An eVTOL vertical take-off and landing path planning and monitoring intelligent assistance system

Abstract

This invention discloses an intelligent auxiliary system for eVTOL (Electric Vertical Take-Off and Landing) path planning and monitoring, relating to the field of electric vertical take-off and landing aircraft control technology. The system comprises six modules: three-dimensional safety zone construction, obstacle detection and modeling, autonomous path generation, human-machine interaction confirmation, dynamic target early warning, and fault emergency handling. It constructs a dual-layer airspace of a core safety zone and a buffer early warning zone using a dynamic grid algorithm; achieves accurate obstacle modeling through multi-source sensor fusion and DBSCAN clustering; generates an initial path that meets energy consumption, smoothness, and wind speed adaptability using an improved A* algorithm; introduces a manual confirmation mechanism to adapt to mission requirements; predicts dynamic target trajectories based on a uniformly variable speed curve model and provides graded early warnings; and ensures safe landing through Kalman filtering, redundant control, and manual emergency paths in case of faults. This system balances automation efficiency with safety and controllability, significantly improving the safety and reliability of electric vertical take-off and landing aircraft.

Description

Technical Field

[0001] This invention relates to the field of electric vertical takeoff and landing (eVTOL) aircraft technology, specifically to an intelligent auxiliary system for eVTOL path planning and monitoring. Background Technology

[0002] With the rise of the concept of urban air mobility, eVTOL (Electric Vertical Take-Off and Landing) aircraft, with its advantages of zero emissions, low noise, vertical take-off and landing, and low dependence on traditional runways, has become an important component of the future urban three-dimensional transportation system. However, the commercial operation of eVTOL in complex urban airspace faces a series of severe technical challenges, the core of which lies in flight safety and the level of intelligence. First, the definition of three-dimensional airspace is static. Existing technologies mostly use fixed-size airspace divisions, without customizing safe zones based on dynamic parameters such as the fuselage size, flight speed, and ambient wind speed of the eVTOL, which can easily lead to wasted airspace resources or insufficient redundancy. Second, existing obstacle perception and avoidance technologies are mostly designed for ground vehicles or high-altitude aircraft, which is difficult to meet the full-dimensional needs of eVTOL urban flight. Single sensors lack reliability under complex weather conditions, varying lighting, or electromagnetic interference, and lack effective prediction and collaborative modeling of dynamic target motion trends, failing to provide accurate environmental model support for real-time path replanning. Furthermore, traditional path planning algorithms, when applied to eVTOL, often only consider the shortest path or shortest time, failing to fully integrate eVTOL's unique flight performance parameters, such as vertical climb rate, hovering capability, energy consumption characteristics, airspace rules, and flight smoothness and passenger comfort requirements. The generated paths may be too close to obstacles or dynamically infeasible, resulting in insufficient safety margins. In addition, there is an imbalance in path control autonomy; traditional systems either rely entirely on fully automated path generation or excessively on manual planning. Such manual planning is inefficient and struggles to cope with complex airspace environments. Finally, for potential power, sensor, or communication system failures during eVTOL flight, existing emergency solutions are mostly simple one-click return or forced landing, lacking an integrated safety framework encompassing dynamic airspace demarcation, accurate obstacle detection, autonomous path generation, and manual confirmation and control.

[0003] Therefore, a new intelligent auxiliary system for eVTOL vertical take-off and landing path planning and monitoring has become a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] This application provides an intelligent auxiliary system for eVTOL vertical takeoff and landing path planning and monitoring, including: a three-dimensional safety zone construction module, configured to construct a three-dimensional safety flight zone containing three-dimensional spatial boundaries based on geographic information data, airspace control rules and performance parameters of electric vertical takeoff and landing aircraft; The obstacle detection and modeling module is communicatively connected to the three-dimensional safe area construction module and is configured to detect static and dynamic obstacles in the three-dimensional safe flight area through a multi-source sensor unit and a data fusion unit, and to establish a three-dimensional obstacle model unit. The autonomous path generation module is communicatively connected to the three-dimensional safe zone construction module and the obstacle detection and modeling module. The autonomous path generation module is configured to generate an initial flight path from the takeoff point to the target point based on the boundary constraints and three-dimensional obstacle model of the three-dimensional safe flight zone and through a path optimization algorithm. The human-computer interaction confirmation module is communicatively connected to the autonomous path generation module and is configured to display the initial flight path and receive manual operation instructions, including path confirmation instructions and path adjustment instructions. The dynamic target warning module is communicatively connected to the three-dimensional safety zone construction module. The dynamic target warning module is configured to monitor external dynamic targets entering the three-dimensional safety flight zone in real time, calculate target motion parameters and collision risk values, and generate a warning signal when the collision risk value exceeds a preset threshold. The fault emergency handling module is communicatively connected to the autonomous path generation module, obstacle detection and modeling module, and human-machine interaction confirmation module, respectively. It is configured to replan an emergency alternative path based on the real-time obstacle 3D model and the performance parameters of the electric vertical take-off and landing aircraft when the electric vertical take-off and landing aircraft malfunctions, and output the alternative path through the human-machine interaction confirmation module.

[0005] Optionally, the three-dimensional safety zone construction module specifically involves: acquiring geographic information data, including terrain elevation data, building distribution data, and no-fly zone coordinate data; Receive the performance parameters of the electric vertical takeoff and landing aircraft, which include maximum flight altitude, minimum turning radius, range and fuselage size; The performance parameters of the electric vertical takeoff and landing aircraft, combined with the geographic information data, are used to construct a three-dimensional safety boundary with height layers.

[0006] Optionally, the obstacle detection and modeling module is equipped with a multi-source sensor unit; The multi-source sensor unit includes a lidar, a millimeter-wave radar, and a visual camera, used to detect obstacle data; The data fusion unit is configured to perform spatiotemporal registration and feature fusion on obstacle data collected by the multi-source sensor unit. The feature fusion includes joint estimation of obstacle position, size, material properties and motion state. The obstacle 3D model unit is configured to construct an obstacle 3D model based on the obstacle data after spatiotemporal registration and feature fusion.

[0007] Optionally, the path optimization algorithm of the autonomous path generation module is configured as an improved A* algorithm, including: A multi-objective optimization function is constructed with path length, energy consumption cost, and safety margin as optimization objectives; The improved A* algorithm is used for path search. The heuristic function expression of the improved A* algorithm is as follows:

[0008] Where g(n) is the actual cost from the starting point to the current node, h(n) is the estimated energy cost from the current node to the target point, and s(n) is the path smoothness factor. and These are the energy consumption weighting coefficient and the smoothness weighting coefficient, respectively. The following constraints are applied during the path search process: the flight path must be located within the core safety zone, the horizontal deviation must not exceed ±0.3m, and the vertical speed range is limited to 0.5-10m / s; The initial path obtained from the search is smoothed to ensure that the path curvature is continuous and does not exceed the maximum turning curvature of the electric vertical takeoff and landing aircraft.

[0009] Optionally, the human-computer interaction confirmation module includes: a display unit configured to display the initial flight path, three-dimensional safety zone and obstacle distribution in a three-dimensional visualization manner, and to support path zooming, rotation and profile viewing; The input unit is configured to receive touch operations or voice commands, wherein the touch operations include dragging and adjusting path segments and adding or removing key points; The feedback unit is configured to generate a path lock signal after receiving a path confirmation command, trigger an audible and visual prompt when no path confirmation command is received and a preset waiting time has been exceeded, and transmit the path command to the flight control module of the electric vertical take-off and landing aircraft via the CAN bus after receiving a path activation command.

[0010] Optionally, the collision risk value of the dynamic target early warning module is calculated as follows: Based on the real-time position and velocity vector of the external dynamic target, predict the target's trajectory within a preset time period in the future; Based on the real-time flight trajectory of the electric vertical takeoff and landing aircraft, calculate the minimum distance between the two and the time to reach the minimum distance; An early warning signal is generated when the minimum distance is less than the safety threshold and the time to reach the minimum distance is less than a preset time.

[0011] Optionally, the fault emergency handling module is configured such that the types of preset faults include power system faults, sensor faults, and communication link faults; When a power system failure, sensor failure, or communication link failure is detected, the remaining performance parameters of the electric vertical takeoff and landing aircraft are collected in real time. The remaining performance parameters of the electric vertical takeoff and landing aircraft include the remaining range, maximum rate of climb, and controllable attitude angle range. Based on the remaining performance parameters of the electric vertical takeoff and landing aircraft and the three-dimensional model of the obstacle, a preset alternate landing point is selected to generate an emergency alternative route. When no alternative landing point is available, a vertical forced landing path is generated.

[0012] The beneficial effects of this application are as follows: 1. This application, through a three-dimensional safety zone construction module, not only incorporates geographic information data and airspace control rules, but also deeply integrates key performance parameters of eVTOL such as maximum flight altitude, minimum turning radius, range, and fuselage dimensions. This transforms the safety flight zone from a fixed-size airspace box into a personalized three-dimensional safety boundary that can be dynamically constructed based on the aircraft's own characteristics and the real-time environment. This dynamic adaptation mechanism avoids the waste of airspace resources caused by excessive safety margins while preventing collision risks due to excessively small safety zones, achieving a balance between safety and operational efficiency. 2. This application incorporates a multi-source sensor unit and a data fusion unit. Through spatiotemporal registration and feature fusion, the system can jointly estimate the position, size, material properties, and motion state of obstacles, constructing an accurate 3D obstacle model. This technique overcomes the limitations of a single sensor under complex weather or lighting conditions, providing a stable, reliable environmental model that includes dynamic motion trends for subsequent path planning. Attached Figure Description

[0013] Figure 1 This is a flowchart of the overall system workflow. Figure 2 Flowchart for calculating the three-dimensional safety zone; Figure 3 Flowchart for obstacle detection and point cloud modeling; Figure 4 Flowchart for autonomous path generation and manual confirmation; Figure 5 This is a flowchart for emergency troubleshooting. Detailed Implementation

[0014] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some 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.

[0015] This application provides an intelligent auxiliary system for eVTOL vertical takeoff and landing path planning and monitoring, including: a three-dimensional safety area construction module, configured to construct a three-dimensional safety flight area containing three-dimensional spatial boundaries based on geographic information data, airspace control rules and electric vertical takeoff and landing aircraft performance parameters; wherein eVTOL represents electric vertical takeoff and landing aircraft. The three-dimensional safety zone construction module is specifically configured to: acquire high-precision geographic information data, including terrain elevation data, building distribution data, and no-fly zone coordinate data; Receive performance parameters of electric vertical takeoff and landing aircraft, including maximum flight altitude, minimum turning radius, range and fuselage size; The airspace control rules are based on the geographic information data and the performance parameters of the electric vertical takeoff and landing aircraft, and construct a three-dimensional safety boundary with height layers.

[0016] It should be noted that the calculation of the three-dimensional security area construction module is based on a dynamic mesh algorithm, where the dynamic mesh partitioning rules and real-time spatial state marking need to be clarified: Firstly, the dynamic mesh generation rule is as follows: based on the core takeoff and landing parameters of the electric vertical takeoff and landing aircraft, the mesh size is dynamically adapted to the airspace range. Input parameters include the takeoff and landing point coordinates: X0, Y0, Z0; and the target point coordinates: X... t Y t Z t Aircraft dimensions, including length L, width W, height H, and flight speed, where flight speed includes horizontal speed. and vertical velocity and ambient wind speed The division logic is as follows:

[0017] Vertical direction: In a typical scenario, the vertical velocity... ≤10m / s, meshed at 10m / layer; high-speed scene, where vertical velocity For speeds greater than 10m / s, such as rapid ascents / descents, the system automatically switches to 5 meters per floor to improve the detection accuracy of vertical obstacles and trajectories.

[0018] Horizontally: In conventional scenarios, a 5m x 5m grid is used; in low-altitude scenarios, the altitude Z is less than or equal to 50m, and low-altitude scenarios are the key areas for take-off and landing; in high-density airspace, the number of electric vertical take-off and landing aircraft is greater than or equal to 5 per 10,000 square meters, refined to 1m x 1m, to adapt to complex low-altitude environments and the needs of multiple aircraft taking off and landing in close proximity.

[0019] Three-dimensional safe airspace delineation: Using the line connecting the take-off and landing point and the target point as the central axis, a two-layer structure is constructed: a core safe zone and a buffer early warning zone. The buffer early warning zone extends outward by 50% from the core safe zone to detect potential threats in advance.

[0020] Where the horizontal safety clearance d= Vertical safety clearance: First, it ensures dynamic correlation with speed and fuselage size; second, the method for real-time marking of airspace status is: updating the grid status based on multi-source data to achieve dynamic airspace management. Marking types include four states: Idle, Pre-occupied, Occupied, and No-fly. Idle state refers to grids without obstacles or dynamic targets; Pre-occupied state is for grids where a target is predicted to enter within the next 5 seconds, with a trigger distance of 1.5 times the safe distance; Occupied state is for grids where electric vertical take-off and landing aircraft are currently located or covered by obstacles; No-fly state is used to mark grids with obstacles, such as tall buildings, cables, or harsh environments, where harsh environments include wind speeds >12m / s.

[0021] Update frequency: Updated synchronously with lidar data to ensure consistency with the actual environment and provide real-time spatial constraints for subsequent autonomous path generation.

[0022] The obstacle detection and modeling module is communicatively connected to the three-dimensional safe flight area construction module and is configured to detect static and dynamic obstacles in the three-dimensional safe flight area through a multi-source sensor unit and a data fusion unit, and to establish a three-dimensional obstacle model unit. The multi-source sensor unit includes a lidar, a millimeter-wave radar, and a visual camera, which detect obstacle data. The data fusion unit is configured to perform spatiotemporal registration and feature fusion on obstacle data collected by the multi-source sensor unit. The feature fusion includes joint estimation of obstacle position, size, material properties and motion state. The obstacle 3D model unit is configured to construct an obstacle 3D model based on the obstacle data after spatiotemporal registration and feature fusion.

[0023] It should be noted that the obstacle detection and modeling module provides obstacle constraints for autonomous path generation, and the method is as follows: LiDAR data acquisition: High-precision lidar is used as the core detection equipment, with the following parameters: detection range of 0-500 meters, accuracy of ±3cm, sampling frequency of 10Hz, covering the entire range of the three-dimensional safety area, and collecting point cloud data of fixed obstacles in the take-off and landing airspace in real time, such as buildings, trees, static small obstacles such as ground equipment, low-altitude cables, and dynamic obstacles such as pedestrians and other low-altitude aircraft, as the obstacle data source for autonomous path generation.

[0024] Point cloud preprocessing and cluster modeling: Preprocessing steps: First, noise points are removed using voxel mesh filtering, where the voxel mesh resolution is 0.1 m × 0.1 m to reduce data redundancy; then, time synchronization and spatial coordinate transformation are used to convert the LiDAR polar coordinates ( To convert ) to global Cartesian coordinates, the formula is: ( This aligns the point cloud data with the 3D dynamic mesh, with a total preprocessing time of less than or equal to 50ms.

[0025] DBSCAN clustering modeling: Algorithm parameters: radius threshold in meters, based on LiDAR accuracy of ±3cm; minimum number of point clouds. Filter out isolated noise points, perform point-by-point clustering on the preprocessed point cloud, and group points with a distance less than [a certain value]. The points are divided into the same cluster to generate the three-dimensional outline of the obstacle; further obstacle features are extracted, including the minimum bounding cube and the centroid coordinates (X, Y, Z). Motion vectors are calculated for dynamic obstacles to achieve accurate identification and modeling of obstacles. The clustering and feature extraction take less than or equal to 80ms, which is accelerated by GPU parallel computing, providing accurate obstacle location and motion information for autonomous path generation.

[0026] The autonomous path generation module is communicatively connected to the three-dimensional safety zone construction module and the obstacle detection and modeling module. The autonomous path generation module is configured to generate an initial flight path from the takeoff point to the target point based on the boundary constraints of the three-dimensional safety zone and the three-dimensional model of the obstacles through a path optimization algorithm. The path optimization algorithm of the autonomous path generation module is configured as an improved A* algorithm, including: A multi-objective optimization function is constructed with path length, energy consumption cost, and safety margin as optimization objectives; The improved A* algorithm is used for path search. The heuristic function expression of the improved A* algorithm is as follows: ; Where g(n) is the actual cost from the starting point to the current node, and h(n) is the estimated energy cost from the current node to the target point. For path smoothness factor, and These are the energy consumption weighting coefficient and the smoothness weighting coefficient, respectively. The following constraints are applied during the path search process: the flight path must be located within the core safety zone, the horizontal deviation must not exceed ±0.3m, and the vertical speed range is limited to 0.5-10m / s; The initial path obtained from the search is smoothed to ensure that the path curvature is continuous and does not exceed the maximum curvature of the electric vertical takeoff and landing aircraft.

[0027] The safe takeoff and landing path construction and alarm parameter calculation are carried out through a combination of autonomous generation and manual confirmation, specifically including three processes: safe takeoff and landing path construction, manual confirmation and path activation, and alarm parameter calculation. 1. Safe takeoff and landing path construction, based on an improved A* algorithm. Based on the grid state and obstacle model of the three-dimensional safety zone, an improved A* algorithm is used to plan the optimal takeoff and landing path. The core optimization directions are as follows: (1) Heuristic function design: The function expression is , in: Actual cost coefficient: Comprehensive flight energy consumption ,in Energy consumption coefficient For eVTOL quality, To improve flight speed, a penalty is added when the turning angle is greater than 15°, ensuring path economy and stability. Manhattan vertical distance: dynamically adjust weights In low-altitude scenarios, Z is less than or equal to 50 meters; weights in low-altitude scenarios. =0.8, increasing the priority of height-oriented planning; weight in high-altitude scenarios. =0.5, balancing the costs of horizontal and vertical paths; Wind speed influence factor: When the angle θ between the wind direction and the flight direction is greater than 90°, it is considered a headwind, and the weight is [not specified]. The dynamic value is increased to 0.4, which increases the cost of headwind paths and prioritizes planning paths with less tailwind or crosswind.

[0028] (2) Path constraints: The path must be located entirely within the core safety zone. The horizontal offset threshold is ±10 meters in normal scenarios and ±3 meters in high-density scenarios. The vertical speed range is 0.5-10 m / s to ensure the feasibility and safety of the path within the dynamic grid. The path generation time is less than or equal to 60 ms to meet the real-time planning requirements. The generated preliminary path is synchronously output to the human interaction interface for subsequent verification.

[0029] 2. Manual confirmation and path activation: (1) Manual verification and adjustment: Technical personnel conduct their work based on three core types of information output by the system: Basic data: 3D safe zone model, including no-fly / pre-occupancy grid, obstacle model, including position, outline, and motion vector; Autonomous path parameters: node coordinates of the preliminary path, velocity / acceleration of each segment, energy consumption data, and wind speed adaptability analysis; Mission requirements: If emergency rescue requires shortening the route time, energy consumption constraints can be appropriately relaxed; urban patrols should prioritize avoiding densely populated areas and adjust the route accordingly. Adjustments: If the initial path has a risk of being close to obstacles, such as the distance to the predicted trajectory of dynamic obstacles being less than 80% of the safe distance, or if the task is not well adapted, such as the emergency rescue path taking too long, technicians will manually adjust the path nodes, such as horizontal offset ±2 meters and vertical height ±5 meters, to ensure that the path takes into account both safety and task requirements.

[0030] (2) Path confirmation and activation: Feasibility verification: Technical personnel reviewed the adjusted path to determine whether the energy consumption was less than the maximum range of the electric vertical takeoff and landing aircraft, and whether the obstacle distance was greater than the safe clearance. Broken After confirming that the speed constraint is within the range of 0.5-10m / s, submit the path confirmation command on the system interface. Activation trigger: After technicians combine real-time airspace status, such as whether there is a sudden no-fly order or whether dynamic obstacles deviate from the predicted trajectory, and determine that the take-off and landing conditions are met, they click the path activation button. The system then transmits the final path command to the flight control module via the CAN bus. The time taken is less than or equal to 10ms, and the electric vertical take-off and landing aircraft is started.

[0031] 3. Alarm parameter calculation: Based on the flight status of the electric vertical takeoff and landing aircraft, the location of obstacles, and the trajectory of dynamic targets, the system automatically sets three types of alarm parameters, clarifies the triggering conditions and response suggestions, and outputs them to the human-interactive interface: Obstacle Alarm: The trigger condition is that an obstacle intrudes into the buffer warning zone. The parameters include the alarm distance and duration. The distance from the center of the obstacle to the boundary of the core safety zone is less than 10 meters. The response suggestion for the duration is to pay attention to the obstacle location and prepare for path fine-tuning. Collision warning: The triggering conditions are that an obstacle intrudes into the core safety zone, or the spatiotemporal grid intersection between the predicted trajectory of the dynamic target and the planned path is greater than or equal to 1, where the horizontal projection overlap area is greater than 20% of the grid area, the parameters include a warning time greater than or equal to 5 seconds and a collision probability greater than 60%, and the response suggestion is to start path replanning / adjust height / adjust speed. Emergency Alarm: The triggering conditions are sensor failure, motor failure, or a dynamic target with a predicted collision probability greater than 90% within 3 seconds. The parameters include an emergency response time of less than or equal to 0.2 seconds and a priority. The response suggestion is to activate an emergency path and push the nearest safe landing point.

[0032] The human-computer interaction confirmation module is communicatively connected to the autonomous path generation module and is configured to display the initial flight path and receive manual operation instructions, including path confirmation instructions or path adjustment instructions. The human-computer interaction confirmation module includes a display unit configured to display the initial flight path, three-dimensional safety zone and obstacle distribution in a three-dimensional visualization manner, supporting path zooming, rotation and cross-sectional viewing; The input unit is configured to receive touch operations or voice commands, wherein the touch operations include dragging and adjusting path segments or adding or removing key points; The feedback unit is configured to generate a path lock signal after receiving a path confirmation command, trigger an audible and visual prompt when no path confirmation command is received and a preset waiting time has been exceeded, and transmit the path command to the flight control module of the electric vertical take-off and landing aircraft via the CAN bus after receiving a path activation command.

[0033] The dynamic target warning module is communicatively connected to the three-dimensional safety zone construction module. It is configured to monitor external dynamic targets entering the three-dimensional safety flight zone in real time, calculate target motion parameters and collision risk values, and generate a warning signal when the collision risk value exceeds a preset threshold. The collision risk value of the dynamic target early warning module is calculated as follows: Based on the real-time position and velocity vector of the external dynamic target, predict the target's trajectory within a preset time period in the future; By combining the real-time flight trajectory of the electric vertical takeoff and landing aircraft, the minimum distance between the two and the time to reach the minimum distance are calculated. An early warning signal is generated when the minimum distance is less than the safety threshold and the time to reach the minimum distance is less than a preset time.

[0034] It is worth noting that dynamic target trend prediction and collision alarm are implemented through a combination of system alerts and manual decision-making, encompassing methods for dynamic target data acquisition and trajectory prediction, as well as collision determination and avoidance measures. First, dynamic target data acquisition and trajectory prediction: Data input: Real-time data of dynamic targets, such as other electric vertical take-off and landing aircraft, including position, speed, acceleration and current path curvature, are collected through lidar and multi-aircraft collaborative communication module and synchronously output to human interactive interface; Kinematic model prediction: Using a uniformly accelerated curvilinear motion model, the trajectory of a dynamic target within the next 5 seconds is predicted. The core formula is:

[0035]

[0036] in / The initial position of the dynamic target. / The initial velocity, / For acceleration, k is the path curvature coefficient, which is calculated based on the improved A* algorithm for path smoothness. The predicted trajectory is mapped to a spatiotemporal grid with a time window of Δt = 1 second and a cell format of 5m × 5m × 5m × 1 second. This yields the grid occupancy set of the dynamic target at each time node, which is simultaneously marked on the 3D visualization interface for manual intuitive judgment of conflict risk.

[0037] Secondly, collision assessment and avoidance measures: S1. Collision Decision Logic: The system compares the spatiotemporal grid set of the planned path of the electric vertical take-off and landing aircraft with the predicted grid set of the dynamic target. If the intersection is greater than or equal to 1 grid, that is, the horizontal projection overlap area is greater than 20% of the grid area, the corresponding alarm is triggered, and the conflicting grid and collision probability are highlighted on the interface. S2. Manual Decision-Making and Execution: Based on the avoidance suggestions provided by the system, such as a vertical grid of Z=20m with no space, suggesting height adjustment; or a horizontal grid with no space, suggesting acceleration of 0.5m / s, and considering task priorities, such as emergency rescue > logistics transportation, technicians select and confirm avoidance measures. S3. Altitude Adjustment: Adjust the altitude of the electric vertical takeoff and landing aircraft by ±5-10 meters, keep the vertical speed less than or equal to 10m / s, and stagger the dynamic target trajectory. S4. Speed ​​Adjustment: Adjust the horizontal speed, i.e., accelerate or decelerate, to change the time to reach the conflict point; S5. Path detour: Call the improved A* algorithm to regenerate the detour path. The detour distance is less than or equal to 1.2 times the original path. The technical staff will verify and confirm the execution. S6. Priority Avoidance: Low-priority eVTOLs perform avoidance to ensure high-priority tasks are protected; S7. Closed-loop feedback: After the avoidance measures are executed, the system provides real-time feedback on the attitude data of the electric vertical take-off and landing aircraft, including position, speed and acceleration. After the technicians confirm that there are no new conflicts, the normal take-off and landing process is resumed.

[0038] The fault emergency handling module is communicatively connected to the autonomous path generation module, obstacle detection and modeling module, and human-machine interaction confirmation module, respectively. It is configured to replan an emergency alternative path based on the real-time obstacle 3D model and the remaining performance parameters of the electric vertical take-off and landing aircraft when a preset fault occurs, and output the alternative path through the human-machine interaction confirmation module.

[0039] The fault emergency handling module is configured such that the preset fault types include power system faults, sensor faults, and communication link faults. When the power system failure, sensor failure and communication link failure are detected, the remaining performance parameters of the electric vertical take-off and landing aircraft are collected in real time. The remaining performance parameters of the electric vertical take-off and landing aircraft include the remaining range, maximum rate of climb and controllable attitude angle range. Based on the remaining performance parameters of the electric vertical takeoff and landing aircraft and the three-dimensional model of the obstacle, a preset alternate landing point is selected to generate an emergency alternative route. When no alternative landing point is available, a vertical forced landing path is generated.

[0040] It should be noted that emergency response is achieved through a combination of system compensation and manually led emergency response procedures, which includes system fault detection and compensation as well as the construction and activation of manually led emergency response procedures. System Fault Detection and Compensation Sensor failure: The system immediately initiates Kalman filter compensation based on historical data from the 10 seconds prior to the failure, such as position and velocity. It estimates the current state, with a position error ≤ 2m and a velocity error ≤ 0.5m / s, allowing time for manual intervention. Motor failure: The system switches to redundant control. If a single motor of the quadcopter fails, the power of the diagonal motors will be increased by 1.5 times to maintain the attitude stability of the electric vertical take-off and landing aircraft, with roll angle ≤3° and pitch angle ≤3°. At the same time, the fault type and current position will be pushed to the manual interface.

[0041] Human-led emergency response path construction and activation Data support: The system automatically pushes 3-5 nearest safe landing points around the take-off and landing point, based on the grid's idle status marking, and prioritizes the coordinates and airspace status of flat and unobstructed areas; Emergency Path Construction: Based on the fault status, such as the need to shorten the path due to sensor failure or reduce speed due to motor failure, and referring to safe landing point data, technicians manually plan emergency paths, such as "current location → emergency landing point, vertical speed reduced to 5m / s, horizontal speed reduced to 8m / s", ensuring that there are no no-fly grids along the path; Confirmation and Activation: After technicians verify the feasibility of the emergency path, they confirm and activate the command. The system controls the electric vertical take-off and landing aircraft to land safely along the emergency path, while recording fault data, including time, type, and error, for subsequent review and optimization.

[0042] In this embodiment, an emergency rescue take-off and landing mission of an electric vertical take-off and landing aircraft in an urban air traffic (UAM) scenario is taken as an example, combined with... Figures 1 to 1 This section explains the execution steps and module collaboration of the core process of autonomous path generation and manual confirmation: I. Preliminary Preparations: Please refer to... Figure 1 As shown, parameter input and system initialization; 1. Manually enter core parameters Input task parameters through the system's interactive interface: Coordinates of take-off and landing points and target points: Take-off and landing point (116.35°E, 39.92°N, 8m), target point (116.38°E, 39.95°N, 35m); Aircraft fuselage and performance parameters: Length L = 7.5m, Width W = 6.2m, Height H = 2.8m, Horizontal speed v =12m / s, vertical velocity v =8m / s; Mission requirements: Emergency rescue, highest priority, must reach the target point within 8 minutes, with appropriate relaxation of energy consumption constraints.

[0043] 2. System Data Acquisition and Initialization Environmental parameters: The system collects real-time wind speed v_wind=6m / s, the angle between wind direction and flight direction θ=60°, at this time it is a crosswind, and there is no severe weather warning; Module startup: Start the lidar, GPS, and weather sensors, and test the CAN bus communication latency to ≤10ms; load voxel filtering and DBSCAN clustering, which are GPU-accelerated and improved A* algorithm modules; detect the status of motors and sensors, and detect that the 4 motors are normal and the sensor signals are complete.

[0044] II. Pre-takeoff and landing phase: Construction of three-dimensional safety zone, obstacle modeling and autonomous path generation 1. Please refer to Figure 2 As shown, a three-dimensional security zone is constructed; Scenario determination: If the takeoff and landing altitude Z is less than or equal to 50m, it is a low-altitude scenario. The airspace density around the takeoff and landing point is greater than or equal to 3 aircraft / 10000㎡, and the grid rule is 1m×1m horizontally and 10m / layer vertically. Horizontal safety clearance calculation: d = min(10 + 0.5 × 12, 30) = 16m, d =min(5+0.2×8,15)=6.6m; Two-layer airspace: core safety zone (X±19.75m, Y±19.1m, Z±8m), buffer warning zone (X±29.6m, Y±29.6m, Z±12m); a 6-story building (22m high) 25m to the east is marked as a no-fly zone.

[0045] 2. Please refer to Figure 3 As shown, obstacle detection and point cloud modeling; Point cloud acquisition: LiDAR scans the core security zone and buffer warning zone, generating 8000 points per frame; Preprocessing: After voxel filtering, 5000 points are retained, which takes 30ms. After coordinate transformation, the points are aligned with the 3D mesh, which takes 15ms. DBSCAN clustering: Clustering is completed within 80ms, identifying 2 obstacles: fixed obstacles, such as building corners, and centroid coordinates. +25m, +18m, Z10~18m, dynamic obstacles, civilian drones, centroid coordinates ( +12m, +10m, Z15m), speed 5m / s, direction 30°, marked as no-fly zone and reserved zone respectively.

[0046] 3. Please refer to Figure 4 As shown, autonomous path generation; Algorithm calculation: The improved A* algorithm is guided by emergency rescue priority, with w1=0.8 and w2=0.3, generating the initial path: Vertical trajectory: 8m→15m→25m→35m, vertical speed 8m / s; Horizontal trajectory: along the X0→Xt main line, with a horizontal offset of +5m when avoiding buildings, and a horizontal offset of -3m 3 seconds in advance when avoiding the predicted trajectory of the drone. Path output: The system outputs the node coordinates of the preliminary path and the energy consumption of each segment. The total energy consumption is 2800J, which is less than or equal to the maximum range of 3000J. The time taken is 7 minutes and 30 seconds, which meets the 8-minute requirement.

[0047] III. Please continue to refer to Figure 1 As shown, the core steps are: manual verification, adjustment, and path activation. 1. Manual verification: Technicians can view the following in the visual interface: Safety verification: All nodes of the preliminary path are located within the core safety zone, with a minimum distance of 9m from buildings, greater than d. =50% of 16m, compared to the UAV's predicted trajectory Z=15m. The vertical distance is 10m, and the verification shows no conflict. Task compatibility: The time taken was 7 minutes and 30 seconds, which meets the requirements, and the energy consumption of 2800J is within the allowable range.

[0048] 2. Path Confirmation and Activation: Manual confirmation: Technicians click the path confirmation button to lock the path parameters; Activation trigger: Observe the real-time airspace status. If there is no sudden no-fly zone and the drone's trajectory is not deviated, click to activate the path. The system will send the path command to the flight control module. The process takes 8ms, and the electric vertical take-off and landing aircraft will start taking off.

[0049] IV. Please continue to refer to Figure 1 As shown, during takeoff and landing: dynamic early warning and manual avoidance: 1. Dynamic target trajectory prediction: The system collects real-time data (location) from drones. +12m, +10m, Z15m; velocity 5m / s, acceleration 0), the trajectory at t=3 seconds is predicted by the uniformly accelerated model: X(t)=12+4.33×3=24.99m, Z(t)=15m, and the mapped spatiotemporal unit ID is (X47+Y52+Z2+10:05:06).

[0050] 2. Collision Alarm and Manual Decision-Making: System trigger: The spatiotemporal unit ID of the planned path of the electric vertical take-off and landing aircraft (X46+Y51+Z2+10:05:06) has an intersection with the prediction unit of the UAV by more than one time, with a collision probability of 72%, triggering a collision warning. The interface suggests that the vertical direction Z=20m is an empty grid and suggests adjusting the altitude. Manual decision-making: The technician confirmed that the Z=20m grid was unobstructed, selected height adjustment, confirmed the command to raise the vertical speed to 10m / s, and raised from 15m to 20m within 3 seconds; Execution feedback: After the electric vertical takeoff and landing aircraft performed the operation, the system reported that the vertical direction Z=20m, the speed 10m / s, there were no new conflicts, and the alarm was cleared.

[0051] V. Please refer to Figure 5 As shown, an abnormal scenario: emergency handling for sensor failure. 1. System Fault Response: When the eVTOL reaches Z=25m, the lidar signal is interrupted, and the system activates Kalman filtering: State estimation: Current position coordinates are obtained based on historical data ( +80m, +60m, Z25m), with an error of 1.8m; Push notification: The interface indicates that the lidar has failed and pushes 3 safe landing points, with the nearest point's coordinates being (116.37°E, 39.94°N, 8m).

[0052] 2. Manual emergency response: Path construction: Technicians select the nearest landing point and plan an emergency path. From the current location to the landing point, the vertical speed is reduced to 5 m / s and the horizontal speed is reduced to 10 m / s. Verification Activation: If there is no no-fly zone on the verification path, click Emergency Activation. The electric vertical take-off and landing aircraft will land safely along the emergency path. The system will record fault data, such as failure at 10:05:42 and position error of 1.8m.

[0053] VI. Task Completed If no malfunction occurs, the electric vertical takeoff and landing aircraft will arrive at the target point along the planned path with a landing error of ≤1m. The system will automatically generate a mission report, including autonomous path parameters, manual confirmation records, and alarm handling logs, and the mission will end.

[0054] The above description is merely a few embodiments of this application and is not intended to limit this application in any way. Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any changes or modifications made by those skilled in the art without departing from the scope of the technical solution of this application using the disclosed technical content are equivalent to equivalent implementation cases and all fall within the scope of the technical solution.

Claims

1. An intelligent auxiliary system for eVTOL vertical takeoff and landing path planning and monitoring, characterized in that, include: The three-dimensional safety zone construction module is configured to construct a three-dimensional safety flight zone with three-dimensional spatial boundaries based on geographic information data, airspace control rules, and performance parameters of electric vertical takeoff and landing aircraft. The obstacle detection and modeling module is communicatively connected to the three-dimensional safe area construction module and is configured to detect static and dynamic obstacles in the three-dimensional safe flight area through a multi-source sensor unit and a data fusion unit, and to establish a three-dimensional obstacle model unit. The autonomous path generation module is communicatively connected to the three-dimensional safe zone construction module and the obstacle detection and modeling module. The autonomous path generation module is configured to generate an initial flight path from the takeoff point to the target point based on the boundary constraints and three-dimensional obstacle model of the three-dimensional safe flight zone and through a path optimization algorithm. The human-computer interaction confirmation module is communicatively connected to the autonomous path generation module and is configured to display the initial flight path and receive manual operation instructions, including path confirmation instructions and path adjustment instructions. The dynamic target warning module is communicatively connected to the three-dimensional safety zone construction module. The dynamic target warning module is configured to monitor external dynamic targets entering the three-dimensional safety flight zone in real time, calculate target motion parameters and collision risk values, and generate a warning signal when the collision risk value exceeds a preset threshold. The fault emergency handling module is communicatively connected to the autonomous path generation module, obstacle detection and modeling module, and human-machine interaction confirmation module, respectively. It is configured to replan an emergency alternative path based on the real-time obstacle 3D model and the remaining performance parameters of the electric vertical take-off and landing aircraft when the aircraft malfunctions, and output the alternative path through the human-machine interaction confirmation module.

2. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The three-dimensional safety zone construction module specifically involves: acquiring geographic information data, including terrain elevation data, building distribution data, and no-fly zone coordinate data; Receive the performance parameters of the electric vertical takeoff and landing aircraft, which include maximum flight altitude, minimum turning radius, range and fuselage size; The performance parameters of the electric vertical takeoff and landing aircraft, combined with the geographic information data, are used to construct a three-dimensional safety boundary with height layers.

3. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The obstacle detection and modeling module is equipped with a multi-source sensor unit; The multi-source sensor unit includes a lidar, a millimeter-wave radar, and a visual camera, used to detect obstacle data; The data fusion unit is configured to perform spatiotemporal registration and feature fusion on obstacle data collected by the multi-source sensor unit. The feature fusion includes joint estimation of obstacle position, size, material properties and motion state. The obstacle 3D model unit is configured to construct an obstacle 3D model based on the obstacle data after spatiotemporal registration and feature fusion.

4. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The path optimization algorithm of the autonomous path generation module is configured as an improved A* algorithm, including: A multi-objective optimization function is constructed with path length, energy consumption cost, and safety margin as optimization objectives; The improved A* algorithm is used for path search. The heuristic function expression of the improved A* algorithm is as follows: ； in Let h(n) be the actual cost from the starting point to the current node, and h(n) be the estimated energy cost from the current node to the target point. For path smoothness factor, and These are the energy consumption weighting coefficient and the smoothness weighting coefficient, respectively. The following constraints are applied during the path search process: the flight path must be located within the core safety zone, the horizontal deviation must not exceed ±0.3m, and the vertical speed range is limited to 0.5-10m / s; The initial path obtained from the search is smoothed to ensure that the path curvature is continuous and does not exceed the maximum turning curvature of the electric vertical takeoff and landing aircraft.

5. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The human-computer interaction confirmation module includes a display unit configured to display the initial flight path, three-dimensional safety zone and obstacle distribution in a three-dimensional visualization manner, and supports path zooming, rotation and profile viewing; The input unit is configured to receive touch operations or voice commands, wherein the touch operations include dragging and adjusting path segments and adding or removing key points; The feedback unit is configured to generate a path lock signal after receiving a path confirmation command, trigger an audible and visual prompt when no path confirmation command is received and a preset waiting time has been exceeded, and transmit the path command to the flight control module of the electric vertical take-off and landing aircraft via the CAN bus after receiving a path activation command.

6. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The collision risk value of the dynamic target early warning module is calculated as follows: Based on the real-time position and velocity vector of the external dynamic target, predict the target's trajectory within a preset time period in the future; Based on the real-time flight trajectory of the electric vertical takeoff and landing aircraft, calculate the minimum distance between the two and the time to reach the minimum distance; An early warning signal is generated when the minimum distance is less than the safety threshold and the time to reach the minimum distance is less than a preset time.

7. The eVTOL vertical takeoff and landing path planning and monitoring intelligent auxiliary system according to claim 1, characterized in that, The fault emergency handling module is configured such that the preset fault types include power system faults, sensor faults, and communication link faults. When a power system failure, sensor failure, or communication link failure is detected, the remaining performance parameters of the electric vertical takeoff and landing aircraft are collected in real time. The remaining performance parameters of the electric vertical takeoff and landing aircraft include the remaining range, maximum rate of climb, and controllable attitude angle range. Based on the remaining performance parameters of the electric vertical takeoff and landing aircraft and the three-dimensional model of the obstacle, a preset alternate landing point is selected to generate an emergency alternative route. When no alternative landing point is available, a vertical forced landing path is generated.

Images