needs to be replanned.

[0003] Existing grid-based flight route replanning schemes typically re-search for affected subsequent flight segments based on updated grid traffic information. Alternative grid paths that bypass obstructed areas are determined based on grid adjacency relationships, path length, and traffic costs. For aircraft motion characteristics, after obtaining alternative grid paths, trajectory smoothing, curvature adjustment, or flight feasibility checks can be performed at turning points along the path. The adjusted alternative segments are then reconnected to the original flight route to continue subsequent inspection tasks.

[0004] However, while the adjacent grids in the alternative path obtained based on grid adjacency are spatially connected, this does not mean that the aircraft can continuously pass through the corresponding grid boundaries from its current flight heading. When the conversion direction of adjacent grids in the alternative path changes significantly, the aircraft may be unable to complete the required heading adjustment within the corresponding grid range due to its turning capability limitations. Furthermore, both ends of the alternative path need to connect with the original flight routes before and after the obstructed segment. Therefore, when determining the grid path first and then adjusting the trajectory, it is possible to obtain a grid path that bypasses the obstructed area, but some grid conversions in this path cannot form a continuous trajectory that satisfies flight motion constraints. This necessitates reselecting the bypass grid or adjusting the access position, thereby increasing path backtracking and repeated searches during the replanning process. Summary of the Invention

[0005] The purpose of this invention is to provide a real-time flight path replanning method for gridded low-altitude inspection, in order to solve the problems mentioned in the background art. Specifically, it addresses the problem that when the original flight path is partially blocked, although the detour path formed according to the grid adjacency relationship can bypass the blocked area, some grid transformations cannot form a flight track that can continuously connect to and rejoin the original route due to the limitations of the aircraft's current heading and minimum turning radius.

[0006] To achieve the above objectives, the present invention aims to provide a real-time flight path replanning method for gridded low-altitude inspection, specifically including the following method steps:

[0007] S1. Obtain the grid accessibility status, original flight path, replanning trigger position, replanning trigger heading, and minimum turning radius of the low-altitude inspection space grid; form the remaining inspection grid sequence from the grid where the replanning trigger position is located along the original flight path to the endpoint; according to the grid accessibility status, connect the first impassable grid in the remaining inspection grid sequence along the grid complete shared edge to form a blocking grid connection block.

[0008] The formation process of the remaining inspection grid sequence specifically includes:

[0009] Each time a replanning is triggered, the grid access status and the original flight route under the same grid state version are read. At the same time, the grid state version is recorded as the initial grid state version. The set of grids that have been visited and the first unfinished grid are determined from the flight control record, so that subsequent processing only applies to the unfinished part of the original flight route.

[0010] Based on the replanning trigger location, determine the replanning starting grid corresponding to the replanning trigger location in the incomplete part of the original flight path;

[0011] Find the first occurrence position of the replanning starting grid in the incomplete part of the original flight route. Starting from that position, retain all grids to the end point according to the existing access order of the original flight route. For the same grid that is repeatedly repeated in the original flight route, only retain the first occurrence record to form the remaining inspection grid sequence. At the same time, retain the original sequence number of each grid in the original flight route and the corresponding heading between the grids in the original flight route.

[0012] The process of forming a blocked mesh connectivity component specifically includes:

[0013] Check the grid accessibility status in the order of accessing the remaining inspection grid sequence, and determine the first impassable grid as the first impassable grid;

[0014] Using the first impassable grid as the connection starting point, read the grids that are adjacent to the currently processed grid through grids that share a complete edge. Add the impassable adjacent grids that have not yet entered the processing set to the processing set, and repeat the connection processing on the newly added grids until the processing set is empty.

[0015] A blocking grid connectivity block will be formed by sharing edge chains between an impassable grid consisting only of impassable grids and the first impassable grid, and connectivity processing will be carried out only within the effective range of the low-altitude inspection space grid.

[0016] Step S1 retains the remaining inspection grid sequence that actually needs to be processed from the incomplete part of the original flight route corresponding to the replanning trigger position. Starting from the first impassable grid, it determines the blocked grid connectivity blocks that are interconnected through fully shared edges. Simultaneously, it retains the original sequence number of each grid in the original flight route and the corresponding heading between grids in the original flight route. This distinguishes the continuous impassable range corresponding to the local obstruction of the original flight route from the already visited route portion, providing a basis for determining the cut-out and cut-in boundaries based on the original flight route and for propagating the replanning trigger heading to the detour area, corresponding to the current flight state, in terms of processing objects and route sequence.

[0017] S2. The cut-out boundary is formed by the complete shared edge of the grid that the original flight path crosses when it first enters the blocked grid connected block, and the cut-in boundary is formed by the complete shared edge of the grid that the original flight path crosses when it leaves the blocked grid connected block for the last time after entering. The passable grid is expanded layer by layer from the periphery of the blocked grid connected block until the expanded grid that contacts the cut-out boundary and the cut-in boundary is first connected by the complete shared edge of the grid, forming the minimum connected bypass grid zone.

[0018] The process of forming the cut boundary specifically includes:

[0019] Following the access order of the remaining inspection grid sequence, inspect adjacent grid pairs from the first grid in the remaining inspection grid sequence along the original flight path to the endpoint;

[0020] When the current grid is outside the blocked grid connected component and the next grid is inside the blocked grid connected component, the adjacent grid pair is determined as a transition into the blocked grid connected component;

[0021] Following the original flight path's access order, the cut-out boundary is formed by the complete shared edge between adjacent grids when the transformation first occurs.

[0022] The formation process of the cutting boundary specifically includes:

[0023] Continue checking the remaining inspection grid sequence from the original flight path position where the cut-out boundary is formed to the endpoint, and record the original sequence number of the grid that belongs to the blocked grid connectivity block in the original flight path;

[0024] The grid corresponding to the maximum value in the original recorded sequence number is determined as the last occupied grid;

[0025] When there is an original flight path grid located outside the blocked grid connected component after the last occupied grid and adjacent to the last occupied grid through a grid-shared edge, the grid-shared edge between the two grids forms the cutting boundary.

[0026] The formation process of the minimum connected bypass mesh band specifically includes:

[0027] Using the boundary mesh fully shared edge of the blocked mesh connected block as the starting edge for expansion, the meshes that are passable but not yet fully visited are expanded layer by layer to the periphery of the blocked mesh connected block according to the adjacency relationship of the fully shared edges, forming an expansion layer, and merging the meshes in the already formed expansion layer to form a cumulative expansion set;

[0028] In the cumulative extended set, determine the meshes that completely coincide with the cut-out boundary and the cut-in boundary on the outside of the blocked mesh connected component, and start from the mesh that completely coincides with the cut-out boundary, propagate the connectivity marker only along the mesh's complete shared edge;

[0029] When a connectivity marker first reaches a mesh that completely coincides with the cut-in boundary, the connectivity components are divided according to the mesh's complete shared edges. The connectivity components that simultaneously contain meshes that completely coincide with both the cut-out boundary and the cut-in boundary are retained, and the retained connectivity components form the minimum connected bypass mesh band.

[0030] Step S2 utilizes the shared edges of the grids corresponding to the first entry and last exit of the blocked grid block from the original flight path to define the cut-out and cut-in boundaries that require local replacement. It then expands the accessible but unvisited grids layer by layer from the periphery of the blocked grid block until the expanded grids outside the cut-out and cut-in boundaries form a shared edge connection for the first time. The resulting minimum connected detour grid band initially limits the detour search to the connected grids surrounding the blocked grid block that can connect the cut-out and cut-in boundaries. This ensures that when subsequently constraining the replanning trigger heading and minimum turning radius, it is unnecessary to propagate the heading range in accessible grids unrelated to the current locally blocked detour, and it provides a limited range for determining whether a truly continuous grid transition can be formed between connected grids.

[0031] S3. Based on the replanning trigger heading and minimum turning radius, propagate into and out of the heading range along the complete shared edge of the minimum connected bypass grid zone. The intersection of the out-of-heading range of the previous grid and the in-heading range of the next grid is used as the continuation condition to form a heading reachable grid sub-zone. Cut-out and cut-in grid pairs are formed by the cut-out grid and cut-in grid that are in contact with the cut-out boundary and cut-in boundary, respectively. The cut-out and cut-in grid pairs are connected along the heading reachable grid sub-zone to form a bypass grid track sequence.

[0032] The formation process of the reachable grid sub-band specifically includes:

[0033] For the minimum connected bypass grid zone, the grids adjacent to each other through the grid complete shared edge are deducted from the two endpoints of the grid complete shared edge along the edge, and the resulting intermediate line segment forms the effective shared edge segment. The length of the effective shared edge segment forms the grid effective span.

[0034] The angle between the direction from the grid center to the endpoint of the effective shared edge and the normal azimuth of the complete shared edge of the grid forms the geometric span half angle, and the geometric span heading range is formed with the normal azimuth of the complete shared edge of the grid as the center according to the geometric span half angle; the maximum heading change of a single grid is determined according to the minimum turning radius and the effective span of the grid;

[0035] The replanning trigger heading is centered on the heading measurement tolerance and extended to both sides to form the replanning trigger heading range. The grids that completely coincide with the cut-out boundary outside the blocked grid connected block are formed into cut-out side candidate grids. The replanning trigger heading range is propagated to the cut-out side candidate grids along the access order of the unblocked grids of the original flight route, using the geometric cross-edge heading range of the original flight route grid between adjacent grids and the corresponding grids' completely shared edges, to form the initial cut-out heading range.

[0036] Starting from the candidate grid on the cut-out side, the entry and exit heading ranges are propagated only along the grid-wide shared edges within the minimum connected bypass grid zone. The intersection of the exit heading range already formed by the previous grid and the geometric cross-edge heading range allowed by the grid-wide shared edges between adjacent grids forms the entry heading range of the next grid. Non-empty entry heading ranges are expanded according to the maximum heading change of a single grid to form the exit heading range.

[0037] Starting from the candidate grid on the cut-out side, a directional reachable grid sub-band will be formed by continuously passing through non-empty grids within the entry heading range.

[0038] The process of forming cut-in mesh pairs specifically includes:

[0039] In the reachable sub-band of the heading, find the grid that completely coincides with the cut-out boundary outside the connected block of the blocking grid, and form a candidate set of cut-out grids; find the grid that completely coincides with the cut-in boundary outside the connected block of the blocking grid and can be continuously connected to the heading of the original flight path grid after the grid, and form a candidate set of cut-in grids.

[0040] For the candidate cutout mesh, search for candidate mesh chains along the mesh complete shared edges that meet the continuation conditions to the candidate cutin mesh. Use the number of mesh complete shared edge transformations of the candidate mesh chain as the first sorting factor, and prioritize the combination of candidate cutout mesh and candidate cutin mesh with fewer mesh complete shared edge transformations.

[0041] When the number of complete shared edges in a grid is the same, the intersection of the departure heading range formed by the current grid and the geometric cross-edge heading range corresponding to the next complete shared edge is determined as the departure heading range corresponding to the next complete shared edge. Based on the already formed entry heading range and the departure heading range, the minimum achievable circular heading change is determined. The minimum circular heading change of each grid in the candidate grid chain is added together to form the cumulative minimum necessary heading change. The combination with the smaller cumulative minimum necessary heading change is selected.

[0042] When the number of complete shared edge transformations and the cumulative minimum necessary heading change are the same, a unique combination is determined sequentially based on the original sequence number of the candidate cutting out grid in the original flight path, the original sequence number of the candidate cutting in grid in the original flight path, and the lexicographical order of the grid numbers. The cutting out grid and the cutting in grid in the unique combination form a cutting out and cutting in grid pair.

[0043] The formation process of the circumnavigation grid track sequence specifically includes:

[0044] Candidate mesh chains are processed sequentially according to their order. Starting from the cut-out mesh, the complete shared edges between adjacent meshes are listed sequentially. The mesh boundary safety distance is deducted from the two endpoints of the complete shared edges to obtain the effective shared edge segments. Candidate cross-edge points are generated on the effective shared edge segments.

[0045] For a complete shared edge of a grid that meets the continuation condition, the intersection of the departure heading range of the previous grid and the geometric cross-edge heading range of the complete shared edge of the grid is used to form the cross-edge heading range, and candidate cross-edge headings are formed based on the normal azimuth of the complete shared edge of the grid and the cross-edge heading range.

[0046] Adjacent grids share the same candidate cross-edge point and candidate cross-edge heading. Starting from the cut-out grid, propagate backward according to the candidate cross-edge point and candidate cross-edge heading. Within each grid, the entrance position and entrance heading and the exit position and exit heading are used as a set of directed endpoints. Enumerate the left arc straight left arc, right arc straight right arc, left arc straight right arc, right arc straight left arc, left arc right arc left arc, and right arc left arc right arc connection configurations. The radius of each arc is taken as the minimum turning radius, and the arc segment and the adjacent straight segment or arc segment are kept in position and tangential continuity at the connection.

[0047] Check whether all points of each connection configuration are located inside the corresponding grid, so that the track, except for the valid shared edge segment as the entrance and the valid shared edge segment as the exit, does not intersect the grid boundary, and the minimum distance of the track to the grid boundary and grid vertex that are not selected as the entrance or exit is not less than the grid boundary safety distance; retain the candidate states that can be propagated backward.

[0048] When a candidate state reaches the position of the original flight path after entering the grid, the preceding relationship of the candidate state is traced back to form a continuous track. The continuous track is associated with the exit grid, intermediate bypass grid, entry grid, cross-edge point, cross-edge heading, and the circular arc segment and straight line segment within the grid, forming a bypass grid track sequence.

[0049] Step S3 propagates the replanning-triggered heading along the unobstructed grid of the original flight path to the candidate grid on the cut-out side. Within the minimum connected bypass grid zone, it combines the geometric cross-edge heading range and the maximum heading change of a single grid determined by the minimum turning radius and the effective grid span, continuously propagating into and out of the heading range. When the out of the heading range of the current grid cannot form a non-empty intersection with the allowed geometric cross-edge heading range between adjacent grids, the corresponding grid transformation will not continue into the heading-reachable grid sub-zone. This further eliminates grid transformations that cannot be continuously completed due to the restrictions of the replanning-triggered heading and the minimum turning radius, based on the complete shared edge topological connectivity of the grid. Subsequently, through the continuous construction and checking of the cut-out and cut-in grid pairs, candidate grid chains, candidate cross-edge points and candidate cross-edge headings, as well as the circular arc segments and straight line segments within the grid, the remaining grid transformations are further mapped to continuous tracks that satisfy the constraints of positional continuity, tangential continuity, and minimum turning radius. This facilitates the formation of a bypass grid track sequence that can cut out from the original flight path and reconnect to the original flight path after bypassing the blocking grid connectivity block.

[0050] S4. Replace the segments in the original flight path located between the cut-out and cut-in grid pairs with the bypass grid track sequence to form a real-time replanned flight path.

[0051] The process of real-time flight route replanning specifically includes:

[0052] The replacement interval is determined based on the original sequence numbers of the cut-out grid and the cut-in grid in the original flight route. The route prefix before the cut-out grid in the original flight route that has not yet been executed is retained. The grids and corresponding segments of the original flight route that are blocked from after the cut-out grid to before the cut-in grid are deleted. The bypass grid track sequence is connected between the cut-out grid and the cut-in grid. The flight continues to fly along the suffix of the original flight route from the cut-in grid to complete the local replacement.

[0053] After completing the local replacement, starting from the replanning trigger point, check all grids involved in the unexecuted route after the replacement; when there are still impassable grids in the unexecuted route, take the route that has completed the current local replacement as the next round of pending routes, and take the first impassable grid that appears as the new first impassable grid. Repeat the process of blocking the formation of grid connectivity, cutting out boundary and cutting in boundary, forming minimum connected bypass grid band, propagating into and out of the heading range, and local replacement, until there are no impassable grids in the unexecuted route or the replanning calculation time limit is reached;

[0054] Before submitting the route, the grid status version is reread to check that the reread grid status version is consistent with the initial grid status version. It is also checked that all grids in the unexecuted route are passable, and that adjacent segments meet the requirements of position continuity, heading continuity, and minimum turning radius. After the checks are passed, the original flight routes executed before the replanning trigger position and the unexecuted routes obtained after the replanning trigger position through local replacement are combined in the execution order to form the real-time replanned flight routes.

[0055] Step S4 uses the original sequence numbers of the cut-out and cut-in grids in the original flight path to replace the corresponding blocked segments with the already formed bypass grid track sequence. Simultaneously, it retains the route prefix before the cut-out grid and the original flight path suffix after the cut-in grid, ensuring that the bypass track that meets the replanning trigger heading and minimum turning radius constraints forms a partial continuation with the original flight path. For inaccessible grids that persist in the unexecuted route after partial replacement, subsequent locally blocked segments can be processed by continuing to execute the formation of blocking grid connectivity blocks, cut-out and cut-in boundaries, minimum connected bypass grid bands, propagation into and out of heading ranges, and local replacement in the order of appearance. Combined with the checks on grid status version, grid accessibility, position continuity, heading continuity, and minimum turning radius before submitting the route, the final real-time replanning flight path bypasses inaccessible grids while reducing the possibility of grid paths being connected but some grid transitions failing to execute continuously according to the replanning trigger heading and minimum turning radius.

[0056] Compared with the prior art, the beneficial effects of the present invention are:

[0057] 1. Based on further analysis and research of existing technical problems, this invention recognizes that when the original flight path is partially blocked, and the detour path is determined solely based on the adjacency relationship of the complete shared edges of the grid, the spatial connectivity between grids does not indicate that the aircraft can continuously complete the conversion between adjacent grids according to the replanning trigger heading and minimum turning radius. Therefore, it is possible that the detour path has avoided the impassable grids, but some grid conversions cannot form a flight track that continuously cuts out from the original flight path and continuously rejoins the original flight path after the detour.

[0058] Based on this, the present invention forms a blocking grid connection block based on the first impassable grid in the remaining inspection grid sequence, and determines the cut-out boundary and cut-in boundary based on the first entry position and the last exit position of the original flight path relative to the blocking grid connection block. Then, the passable grid is expanded layer by layer from the periphery of the blocking grid connection block until the expanded grids that are in contact with the cut-out boundary and the cut-in boundary respectively are connected by the grid sharing edge for the first time, forming a minimum connected bypass grid zone.

[0059] On the other hand, based on the replanning trigger heading and minimum turning radius, the heading range is propagated into and out of the heading range along the grid complete shared edge of the minimum connected bypass grid zone. The intersection of the out-heading range of the previous grid and the in-heading range of the next grid is used as the continuation condition to form a heading reachable grid sub-zone. Then, the cutting-out and cutting-in grid pairs are connected along the heading reachable grid sub-zone to form a bypass grid track sequence.

[0060] Therefore, the detour range can be limited to a locally connected region that can bypass the blocking grid connected blocks and connect the cut-out boundary and the cut-in boundary. Then, the replanning trigger heading and minimum turning radius are used to perform continuous reachability screening of the grid transformations in the locally connected region. This prevents grid transformations that only have grid adjacency relationships but cannot meet the continuous turning requirements from continuing to participate in the formation of the detour track. This reduces the possibility of detour results where the grid path can bypass the blocked area but cannot actually fly continuously. It also makes the formed detour grid track sequence more conducive to continuously cutting out from the original flight route, bypassing the blocking grid connected blocks, and reconnecting to the original flight route continuously, thus improving the actual executability of real-time replanning flight routes.

[0061] 2. Furthermore, based on the formation of reachable grid sub-strips and cut-out and cut-in grid pairs, the present invention generates candidate cross-edge points on the effective shared edge segments between adjacent grids and forms candidate cross-edge directions according to the cross-edge direction range, so that adjacent grids can share the same candidate cross-edge points and candidate cross-edge directions.

[0062] Meanwhile, within each grid, the entrance position and entrance heading, and the exit position and exit heading are used as a set of directed endpoints. The configurations of left arc straight left arc, right arc straight right arc, left arc straight right arc, right arc straight left arc, left arc right arc left arc, and right arc left arc right arc connection are enumerated. The candidate states are checked and propagated backward by combining the minimum turning radius, position continuity, tangential continuity, grid boundary, and grid boundary safety distance.

[0063] Therefore, the inter-grid reachability relationship represented by entering and leaving the heading range can be further realized as a continuous track within the grid, so that adjacent grids are consistent in the cross-edge position and cross-edge heading, and candidate states that only meet the continuation conditions at the heading range level but are difficult to form actual tracks that meet the minimum turning radius and boundary constraints within the corresponding grid can be further screened out, thereby further improving the consistency between the orbiting grid track sequence and the actual motion constraints of the aircraft. Attached Figure Description

[0064] Figure 1 This is a schematic diagram of the overall method steps of the present invention;

[0065] Figure 2This is a schematic diagram of the core process of the overall method of the present invention;

[0066] Figure 3 This is a schematic diagram of the core process of forming the remaining inspection grid sequence and the blocking grid connectivity block in step S1 of the present invention;

[0067] Figure 4 This is a schematic diagram of the core process of forming the cut-out boundary, cut-in boundary, and minimum connected bypass mesh zone in step S2 of the present invention;

[0068] Figure 5 This is a schematic diagram of the core process of forming the heading reachable grid sub-strip, cutting out and cutting in grid pairs, and the bypass grid track sequence in step S3 of the present invention;

[0069] Figure 6 This is a schematic diagram of the core process of real-time flight route replanning in step S4 of the present invention. Detailed Implementation

[0070] The technical solutions in 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 embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] Next, please refer to Figure 1-2 The purpose of this embodiment is to provide a real-time flight route replanning method for gridded low-altitude inspections, targeting inspection tasks that have completed the low-altitude inspection space grid division and formed the original flight route. The low-altitude inspection space grid is represented by a square grid on the ground projection plane, with each grid corresponding to a low-altitude space cylinder. The grid access status is used to characterize whether the corresponding low-altitude space cylinder allows the UAV to enter at the time of this replanning. The original flight route consists of the grid sequence arranged according to the order of access and the inter-grid headings of the original flight route between adjacent grids. The replanning trigger position is represented by the coordinates of the UAV reference point in a unified plane coordinate system.

[0072] This embodiment sets up a unified Cartesian coordinate system. The horizontal and vertical directions of the low-altitude inspection space grid are parallel to... shaft and Axis; grid side length is denoted as The safety distance of the grid boundary is denoted as The position determination tolerance is denoted as The heading measurement tolerance is denoted as The grid state version is denoted as The replanning trigger course is recorded as The minimum permissible planar turning radius of the unmanned aerial vehicle under the load and speed conditions of this mission is denoted as . ; The grid number is recorded as ,in Indicates a horizontal index. Indicates a vertical index; exist The axial direction corresponds to from to The coordinate range, in The axial direction corresponds to from to The coordinate range; the same symbol in this embodiment always represents the same technical quantity or the same technical object.

[0073] In this embodiment, the grid side length Take 20 meters as the safety distance at the grid boundary. Take 2 meters, and determine the position tolerance. Take 1.5 meters, heading measurement tolerance Take 8 degrees; the grid boundary safety distance is less than half the grid side length, so that each grid fully shared edge retains a non-empty valid shared edge segment after deducting the grid boundary safety distance at both ends.

[0074] The real-time flight path replanning method for gridded low-altitude inspection includes steps S1 to S4, wherein:

[0075] S1. Obtain the grid accessibility status, original flight path, replanning trigger position, replanning trigger heading, and minimum turning radius of the low-altitude inspection space grid; form the remaining inspection grid sequence from the grid where the replanning trigger position is located along the original flight path to the endpoint; according to the grid accessibility status, connect the first impassable grid along the complete shared edge of the grid to form a blocking grid connection block;

[0076] S2. The cut-out boundary is formed by the complete shared edge of the grid that the original flight path crosses when it first enters the blocked grid connected block, and the cut-in boundary is formed by the complete shared edge of the grid that the original flight path crosses when it leaves the same blocked grid connected block for the last time after entering. The passable grid is expanded layer by layer from the periphery of the blocked grid connected block until the expanded grid that contacts the cut-out boundary and the cut-in boundary is first connected by the complete shared edge of the grid, forming the minimum connected bypass grid zone.

[0077] S3. Based on the replanning trigger heading and minimum turning radius, propagate into and out of the heading range along the complete shared edge of the minimum connected bypass grid zone. The intersection of the out-of-heading range of the previous grid and the in-heading range of the next grid is used as the continuation condition to form a heading reachable grid sub-zone. Cut-out and cut-in grid pairs are formed by the cut-out grid and cut-in grid that are in contact with the cut-out boundary and cut-in boundary respectively. The cut-out and cut-in grid pairs are connected along the heading reachable grid sub-zone to form a bypass grid track sequence.

[0078] S4. Replace the segments in the original flight path located between the cut-out and cut-in grid pairs with the bypass grid track sequence to form a real-time replanned flight path.

[0079] Specifically, the methods and steps are as follows:

[0080] Please see Figure 3 S1. Obtain the grid accessibility status, original flight path, replanning trigger position, replanning trigger heading, and minimum turning radius of the low-altitude inspection space grid; from the incomplete position of the original flight path corresponding to the replanning trigger position, follow the original flight path to the endpoint to form the remaining inspection grid sequence; according to the grid accessibility status, connect the first impassable grid in the remaining inspection grid sequence along the grid complete shared edge to form the blocking grid connection block, specifically including the formation process of the remaining inspection grid sequence and the formation process of the blocking grid connection block.

[0081] The formation process of the remaining inspection grid sequence specifically includes:

[0082] Read the same grid state version on each replanning trigger. The grid's current traffic status and original flight path are recorded, and this grid status version is also recorded as the initial grid status version; for any grid... The grid access status is recorded as Where 0 indicates passage is permitted and 1 indicates passage is prohibited; the validity period of the status is denoted as . When grid state is missing, or the difference between the grid state timestamp and the replanning trigger time exceeds [a certain threshold], [the situation will occur]. If the height range of the grid does not intersect with the height of this inspection, the grid will be treated as impassable in this replanning.

[0083] Read the original flight path and verify whether any two adjacent grids in the original flight path have a complete shared edge. If two adjacent grids only touch through a corner, the grid index crosses more than one grid, or the grid number does not exist, generate a route topology anomaly and terminate this replanning. At the same time, determine the set of grids that have been visited from the flight control record and record them as follows. At the same time, the first incomplete grid is identified so that subsequent processing only applies to the unexecuted portion of the original flight path;

[0084] When determining the grid where the replanning trigger point is located, first determine the distance between the replanning trigger point and the grid vertex; the distance to any grid vertex is no greater than... When comparing candidate meshes, only those containing the vertex and appearing in the incomplete portion of the original flight path are compared; if the vertex condition is not met, the distance to the mesh boundary is then considered; the distance to a mesh boundary is no greater than When the boundary is incomplete, only candidate grids that are both on either side of the boundary and appear simultaneously in the incomplete part of the original flight path are compared; otherwise, a single grid containing the replanning trigger location is used.

[0085] If there are more than two candidate grids, read the heading between the original flight path grids corresponding to each candidate grid, and calculate the minimum circular angle between the heading and the replanning trigger heading; prioritize the candidate grid with the smallest circular angle not greater than 90 degrees; if the angles are the same, select the candidate grid with the earlier sequence number in the incomplete part of the original flight path; terminate the replanning if no candidate grid meets the 90-degree condition; thus forming a unique replanning starting grid;

[0086] Find the first occurrence position of the replanning starting grid in the incomplete part of the original flight route, and retain all grids to the end point according to the existing access order of the original flight route from that position to form the remaining inspection grid sequence; only the first occurrence record of the same grid that is repeatedly repeated in the original flight route is retained; each grid in the remaining inspection grid sequence retains its original sequence number in the original flight route and the corresponding heading between the grids in the original flight route.

[0087] When the starting grid for replanning is the last grid of the original flight route, the remaining inspection grid sequence only includes that grid; if the grid is passable, no route replacement is required; if the grid is impassable, since there is no original flight route grid after the impassable grid, the process enters the safety hold process; if there are no incomplete grids, it means that the original flight route has been completed and replanning is not performed.

[0088] The formation process of the remaining inspection grid sequence yields an ordered grid sequence from the replanning starting grid to the end point of the original flight route, while retaining the original sequence number of each grid and the heading between the original flight route grids, providing a sequential basis for the formation of blocking grid connectivity blocks, cut-out boundaries, and cut-in boundaries.

[0089] The process of forming a blocked mesh connectivity component specifically includes:

[0090] Check the access order along the remaining inspection grid sequence The first one to be satisfied The grid in question is identified as the first impassable grid. If there are no impassable grids in the remaining inspection grid sequence, the unexecuted portion of the original flight route remains unchanged, and the replanning ends.

[0091] Using the first impassable grid as the starting point for connectivity, read its four adjacent grids (upper, lower, left, and right) that share a complete edge; adjacent grids satisfy... If a grid has not yet been added to the set to be processed, add it to the set. Repeat the above process for newly added grids until the set to be processed is empty. The resulting set of impassable grids is denoted as a blocking grid connectivity block. Each grid in a blocked grid connectivity block can be fully connected to the first impassable grid through a grid consisting only of impassable grids, sharing an edge chain.

[0092] Each grid is processed only once during the grid connectivity process; grid indices outside the effective range of the low-altitude inspection space grid are not included in connectivity; when a blocking grid connectivity block contacts the boundary of the inspection area, only the impassable grids within the inspection area are retained, and the space outside the inspection area is not included in the minimum connectivity bypass grid zone.

[0093] When the grid state version is detected during connectivity processing, When a change occurs, immediately discard the already formed blocked mesh connectivity block and reread the mesh access status of the same new version; if the version change still occurs after two consecutive rereads, proceed to safe hold process;

[0094] The formation process of the blocked grid connection block yields a set of impassable grids that are fully connected to the first impassable grid in the remaining inspection grid sequence through grid-shared edge chains. This provides processing objects for step S2 to determine the spatial boundaries of the original flight path entering and leaving the blocked grid connection block and to expand the traversable grids.

[0095] Step S1 obtains the remaining inspection grid sequence and the blocked grid connectivity block, and passes the replanning trigger position, replanning trigger heading, minimum turning radius, grid side length, grid boundary safety distance, grid status version, original sequence number and the heading between the original flight route grids to step S2.

[0096] Please see Figure 4 S2. The cut-out boundary is formed by the complete shared edge of the grid that the original flight path crosses when it first enters the blocked grid connected block, and the cut-in boundary is formed by the complete shared edge of the grid that the original flight path crosses when it leaves the same blocked grid connected block for the last time after entering. The passable grid is expanded layer by layer from the periphery of the blocked grid connected block until the expanded grid that contacts the cut-out boundary and the cut-in boundary is first connected by the complete shared edge of the grid, forming the minimum connected bypass grid zone. Specifically, this includes the formation process of the cut-out boundary, the formation process of the cut-in boundary, and the formation process of the minimum connected bypass grid zone.

[0097] The process of forming the cut boundary specifically includes:

[0098] Based on the original sequence number retained in the remaining inspection grid sequence, check adjacent grid pairs from the replanning starting grid along the original flight path to the endpoint; the current grid is located in a blocked grid connectivity block. Outside and the next grid is located Within this range, the adjacent mesh pair is identified as a transition to the blocked mesh connected component; when this transition occurs for the first time according to the original flight path access order, the complete shared edge between the two meshes forms a cut-out boundary, and the cut-out boundary is denoted as... ;

[0099] Cut out the boundary The coordinates of the two endpoints of the shared edge in the mesh are used, as well as the coordinates of the midpoint of the edge and the direction pointed from the outside of the blocked mesh connectivity block. The inner normal orientation is indicated; with lie in The outer grid, which belongs to the unfinished part of the original flight path, forms the candidate grid for the cut-out side;

[0100] If the original flight path has already entered the blocked grid connectivity block before the incomplete position corresponding to the replanning trigger position, it means that the cut-out position used for normal detour is already located in the executed segment. In this embodiment, the already executed grid is not used as the cut-out grid in reverse, but instead the safety hold process is entered.

[0101] The process of forming the cut-out boundary yields the unique complete shared edge of the grid corresponding to the first entry of the unexecuted part of the original flight path into the blocking grid block from the outside, providing a spatial reference for the contact conditions of the cut-out end of the minimum connected bypass grid band and the determination of the cut-out grid.

[0102] The formation process of the cutting boundary specifically includes:

[0103] From the formation of the cutting boundary Continue from the original flight path position towards the endpoint to check the remaining inspection grid sequence and record it. All subsequent blocks belonging to the blocking grid connectivity The original grid indices; the grid corresponding to the maximum value in these original indices is determined as the original flight path pair. The last grid cell occupied;

[0104] If after the grid is finally occupied, there is a location located in In addition to the original flight path grid that is adjacent to the last occupied grid through a fully shared edge, the fully shared edge between the two grids forms the cutting boundary, and the cutting boundary is denoted as... Determined by the last occupied grid. This caused the original flight path to enter for the first time. Even if you briefly leave and then re-enter the same blocked grid connected component, subsequent route replacements can cross the entire blocked grid connected component in one go.

[0105] The cut-in boundary is represented by the coordinates of the two endpoints of the fully shared edge of the mesh, the coordinates of the midpoint of the edge, and the normal orientation pointing from the inside to the outside of the blocking mesh connected block; if the last occupied mesh is the end point of the original flight path, or there is no mesh outside the blocking mesh connected block after it, then the state of no cut-in boundary is recorded and the process enters the safe hold process.

[0106] The process of forming the cutting boundary yields the unique complete shared edge of the grid corresponding to the last departure from the same blocking grid block after the original flight path first enters the blocking grid block, providing a spatial reference for the contact conditions of the cutting end of the minimum connected bypass grid zone and the determination of the cutting grid.

[0107] The formation process of the minimum connected bypass mesh band specifically includes:

[0108] To block mesh connectivity All boundary meshes share a common edge for the expansion; the expansion level is denoted as... From the grid To block the grid connectivity The four-adjacent distance is denoted by the full shared edge count of the mesh. The upper limit of the expansion level is denoted as ;No. The extended layer is denoted as Only includes grid access statuses that are passable, incomplete access, and have reached [the specified location]. The four-adjacent distance is equal to The grid; the first The formula for the extended layer is as follows:

[0109] , In the set construct, the colon indicates that each subsequent term is a mesh. Entering the The conditions that must be met simultaneously when extending the layer; this formula represents the first... The extended layer is defined by the grid being passable and the distance to the four adjacent blocks of the blocking grid being equal to... And it does not belong to the set of grids that have been visited. The grid is composed of layers, with the expansion levels increasing sequentially from 1 to... ;

[0110] When forming each extension layer, only the grid indexes within the effective range of the low-altitude inspection space grid are traversed; the same grid enters only one extension layer; each extension layer records the grid number, the extension layer level to which it belongs, and the complete shared edge adjacency relationship between the grids in the same layer and adjacent layers.

[0111] Extend the first layer to the second layer. The total number of meshes in the extended layer constitutes the cumulative extended set. ;exist In the middle, and the cut-out boundary exist The set of meshes that completely overlap on the outer side is denoted as , and the cutting boundary exist The set of meshes that completely overlap on the outer side is denoted as The formula for the cumulative extended set is as follows:

[0112] Among the symbols This represents the union operation of sets; the formula represents the cumulative expansion of sets. From the first extension layer To the Extension layer The result is obtained by merging all the meshes in the extended range, and is used to determine whether the cut-out end and the cut-in end are already connected within the extended range.

[0113] In the cumulative extended set From Starting from the grid in the middle, the connectivity marker is propagated only along the complete shared edges of the grid; the connectivity marker reaches... When any grid in the array is selected, determine the connectivity between the two ends. Set to 1 otherwise set to 0, where This represents the connectivity determination operation, returning 1 if the two ends are fully connected through the mesh in the given cumulative expansion set, and returning 0 if they are not connected; the smallest expansion level that first obtains a connectivity value of 1 is denoted as . , This indicates taking the minimum value among candidate values ​​that meet the given conditions, as will appear later in the text. Using this meaning, the formula for the first connection of the extended levels is expressed as follows:

[0114] This formula represents all possible... and In the cumulative extended set Extension levels of inner connections Take the minimum value and determine the resulting minimum level as the first connected expansion level. This limits the formation of subsequent minimum connected bypass mesh zones only within the extended range where the first two ends are connected.

[0115] If the expansion level reaches If a connectivity value of 1 cannot be obtained, record the state of no bypass connectivity space and enter the safety hold process; if a connectivity value of 1 is obtained... Then for Connected components are partitioned according to the complete shared edges of the mesh; those containing at least one edge are retained. Grid and at least one All connected components of the mesh are combined, and the union of these connected components is used to form a minimum connected bypass mesh band. ;

[0116] Minimum Connectivity Bypass Mesh Each grid record in the process belongs to its extended level, its four adjacent distances to the blocked grid connected components, and the adjacent grid numbers of the grids sharing a complete edge; step S3 is only allowed in Internal propagation into and out of the heading range;

[0117] The formation process of the minimum connected bypass mesh band obtains the traversable mesh space range that simultaneously connects the cut-out end and the cut-in end within the first extended connected level, directly defining the spatial boundary for the heading range propagation and bypass mesh chain search in step S3.

[0118] Step S2 obtains the cutout boundary, cut-in boundary, and minimum connected bypass grid zone, and transfers the replanning trigger position, replanning trigger heading, minimum turning radius, heading between original flight path grids, and grid boundary safety distance to step S3.

[0119] Please see Figure 5 S3. Based on the replanning trigger heading and minimum turning radius, propagate along the complete shared edge of the grid within the minimum connected bypass grid zone to enter and leave the heading range, forming a heading reachable grid sub-zone; within the heading reachable grid sub-zone, determine the cutting-out grid, the cutting-in grid, and the bypass grid chain connecting the two, and form a continuous bypass grid track sequence based on the effective shared edge segment, heading range, and minimum turning radius, specifically including the formation process of the heading reachable grid sub-zone, the formation process of the cutting-out and cutting-in grid pairs, and the formation process of the bypass grid track sequence.

[0120] The formation process of the reachable grid sub-band specifically includes:

[0121] Heading angles are uniformly converted to radians and in negative... to The range is represented in a ring shape, where Represents pi; for the least connected loop mesh zone Two adjacent grids that fully share edges through the grid. and ,in Indicates the propagation index of the previous grid. Indicates the propagation index of the next grid, from point to The complete shared edge normal orientation of the mesh is denoted as Subtract the grid boundary safety distance from each of the two endpoints of any fully shared edge in the grid. The resulting intermediate line segment is defined as the valid shared edge segment;

[0122] The length of the effective shared edge segment is defined as the effective span of the grid. The effective span of the grid only represents the length of the effectively shared edge segments, and no longer represents the length of other geometric lengths within the grid; the formula for the effective span of the grid is as follows:

[0123] This formula represents a length of... The mesh complete shared edge is deducted from both ends of the mesh boundary safety distance. After that, the length is obtained as The effective shared edge segment; the resulting effective span of the mesh is used to subsequently determine the geometric span half-angle and the maximum heading change of a single mesh;

[0124] The normal distance from the mesh center to the fully shared edge of the mesh is Effective shared edge segments extend from the midpoint of the complete shared edge of the mesh to both ends. The angle between the direction from the grid center to the endpoint of the effective shared edge and the normal of the complete shared edge of the grid forms the geometric half-angle of the edge, and this half-angle is denoted as . ; This represents the arctangent function, which returns the angle whose tangent value equals a given ratio; the formula for the geometric span half-angle is as follows:

[0125] This formula represents the determination of the geometric span half-angle using the ratio of the effective shared edge half-length to the normal distance from the mesh center to the complete shared edge of the mesh. Since the two are respectively and The ratio is simplified to The result Used to define the range of geometric headings when crossing a fully shared edge between adjacent grids;

[0126] Shared edge normal orientation with complete mesh Extend the geometric span half-angle to both sides from the center. The geometric cross-side heading range used for mesh chain search prediction is obtained, and this range is denoted as... The square brackets in the formula indicate a closed heading range including both endpoints; the formula for the geometric cross-side heading range is as follows:

[0127] This formula represents the change from the previous grid. Enter the next grid At that time, the azimuth of the fully shared edge is used. As the central heading, and with geometric spanning half-angle By defining the permissible deviation range, a geometric cross-edge heading range is formed for predicting shared edge continuation. ;

[0128] Based on minimum turning radius and grid effective span Determine the maximum heading change per grid cell used for heading propagation prediction, and denot it as... This quantity is only used to exclude continuation relationships that obviously cannot meet the turning capability during the grid chain search phase. The actual track is still verified again according to the actual position and minimum turning radius during the formation of the grid track sequence. This represents the arcsine function, which returns the angle whose sine value equals a given value; the formula for the maximum change in heading is as follows:

[0129] This formula represents the effective span of the grid. With diameter By comparing the values ​​and using the arcsine relation, the maximum heading change for single-grid propagation prediction under the minimum turning radius constraint is obtained. The smaller value between the ratio and 1 is used to ensure that the input of the arcsine function is not greater than 1.

[0130] Replanning triggers course The heading measurement tolerance is centered and extends to both sides. The replanning trigger heading range is formed and recorded as The formula for determining the course range triggered by replanning is as follows:

[0131] The formula represents a course replanning trigger. Centered on the yoke, heading measurement tolerances are added to both sides. This forms a replanning trigger heading range that includes the measurement error range. This range serves as the initial heading condition for subsequent propagation to the candidate mesh on the cut-out side.

[0132] If the starting grid for replanning is the same as the candidate grid for the cut-out side, then... Directly used as the initial heading range If the replanning starting grid is located before the candidate grid on the cut-out side, then the propagation will proceed according to the access order of the unobstructed grids on the original flight path, using the inter-grid headings of the original flight path between adjacent grids and the geometric cross-edge heading range of the corresponding grid's fully shared edges. Until the candidate grid on the cut-out side is reached, and the course range that can still be maintained when reaching that grid is formed. If the heading range is empty during propagation, then enter the safety hold process;

[0133] Both entering and leaving a heading range are allowed to consist of one or more non-overlapping closed heading ranges; for crossing negative heading ranges... and The closed heading range of the boundary is first divided into those located in the negative direction. to Two closed heading ranges on continuous coordinate axes are then intersected; closed heading ranges that overlap or have their endpoints connected are merged.

[0134] For the candidate mesh to be cut out, first cut out the initial heading range. According to the maximum change of course Expand to form the range of the departure heading of the grid; for a general propagation process, the previous grid is extended. The established departure heading range is denoted as , the next grid The range of entry headings allowed through the fully shared edge of the mesh between the two is denoted as ;symbol This represents the intersection of two heading ranges, that is, the set of headings that simultaneously belong to both heading ranges; the formula for entering a heading range is as follows:

[0135] This formula represents the next grid. Entering the course range Keep only the previous grid The departure course range that has already been formed The geometric cross-edge heading range allowed by the complete shared edge across both meshes. The common part; when the common part is empty, the two cannot form a continuous grid connection in the heading;

[0136] when When empty, it is not allowed to start from... Continue to ;when When not empty, for Each closed heading range expands to both sides. And perform ring splitting and overlapping merging to form leaving the course range ;in This indicates that for each closed heading range indicated by the first parameter, the range is extended to both sides by the angle indicated by the second parameter, and the range extension operation of ring splitting and overlapping merging is performed; the formula for leaving the heading range is as follows:

[0137] This formula represents the next grid. Entering the course range Based on this, the maximum heading change in a single grid Expanding to both sides within the permitted range, forming The departure heading range available for the next grid continuation This result continues to serve as the condition for the previous grid to leave the course when entering the course range for the next intersection.

[0138] The `expand` operation is performed separately for each closed heading range; when the sum of the original width of a closed heading range and the expansion amounts on both sides reaches or exceeds 2... At that time, directly set the range to the complete heading range; after expansion, it crosses the negative range. and At the boundary, it is split into two closed heading ranges; after all expansion results are completed, the overlapping ranges are merged again.

[0139] Starting from the candidate mesh on the cut-out side, repeat the above propagation, only circling the mesh band along the least connected side. The inner grid fully shares edges and expands outwards; the same grid can retain multiple non-overlapping inbound and outbound heading ranges; the maximum number of closed heading ranges retained by each grid is denoted as . When the quantity exceeds The width of the closed heading range is retained from largest to smallest. When the widths are the same, priority is given to retaining the range that can be fully reached through the transformation of shared edges using fewer meshes;

[0140] Starting from the candidate grid on the cut-out side, all grids that can be continuously reached through the non-airway entry heading range are formed into a heading reachable grid sub-band; when propagating to the grid outside the cut-in boundary, the exit heading range of the grid is also intersected with the heading tolerance range between the original flight path grids after the grid, and only grids with non-airway intersections are retained as candidate grids on the cut-in side.

[0141] The formation process of the reachable grid sub-band of the course obtains the set of grids that can be continuously propagated within the minimum connected bypass grid band under the constraints of replanning triggered course and minimum turning radius prediction. It also preserves the entry and exit ranges and previous grid relationships for the complete shared edges of each course-connectable grid, providing the course boundary for the subsequent determination of cut-out and cut-in grid pairs and the actual bypass track.

[0142] The process of forming cut-in mesh pairs specifically includes:

[0143] Find and cut out the boundary in the reachable grid sub-strip. For meshes that completely overlap outside the blocked mesh connectivity component, form a candidate set of cutout meshes; find and cut into the boundary. A set of candidate cutting grids is formed by grids that completely overlap outside the blocking grid and can be continuously connected in the heading between the original flight path grids following the grid.

[0144] For each candidate grid to be cut out, search for a grid chain along the grid complete shared edge that has been recorded as a continuous heading. The first sorting factor of the grid chain is the number of grid complete shared edge transformations. When the number of transformations is the same, at each grid, the intersection of the departure heading range formed by the current grid and the geometric cross-edge heading range corresponding to the next grid complete shared edge is determined as the departure heading range corresponding to the next grid complete shared edge. Then, based on the already formed entry heading range and the departure heading range, determine the minimum circumferential heading change that can be achieved between the two. Add the minimum circumferential heading changes of each grid in the entire grid chain to form the cumulative minimum necessary heading change.

[0145] Prioritize selecting candidate cutout grids and candidate in-grids with fewer complete shared edge transformations; if the transformation counts are the same, select the combination with the smaller cumulative minimum necessary heading change; if both are the same, select the combination with the smaller original cutout grid number, then select the combination with the smaller original in-grid number; if they are still the same, determine the unique combination according to the lexicographical order of the grid numbers.

[0146] The selected cutout grid and cutin grid form a cutout-cutin grid pair; both the cutout grid and the cutin grid are located outside the blocking grid connectivity block, and there is at least one directionally connectable grid chain between them that is completely located in the directionally reachable grid sub-band; if there is no combination that meets the conditions, then proceed to the safety holding process;

[0147] The process of forming the cut-out and cut-in grid pairs yields unique grid pairs corresponding to the cut-out and cut-in ends, and determines the grid chain comparison order without using specific representative headings in advance, providing a definite starting grid, ending grid, and candidate grid chain order for the formation of the circumnavigation grid track sequence.

[0148] The formation process of the circumnavigation grid track sequence specifically includes:

[0149] According to the candidate grid chains determined during the formation process of the cut-out and cut-in grids, the candidate grid chains with the fewest shared edge transformations and the smallest cumulative necessary heading changes are processed in turn. For each candidate grid chain, starting from the cut-out grid, the complete shared edges between adjacent grids are listed in turn, and the grid boundary safety distance is deducted from the two endpoints of each complete shared edge to obtain the effective shared edge segment corresponding to the complete shared edge of the grid.

[0150] For each valid shared edge segment, candidate cross-edge points are generated; the candidate cross-edge points are first taken as the midpoint of the valid shared edge segment, and then the position tolerance is not greater than the value of the cross-edge point. The spacing is calculated by taking points from the midpoint towards both ends, retaining the two endpoints of the valid shared edge segment; candidate cross-edge points are sorted in ascending order of distance from the midpoint, and if the distances are the same, they are sorted according to... Coordinates then according to Coordinate sorting; this yields a finite set of candidate cross-edge points with a defined order;

[0151] For any heading that can be connected to a complete shared edge of the grid, the intersection of the heading range of the previous grid leaving the heading range and the heading range of the complete shared edge of the grid is taken as the heading range of the cross-edge. For each closed heading range of the heading range of the cross-edge, if the normal azimuth of the complete shared edge of the grid is within the closed heading range of the heading, then the first candidate heading of the cross-edge is formed with the normal azimuth; if it is not within the closed heading range of the heading, then the first candidate heading of the cross-edge is formed with the endpoint of the closed heading range that is closer to the normal azimuth; the remaining endpoints are arranged in ascending order of the angle with the normal azimuth of the complete shared edge of the grid, forming a finite set of candidate headings of the cross-edge.

[0152] For a candidate grid chain, two adjacent grids share the same candidate cross-edge point and the same candidate cross-edge heading. Starting from the replanning trigger position of the cut-out grid or the previous original flight route continuation state, propagation proceeds backward according to the predetermined order of candidate cross-edge points and candidate cross-edge headings. Within each grid, the entrance position and entrance heading, and the exit position and exit heading are treated as a set of directed endpoints. Enumerate the left arc straight-left arc, right arc straight-right arc, left arc straight-right arc, right arc straight-left arc, left arc right arc left arc, and right arc left arc right arc connection configurations composed of left-turn arcs, right-turn arcs, and straight segments. The radius of each arc is taken as the minimum turning radius, and the arc segments and adjacent straight segments or arc segments maintain positional continuity and tangential continuity at the connection points.

[0153] For each connection configuration, check whether all its points are located inside the corresponding grid; except for the valid shared edge segments at the entrance and exit, the track must not intersect the grid boundary; the minimum distance from the track to the grid boundary that is not selected as an entrance or exit and to the grid vertex must not be less than the grid boundary safety distance; connection configurations that meet the above conditions are sorted from shortest to longest total track length, and when the total track lengths are the same, they are selected in the predetermined order of left arc straight left arc, right arc straight right arc, left arc straight right arc, right arc straight left arc, left arc right arc left arc, right arc left arc, right arc left arc right arc;

[0154] If an entry state can reach a candidate cross-edge state on the next grid's complete shared edge through at least one connection configuration that meets the conditions, then the cross-edge state, the corresponding preceding state, and the connecting segment within the grid are recorded and propagated forward; candidate states that cannot reach the next grid's complete shared edge are deleted; when propagating to the cutting grid, the same connection check is performed using the next grid's complete shared edge of the original flight path after the cutting grid and the inter-grid heading of the original flight path as the exit constraint;

[0155] When a candidate state first reaches the continuation position of the original flight path after the cut-in grid, its previous state is traced back according to the grid chain sorting, candidate cross-edge point sorting, candidate cross-edge heading sorting, and grid connection configuration sorting to form a unique continuous track. This continuous track is sequentially associated with the cut-out grid, intermediate bypass grid, cut-in grid, cross-edge points on each effective shared edge segment, cross-edge heading, and circular arc segments and straight line segments within the grid to form a bypass grid track sequence. If the current candidate grid chain does not have a state that can completely reach the cut-in end, the next candidate grid chain is processed. When all candidate grid chains fail, the safety hold process is entered.

[0156] The formation process of this bypass grid track sequence results in a bounded curvature track that continuously reaches the cutting grid from the cutting grid and reconnects to the original flight path. Each crossing point is located within an effective shared edge segment, each arc radius is not less than the minimum turning radius, and the track within the grid does not cross the grid boundary that has not been selected as an entrance or exit. Therefore, it can be directly used as the flight path object for replacing the blocked segment in step S4.

[0157] Step S3 obtains the heading reachable grid sub-strip, cut-out and cut-in grid pairs, bypass grid chains, and bypass grid track sequences, and passes the original sequence numbers of the cut-out and cut-in grids and the bypass grid track sequences to step S4.

[0158] Please see Figure 6 S4. Replace the obstructed segments between the cut-out and cut-in grids in the original flight route with the bypass grid track sequence, check the grid access status and grid status version of the route that was not executed after replacement, and form a real-time replanned flight route, specifically including the formation process of the real-time replanned flight route.

[0159] The process of real-time flight route replanning specifically includes:

[0160] The replacement interval is determined based on the original sequence numbers of the cut-out grid and the cut-in grid in the original flight route; the route prefix before the cut-out grid in the original flight route that has not yet been executed is retained, the obstructed grids and corresponding segments of the original flight route from after the cut-out grid to before the cut-in grid are deleted, the bypass grid track sequence is connected between the cut-out grid and the cut-in grid, and the flight continues along the suffix of the original flight route from the cut-in grid; the cut-out boundary and the cut-in boundary are only used to determine the two ends of the space corresponding to the minimum connected bypass grid band and the cut-out grid and the cut-in grid, and it is not required that the bypass grid track sequence actually crosses the cut-out boundary or the cut-in boundary;

[0161] After completing a local replacement, starting from the replanning trigger point, check all grids involved in the unexecuted route after the replacement; if there are still impassable grids, take the route that has completed the current local replacement as the next round of pending routes, and take the first impassable grid that appears as the new first impassable grid, and repeat the formation of blocking grid connectivity, cutting-out and cutting-in boundary formation, minimum connectivity bypass grid band formation, heading propagation and local replacement; repeat the process until there are no impassable grids in the unexecuted route or the replanning calculation time limit is reached;

[0162] This embodiment sets a time limit for replanning calculations. Read and record the initial grid state version Start timing; if during any round of processing, if the current mesh state version is found to be inconsistent with the initial mesh state version, discard all uncommitted local replacement results and re-execute steps S1 to S4 according to the new mesh state version, while using the new mesh state version as the new initial mesh state version; when the cumulative processing time reaches If no unexecuted route is found that does not contain an impassable grid, proceed to safety hold mode.

[0163] Before submitting the route, reread the grid state version and check that it is consistent with the initial grid state version; check again whether the position corresponding to the cut-out grid is still within the segment that the UAV has not yet executed; if the UAV has already passed the position during the calculation, do not submit the old replanning result, but re-execute this method with the new replanning trigger position and replanning trigger heading; at the same time, check that all grids in the unexecuted route are passable, and check the positional continuity, heading continuity and minimum turning radius of adjacent segments;

[0164] When all the above checks pass, the original flight routes that have been executed before the replanning trigger position and the unexecuted routes obtained after one or more local replacements after the replanning trigger position are combined in the execution order to form a real-time replanning flight route, and the new route version and the new route effective position are sent to the flight control terminal.

[0165] The process of forming the real-time replanning flight route results in unexecuted routes that have completed local detour replacements for all discovered impassable grids under the same grid state version, and ensures that route replacement only applies to segments that the UAV has not yet executed.

[0166] In this embodiment, the grid side length is determined based on the spatial interval of the inspection object, the upper limit of the positioning error, and the turning capability of the UAV; the grid boundary safety distance is determined based on the upper limit of the positioning error, the upper limit of the lateral error of the track tracking, and the grid boundary safety margin; the heading measurement tolerance is determined based on the heading sensor calibration error and the flight control heading tracking error; the minimum turning radius is determined by the flight control envelope under the current inspection speed and load conditions; when the above parameters change, the parameter values ​​of the current inspection task are reset according to the corresponding physical quantities.

[0167] In this embodiment, when the replanning trigger position is missing, the replanning trigger heading is missing, the minimum turning radius is a non-positive number, the grid state version cannot be read, the cutting boundary does not exist, the upper limit of the expansion level is reached but the two ends cannot be connected, the heading reachable grid sub-band cannot reach the cutting end, all candidate grid chains cannot form a continuous track that meets the minimum turning radius and boundary safety conditions, or the replanning calculation time reaches the replanning calculation time limit, the safety maintenance process is entered, and the route is not forcibly formed by relaxing the minimum turning radius, using impassable grids, or using grids outside the minimum connected bypass grid band.

[0168] In this embodiment, grid connectivity, heading range propagation, candidate directed endpoint connection configuration solving, route replacement, and version consistency checking can all be performed by the flight mission computer according to the above-mentioned determined input, traversal order, and judgment conditions; grid access status can be generated by no-fly zone configuration, obstacle perception results, or airspace restriction updates, but after entering this method, the grid access status of the low-altitude inspection space grid will be used for processing.

[0169] To further illustrate the working process of this embodiment, a specific operation process is given below:

[0170] Grid side length Take 20 meters as the safety distance at the grid boundary. Take 2 meters, and determine the position tolerance. Take 1.5 meters, heading measurement tolerance Take 8 degrees, valid state time Set the time to 0.4 seconds and increase the upper limit of the expansion level. Take 12 floors, maximum number of floors in the heading range Take 8, and recalculate the time limit. Set the time to 0.30 seconds; the grid coordinate origin is located at the southwest corner of the inspection area. Axis to the east, The axis is north; the heading propagation calculation uses radians, and the following angle values ​​are converted to degrees for ease of description.

[0171] The original flight route passed through in sequence , , , , , , and The replanning trigger location is (82 meters, 90 meters), located at... Internally, replanning triggers course Take 0 degrees as the minimum turning radius. Take 9 meters; replan the trigger position to The distances at all boundaries are greater than the location determination tolerance; therefore, the starting grid for replanning is uniquely determined. .

[0172] set up , , and The first impassable grid is the one included in this replanning; the remaining grids involved are traversable. ;from The mesh can be connected sequentially along the complete shared edges. , and Therefore, the above four grids form a blocking grid connectivity block.

[0173] The original flight route was Enter When entering the blocking mesh connected component for the first time, the vertical meshes between the two completely share edges, forming a cut-out boundary. The original flight path was last occupied by a grid in the blocked grid connected component. After that Located outside the blocking mesh connectivity block, therefore and The vertical grids between them share edges to form a cutting boundary. Cut out the boundary The coordinates are 100 meters, and the boundary is cut off. The coordinates are 140 meters, and the two sides are... The coordinates are all within the range of 80 meters to 100 meters.

[0174] In the first extension layer, Contact with the cut-out boundary, It contacts the cutting boundary, but the two cannot fully share edge connectivity through the mesh within the first extended layer's cumulative set; after forming the second extended layer, the bottom side appears , , , , and The continuous mesh fully shares edge paths, and there is another path connecting the cut-out end and the cut-in end on the top side, thus the first connection extends the level. It equals 2.

[0175] According to the formula for the effective span of the grid, by Mihe Rice obtained Meter; Geometric span half angle Approximately 38.66 degrees; in Maximum heading change in meters Approximately 125.47 degrees; replanning trigger heading range The range is from -8 degrees to 8 degrees; since the initial grid for reprogramming is the same as the side candidate grid, the initial heading range is thus defined. and same.

[0176] exist In the middle, after expanding the initial heading range according to the maximum heading change, it is combined with... Entering from the south The intersection of the geometric cross-side heading ranges yields the non-empty entry heading range; this entry heading range continues within... After internal expansion, it can enter from the east. The geometric cross-side heading ranges intersect; subsequently to , to All of them formed an eastward non-airborne entry into the flight path range. Head north At the same time, it also forms a non-airborne entry into the course range, and Able to follow the original flight route Enter The eastward grid directions are continuous.

[0177] Following the rules of prioritizing the number of complete shared edge transformations, followed by the minimum necessary cumulative heading change, and then prioritizing the lexicographical order of grid numbers, the bottom-side grid chain was determined as the first candidate grid chain, and its grid order is as follows: , , , , and .

[0178] On the first candidate mesh chain, select and The points crossing the edge (91m, 80m) on the effectively shared edge segment are selected. and The UAV effectively shares the cross-point on the edge segment (100m, 71m); starting from the replanning trigger position (82m, 90m), the UAV enters a right-turn arc with a radius of 9m eastward, with the center of the arc at (82m, 81m), obtains a southward heading at (91m, 81m), and then travels along a southward straight line to (91m, 80m); Then, a left-turning arc with a radius of 9 meters is formed with (100 meters, 80 meters) as the center, and an eastward heading is obtained when it reaches (100 meters, 71 meters).

[0179] The unmanned aerial vehicle then followed The eastward straight line of meters passes through successively and and enter from the cross point (140 meters, 71 meters). ;exist Inside, form a left-turning arc with a radius of 9 meters centered at (140 meters, 80 meters), obtain a north heading at (149 meters, 80 meters) and enter... ;exist Inside, form a right-turn arc with a radius of 9 meters centered at (158 meters, 80 meters), obtain an eastward heading at (158 meters, 89 meters), and then proceed along a straight eastward path to (160 meters, 89 meters), thus entering the original flight path. The eastward grid directions are continuous.

[0180] Each of the above-mentioned cross-edge points is at least 2 meters away from the corresponding complete shared edge endpoint of the grid, each arc radius is equal to 9 meters and not less than the minimum turning radius, each arc and straight line segment is located inside the corresponding passable grid, and crosses the grid boundary only at the selected valid shared edge segment. Therefore, the first candidate grid chain can form a continuous bypass grid track sequence.

[0181] Step S4: Delete the original flight path After that Previous and The blocked segment, and after passing through , , and Connection of the circumnavigation grid track sequence and Cut into the boundary For determination only To enter the grid, the sequence of circumnavigating the grid starts from... Enter from the north side No need to cross and The cut-in boundary between .

[0182] After completing the partial replacement, start checking for new unexecuted routes from the replanning trigger point. , , , , , , and All grids are passable, and the reread grid state version is consistent with the initial grid state version. Therefore, this replanning does not need to continue processing the second blocking grid connectivity block, and the resulting real-time replanned flight path can be submitted to the flight control terminal.

[0183] As can be seen from the above description, the real-time flight path replanning method for gridded low-altitude inspection provided in this embodiment has the following technical effects:

[0184] After the original flight path is partially blocked, the first impassable grid in the remaining inspection grid sequence is used to form a blocking grid connected block. The cut-out boundary and cut-in boundary are used to limit the two ends of the blocked part of the original flight path. Then, the passable grid is expanded layer by layer from the periphery of the blocking grid connected block to form a minimum connected detour grid band, so that the detour processing is concentrated in the local connected area that can connect the cut-out end and the cut-in end.

[0185] Based on this, the replanning trigger heading and minimum turning radius propagate into and out of the heading range. By forming heading reachable grid sub-bands through continuity conditions, grid transformations that only have complete shared edge adjacency relationships but cannot meet the continuous turning requirements will no longer participate in the formation of candidate grid chains. Furthermore, by effectively sharing candidate cross-edge points, candidate cross-edge headings, and grid connection configurations, the heading reachability relationships between grids are implemented into continuous tracks that meet the requirements of positional continuity, tangential continuity, minimum turning radius, and grid boundary safety distance. This forms a sequence of bypass grid tracks that can continuously cut out from the original flight route, bypass blocked grid connected blocks, and reconnect to the original flight route. This reduces the possibility that although the bypass path can avoid inaccessible grids, some grid transformations cannot be continuously executed due to replanning trigger heading and minimum turning radius restrictions. It also improves the consistency and actual executability between real-time replanning flight routes and the actual motion constraints of the UAV.

[0186] The above embodiments illustrate a complete implementation of the method of the present invention. While keeping the formation and calling relationship between the remaining inspection grid sequence, blocked grid connected blocks, cut-out boundary, cut-in boundary, minimum connected bypass grid band, heading reachable grid sub-band, cut-out and cut-in grid pairs, bypass grid track sequence and real-time replanning flight route unchanged, specific parameters can be adjusted according to the coordinate range of the inspection area and the flight envelope of the unmanned aerial vehicle.

[0187] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A real-time flight path replanning method for gridded low-altitude inspection, characterized in that, The methods and steps include the following: S1. Obtain the grid accessibility status, original flight path, replanning trigger position, replanning trigger heading, and minimum turning radius of the low-altitude inspection space grid; form a remaining inspection grid sequence from the grid where the replanning trigger position is located along the original flight path to the endpoint; according to the grid accessibility status, connect the first impassable grid in the remaining inspection grid sequence along the complete shared edge of the grid to form a blocking grid connection block; S2. A cut-out boundary is formed by the complete shared edge of the grid that the original flight path crosses when it first enters the blocked grid connected block, and an inbound boundary is formed by the complete shared edge of the grid that the original flight path crosses when it leaves the blocked grid connected block for the last time after entering. The passable grid is expanded layer by layer from the periphery of the blocked grid connected block until the expanded grid that contacts the cut-out boundary and the inbound boundary is first connected by the complete shared edge of the grid, forming a minimum connected bypass grid band. S3. Based on the replanning trigger heading and the minimum turning radius, propagate into and out of the heading range along the complete shared edge of the minimum connected bypass grid zone. The intersection of the out of the heading range of the previous grid and the entry heading range of the next grid is used as the continuation condition to form a heading reachable grid sub-zone. A cut-out and cut-in grid pair is formed by the cut-out grid and the cut-in grid that respectively contact the cut-out boundary and the cut-in boundary. The cut-out and cut-in grid pairs are connected by the reachable grid sub-band along the flight direction to form a bypass grid track sequence. S4. Replace the segments in the original flight route located between the cut-out and cut-in grid pairs with the bypass grid track sequence to form a real-time replanned flight route.

2. The real-time flight path replanning method for gridded low-altitude inspection according to claim 1, characterized in that, The formation process of the remaining inspection grid sequence specifically includes: Each time a replanning is triggered, the grid access status and the original flight route under the same grid state version are read, and the set of grids that have been visited and the first unfinished grid are determined from the flight control record, so that subsequent processing only applies to the unfinished part of the original flight route; Based on the replanning trigger position, determine the replanning starting grid corresponding to the replanning trigger position in the incomplete part of the original flight path; The first occurrence position of the replanning starting grid is found in the incomplete part of the original flight route. Starting from that position, all grids to the end point are retained according to the existing access order of the original flight route. For the same grid that is repeatedly repeated in the original flight route, only the first occurrence record is retained to form the remaining inspection grid sequence. At the same time, the original sequence number of each grid in the original flight route and the corresponding heading between the grids in the original flight route are retained.

3. The real-time flight path replanning method for gridded low-altitude inspection according to claim 2, characterized in that, The process of forming the blocked grid connectivity block specifically includes: The access status of the grid is checked along the access order of the remaining inspection grid sequence, and the first impassable grid is determined as the first impassable grid; Using the first impassable grid as the connection starting point, read the grids that are completely adjacent to the currently processed grid through the grid sharing the edge. Add the impassable adjacent grids that have not yet entered the processing set to the processing set, and repeat the connection processing on the newly added grids until the processing set is empty. The blocked grid connectivity block is formed by connecting the first impassable grid with an impassable grid that consists only of impassable grids through a complete shared edge chain, and the connectivity processing is carried out only within the effective range of the low-altitude inspection space grid.

4. The real-time flight path replanning method for gridded low-altitude inspection according to claim 1, characterized in that, The process of forming the cut-out boundary specifically includes: According to the access order of the remaining inspection grid sequence, check the adjacent grid pairs from the first grid of the remaining inspection grid sequence along the original flight path to the endpoint; When the current grid is outside the blocked grid connection block and the next grid is inside the blocked grid connection block, the adjacent grid pair is determined as a transition into the blocked grid connection block; The cut-out boundary is formed by the complete shared edge between adjacent grids when the transformation first occurs, according to the access order of the original flight route.

5. The real-time flight path replanning method for gridded low-altitude inspection according to claim 4, characterized in that, The process of forming the cutting boundary specifically includes: Continue checking the remaining inspection grid sequence from the original flight path position where the cut-out boundary is formed to the endpoint, and record the original sequence number of the grid that belongs to the blocking grid connectivity block in the original flight path; The grid corresponding to the maximum value in the original recorded sequence number is determined as the last occupied grid; When there is an original flight path grid located outside the blocking grid connected block after the last occupied grid and adjacent to the last occupied grid through the grid's complete shared edge, the cutting boundary is formed by the grid's complete shared edge between the two.

6. The real-time flight path replanning method for gridded low-altitude inspection according to claim 1, characterized in that, The formation process of the minimum connected bypass mesh band specifically includes: Using the boundary mesh fully shared edge of the blocked mesh connected block as the starting edge for expansion, the meshes that are passable but not yet fully visited are expanded layer by layer to the periphery of the blocked mesh connected block according to the adjacency relationship of the fully shared edge of the mesh, forming an expansion layer, and the meshes in the already formed expansion layer are merged to form a cumulative expansion set; In the cumulative extended set, determine the meshes that completely coincide with the cutout boundary and the cutin boundary outside the blocked mesh connected block, and start from the mesh that completely coincides with the cutout boundary, propagate the connectivity marker only along the complete shared edge of the mesh; When the connectivity marker first reaches a grid that completely coincides with the cut-in boundary, the connectivity components are divided according to the complete shared edges of the grid. The connectivity components that simultaneously contain grids that completely coincide with the cut-out boundary and grids that completely coincide with the cut-in boundary are retained, and the retained connectivity components form the minimum connectivity bypass grid band.

7. The real-time flight path replanning method for gridded low-altitude inspection according to claim 1, characterized in that, The formation process of the heading reachable grid sub-band specifically includes: For the grids adjacent to each other through the complete shared edge in the minimum connected bypass grid zone, the grid boundary safety distance is deducted from the two endpoints of the complete shared edge along the edge, and the resulting intermediate line segment forms an effective shared edge segment. The length of the effective shared edge segment forms the effective span of the grid. The angle between the direction from the grid center to the endpoint of the effective shared edge segment and the normal azimuth of the complete shared edge of the grid forms a geometric span half-angle, and the geometric span heading range is formed with the normal azimuth of the complete shared edge of the grid as the center according to the geometric span half-angle; the maximum heading change of a single grid is determined according to the minimum turning radius and the effective span of the grid; Centered on the replanning trigger heading, the heading measurement tolerance is extended to both sides to form the replanning trigger heading range; the grids that completely coincide with the cutout boundary outside the blocked grid connected block are formed into cutout side candidate grids, and the replanning trigger heading range is propagated to the cutout side candidate grids along the access order of the unblocked grids of the original flight route, using the original flight route grid inter-grid headings between adjacent grids and the geometric cross-edge heading range of the corresponding grid complete shared edge, to form the initial cutout heading range; Starting from the candidate grid on the cut-out side, the entry heading range and the exit heading range are propagated only along the grid complete shared edges within the minimum connected bypass grid band. The intersection of the exit heading range already formed by the previous grid and the geometric cross-edge heading range allowed by the grid complete shared edges between adjacent grids forms the entry heading range of the next grid. The non-empty entry heading range is expanded according to the maximum heading change of a single grid to form the exit heading range. The heading reachable grid sub-band is formed by continuously passing through non-empty grids within the heading range starting from the candidate grid on the cut-out side.

8. The real-time flight path replanning method for gridded low-altitude inspection according to claim 7, characterized in that, The formation process of the cut-out and cut-in mesh pairs specifically includes: In the reachable sub-band of the heading, find the grid that completely coincides with the cut-out boundary outside the block of the blocking grid, and form a candidate set of cut-out grids; find the grid that completely coincides with the cut-in boundary outside the block of the blocking grid and can be continuously connected in heading with the original flight path grid after the grid, and form a candidate set of cut-in grids. For the candidate cutout mesh, search for candidate mesh chains along the mesh complete shared edge that satisfies the continuation condition to the candidate cutin mesh. Use the number of mesh complete shared edge transformations of the candidate mesh chain as the first sorting factor, and preferentially select the combination of candidate cutout mesh and candidate cutin mesh with fewer mesh complete shared edge transformations. When the number of complete shared edges of the mesh is the same, the intersection of the departure heading range already formed by the current mesh and the geometric cross-edge heading range corresponding to the next complete shared edge of the mesh is determined as the departure heading range corresponding to the next complete shared edge of the mesh. Based on the already formed entry heading range and departure heading range, the minimum achievable circular heading change is determined. The minimum circular heading change of each mesh in the candidate mesh chain is added together to form the cumulative minimum necessary heading change. The combination with the smaller cumulative minimum necessary heading change is selected. When the number of complete shared edge transformations of the grid and the cumulative minimum necessary heading change are the same, a unique combination is determined sequentially based on the original sequence number of the cut-out grid candidate in the original flight path, the original sequence number of the cut-in grid candidate in the original flight path, and the lexicographical order of the grid numbers, and the cut-out grid and the cut-in grid in the unique combination form the cut-out and cut-in grid pair.

9. The real-time flight path replanning method for gridded low-altitude inspection according to claim 8, characterized in that, The formation process of the bypass grid track sequence specifically includes: The candidate mesh chains are processed sequentially according to their sorting. Starting from the cut-out mesh, the complete shared edges between adjacent meshes are listed sequentially. The mesh boundary safety distance is deducted from the two endpoints of the complete shared edges to obtain the effective shared edge segments. Candidate cross-edge points are generated on the effective shared edge segments. For a complete shared edge of the mesh that satisfies the continuation condition, the intersection of the departure heading range of the previous mesh and the geometric cross-edge heading range of the complete shared edge of the mesh is used to form a cross-edge heading range, and a candidate cross-edge heading is formed based on the normal azimuth of the complete shared edge of the mesh and the cross-edge heading range. Adjacent grids share the same candidate cross-edge point and candidate cross-edge heading. Starting from the cut-out grid, propagate backward according to the candidate cross-edge point and candidate cross-edge heading. Within each grid, the entrance position and entrance heading, and the exit position and exit heading are used as a set of directed endpoints. Enumerate left arc straight left arc, right arc straight right arc, left arc straight right arc, right arc straight left arc, left arc right arc left arc, and right arc left arc right arc connection configurations. The radius of each arc is taken as the minimum turning radius. And keep the arc segment and the adjacent straight segment or arc segment positionally and tangentially continuous at the connection. Check whether all points of each connection configuration are located inside the corresponding grid, so that the track, except for the valid shared edge segment as the entrance and the valid shared edge segment as the exit, does not intersect with the grid boundary, and the minimum distance of the track to the grid boundary and grid vertex that are not selected as the entrance or exit is not less than the grid boundary safety distance; retain the candidate states that can be propagated backward. When the candidate state reaches the original flight path continuation position after the cut-in grid, the preceding relationship of the candidate state is traced back to form a continuous track. The continuous track is associated with the cut-out grid, the intermediate bypass grid, the cut-in grid, the cross-edge point, the cross-edge heading, and the circular arc segment and straight line segment within the grid to form the bypass grid track sequence.

10. The real-time flight path replanning method for gridded low-altitude inspection according to claim 1, characterized in that, The process of forming the real-time replanning flight route specifically includes: At the start of this replanning, the initial grid state version corresponding to the grid accessibility status is read and recorded; the replacement interval is determined based on the original sequence numbers of the cut-out grid and the cut-in grid in the original flight route; the route prefix before the cut-out grid in the original flight route that has not yet been executed is retained; the obstructed grids and corresponding segments of the original flight route from after the cut-out grid to before the cut-in grid are deleted; the bypass grid track sequence is connected between the cut-out grid and the cut-in grid; and the flight continues along the suffix of the original flight route from the cut-in grid to complete the local replacement. After the local replacement is completed, starting from the replanning trigger position, check all the grids involved in the unexecuted route after the replacement; when there are still impassable grids in the unexecuted route, the route that has completed the current local replacement is taken as the next round of pending routes, and the first impassable grid that appears is taken as the new first impassable grid. Repeat the process of blocking the formation of grid connectivity blocks, cutting out boundaries and cutting in boundaries, forming minimum connected bypass grid bands, propagating into and out of the heading range, and local replacement, until there are no impassable grids in the unexecuted route or the replanning calculation time limit is reached; Before submitting the route, the grid status version is reread, and it is checked that the reread grid status version is consistent with the initial grid status version. It is also checked that all grids in the unexecuted route are passable, and whether adjacent flight segments meet the requirements of position continuity, heading continuity, and minimum turning radius. After the checks are passed, the original flight route that has been executed before the replanning trigger position and the unexecuted route obtained after the replanning trigger position through local replacement are combined in the execution order to form the real-time replanning flight route.

Images