A Dynamic Inspection Method and System for Substations Using Unmanned Aerial Vehicles Based on Underpass Constraints

Abstract

This invention discloses a dynamic inspection method and system for substations using UAVs based on underpass constraints, belonging to the field of intelligent UAV inspection technology for substations. The method includes: acquiring substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data; and marking the inspection waypoints with attributes based on the obstacle data above the waypoint and the no-fly zone boundary data to obtain ordinary inspection points, underpass points, clearance turning points, and no-fly adjacent points. This dynamic inspection method and system for substations using UAVs based on underpass constraints expands traditional single-coordinate waypoints into topological nodes with safety semantics, enabling dynamic inspection path planning to simultaneously consider equipment inspection needs, overhead obstacle distribution, no-fly zone boundaries, and return-to-home safety requirements, thus improving the completeness of waypoint representation in the complex spatial environment of substations.

Description

Technical Field

[0001] This invention relates to the field of intelligent inspection of substations by drones, and more specifically, to a dynamic inspection method and system for substations by drones based on underpass constraints. Background Technology

[0002] With the development of intelligent operation and maintenance of substations, drone inspections have been used for image acquisition of high-altitude equipment, meters, insulators, busbar structures, and areas adjacent to conductors. In existing technologies, CN113077561A discloses an intelligent drone inspection system that can perform complex trajectory planning based on 3D maps and automatically select equipment photo points using deep learning algorithms, while simultaneously planning the drone's position, orientation, and camera gimbal angle. CN116594426A discloses a method and system for planning drone inspection routes in substations, which acquires 3D point cloud data of the substation, divides it into sub-regions, extracts inspection points, and generates waypoints and safety passages to form dynamic inspection routes. While these solutions improve the efficiency of waypoint generation and route planning, they still primarily focus on 3D maps, point cloud models, waypoint extraction, and path generation, lacking specific handling of the return-to-base risk at underpasses within the substation. Substations have densely distributed conductors, busbars, and structures, and some inspection flight points are located below overhead obstacles. If a drone, when low on battery, experiencing communication failures, or with a mission interruption, simply ascends vertically to its return altitude using conventional strategies, it is prone to collisions with overhead obstacles. Therefore, there is an urgent need for a drone inspection method and system that incorporates underpass constraints into dynamic inspection path planning and binds the underpass point to a safe, accessible turning point. Summary of the Invention

[0003] The purpose of this invention is to provide a dynamic inspection method and system for substations using unmanned aerial vehicles (UAVs) based on underpass constraints, which solves the problem of collision risk when existing UAVs directly ascend and return to base from underpass points.

[0004] This invention achieves the above objective through the following technical solution: a dynamic inspection method for substations using unmanned aerial vehicles (UAVs) based on underpass constraints, the method comprising:

[0005] Acquire substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data.

[0006] Based on the obstacle data above the waypoints and the no-fly zone boundary data, the inspection waypoints are marked with attributes to obtain ordinary inspection points, underpass points, clear airspace turning points and no-fly adjacent points.

[0007] Using the inspected waypoints as topology nodes and the waypoint connections that meet the conditions of safe distance, altitude difference and no-fly zone detour as topology edges, a waypoint safety topology graph is constructed.

[0008] In the waypoint safety topology map, an underpass constraint is configured for the underpass point. The underpass constraint is used to prevent the UAV from directly pulling up to the return altitude in the vertical direction at the underpass point.

[0009] For the underpass point, match the airspace turning point in the waypoint safety topology map and establish a binding relationship between the underpass point and the airspace turning point;

[0010] In response to the selected waypoint to be inspected, a dynamic inspection path is generated based on the waypoint safety topology map, the underpass constraint, and the binding relationship;

[0011] Control the drone to perform inspection tasks according to the dynamic inspection path, and monitor the drone's current location, battery status, communication status and task status;

[0012] When a low battery return-to-home, communication failure return-to-home, mission interruption return-to-home, or manual return-to-home is triggered, the system determines whether the drone's current position or return-to-home path involves an underpass. If an underpass is involved, the drone is controlled to fly to the clearance turning point bound to the underpass, and then rise to the return-to-home altitude from the clearance turning point and return to home. If no underpass is involved, the drone is controlled to return to home along the normal return-to-home path.

[0013] After the drone returns to base, needs a battery swap, or resumes its mission, a continued flight inspection path is generated based on the inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and the aforementioned binding relationship.

[0014] Furthermore, the acquisition of substation inspection route data includes:

[0015] Acquire panoramic images of the equipment to be inspected, collected by the drone;

[0016] The equipment components and their locations in the panoramic image are identified using an equipment component recognition model, and candidate inspection points are generated.

[0017] Based on the real-time pose of the UAV, the gimbal angle, the shooting magnification, and the target box position of the candidate inspection point in the panoramic image, the candidate inspection point is converted into a re-flying inspection waypoint.

[0018] Based on the offset between the center of the target frame and the center of the image, and the ratio of the target frame area to the image area, the UAV pose, gimbal angle, or shooting magnification are corrected to obtain the inspection waypoint data.

[0019] Furthermore, the step of marking the inspection waypoints with attributes based on the obstacle data above the waypoints and the no-fly zone boundary data includes:

[0020] When the vertical support column formed between the shooting height and the return height of the inspection point intersects with the conductor, busbar or structure, the inspection point is marked as the underpass point;

[0021] When there are no conductors, busbars and structures in the vertical lifting column formed between the shooting height and the return height of the inspection point, and there is a passable topological edge between the inspection point and the underpass point, the inspection point is marked as the clearance turning point.

[0022] When an inspection point falls within the safe distance beyond the boundary of the no-fly zone, the inspection point shall be marked as a no-fly adjacent point.

[0023] When an inspection point is not marked as an underpass point, a clear zone bend point, or a no-fly adjacent point, the inspection point shall be marked as a regular inspection point.

[0024] Furthermore, the construction of a waypoint safety topology graph, using the inspected waypoints as topology nodes and waypoint connections that meet the conditions for safe distance, altitude difference, and no-fly zone detours as topology edges, includes:

[0025] Calculate the horizontal distance, altitude difference, and whether the line connecting any two inspection waypoints crosses a no-fly zone;

[0026] When the horizontal distance is not greater than the waypoint adjacency distance threshold, the altitude difference is not greater than the altitude change threshold, and the line connecting the two inspection waypoints does not cross the no-fly zone, a topological edge is established between the two inspection waypoints.

[0027] Based on the waypoint attributes connected by the topological edge, the topological edge is marked as a regular passage edge, an underpass passage edge, a clearance enhancement edge, or a no-fly detour edge;

[0028] Configure the topology edge connecting the underpass point as the underpass passage edge, and configure the topology edge connecting the clearance turning point and allowing the drone to lift as the clearance lift edge.

[0029] Furthermore, the step of matching the airspace turning point in the waypoint safety topology map for the underpass point and establishing a binding relationship between the underpass point and the airspace turning point includes:

[0030] Using the underpass point as the search starting point, candidate clearance turning points with passable topological edges between them are obtained in the waypoint safety topology map;

[0031] Determine whether the candidate clearance turning point meets the clearance lifting condition. The clearance lifting condition is that there are no conductors, busbars and frames in the vertical lifting column formed between the shooting height and the return height of the candidate clearance turning point.

[0032] The binding cost of the candidate airspace turning point is calculated based on the flight distance between the candidate airspace turning point and the underpass point, the flight distance from the candidate airspace turning point to the return point, the airspace height of the candidate airspace turning point, and the return energy consumption corresponding to the candidate airspace turning point.

[0033] The candidate airspace turning point with the lowest binding cost and that meets the airspace increase conditions is determined as the binding airspace turning point.

[0034] The waypoint number of the underpass point is associated with and stored with the waypoint number of the bound clearance turning point.

[0035] Furthermore, in response to the selected waypoint to be inspected, a dynamic inspection path is generated based on the waypoint safety topology map, the underpass constraint, and the binding relationship, including:

[0036] Receive the list of waypoints to be inspected;

[0037] Extract a task sub-topology map from the waypoint safety topology map, which includes the set of waypoints to be inspected, underpass points, and bound airspace turning points;

[0038] Calculate the path cost of each topological edge in the task sub-topological graph based on the topological edge distance, altitude change, number of shooting attitude switching, no-fly detour status, underpass passage status, and return-to-home energy consumption.

[0039] Search the inspection path with the minimum sum of path costs in the task sub-topology graph;

[0040] When the inspection path passes through the underpass point, the airspace turning point bound to the underpass point will be inserted into the inspection path, so that the drone can enter or exit the underpass point via the airspace turning point.

[0041] Furthermore, the monitoring of the drone's current location, battery status, communication status, and mission status includes:

[0042] Obtain the real-time location of the UAV and map the real-time location to the current topology node or current topology edge in the waypoint safety topology graph;

[0043] When the current topology node is a passing point or the current topology edge is a passing edge, it is determined that the UAV is in the passing constraint region;

[0044] Based on the estimated energy consumption from the drone's current location to the bound airspace turning point, the estimated energy consumption from the bound airspace turning point to the return point, the estimated energy consumption from the bound airspace turning point to the return altitude, and the safety margin, calculate the dynamic low battery threshold.

[0045] When the drone's current battery level is lower than the dynamic low battery threshold, a low battery return-to-home mechanism is triggered.

[0046] Furthermore, if an underpass point is involved, the drone is controlled to first fly to the clearance turning point associated with the underpass point, and then ascend from the clearance turning point to the return altitude and return, including:

[0047] Determine the underpass point corresponding to the drone's current location, or the underpass point on the return path that is closest to the drone's current location;

[0048] Read the bound clearance turning point associated with the underpass point;

[0049] Generate a low-altitude retreat path from the drone's current position to the bound airspace turning point;

[0050] Control the drone to fly along the low-altitude withdrawal path, and limit the drone's flight altitude in the low-altitude withdrawal path to not exceed the safe lowering altitude;

[0051] Once the drone reaches the designated airspace turning point, control the drone to ascend from the designated airspace turning point to the return altitude and return to the return point.

[0052] Furthermore, after the drone returns to base, undergoes battery swapping, or resumes its mission, a continued flight inspection path is generated based on inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and the aforementioned binding relationship. This includes:

[0053] Record the waypoints where image acquisition has been completed, and obtain the set of waypoints that have been inspected;

[0054] The set of inspected waypoints is deleted from the original set of waypoints to be inspected to obtain the set of uninspected waypoints.

[0055] The take-off location or mission resumption location of the drone after battery swapping shall be used as the starting point for continued flight.

[0056] Based on the set of uninspected waypoints, the waypoint safety topology map, binding relationships, remaining battery power, and return safety margin, the cost of the continued flight path is recalculated.

[0057] Generate a continuous flight inspection path that satisfies the underpass constraint, no-fly detour constraint, and return-to-base energy consumption constraint.

[0058] A substation UAV dynamic inspection system based on underpass constraints is used to execute the aforementioned substation UAV dynamic inspection method based on underpass constraints. The system includes:

[0059] The waypoint data acquisition module is used to acquire substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data.

[0060] The waypoint attribute marking module is used to mark the inspection waypoints according to the obstacle data above the waypoints and the no-fly zone boundary data, so as to obtain ordinary inspection points, underpass points, clear airspace turning points and no-fly adjacent points;

[0061] The safety topology construction module is used to construct a waypoint safety topology graph using the inspected waypoints as topology nodes and the waypoint connection relationships that meet the safety distance, altitude difference and no-fly zone detour conditions as topology edges.

[0062] The underpass constraint configuration module is used to configure underpass constraints for underpass points in the waypoint safety topology map. The underpass constraints are used to prevent the UAV from directly pulling up to the return altitude in the vertical direction at the underpass point.

[0063] The binding relationship generation module is used to match the clearance turning point in the waypoint safety topology map for the underpass point, and establish the binding relationship between the underpass point and the clearance turning point.

[0064] The dynamic path generation module is used to generate a dynamic inspection path in response to the selected waypoint to be inspected, based on the waypoint safety topology map, the underpass constraint and the binding relationship;

[0065] The inspection status monitoring module is used to control the drone to perform inspection tasks according to the dynamic inspection path, and to monitor the drone's current location, battery status, communication status and task status.

[0066] The abnormal return-to-home control module is used to determine whether the current position or return-to-home path of the drone involves a point of passage when low battery return-to-home, communication abnormal return-to-home, mission interruption return-to-home, or manual return-to-home is triggered. If a point of passage is involved, the module controls the drone to fly to the clearance turning point bound to the point of passage first, and then pull up to the return-to-home height from the clearance turning point and return to home.

[0067] The flight path generation module is used to generate a flight path after the UAV returns to base, has its battery swapped, or resumes its mission, based on the inspected waypoints, uninspected waypoints, the UAV's current location, remaining battery power, and the binding relationship.

[0068] The beneficial effects of this invention are as follows:

[0069] 1. By marking the attributes of inspection waypoints as ordinary inspection points, underpass points, clearance turning points, and no-fly adjacent points, the traditional single-coordinate waypoints are expanded into topological nodes with safety semantics. This enables dynamic inspection path planning to simultaneously consider equipment inspection needs, overhead obstacle distribution, no-fly zone boundaries, and return safety requirements, thereby improving the completeness of waypoint representation in the complex spatial environment of substations.

[0070] 2. This invention establishes a binding relationship between the underpass point and the clearance turning point for the underpass point of the conductor, busbar and the structure. This allows the UAV to retreat to the clearance turning point that meets the clearance conditions for take-off before returning to base, instead of directly vertically ascending from the underpass point when the UAV is low on power, has communication problems, is interrupted, or is manually returned to base. This reduces the risk of the UAV colliding with obstacles above it.

[0071] 3. This invention introduces underpass constraints, no-fly detour constraints, and return-to-base energy consumption constraints when generating dynamic inspection paths. This allows the system to automatically generate inspection paths that meet safety conditions after the user selects some waypoints to be inspected, thereby decoupling waypoints from routes and improving the flexibility of inspection task configuration.

[0072] 4. This invention determines a dynamic low battery threshold based on the drone's current location, bound airspace turning point, return point, battery health, and return safety margin, so that the low battery return triggering time matches the safe retreat requirements of the underpass area, avoiding the problem that a fixed battery threshold cannot adapt to different locations and different return paths.

[0073] 5. After the UAV returns to base, undergoes battery swapping, or resumes its mission, this invention regenerates the continued flight inspection path based on the binding relationship between inspected waypoints, uninspected waypoints, remaining battery power, and underpass points. This avoids repeated inspections of completed waypoints, improving the utilization rate of UAV endurance resources and the efficiency of inspection mission recovery. Attached Figure Description

[0074] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0075] Figure 1 This is a flowchart illustrating the overall method of the present invention;

[0076] Figure 2 This is a schematic diagram of the substation inspection navigation point attribute marking of the present invention;

[0077] Figure 3 This is a schematic diagram illustrating the dynamic inspection path and the binding relationship between the underpass point and the clearance turning point of the present invention.

[0078] Figure 4 This is a flowchart of the abnormal return-to-home control process of the present invention;

[0079] Figure 5 This is a system block diagram of the present invention. Detailed Implementation

[0080] The present application will now be described in further detail with reference to the accompanying drawings. It should be noted that the following specific embodiments are only used to further illustrate the present application and should not be construed as limiting the scope of protection of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.

[0081] Example 1:

[0082] Please see Figures 1-4 This invention provides a dynamic inspection method for substations using unmanned aerial vehicles (UAVs) based on underpass constraints. This method is applied to UAV inspection tasks of transformers, circuit breakers, disconnect switches, instrument transformers, surge arresters, meters, insulators, busbars, and structural areas within substations. The method acquires inspection waypoint data, marks waypoint attributes, constructs a waypoint safety topology map, establishes a binding relationship between underpass points and clearance turning points, and controls the UAV to first retreat to a clearance turning point before ascending back to base when returning due to low battery, communication failure, mission interruption, or manual return. This avoids collisions caused by the UAV directly ascending vertically in the underpass area.

[0083] In this embodiment, distance, height, and radius are uniformly adopted. As a unit, energy consumption is uniformly adopted. As a unit, electricity consumption is uniformly expressed as... As a unit of measurement, wind speed is uniformly adopted. As a unit, the angle is uniformly adopted. As a unit.

[0084] S1. Obtain substation inspection route data.

[0085] The system acquires substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data. Waypoint coordinates are represented using a unified coordinate system within the substation, including horizontal coordinates, vertical coordinates, and the waypoint shooting altitude.

[0086] In one implementation, the UAV first acquires panoramic images of the equipment to be inspected, and then uses an equipment component recognition model to identify the equipment components and their positions in the panoramic image, generating candidate inspection points. Based on the UAV's real-time pose, gimbal angle, shooting magnification, and the target bounding box position of the candidate inspection points in the panoramic image, the system converts the candidate inspection points into re-flyable inspection waypoints.

[0087] To ensure that the inspection images meet the requirements for subsequent defect identification, the system verifies the target bounding box center offset and the target bounding box area ratio. The normalized offset threshold between the target bounding box center and the image center is set to... to When the inspection target is a meter, bushing end, or disconnector contact, the offset threshold is taken as follows: When the inspection target is the overall appearance of the equipment, the offset threshold is taken as follows: The threshold is based on the following: meters, bushing ends, and knife switch contacts are delicate components, requiring high precision in centering the image; the overall appearance of the equipment is a contour-based target, requiring relatively lower precision in centering the image.

[0088] The threshold for the proportion of the target bounding box area to the image area is taken as follows. to When the inspection target is a time meter, take... to When the inspection target is an insulator string, take... to When the inspection target is the overall appearance of the equipment, take... to This threshold ratio is determined based on the effective proportion of target components in historical defect identification samples and is updated according to the statistical results of the identification accuracy of similar devices. When the center offset of the target frame exceeds the corresponding threshold, or the area proportion of the target frame is lower than the corresponding threshold, the system adjusts the lateral movement distance, forward and backward distance, gimbal pitch angle, gimbal yaw angle, or shooting magnification of the UAV, and updates the corresponding inspection waypoint data.

[0089] S2. Mark the attributes of the inspection waypoints.

[0090] The system marks the inspection points with attributes based on obstacle data above the waypoints and no-fly zone boundary data, resulting in ordinary inspection points, underpass points, clear airspace turning points, and no-fly adjacent points.

[0091] For each inspection waypoint, the system constructs a vertical lifting column, starting from the shooting altitude of that waypoint and ending at the return altitude. This vertical lifting column is used to determine whether the UAV will spatially interfere with power lines, busbars, or structures if it directly lifts vertically to the return altitude at that waypoint. The radius of the vertical lifting column is determined according to Formula 1:

[0092]

[0093] in, Indicates the radius of the vertically stretched column, in units of... ; This represents the circumcircle radius of the UAV rotor, in units of 1. It is determined by the parameters of the UAV model; This represents the positioning error margin, in units of... Prioritize the most recent Secondary inspection of flight positioning error Percentile values ​​are determined; Indicates attitude perturbation margin, in units of It is determined by the statistical values ​​of lateral drift when the drone hovers, turns, and flies at low speed; Indicates environmental safety margin, in units of The distance is determined by the safety distance requirements within the station and the density of equipment.

[0094] In this embodiment, Pick to ;when Stable positioning and positioning error no greater than hour, Pick to ;when When the signal is blocked or visual positioning is involved in the fusion, Pick to . Pick to When the on-site wind speed is no greater than hour, Pick to When the on-site wind speed is greater than hour, Pick to . Pick to When a waypoint is near a guide line, busbar, or structure, Pick to When the waypoint is in an open passage, Pick to .

[0095] When the vertical lifting column of an inspection waypoint intersects with the spatial envelope of a guide wire, busbar, or framework, the system marks the inspection waypoint as a crossing point. A crossing point indicates that the UAV can perform low-altitude inspection or photography operations at that point, but must not lift directly vertically to the return-to-home altitude at that point.

[0096] When there are no conductors, busbars, or structures within the vertical lifting column of an inspection waypoint, and there is a passable topological edge between the inspection waypoint and the underpass point, the system marks the inspection waypoint as a clearance turning point. Clearance turning points are used for entering and exiting the underpass area, path turning, and safe lifting and return.

[0097] When an inspection point falls within the safe distance extending beyond the no-fly zone boundary, the system marks that inspection point as a no-fly adjacent point. The safe distance extending beyond the no-fly zone boundary is taken as... to When the no-fly zone is adjacent to the area above oil-filled equipment, the area near lightning rods, or the area with dense structures, take... to When a no-fly zone is adjacent to an open passageway, take... to This threshold is determined based on the UAV's dimensions, positioning error margin, in-station safety procedures, and manual verification results.

[0098] When an inspection point is not marked as an underpass point, a clear zone bend point, or a no-fly adjacent point, the system marks the inspection point as a regular inspection point.

[0099] S3. Construct a waypoint safety topology map and configure underpass constraints.

[0100] The system uses inspection waypoints as topology nodes and waypoint connections that meet the conditions of safe distance, altitude difference, and no-fly zone detour as topology edges to construct a waypoint safety topology graph.

[0101] Specifically, the system calculates the horizontal distance, altitude difference, and whether the connecting line between any two inspection waypoints crosses a no-fly zone. When the horizontal distance between two inspection waypoints is no greater than the waypoint adjacency distance threshold, the altitude difference is no greater than the altitude change threshold, and the connecting line between the two inspection waypoints does not cross a no-fly zone, a topological edge is established between the two inspection waypoints.

[0102] Waypoint adjacency distance threshold to ; substation to , substation to , substation to This threshold is determined based on the spacing between equipment within the station, the stable single-segment flight distance of the UAV, and the width of the on-site passageway. The altitude change threshold is taken as... to When the inspection task mainly involves meter reading and close-range image acquisition, take... to When the inspection task includes data collection from the upper part of the structure or the busbar area, take... to This threshold is determined based on the permissible climb or descent range between adjacent waypoints, gimbal compensation capabilities, and flight stability requirements.

[0103] The system marks topological edges as regular access edges, underpass access edges, clearance rise edges, or no-fly detour edges based on the waypoint attributes they connect. Topological edges connecting ordinary inspection points without involving underpasses are marked as regular access edges; topological edges connecting underpasses are marked as underpass access edges; topological edges connecting clearance bends and allowing UAVs to rise are marked as clearance rise edges; and topological edges used to bypass the safe distance range outside the no-fly zone boundary are marked as no-fly detour edges.

[0104] The system configures underpass constraints for underpass points in the waypoint safety topology map. Underpass constraints include: the UAV must not directly ascend to the return-to-home altitude in the vertical direction at the underpass point; when the UAV passes through the underpass passage edge, its flight altitude must not exceed the underpass safety altitude; when the UAV triggers a low battery return-to-home, communication failure return-to-home, mission interruption return-to-home, or manual return-to-home, if the current location is an underpass point or the return-to-home path involves an underpass point, it must first fly to the bound clearance turning point before performing the ascent and return-to-home.

[0105] Return altitude to And higher than the height of the highest conductor, highest busbar, highest structure, or highest lightning rod within the inspection flight area. The safety height margin is taken as... to When the height of the obstacle is from When obtained by measurement or three-dimensional measurement, the safety height margin is taken as... to When the obstacle height is obtained from manual input or historical records, the safety height margin is taken as... to .

[0106] S4. Establish the binding relationship between the underpass point and the clearance turning point.

[0107] For each underpass point, the system matches the airspace turning point in the waypoint safety topology map and establishes a binding relationship between the underpass point and the airspace turning point.

[0108] The system uses the underpass point as the starting point for the search and obtains candidate clearance bend points in the waypoint safety topology map that have a passable topological edge to the underpass point. Candidate clearance bend points must meet the clearance lifting condition, that is, there are no conductors, busbars and structures in the vertical lifting column of the candidate clearance bend point.

[0109] To avoid unclear path selection rules due to multiple weightings, this embodiment adopts a hierarchical selection rule of "constraint screening first, then minimum energy consumption, and finally distance verification." The system first eliminates candidate airspace turning points that do not meet the airspace lift conditions, no-fly detour conditions, and return-to-base energy consumption conditions. Then, it selects the point with the minimum total return-to-base energy consumption from the remaining candidate airspace turning points as the bound airspace turning point. If the difference in total return-to-base energy consumption between two candidate airspace turning points does not exceed... If so, then select the candidate clearance turning point that is shorter in horizontal distance from the underpass point.

[0110] The net clearance turning point is determined according to formula 2:

[0111] ;

[0112] ;

[0113] in, Indicates the first Each underpass point corresponds to a binding clearance turning point; Indicates the first The set of candidate clearance turning points corresponding to each underpass point; Indicates the candidate clearance turning point; Indicates that the drone is from the first The underpass point passes through the candidate clearance turning point. The estimated energy consumption required for retreat, ascent, and return, in units of ; Indicates the candidate net clearance turning point The interference state of the vertically stretched column, when there are no conductors, busbars, or frames inside the vertically stretched column. Pick Otherwise take ; Indicates the first Underpass point and candidate clearance turning point The topologically accessible state between two entities, when there is a accessible topological edge between them. Pick Otherwise take ; This indicates the drone's current available battery energy, in units of... .

[0114] The battery's rated energy at full charge, battery health, and current charge percentage are used to determine its energy level. Battery health is output by the flight controller's battery management module or determined by the most recent battery charge / discharge test result. If the flight controller system has already output the current available battery energy, then the flight controller system's output value is used directly. Formula 2 does not use multi-weighted summation; instead, it first uses... , and Feasibility screening is conducted, and then the airspace turning point with the lowest expected energy consumption is selected from the feasible set, making the rules for determining the binding relationship clearer.

[0115] After the system determines the binding clearance turning point, it associates and stores the waypoint number of the underpass point with the waypoint number of the binding clearance turning point, and writes this association into the waypoint safety topology map. Subsequent inspection path generation, abnormal return control, and follow-up flight path reconstruction all call upon this binding relationship.

[0116] S5. Generate dynamic inspection paths.

[0117] The system responds to the selected waypoints to be inspected and generates a dynamic inspection path based on the waypoint safety topology map, underpass constraints, and binding relationships.

[0118] Specifically, after the user selects the set of waypoints to be inspected, the system extracts a task sub-topology map from the waypoint safety topology map, containing the set of waypoints to be inspected, underpass points, and bound clearance turning points. The system first eliminates paths crossing no-fly zones, paths violating underpass safety altitudes, and paths that do not meet return-to-home energy consumption constraints; then, it selects the path with the shortest flight distance from the remaining paths; if the difference in flight distance between two paths does not exceed... If the number of shooting posture changes is the same, then choose the path with fewer shooting posture changes; if the number of shooting posture changes is the same, then choose the path with fewer passing edges.

[0119] Flight distance difference not exceeding The rules are designed to avoid frequent path switching due to extremely small distance differences. The threshold is determined by statistical results of the impact of path distance fluctuations on flight time in historical inspection missions. If in continuous... During this inspection, the change in path distance had a smaller impact on the total flight time than Then maintain The threshold for judgment; if the change in path distance has an impact on the total flight time of not less than Then the judgment threshold is adjusted to .

[0120] When the inspection path passes through an underpass, the system inserts the airspace turning point bound to that underpass into the inspection path, allowing the drone to enter or exit the underpass via the airspace turning point. If there are two feasible airspace turning points for the same underpass, the bound airspace turning point determined by Formula 2 will be used first.

[0121] S6. Perform inspection tasks and monitor flight status.

[0122] The drone performs inspection tasks according to a dynamic inspection path. The system continuously acquires the drone's real-time location, battery status, communication status, and task status, and maps the drone's real-time location to the current topology node or edge in the waypoint safety topology map.

[0123] When the current topology node is a downpass point, or the current topology edge is a downpass passage edge, the system determines that the drone is in the downpass constraint region. The system calculates the dynamic low battery threshold based on the drone's current position, bound clearance turning point, return point, battery status, and safety margin. The dynamic low battery threshold is determined according to Formula 3:

[0124] ;

[0125] in, Indicates the dynamic low battery threshold, in units of ; This indicates the estimated energy consumption from the drone's current location to the designated airspace turning point, in units of... ; This represents the estimated energy consumption of the drone as it ascends from the designated airspace turning point to its return altitude, expressed in units of... ; This indicates the estimated energy consumption of the drone from the designated airspace turning point to the return point, in units of... ; This indicates the safety margin for returning to base, in units of... ; This indicates the battery's rated energy when fully charged, in units of... ; Indicates battery health, dimensionless, with a value range of [value missing]. to .

[0126] , and The energy consumption per unit horizontal distance, energy consumption per unit altitude climb, current wind speed correction factor, and current load status are determined from the UAV's historical flight logs, or can be estimated in real time by the flight control system. Pick , and sum of to When battery health is below The on-site wind speed is greater than Or the ambient temperature is lower than hour, Pick to When the battery health level is not lower than The on-site wind speed is no greater than And the ambient temperature is not lower than hour, Pick to The above rules are based on the fact that low temperatures, wind disturbances, and battery degradation increase the uncertainty of return-to-home energy consumption, thus increasing the safety margin for return-to-home under these conditions.

[0127] When the drone's current battery level is low At this time, the system triggers a low battery return-to-home function. The numerator in Formula 3 is... The denominator is After unifying the dimensions of the two, a dimensionless ratio is obtained, which is then multiplied by... Converted to percentage. Compared to a fixed low battery threshold, this dynamic low battery threshold incorporates the energy consumption for retreating from the current location to the bound clearance turning point, the energy consumption for pulling up from the clearance turning point, and the energy consumption from the clearance turning point to the return point into the same judgment process, so that the low battery triggering timing matches the safe return requirements of the underpass area.

[0128] S7. Execute abnormal return control based on underpass constraint.

[0129] When a low battery return-to-home, communication failure return-to-home, mission interruption return-to-home, or manual return-to-home is triggered, the system determines whether the drone's current location or return-to-home path involves an underpass point.

[0130] If the drone's current location is not at an underpass point and the return path does not pass through an underpass point, the system controls the drone to return via the normal return path. If the drone's current location is at an underpass point, or an underpass point exists in the return path, the system determines the underpass point corresponding to the drone's current location, or the closest underpass point to the drone's current location in the return path; reads the bound clearance turning point associated with that underpass point; generates a low-altitude retreat path from the drone's current location to the bound clearance turning point; controls the drone to fly along the low-altitude retreat path, and restricts the drone's flight altitude in the low-altitude retreat path to not exceed the underpass safety altitude.

[0131] Once the drone reaches the designated airspace turning point, the system again checks whether there are any conductors, busbars, or frames within the vertical lifting column of that point. If not, the system controls the drone to lift from the designated airspace turning point to the return-to-home altitude and return to the return-to-home point; if so, the system re-executes the airspace turning point selection process in Formula 2, selects an alternative airspace turning point, and updates the return-to-home path.

[0132] In this way, when the drone triggers its return-to-home action at the underpass point or underpass passage, it will not directly perform a vertical ascent in the underpass area. Instead, it will first retreat to a clear turning point with clear airspace conditions before performing the return-to-home altitude ascent, thereby reducing the risk of collision with conductors, busbars, or structures.

[0133] S8. Generate a continued flight inspection path.

[0134] After the drone returns to base, needs a battery swap, or resumes its mission, the system generates a follow-up flight inspection path based on inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and binding relationships.

[0135] The system records the waypoints for which image acquisition has been completed, thus obtaining a set of inspected waypoints; it deletes the set of inspected waypoints from the original set of waypoints to be inspected, thus obtaining a set of uninspected waypoints; and it uses the take-off position or mission resumption position of the UAV after battery swap as the starting point for continued flight.

[0136] Based on the set of uninspected waypoints, the waypoint safety topology map, binding relationships, remaining battery power, and return-to-home safety margin, the system regenerates a continued flight inspection path that satisfies underpass constraints, no-fly detour constraints, and return-to-home energy consumption constraints. The rules for generating the continued flight path are consistent with S5, namely, first satisfying safety constraints, then satisfying energy consumption constraints, and finally selecting the path with the shorter flight distance, fewer shooting attitude changes, and fewer underpass crossing edges from the feasible paths.

[0137] Example 2:

[0138] Please see Figure 5 This embodiment provides a dynamic inspection system for substations using unmanned aerial vehicles (UAVs) based on underpass constraints, used to execute the method of Embodiment 1. The system includes a waypoint data acquisition module, a waypoint attribute marking module, a safety topology construction module, an underpass constraint configuration module, a binding relationship generation module, a dynamic path generation module, an inspection status monitoring module, an abnormal return-to-home control module, and a continued flight path generation module.

[0139] The waypoint data acquisition module is used to acquire substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data. This module also generates candidate inspection points from panoramic images using an equipment component recognition model, and generates re-flyable inspection waypoints based on the UAV's real-time pose, gimbal angle, shooting magnification, and target box position.

[0140] The waypoint attribute marking module is used to mark the attributes of inspection waypoints based on obstacle data above the waypoints and no-fly zone boundary data, resulting in ordinary inspection points, underpass points, clearance turning points, and no-fly adjacent points. This module is also used to determine underpass points based on the spatial interference between the vertical support column and the conductor, busbar, or framework, and to determine clearance turning points based on whether the vertical support column meets clearance requirements.

[0141] The safety topology construction module is used to construct a waypoint safety topology graph, with inspection waypoints as topology nodes and waypoint connections that meet the conditions of safe distance, altitude difference, and no-fly zone detour as topology edges.

[0142] The underpass constraint configuration module is used to configure underpass constraints for underpass points in the waypoint safety topology map. Underpass constraints are used to prevent UAVs from directly ascending to the return-home altitude in the vertical direction at the underpass point, and to limit the flight altitude of UAVs in the underpass passage edge to not exceed the underpass safety altitude.

[0143] The binding relationship generation module is used to match clearance turning points in the waypoint safety topology map for underpass points and establish binding relationships between underpass points and clearance turning points. This module determines the binding clearance turning points based on the vertical lifting column interference state, topological passability state, and expected energy consumption of candidate clearance turning points.

[0144] The dynamic path generation module is used to generate a dynamic inspection path in response to the selected waypoint to be inspected, based on the waypoint safety topology map, underpass constraints, and binding relationships.

[0145] The inspection status monitoring module controls the drone to perform inspection tasks according to a dynamic inspection path and monitors the drone's current location, battery status, communication status, and task status. This module also calculates a dynamic low battery threshold based on the estimated energy consumption from the drone's current location to the bound airspace turning point, the estimated energy consumption from the bound airspace turning point to the return altitude, the estimated energy consumption from the bound airspace turning point to the return point, and the return safety margin.

[0146] The abnormal return-to-home control module is used to determine whether the current position or return-to-home path of the drone involves a crossing point when a low battery return-to-home, communication abnormal return-to-home, mission interruption return-to-home, or manual return-to-home is triggered. If a crossing point is involved, the module controls the drone to fly to the clearance turning point bound to the crossing point first, and then pull up to the return-to-home altitude from the clearance turning point and return to home.

[0147] The flight path generation module is used to generate a flight path after the drone returns to base, has its battery swapped, or resumes its mission, based on inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and binding relationships.

[0148] The modules described above can be deployed in UAV flight control terminals, edge computing terminals, inspection task management platforms, or cloud servers, or they can be executed collaboratively by multiple devices. The modules interact with each other via wired communication, wireless communication, or a dedicated intra-site network.

[0149] Example 3:

[0150] Please see Figures 1-5 This embodiment uses a single The specific execution process of this invention is explained using a drone-based meter inspection task in a substation as an example. The substation is equipped with circuit breakers, disconnect switches, current transformers, voltage transformers, surge arresters, busbar structures, and meter inspection areas. The system pre-establishes a unified coordinate system within the substation, with coordinate units of [missing information]. Energy consumption unit is The unit of electricity is .

[0151] In this embodiment, the circumcircle radius of the UAV rotor is... Pick The drone uses The fusion of localization and visual localization methods has recently become a popular approach. Secondary inspection of flight positioning error percentile value Therefore, the positioning error margin Pick The on-site wind speed was attitude perturbation margin Pick Because the area to be inspected is adjacent to the busbar structure, the environmental safety margin is limited. Pick According to Formula 1, the radius of the vertically stretched column is calculated. for:

[0152] ;

[0153] Therefore, in this embodiment, the radius of the vertically stretched column is... for This radius is used to determine whether the safety envelope of the UAV's body interferes spatially with the conductor, busbar, or framework when the UAV is vertically pulled up directly at a waypoint.

[0154] After acquiring the inspection route data, the system identified a circuit breaker meter inspection route. Its waypoint coordinates are ,in The system sets the return altitude for this mission to be [height for shooting]. The return altitude was higher than the highest busbar height in the station's inspection area, and was superimposed on... Safety margin. The system is based on... Shooting height Starting from the return altitude With the endpoint as the construction radius as The vertically extending column. Obstacle data comparison revealed that this vertically extending column intersects with a section of the busbar spatial envelope, the lowest height of which is... Therefore, the system will Marked as the underpass point.

[0155] exist The system searches the surrounding area from the waypoint safety topology map. The candidate net airspace turning points are as follows: , and .in, Coordinates are , Coordinates are , Coordinates are The system is built separately. The candidate clearance turning point is a vertically raised column from the shooting height to the return height, and it is determined whether there are conductors, busbars or structures inside the column.

[0156] Based on the judgment, There are no conductors, busbars, or frames inside the vertical lifting column. ; There are no conductors, busbars, or frames inside the vertical lifting column. ; The vertically stretched columns intersect with the spatial envelope of the framework's edge. .therefore, If the conditions for increasing airspace are not met, the system will remove the entry.

[0157] Furthermore, the system makes a judgment Does a passable topological edge exist between the candidate clearance turning point and the candidate clearance turning point? and The horizontal distance between them is The height difference is The connecting lines must not cross no-fly zones and must meet the waypoint adjacency distance threshold and altitude change threshold. . and The horizontal distance between them is The height difference is The connection must not cross the no-fly zone. .therefore, and All of them entered the feasible candidate set.

[0158] In this embodiment, the fully charged rated energy of the drone battery... for Battery health for Current battery level is The drone currently has available battery energy. for The system estimates based on flight control history logs and real-time wind speed. The energy consumption required for retreat, ascent, and return are as follows: Estimated energy consumption for return flight for ,through Estimated energy consumption for return flight for Both are less than the energy currently available in batteries. This meets the energy consumption requirements for returning to base.

[0159] According to formula 2, since and All meet , and ,and Less than The system will Determined as The corresponding binding clearance turning point, and waypoint numbers and The waypoint numbers are associated with the storage.

[0160] Subsequently, the maintenance personnel selected the inspection task interface. , , , , and common There are several waypoints to be inspected. The system extracts a task sub-topology map from the waypoint safety topology map, which includes the aforementioned waypoints to be inspected, related underpass points, and bound clearance turning points. The system first eliminates paths that cross no-fly zones, violate underpass safety altitudes, and do not meet return-to-home energy consumption constraints, and then selects the path with the shortest flight distance from the remaining paths. Due to... As the underpass point, the system automatically... Insert it into the inspection path so that the drone follows... , , Entering and exiting the underpass area in a safe sequence, rather than in... The aircraft will either directly ascend or directly transition to a high-altitude return phase.

[0161] During the inspection mission performed by the drone, when the drone completes... Shooting and in to When crossing the underpass, the system continuously monitors the drone's battery level and current position. At this time, the drone's current position is... Projected energy consumption for , Expected energy consumption for climbing to return altitude for , Estimated energy consumption to return point for Due to the wind speed at the scene... The ambient temperature is Battery health is The system will have a return safety margin. Take the sum of the energy consumption of the above three segments. ,Right now .

[0162] The system calculates the dynamic low battery threshold according to Formula 3:

[0163] ;

[0164] Therefore, the system determines the dynamic low-power threshold within the current underpass constraint region. for When the drone's current battery level is below At that time, the system triggered a low battery return-to-home mechanism.

[0165] In this embodiment, the drone flies to The current battery level then dropped to Below the dynamic low battery threshold The system determines that the drone is currently in the underpass constraint area, and its current position involves an underpass point. Therefore, a direct ascent and return to base is not executed. The system reads and... Bounded clearance turning point Generate from the current position to A low-altitude withdrawal path was adopted, and the drone's flight altitude was limited to no higher than the safe lowering altitude during the withdrawal process. The drone arrived... Then, the system makes another judgment. The vertical lifting column contains no conductors, busbars, or frames, meeting the clearance requirements for lifting. Therefore, the drone is controlled by... Rise to Return to altitude and return to the return point.

[0166] After the drone's battery is swapped, the system reads the set of waypoints inspected for this mission. If , and Once image acquisition is complete, the system will remove the aforementioned points from the original set of waypoints to be inspected. From the waypoints, we obtain the set of uninspected waypoints. , and The system uses the takeoff location after battery swapping as the new starting point for continued flight. It re-accesses the waypoint safety topology map, the binding relationship between the underpass point and the clearance turning point, the remaining battery power, and the return-to-home safety margin to generate a continued flight inspection path that satisfies underpass constraints, no-fly zone detour constraints, and return-to-home energy consumption constraints. The drone then only... , and Perform follow-up inspections without repeating those already completed. , and This improves flight endurance.

[0167] As can be seen from this embodiment, the present invention does not simply perform a conventional return-to-home operation when the battery is low, nor does it generate an inspection route solely based on the shortest path. Instead, it identifies underpass points during the waypoint attribute marking stage, inserts clearance turning points during the path planning stage, determines the trigger timing through a dynamic low battery threshold during the return-to-home control stage, and controls the UAV to first retreat to the bound clearance turning point before pulling up for the return. This avoids the UAV directly colliding with conductors, busbars, or structures during vertical ascent in the underpass area, while simultaneously ensuring efficient dynamic point selection for inspection and battery swapping for continued flight.

[0168] The above embodiments provide a detailed description of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A dynamic inspection method for substations using unmanned aerial vehicles (UAVs) based on underpass constraints, characterized in that, The method includes: Acquire substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data. Based on the obstacle data above the waypoints and the no-fly zone boundary data, the inspection waypoints are marked with attributes to obtain ordinary inspection points, underpass points, clear airspace turning points and no-fly adjacent points. Using the inspected waypoints as topology nodes and the waypoint connections that meet the conditions of safe distance, altitude difference and no-fly zone detour as topology edges, a waypoint safety topology graph is constructed. In the waypoint safety topology map, an underpass constraint is configured for the underpass point. The underpass constraint is used to prevent the UAV from directly pulling up to the return altitude in the vertical direction at the underpass point. For the underpass point, match the airspace turning point in the waypoint safety topology map and establish a binding relationship between the underpass point and the airspace turning point; In response to the selected waypoint to be inspected, a dynamic inspection path is generated based on the waypoint safety topology map, the underpass constraint, and the binding relationship; Control the drone to perform inspection tasks according to the dynamic inspection path, and monitor the drone's current location, battery status, communication status and task status; When a low battery return-to-home, communication failure return-to-home, mission interruption return-to-home, or manual return-to-home is triggered, the system determines whether the drone's current position or return-to-home path involves an underpass. If an underpass is involved, the drone is controlled to fly to the clearance turning point bound to the underpass, and then rise to the return-to-home altitude from the clearance turning point and return to home. If no underpass is involved, the drone is controlled to return to home along the normal return-to-home path. After the drone returns to base, needs a battery swap, or resumes its mission, a continued flight inspection path is generated based on the inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and the aforementioned binding relationship.

2. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 1, characterized in that, The acquisition of substation inspection route data includes: Acquire panoramic images of the equipment to be inspected, collected by the drone; The equipment components and their locations in the panoramic image are identified using an equipment component recognition model, and candidate inspection points are generated. Based on the real-time pose of the UAV, the gimbal angle, the shooting magnification, and the target box position of the candidate inspection point in the panoramic image, the candidate inspection point is converted into a re-flying inspection waypoint. Based on the offset between the center of the target frame and the center of the image, and the ratio of the target frame area to the image area, the UAV pose, gimbal angle, or shooting magnification are corrected to obtain the inspection waypoint data.

3. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 1, characterized in that, The step of marking the inspection waypoints with attributes based on the obstacle data above the waypoints and the no-fly zone boundary data includes: When the vertical support column formed between the shooting height and the return height of the inspection point intersects with the conductor, busbar or structure, the inspection point is marked as the underpass point; When there are no conductors, busbars and structures in the vertical lifting column formed between the shooting height and the return height of the inspection point, and there is a passable topological edge between the inspection point and the underpass point, the inspection point is marked as the clearance turning point. When an inspection point falls within the safe distance beyond the boundary of the no-fly zone, the inspection point shall be marked as a no-fly adjacent point. When an inspection point is not marked as an underpass point, a clear zone bend point, or a no-fly adjacent point, the inspection point shall be marked as a regular inspection point.

4. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 3, characterized in that, The process of constructing a waypoint safety topology graph, using the inspected waypoints as topology nodes and waypoint connections that meet the conditions for safe distance, altitude difference, and no-fly zone detours as topology edges, includes: Calculate the horizontal distance, altitude difference, and whether the line connecting any two inspection waypoints crosses a no-fly zone; When the horizontal distance is not greater than the waypoint adjacency distance threshold, the altitude difference is not greater than the altitude change threshold, and the line connecting the two inspection waypoints does not cross the no-fly zone, a topological edge is established between the two inspection waypoints. Based on the waypoint attributes connected by the topological edge, the topological edge is marked as a regular passage edge, an underpass passage edge, a clearance enhancement edge, or a no-fly detour edge; Configure the topology edge connecting the underpass point as the underpass passage edge, and configure the topology edge connecting the clearance turning point and allowing the drone to lift as the clearance lift edge.

5. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 4, characterized in that, The step of matching the underpass point with the airspace turning point in the waypoint safety topology map and establishing a binding relationship between the underpass point and the airspace turning point includes: Using the underpass point as the search starting point, candidate clearance turning points with passable topological edges between them are obtained in the waypoint safety topology map; Determine whether the candidate clearance turning point meets the clearance lifting condition. The clearance lifting condition is that there are no conductors, busbars and frames in the vertical lifting column formed between the shooting height and the return height of the candidate clearance turning point. The binding cost of the candidate airspace turning point is calculated based on the flight distance between the candidate airspace turning point and the underpass point, the flight distance from the candidate airspace turning point to the return point, the airspace height of the candidate airspace turning point, and the return energy consumption corresponding to the candidate airspace turning point. The candidate airspace turning point with the lowest binding cost and that meets the airspace increase conditions is determined as the binding airspace turning point. The waypoint number of the underpass point is associated with and stored with the waypoint number of the bound clearance turning point.

6. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 5, characterized in that, In response to the selected waypoint to be inspected, a dynamic inspection path is generated based on the waypoint safety topology map, the underpass constraint, and the binding relationship, including: Receive the list of waypoints to be inspected; Extract a task sub-topology map from the waypoint safety topology map, which includes the set of waypoints to be inspected, underpass points, and bound airspace turning points; Calculate the path cost of each topological edge in the task sub-topological graph based on the topological edge distance, altitude change, number of shooting attitude switching, no-fly detour status, underpass passage status, and return-to-home energy consumption. Search the inspection path with the minimum sum of path costs in the task sub-topology graph; When the inspection path passes through the underpass point, the airspace turning point bound to the underpass point will be inserted into the inspection path, so that the drone can enter or exit the underpass point via the airspace turning point.

7. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 5, characterized in that, The monitoring of the drone's current location, battery status, communication status, and mission status includes: Obtain the real-time location of the UAV and map the real-time location to the current topology node or current topology edge in the waypoint safety topology graph; When the current topology node is a passing point or the current topology edge is a passing edge, it is determined that the UAV is in the passing constraint region; Based on the estimated energy consumption from the drone's current location to the bound airspace turning point, the estimated energy consumption from the bound airspace turning point to the return point, the estimated energy consumption from the bound airspace turning point to the return altitude, and the safety margin, calculate the dynamic low battery threshold. When the drone's current battery level is lower than the dynamic low battery threshold, a low battery return-to-home mechanism is triggered.

8. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 7, characterized in that, If an underpass point is involved, the drone is controlled to first fly to the clearance turning point associated with the underpass point, and then ascend from the clearance turning point to the return altitude and return, including: Determine the underpass point corresponding to the drone's current location, or the underpass point on the return path that is closest to the drone's current location; Read the bound clearance turning point associated with the underpass point; Generate a low-altitude retreat path from the drone's current position to the bound airspace turning point; Control the drone to fly along the low-altitude withdrawal path, and limit the drone's flight altitude in the low-altitude withdrawal path to not exceed the safe lowering altitude; Once the drone reaches the designated airspace turning point, control the drone to ascend from the designated airspace turning point to the return altitude and return to the return point.

9. The method for dynamic inspection of substations by UAV based on underpass constraints according to claim 8, characterized in that, After the drone returns to base, undergoes a battery swap, or resumes its mission, a continued flight inspection path is generated based on inspected waypoints, uninspected waypoints, the drone's current location, remaining battery power, and the aforementioned binding relationship. This includes: Record the waypoints where image acquisition has been completed, and obtain the set of waypoints that have been inspected; The set of inspected waypoints is deleted from the original set of waypoints to be inspected to obtain the set of uninspected waypoints. The take-off location or mission resumption location of the drone after battery swapping shall be used as the starting point for continued flight. Based on the set of uninspected waypoints, the waypoint safety topology map, binding relationships, remaining battery power, and return safety margin, the cost of the continued flight path is recalculated. Generate a continuous flight inspection path that satisfies the underpass constraint, no-fly detour constraint, and return-to-base energy consumption constraint.

10. A dynamic inspection system for substations using unmanned aerial vehicles (UAVs) based on underpass constraints, characterized in that: The system, applied to the substation UAV dynamic inspection method based on underpass constraint as described in any one of claims 1-9, comprises: The waypoint data acquisition module is used to acquire substation inspection waypoint data, which includes waypoint coordinates, corresponding equipment components, shooting attitude, gimbal angle, shooting magnification, obstacle data above the waypoint, and no-fly zone boundary data. The waypoint attribute marking module is used to mark the inspection waypoints according to the obstacle data above the waypoints and the no-fly zone boundary data, so as to obtain ordinary inspection points, underpass points, clear airspace turning points and no-fly adjacent points; The safety topology construction module is used to construct a waypoint safety topology graph using the inspected waypoints as topology nodes and the waypoint connection relationships that meet the safety distance, altitude difference and no-fly zone detour conditions as topology edges. The underpass constraint configuration module is used to configure underpass constraints for underpass points in the waypoint safety topology map. The underpass constraints are used to prevent the UAV from directly pulling up to the return altitude in the vertical direction at the underpass point. The binding relationship generation module is used to match the clearance turning point in the waypoint safety topology map for the underpass point, and establish the binding relationship between the underpass point and the clearance turning point. The dynamic path generation module is used to generate a dynamic inspection path in response to the selected waypoint to be inspected, based on the waypoint safety topology map, the underpass constraint and the binding relationship; The inspection status monitoring module is used to control the drone to perform inspection tasks according to the dynamic inspection path, and to monitor the drone's current location, battery status, communication status and task status. The abnormal return-to-home control module is used to determine whether the current position or return-to-home path of the drone involves a point of passage when low battery return-to-home, communication abnormal return-to-home, mission interruption return-to-home, or manual return-to-home is triggered. If a point of passage is involved, the module controls the drone to fly to the clearance turning point bound to the point of passage first, and then pull up to the return-to-home height from the clearance turning point and return to home. The flight path generation module is used to generate a flight path after the UAV returns to base, has its battery swapped, or resumes its mission, based on the inspected waypoints, uninspected waypoints, the UAV's current location, remaining battery power, and the binding relationship.

Images