A pre-takeoff conflict-free path planning method for low-altitude aircraft
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CIVIL AVIATION UNIV OF CHINA
- Filing Date
- 2026-03-18
- Publication Date
- 2026-06-16
Smart Images
Figure CN121905022B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of low-altitude aircraft operation safety technology, and in particular to a method for conflict-free flight path planning before takeoff for low-altitude aircraft. Background Technology
[0002] Urban Air Mobility (UAM) can effectively alleviate urban ground traffic congestion. To maximize the effectiveness of UAM, the operational safety of low-altitude aircraft must be ensured first. This involves addressing conflicts in pre-set flight paths before takeoff to guarantee the actual operational safety of low-altitude aircraft.
[0003] The future operational scenario for low-altitude aircraft will involve a mix of manned and unmanned low-altitude aircraft, with integrated operations in both urban and suburban areas. When addressing conflicts in pre-set flight paths before takeoff, it is necessary to consider the differences in the three-dimensional dynamic safety boundary dimensions of low-altitude aircraft operating in different regions using different flight modes.
[0004] Existing research, such as the Chinese invention patent application CN 120892946 A, "A method, system, terminal and storage medium for low-altitude UAV conflict detection based on trajectory prediction," ignores the differences in basic protection intervals in the horizontal and vertical directions between manned and unmanned low-altitude aircraft; it does not consider the performance differences of navigation and communication equipment in urban and suburban areas, and ignores the range of aerodynamic disturbances, resulting in an undersized three-dimensional dynamic safety boundary, which may lead to missed conflict detection and affect the practicality and accuracy of optimal adjustment commands. Meanwhile, the Chinese invention patent applications CN 119645113 A, "A method for UAV conflict resolution based on reinforcement learning and self-attention mechanism," and CN121053823 A, "A method for flight conflict resolution considering uncertainties in human-machine interaction and air-to-ground communication," only consider route conflicts and lack consideration for take-off and landing field conflicts, making the optimal adjustment commands incomplete. Summary of the Invention
[0005] This invention aims to at least solve one of the technical problems existing in related technologies. To this end, this invention provides a conflict-free flight path planning method for low-altitude aircraft before takeoff, which realizes the detection and resolution of conflicts in the preset flight path of low-altitude aircraft before takeoff, and can be used to formulate conflict-free preset flight paths and flight paths for low-altitude aircraft before takeoff.
[0006] This invention provides a method for conflict-free pre-takeoff trajectory planning for low-altitude aircraft, comprising:
[0007] S1: Calculate the basic protection interval of the low-altitude aircraft based on its performance characteristics; solve the range of aerodynamic disturbances generated by the low-altitude aircraft during flight by taking the roll moment coefficient generated by the wake vortex of the low-altitude aircraft as the critical condition that it is equal to the standard roll moment coefficient; determine the three-dimensional dynamic safety boundary of the low-altitude aircraft based on the basic protection interval of the low-altitude aircraft and the range of aerodynamic disturbances generated by the low-altitude aircraft during flight.
[0008] S2: Discretize the airspace into a regular three-dimensional grid array based on the three-dimensional dynamic safety boundary of low-altitude aircraft;
[0009] S3: Construct the pre-takeoff spatiotemporal path of each low-altitude aircraft, map the pre-takeoff spatiotemporal path of each low-altitude aircraft onto a three-dimensional grid array, and identify all pre-takeoff spatiotemporal paths with spatiotemporal overlap by analyzing the expected spatiotemporal occupancy of the low-altitude aircraft in the three-dimensional grid array.
[0010] S4: Construct a spatiotemporal path optimization algorithm and set priority decision rules for various types of low-altitude aircraft;
[0011] S5: Based on the priority decision rules for various types of low-altitude aircraft and the spatiotemporal path optimization algorithm, generate the spatiotemporal path adjustment command with the minimum comprehensive adjustment cost for the pre-set spatiotemporal path before takeoff where there is spatiotemporal overlap, and integrate all spatiotemporal path adjustment commands to generate a conflict-free flight track before takeoff.
[0012] Furthermore, the basic protection interval for low-altitude aircraft, calculated based on their performance characteristics, includes:
[0013] S111: Determine the reaction time based on the driving mode and the typical communication latency based on the operating environment;
[0014] S112: Calculate flight technical error based on the maximum vertical speed that the human body can withstand, the maximum vertical acceleration that the human body can withstand, and the passenger's reaction time to the change in acceleration.
[0015] S113: Calculate the basic horizontal protection interval for low-altitude aircraft based on navigation error, flight technical error, reaction time, typical communication delay, control center response delay, and maneuvering delay.
[0016] S114: Calculate the vertical foundation protection interval based on navigation error and flight technical error;
[0017] The basic protection interval includes a horizontal basic protection interval and a vertical basic protection interval.
[0018] Furthermore, in step S111, determining the reaction time based on the driving mode includes:
[0019] For manned flight mode, the pilot's typical reaction time is used as the reaction time;
[0020] For unmanned driving mode, the typical response delay of the flight control system is used as the reaction time;
[0021] Typical communication latency is determined based on the operating environment, including:
[0022] For urban environments, typical navigation errors in urban environments are used as typical communication delays;
[0023] For suburban environments, typical navigation errors in suburban environments are used as typical communication delays.
[0024] Furthermore, the horizontal dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the horizontal direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the horizontal direction; the vertical dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the vertical direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the vertical direction.
[0025] Furthermore, step S2 includes:
[0026] In the horizontal direction, a standard map projection method is used to map the geographic region to a Cartesian coordinate system and construct a regular planar grid array. The planar grid is iteratively subdivided at a fixed ratio until the size of the finest level planar grid unit is equal to the minimum scale of the three-dimensional dynamic safety boundary.
[0027] A height dimension is added vertically to each planar grid to form a three-dimensional grid; the height of a single three-dimensional grid is planned to be the upper limit of the spatial height. , The height of a single 3D mesh is any integer; the height of a single mesh matches the size of the finest level planar mesh on the horizontal plane.
[0028] Furthermore, constructing the pre-set spatiotemporal path before takeoff includes:
[0029] S31: Define the starting takeoff and landing field, the end point of the vertical takeoff phase, the end point of the climb phase, the end point of the cruise phase, and the starting and ending takeoff and landing fields of the vertical landing phase as key flight phase transition points and mark them as target points.
[0030] S32: Assign key state parameters to each target point for determining spatiotemporal overlap. Key state parameters include the target point's spatial coordinates, instantaneous velocity, and remaining battery power.
[0031] S33: Connect the target points according to the key status parameters to form a pre-set spatiotemporal path before takeoff.
[0032] Furthermore, in step S3, the expected spatiotemporal occupancy range of the three-dimensional mesh by the low-altitude aircraft includes the expected occupancy range of the three-dimensional mesh coordinates and the expected effective time window.
[0033] Based on the interpolation principle and the geographic coordinates of the target point, calculate the current geographic coordinates of the low-altitude aircraft.
[0034] According to the spatial mapping rules, the current geographic coordinates of the low-altitude aircraft and the three-dimensional dynamic safety boundary of the low-altitude aircraft at the current location are converted into multiple continuous three-dimensional grid coordinate ranges that the low-altitude aircraft is expected to occupy.
[0035] The expected start time of the effective time window is calculated by integrating a preset motion model from takeoff time to arrival at the current geographical location;
[0036] The expected end time of the effective time window is calculated based on the start time of the expected effective time window, taking into account the expected motion state of the current geographical location and the projected length of the three-dimensional dynamic safety boundary of the low-altitude aircraft in the motion direction.
[0037] Furthermore, the pre-set spatiotemporal path identification method for takeoff with spatiotemporal overlap is as follows: the expected spatiotemporal occupancy segments of the two low-altitude aircraft have an overlap of effective time windows on at least one three-dimensional grid.
[0038] Furthermore, the spatiotemporal path optimization algorithm includes a policy-value network, where the policy in the policy network represents the logic of the agent selecting adjustment instructions, and the value in the value network is the sum of the current reward value obtained by adopting the policy and the expected future reward value. The spatiotemporal path optimization algorithm is trained through a progressive training environment spectrum from easy to difficult. After training, the spatiotemporal path optimization algorithm generates the spatiotemporal path adjustment instructions with the minimum overall adjustment cost by comprehensively considering the take-off and landing field capacity, airspace structure, and operational rule constraints.
[0039] Furthermore, spatiotemporal path adjustment instructions include takeoff and landing site adjustment instructions and flight route adjustment instructions;
[0040] Landing site adjustment instructions include delays and circling with a waiting procedure;
[0041] Route adjustment instructions include acceleration, deceleration, climb, descent, yaw, and delay;
[0042] The comprehensive adjustment cost is a multi-objective optimization function that comprehensively considers the total change in altitude, total change in heading, total power consumption, total delay time, and total change in speed, and solves for the overall minimum value by weighted summation.
[0043] The above-described one or more technical solutions in the embodiments of the present invention have at least one of the following technical effects:
[0044] This invention calculates the three-dimensional dynamic safety boundary of low-altitude aircraft considering the aerodynamic disturbance range, and introduces the three-dimensional dynamic safety boundaries of manned and unmanned low-altitude aircraft in urban and suburban areas, improving the accuracy of the three-dimensional dynamic safety boundary dimensions of low-altitude aircraft. This invention resolves conflicts through a spatiotemporal path optimization algorithm, seeks the optimal adjustment scheme, and obtains a conflict-free flight path of the low-altitude aircraft before takeoff. It has high computational efficiency, short computation time, and strong robustness.
[0045] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description
[0046] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.
[0047] Figure 1 This is a flowchart illustrating a conflict-free flight path planning method for low-altitude aircraft before takeoff, provided by the present invention.
[0048] Figure 2 This is the convergence curve of the spatiotemporal path optimization algorithm provided by this invention.
[0049] Figure 3 This is a schematic diagram of the route adjustment command calculated according to an embodiment of the present invention.
[0050] Figure 4 This is a schematic diagram of the takeoff and landing field adjustment command calculated according to an embodiment of the present invention. Detailed Implementation
[0051] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention. The following embodiments are used to illustrate this invention but cannot be used to limit the scope of this invention.
[0052] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.
[0053] The following is combined Figures 1 to 4 This invention describes a conflict-free flight path planning method for low-altitude aircraft before takeoff.
[0054] like Figure 1 As shown, a conflict-free flight path planning method for low-altitude aircraft before takeoff includes:
[0055] S1: Calculate the basic protection interval of the low-altitude aircraft based on its performance characteristics; solve the range of aerodynamic disturbances generated by the low-altitude aircraft during flight by taking the roll moment coefficient generated by the wake vortex of the low-altitude aircraft as the critical condition that it is equal to the standard roll moment coefficient; determine the three-dimensional dynamic safety boundary of the low-altitude aircraft based on the basic protection interval of the low-altitude aircraft and the range of aerodynamic disturbances generated by the low-altitude aircraft during flight.
[0056] Based on the pilot's reaction time, the flight control system's reaction time, and the performance characteristics of low-altitude aircraft such as communication delays between urban and suburban areas and navigation errors, the basic protection interval of low-altitude aircraft is calculated.
[0057] S111: Determine the reaction time based on the driving mode and the typical communication latency based on the operating environment;
[0058] For manned flight mode, the pilot's typical reaction time is used as the reaction time;
[0059] For unmanned driving mode, the typical response delay of the flight control system is used as the reaction time;
[0060] Typical communication latency is determined based on the operating environment, including:
[0061] For urban environments, typical navigation errors in urban environments are used as typical communication delays;
[0062] For suburban environments, typical navigation errors in suburban environments are used as typical communication delays.
[0063] S112: Based on the maximum vertical speed that the human body can withstand. The maximum vertical acceleration that the human body can withstand. Passenger reaction time to changes in acceleration Calculate flight technical error; the calculation expression is:
[0064]
[0065] in, This is due to flight technical errors.
[0066] S113: Calculate the basic horizontal protection interval for low-altitude aircraft based on navigation error, flight technical error, reaction time, typical communication delay, control center response delay, and maneuver delay. The calculation expression is as follows:
[0067]
[0068] in, The basic protection interval is in the horizontal direction. For horizontal navigation error, For flight technical errors, Let be the radius of the circumscribed sphere. For low-altitude aircraft cruising speed, Typical communication delay, Due to the delay in the control center's response, During the reaction, For maneuver delay;
[0069] S114: Calculate the vertical foundation protection interval based on navigation error and flight technical error. The calculation expression is as follows:
[0070]
[0071] in, The vertical protection interval is the foundation. This represents the navigation error in the vertical direction.
[0072] Using the condition that the roll moment coefficient generated by the wake vortex of a low-altitude aircraft equals the standard roll moment coefficient as the critical condition, the range of aerodynamic disturbances generated by a low-altitude aircraft during flight is calculated, including:
[0073] S121: Calculate the rolling moment coefficients of low-altitude aircraft generated by wake vortices in the horizontal and vertical directions. The calculation expression is:
[0074]
[0075]
[0076] in, This is the rolling moment coefficient of a low-altitude aircraft in the horizontal direction generated by its wake vortex. For the weight of low-altitude aircraft, Gravitational acceleration, These are atmospheric turbulence parameters. This refers to the range of aerodynamic disturbances generated horizontally by low-altitude aircraft. air density, For low-altitude aircraft cruising speed, The total area of all rotor blades, The rotor radius is... This refers to the rotor speed of a low-altitude aircraft during cruise. This is the rolling moment coefficient of a low-altitude aircraft in the vertical direction caused by its wake vortex. This refers to the range of aerodynamic disturbances generated vertically by low-altitude aircraft.
[0077] S122: Taking the horizontal and vertical roll moment coefficients equal to the standard roll moment coefficient as the critical state, solve the range of aerodynamic disturbances generated by low-altitude aircraft in the horizontal and vertical directions during flight.
[0078] when and Greater than the standard rolling moment coefficient At this time, the wake of the preceding aircraft will pose a significant safety threat to the following aircraft. The critical state is defined as the horizontal and vertical roll moment coefficients equaling the standard roll moment coefficient. Substituting this into the formula in step S121, we obtain... as well as .
[0079] The horizontal dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the horizontal direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the horizontal direction; the vertical dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the vertical direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the vertical direction.
[0080] In some specific embodiments of the present invention, two types of low-altitude aircraft, A and E, are selected, with parameter values shown in Table 1, and environmental parameters such as navigation error and communication delay shown in Table 2. Based on existing research... The value is set to 0.3. Based on the parameters in Tables 1 and 2, the dimensions of the three-dimensional dynamic safety boundary of type A and type E low-altitude aircraft in the horizontal and vertical directions are calculated as shown in Table 3.
[0081] Table 1. Parameter values for Type A and Type E low-altitude aircraft
[0082]
[0083] Table 2 Environmental Parameter Values
[0084]
[0085] Table 3 Three-dimensional dynamic safety boundary dimensions for Type A and Type E low-altitude aircraft
[0086]
[0087] S2: Discretize the airspace into a regular three-dimensional grid array based on the three-dimensional dynamic safety boundary of low-altitude aircraft;
[0088] This is achieved by establishing a multi-level discrete spatial structure, including:
[0089] S21: In the horizontal direction, a standard map projection method is used to map the geographical region to a Cartesian coordinate system, and a regular planar grid array is constructed. The planar grid is iteratively subdivided at a fixed ratio until the size of the finest level planar grid cell is equal to the minimum scale of the 3D dynamic safety boundary; the hierarchy of the finest level planar grid... Size of the finest planar grid cell on the horizontal plane The relationship that satisfies this is:
[0090]
[0091] A regular planar grid array serves as the discrete framework for subsequent track collision detection.
[0092] S22: Add a height dimension to each planar grid in the vertical direction to form a three-dimensional grid; the height of a single three-dimensional grid is planned as the upper limit of the spatial height. , The height of a single 3D mesh is any integer; the height of a single mesh matches the size of the finest level planar mesh on the horizontal plane.
[0093] In some specific embodiments of the present invention, according to Table 3, the minimum value of the three-dimensional dynamic safety boundary dimensions of low-altitude aircraft in the horizontal and vertical directions is 11.16 m. Therefore, the size of the finest planar grid cell at the horizontal level is... The order of magnitude should be 1. The finest level is four. The upper limit of the usable low-altitude airspace is 64 m. Set to 1000 m. Height of a single 3D mesh. Should be with They are of the same order of magnitude, both being 1, therefore It is 40. The final size of a single Level 4 airspace grid is 64m × 64m × 25m.
[0094] S3: Construct the pre-takeoff spatiotemporal path of each low-altitude aircraft, map the pre-takeoff spatiotemporal path of each low-altitude aircraft onto a three-dimensional grid array, and identify all pre-takeoff spatiotemporal paths with spatiotemporal overlap by analyzing the expected spatiotemporal occupancy of the low-altitude aircraft in the three-dimensional grid array.
[0095] Constructing the pre-set spatiotemporal path before takeoff includes:
[0096] S31: Define the starting takeoff and landing field, the end point of the vertical takeoff phase, the end point of the climb phase, the end point of the cruise phase, and the starting and ending takeoff and landing fields of the vertical landing phase as key flight phase transition points and mark them as target points.
[0097] The starting takeoff and landing field, the end point of the vertical takeoff phase, the end point of the climb phase, the end point of the cruise phase, the starting point of the vertical landing phase, and the end takeoff and landing field are defined as key flight phase transition points, and are marked as target points Q1 to Q6 in sequence.
[0098] S32: Assign key state parameters to each target point for determining spatiotemporal overlap. Key state parameters include the target point's spatial coordinates (longitude, latitude, altitude), instantaneous velocity, and remaining battery power.
[0099] Define the latitude and longitude coordinates of target point Q1 ( , ) and the latitude and longitude coordinates of the starting and landing fields ( , The instantaneous velocity of Q1 is the same. The height of Q1 All values are zero; the remaining battery capacity in Q1 is equal to the total battery capacity. Define the latitude and longitude coordinates of target point Q6. , ) and the latitude and longitude coordinates of the final take-off and landing site ( , The instantaneous velocity of Q6 is the same. The height of Q6 All values are zero; the remaining battery charge at target Q6 is zero. Based on energy model predictions, the remaining capacity of the target Q6 battery is estimated. Total battery capacity After deducting the total energy consumption for each phase of vertical takeoff, climb, cruise, descent, and vertical landing, the calculation expression is:
[0100]
[0101] in, This refers to the battery discharge power during the vertical takeoff phase. This refers to the battery discharge power during the climb phase. This refers to the battery discharge power during the cruise phase. This refers to the battery discharge power during the descent phase. This refers to the time required for the vertical takeoff phase. This refers to the time required for the ascent phase. The time required for the descent phase. The total length of the route. This represents the ground distance traversed during the ascent phase. This represents the distance traveled across the ground during the descent phase. This refers to the cruising speed of low-altitude aircraft.
[0102] The takeoff process, which includes the vertical takeoff phase and the climb phase, and the landing process, which includes the vertical landing phase and the descent phase, are respectively regarded as uniformly accelerated and uniformly decelerated motions; the landing process is regarded as the reverse process of the takeoff process.
[0103] , , The calculation expressions are as follows:
[0104]
[0105]
[0106] in, This refers to the height during the vertical takeoff and landing phase. The acceleration during takeoff and landing. Cruise altitude This refers to the climb angle or glide angle during the climb and descent phases. Since descent is the reverse of takeoff, the climb and glide angles are equal during both the climb and descent phases; the magnitudes of acceleration are equal during both the climb and descent phases; and the altitudes during vertical takeoff and vertical landing are equal.
[0107] , , , The calculation expressions are as follows:
[0108]
[0109]
[0110]
[0111]
[0112] in, For the magnitude of resistance, For rotor motor efficiency, The figure of merit coefficient, The rotor disk area,
[0113] The calculation expression is:
[0114]
[0115] in, The parasitic drag coefficient, For aspect ratio, is the lift coefficient, and O is the Ostwald coefficient.
[0116] Define the latitude and longitude coordinates of target point Q2 ( , ) and the latitude and longitude coordinates of the starting and landing fields ( , The height of Q2 is the same. From the vertical takeoff stage altitude Confirmed. Instantaneous velocity of target point Q2. From vertical takeoff and landing time and acceleration during takeoff and landing Jointly determined. The remaining battery capacity at target Q2. Total battery capacity Excluding the total energy consumption during the vertical takeoff phase. Define the latitude and longitude coordinates of target point Q5 ( , ) equals the latitude and longitude coordinates of the final take-off and landing field ( , Q5 height From the vertical takeoff stage altitude Confirmed. Treating the landing process as the reverse of the takeoff process, the instantaneous velocity of target point Q5 is... instantaneous velocity of target Q2 Equal, based on vertical takeoff and landing time and acceleration during takeoff and landing Calculate the remaining battery charge at target Q5. Total battery capacity Excluding the total energy consumption during vertical takeoff, climb, cruise, and descent phases. , , , The calculation formulas are as follows:
[0117]
[0118]
[0119] The climb phase endpoint and the cruise phase endpoint are defined as target point Q3 and target point Q4, respectively, and their latitude and longitude coordinates are determined according to the following principles: Based on the total flight distance of the mission, a preset horizontal range budget is allocated for each of the climb and descent phases; the climb phase endpoint is located at the horizontal range budget of the climb phase flight along the planned heading from the starting point / landing field. The final position; the cruise phase endpoint is located at the horizontal range budget of the descent phase of the flight against the planned heading from the endpoint takeoff and landing field. Location;
[0120] The latitude and longitude coordinates of Q3 and Q4 need to be determined based on the Earth's spherical geometric model, using the coordinates of the starting and ending takeoff and landing fields, the planned heading, and the allocated flight budget. The calculation expression is as follows:
[0121]
[0122] in, The longitude of target point Q3, The target point is at latitude Q3; The longitude of target point Q4 The latitude of target point Q4 The longitude of the starting and landing field. The latitude of the starting and landing field. The longitude of the final take-off and landing site. The latitude of the final take-off and landing site. This refers to the azimuth angle from the starting take-off and landing field to the ending take-off and landing field, measured clockwise from due north. This is the azimuth angle calculated clockwise from due north from the final landing field to the starting landing field. For the distance from the starting point take-off and landing field The corresponding large corner radius, For the distance from the terminal take-off and landing field Corresponding large corner radius; , , , The calculation formulas are as follows:
[0123]
[0124]
[0125]
[0126] .
[0127] Height of target Q3 Height of target Q4 All are cruising altitudes The instantaneous velocity of target Q3 instantaneous velocity of target Q4 Cruise speed of low-altitude aircraft The remaining battery capacity at target Q3 Total battery capacity After deducting the total energy consumption during vertical takeoff and climb, the remaining battery capacity at target Q4 is... Total battery capacity Excluding the total energy consumption during vertical takeoff, climb, and cruise phases. and The calculation expression is:
[0128]
[0129] .
[0130] S33: Connect the target points according to the key status parameters to form a pre-set spatiotemporal path before takeoff.
[0131] The pre-flight spatiotemporal paths of each low-altitude aircraft are mapped onto a unified discretized 3D mesh array. By analyzing the expected spatiotemporal occupancy of the 3D mesh by the low-altitude aircraft, all pre-flight spatiotemporal paths with spatiotemporal overlap are identified. The expected spatiotemporal occupancy of the 3D mesh by the low-altitude aircraft includes the expected occupancy range of the 3D mesh coordinates and the expected effective time window.
[0132] Based on the interpolation principle and the geographic coordinates of the target point, calculate the current geographic coordinates of the low-altitude aircraft.
[0133] According to spatial mapping rules, the current geographic coordinates of the low-altitude aircraft and the three-dimensional dynamic safety boundary of the low-altitude aircraft at its current location are converted into multiple continuous three-dimensional grid coordinate ranges that the low-altitude aircraft is expected to occupy. The coordinate range of the three-dimensional grids that the low-altitude aircraft is expected to occupy on the x-axis is... and the range of coordinates on the y-axis The calculation expressions are as follows:
[0134]
[0135]
[0136] in, The current longitude coordinates of the low-altitude aircraft; The current latitude coordinates of the low-altitude aircraft. For the hierarchy of planar grids, The dimension of the three-dimensional dynamic safety boundary of low-altitude aircraft in the horizontal direction. This refers to the size of the finest planar grid cell at the horizontal level. This is for rounding down.
[0137] The expected occupancy of the 3D mesh by low-altitude aircraft in the coordinate range of the z-axis. Through the first Low-altitude aircraft on the flight path Current height Divide by the height of a single 3D mesh The result is obtained by rounding down to the nearest integer. The calculation expression is:
[0138]
[0139] in, This refers to the vertical dimension of the three-dimensional dynamic safety boundary for low-altitude aircraft.
[0140] Expected start time of the effective time window It is calculated by integrating a preset motion model from takeoff time to arrival at the current geographical location; it needs to be calculated according to the current flight phase, and the calculation expression is:
[0141]
[0142] in, For the takeoff time of low-altitude aircraft This refers to the distance of a low-altitude aircraft from its origin and landing field. The climb angle or glide angle during the climb and descent phases.
[0143] Based on the latitude and longitude coordinates of the origin and landing field of low-altitude aircraft ( , ) Calculate, the calculation expression is:
[0144]
[0145] Expected end time of the effective time window Based on the start time of the expected effective time window, the calculation is performed according to the expected motion state at the current geographical location and the projected length of the low-altitude aircraft's three-dimensional dynamic safety boundary in the direction of motion. This calculation needs to be based on the current flight phase, and the expression is:
[0146] Climbing phase:
[0147] Cruise phase:
[0148] in, This is the current speed of the low-altitude aircraft;
[0149] The calculation expression is:
[0150] Climbing phase:
[0151] Cruise phase:
[0152] The pre-set time-space path before takeoff with spatiotemporal overlap is identified by the following method: the expected time-space occupancy segments of the two low-altitude aircraft have an overlap of effective time windows on at least one three-dimensional grid.
[0153] In some specific embodiments of the present invention, this embodiment, based on EASA research, defines the height of the vertical takeoff and landing phase. Set to 30 m, cruising altitude The ascent and descent angles for the 300 m ascent and descent phases. The angle is 7.125°. The air density is 1.21 kg / m³; the acceleration due to gravity g is 10 m / s². 2 Based on the above parameters and the parameter values for Type A and Type E low-altitude aircraft in Table 1, conflicts in the pre-flight preset paths established in existing studies were detected. These pre-flight preset paths involved 177 Type A low-altitude aircraft and 170 Type E low-altitude aircraft. The two types of low-altitude aircraft were numbered separately according to their respective categories: Type A low-altitude aircraft were numbered A-001~177, and Type E low-altitude aircraft were numbered E-001~170. Type A low-altitude aircraft numbered A-001~040 and Type E low-altitude aircraft numbered E-131~170 were unmanned; the rest were manned. A total of 194 spatiotemporal overlaps were detected in the preset spatiotemporal paths. The aircraft numbers, 3D grid coordinates, spatiotemporal overlap start time, spatiotemporal overlap end time, and piloting methods involved in four of these overlaps are shown in Table 4.
[0154] Table 4. Information related to the four conflicts
[0155]
[0156] S4: Construct a spatiotemporal path optimization algorithm and set priority decision rules for various types of low-altitude aircraft;
[0157] To intelligently adjust pre-set spatiotemporal paths with spatiotemporal overlaps identified during takeoff, a spatiotemporal path optimization algorithm is constructed. This algorithm aims to generate spatiotemporal path adjustment commands with the lowest overall adjustment cost through simulation learning, under the constraints of takeoff and landing field capacity, airspace structure, and operational rules. Before implementation, the spatiotemporal path optimization algorithm requires configuring a progressive training environment hierarchy, designed according to the following configuration logic: starting with a small number of simulated flights and a large static 3D safety boundary, gradually transitioning to high-density simulated flights and refined 3D dynamic safety boundaries; starting with simple flight missions containing fewer target points, gradually transitioning to complex flight missions containing target points, and progressively introducing static obstacle constraints such as simulated buildings and airspace restricted areas.
[0158] The spatiotemporal path optimization algorithm includes a policy-value network. In the policy network, the policy represents the logic by which the agent selects and adjusts instructions, and in the value network, the value is the sum of the current reward obtained using the policy and the expected future reward. Each training environment consists of M training rounds, and each training round includes T time series. Training begins with a simulation environment containing a small number of simulated flights and a large static three-dimensional safety boundary, initializing the policy-value network. At the start of each training round, the spatiotemporal path optimization algorithm uses the current policy network. The system interacts with the simulation environment, and the interaction results include adjustment instructions for the policy network selection, the spatiotemporal overlap state of the spatiotemporal path after adjustment, and the reward value after adjustment. B samples are collected from the interaction results, each sample including the interaction result of one time series from T time series. Subsequently, the value network parameters are updated to accurately guide the policy network parameter update direction. The value network parameters are updated by minimizing the mean squared error and using gradient descent. The calculation expression is:
[0159]
[0160] in, To minimize the gradient of the mean square error, The learning rate of the value network; For old value network parameters, For new value network parameters, To minimize the mean square error;
[0161] The calculation expression is:
[0162]
[0163] in, For sample b in time series status The predicted value of the lower value network, To obtain the expected value, For sample b in time series status The target value below;
[0164] The calculation expression is:
[0165]
[0166] in, For time sequence The reward value at that time, For sample b in time series status The value network prediction value is below. This is the discount factor.
[0167] The calculation expression is:
[0168]
[0169] in, For time sequence The reward value at that time, For future time series Relative to the current time series The offset;
[0170] After the value network parameters are updated, time series calculations are performed using the value network. Generalized advantage estimation at time As a guiding signal for updating the policy network parameters, the calculation expression for generalized dominance estimation is:
[0171]
[0172] in, For parameters of generalized dominance estimation, For future time series The difference error is calculated using the following expression:
[0173]
[0174] in, For future time series The difference error.
[0175] The strategy network parameters are updated using a pruning mechanism and gradient ascent method. The calculation expression is:
[0176]
[0177] in, Network parameters for the old strategy The gradient after clipping, To update the value network parameters, The learning rate of the policy network. The parameters for the old strategy network after pruning;
[0178] The calculation expression is:
[0179]
[0180] Where B is the sample size. The clipping factor is 'clip', and the clipping mechanism is 'clip', which puts the old and new policy networks of sample b in the same state. Select the same adjustment command below The probability ratio Limited to Within the range of values, For sample b, the old and new policy networks are in the same state. Select the same adjustment command below The probability ratio is calculated using the following expression:
[0181]
[0182] in, For the new policy network in the same state Select the same adjustment command below The probability, For the old policy network in the same state Select the same adjustment command below The probability of.
[0183] After the policy network and value network are updated, the spatiotemporal path optimization algorithm stores the training results for each time series and switches to the next training round to continue training. When the spatiotemporal path optimization algorithm obtains the optimal adjustment instruction in one training environment spectrum, it switches to the next training environment spectrum. After completing the four stages of training, real data including the location of the airspace restricted area and the pre-set spatiotemporal path before takeoff is input into the trained agent to obtain the optimal adjustment instruction for the pre-set spatiotemporal path before takeoff under real operating conditions.
[0184] The post-training spatiotemporal path optimization algorithm, taking into account the takeoff and landing field capacity, airspace structure, and operational rule constraints, generates the spatiotemporal path adjustment command with the lowest overall adjustment cost.
[0185] In some specific embodiments of the present invention, the values of various parameters of the spatiotemporal path optimization algorithm are shown in Table 5.
[0186] Table 5. Parameter values for the spatiotemporal path optimization algorithm
[0187]
[0188] Based on the parameter values in Table 5, the convergence curve of the spatiotemporal path optimization algorithm is as follows: Figure 2 The average computation time for the spatiotemporal path optimization algorithm is 73.88 seconds. For example... Figure 2 As shown in Figure (a), the algorithm exhibits a high loss value in the initial stage, indicating that the agent is exploring and trying various adjustment instructions. As training progresses, the loss value decreases, while the average reward value increases and eventually stabilizes, as... Figure 2 As shown in Figure (b), this trend indicates that the agent is gradually learning more effective adjustment instructions. The average reward and loss values converge after approximately 100 training epochs, demonstrating the high computational efficiency of the spatiotemporal path optimization algorithm. The absence of severe oscillations or collapses in the average reward and loss values highlights the enhanced robustness of the spatiotemporal path optimization algorithm to its parameters and training environment, confirming the successful stable convergence of the algorithm.
[0189] The priority decision rules for each type of low-altitude aircraft are as follows: unmanned low-altitude aircraft give way to manned low-altitude aircraft; if the piloting methods are the same, low-altitude aircraft in the landing phase give way to low-altitude aircraft taking off; if the takeoff and landing phases are the same, smaller low-altitude aircraft give way to larger ones; if all the above conditions are the same, low-altitude aircraft with larger serial numbers give way to low-altitude aircraft with smaller serial numbers; the numbering method for low-altitude aircraft is as follows: each type of low-altitude aircraft is individually numbered according to its own category, starting with the number 001.
[0190] S5: Based on the priority decision rules for various types of low-altitude aircraft and the spatiotemporal path optimization algorithm, generate the spatiotemporal path adjustment command with the minimum comprehensive adjustment cost for the pre-set spatiotemporal path before takeoff where there is spatiotemporal overlap; integrate all spatiotemporal path adjustment commands to generate a conflict-free flight track before takeoff.
[0191] Spatiotemporal path adjustment instructions include takeoff and landing site adjustment instructions and flight route adjustment instructions;
[0192] Landing site adjustment instructions include delays and circling with a waiting procedure;
[0193] The hovering time is determined by the hovering speed. With the radius of rotation The calculation expression is:
[0194]
[0195] in, For the time it takes for the program to hover, The number of rotations is any positive number.
[0196] Spiral speed Based on the dynamic formula of the takeoff and landing process, the calculation expression is as follows:
[0197]
[0198] in, The hovering height is for waiting programs.
[0199] hovering radius Based on the downwash diffusion distance during low-altitude aircraft takeoff and landing The ascent and descent angles during the ascent and descent phases The calculation expression is:
[0200]
[0201] Downwash diffusion distance during low-altitude aircraft takeoff and landing The calculation expression is:
[0202]
[0203] in, For the number of rotors of low-altitude aircraft, The diffusion velocity of the downwash stream;
[0204] The calculation expression is:
[0205]
[0206] Route adjustment instructions include acceleration, deceleration, climb, descent, yaw, and delay;
[0207] Applying any route adjustment command will affect the spatiotemporal attributes, takeoff time, and instantaneous speed of the low-altitude aircraft's current position. The low-altitude aircraft's current longitude... ,latitude ,high ,speed and departure time Adjustments for longitude, latitude, altitude, speed, and time , , , , Adding them together gives the adjusted takeoff time of low-altitude aircraft. With longitude ,latitude ,high ,speed Departure time .
[0208] The comprehensive adjustment cost is a multi-objective optimization function that comprehensively considers the total change in altitude. Total change in heading Total power consumption Total delay time Total change in velocity The overall minimum value is solved using a weighted summation method; the calculation expression is:
[0209]
[0210] in, In order to comprehensively adjust the costs, As a weighting factor for the total change in height, As a weighting factor for the total change in heading, As a weighting factor for total power consumption, As a weighting factor for total delay time, As a weighting factor for the total change in velocity, Eliminate the reward value given for all overlapping time-space situations. The penalty value assigned to a pre-takeoff spatiotemporal path that violates any constraint for a final conflict-free takeoff.
[0211] From all feasible combinations of adjustment instructions, the optimal solution is selected as the one that satisfies all constraints and has the lowest overall adjustment cost.
[0212] The time delay is calculated by having low-altitude aircraft use a waiting procedure to circle and resolve the conflict; a reward is given if all spatiotemporal overlaps are eliminated, and a penalty is given if the final conflict-free pre-takeoff spatiotemporal path violates any constraint.
[0213] The formulas for calculating the total change in altitude, total change in heading, total power consumption, total delay time, and total change in speed are as follows:
[0214]
[0215]
[0216]
[0217]
[0218]
[0219] in, For the change in height, This represents the change in longitude. To delay the change in time, For speed changes For the first Low-altitude aircraft on the flight path Total power consumption To adjust the latitude of the low-altitude aircraft;
[0220] The calculation expression is:
[0221]
[0222] in, For rotor motor efficiency, This is the total length of the cruise phase of the conflict-free spatiotemporal path after adjustment. Budget for the horizontal range during the climb phase; For the horizontal range budget during the descent phase, The radius of rotation is given, and the rotation process is a uniform motion.
[0223] Takeoff time of low-altitude aircraft after optimal adjustment instructions No less than the start of operation time It must not exceed the end of the operating time. Adjusted altitude of low-altitude aircraft It must not exceed the upper limit of the airspace for low-altitude aircraft. and the maximum cruising altitude of low-altitude aircraft The calculation expression is:
[0224]
[0225]
[0226] In some specific embodiments of the present invention, , , , , , , Set the air density to -0.8, -0.4, -0.2, -0.5, -0.1, 400, and -300 respectively. Set at 1.21 kg / m 3 The parameters for low-altitude aircraft are shown in Table 1. The hovering altitude for the waiting procedure will be determined later. Set at 152 m, the ascent and descent angles during the ascent and descent phases. Set to 7.125°, cruising altitude for low-altitude aircraft The value is 300 m. Using the above parameters and the low-altitude aircraft parameters in Table 1, the turning radius in this embodiment is calculated to be 1373.56 m.
[0227] In this embodiment, the four adjustment instructions listed in Table 4 are as follows: Figure 3The adjustment instructions for P1 and P2 in Table 4 are as follows: Figure 3 As shown. Figure 3 Figure (a) shows the adjusted flight paths of A-001 and A-002, with stars indicating the original conflict positions of the two unmanned A-001 and A-002 aircraft. Figure 3 As shown in Figure (a), the two unmanned aircraft, A-001 and A-002, overlapped in time and space at a cruising altitude of 300 m. According to the priority decision rule, unmanned aircraft A (numbered A-002) deviated its course 3.5° to the left, while A-001 maintained its original trajectory. Figure 3 As shown in Figure (b), A-002 began adjusting its course 30 seconds before the spatiotemporal overlap occurred (6:55:48.51). After the adjustment, the relative distance between the two low-altitude aircraft increased to over 2000 m, thus avoiding a collision. A-002 then resumed its original course. After the adjustment, the relative distance between the two low-altitude aircraft remained greater than the horizontal three-dimensional dynamic safety boundary dimension of 1059.9 m for A.
[0228] Figure 3 In Figure (c), a manned aircraft E (serial number E-002) takes off at 6:50:00, enters the cruise phase at 6:51:00 (cruise altitude 300m), and overlaps with unmanned aircraft A (serial number A-040) in time and space at 6:51:20. Although A is larger, according to the priority decision rules for unmanned aircraft to avoid manned aircraft, such as... Figure 3 As shown in Figure (d), A-040 began decelerating at 6:51:00 and accelerated to its original cruise speed at 6:53:10. During the acceleration phase to its original cruise speed, the aircraft began decelerating for landing, decreasing its altitude. Due to the initial deceleration, A-040's originally scheduled landing time of 6:53:30 was delayed to 6:56:00.
[0229] The solutions for P3 and P4 in Table 4 are as follows: Figure 4 The adjusted flight path is as follows: Figure 4 In Figure (a), at 7:35:52:47, E (serial number E-006) in the takeoff phase and A (serial number A-044) in the landing phase collided within the terminal airspace of landing field C. According to the priority decision rule, A-044 in the landing phase adopted a waiting procedure and circled until E-006 left landing field C. Figure 4 As shown in Figure (a), A-044 descended from a cruising altitude of 300 m to 152 m and began circling and waiting. During the waiting period, E-006 took off, and the relative distance between the low-altitude aircraft gradually decreased; after it left the takeoff and landing field C, the relative distance increased again, and A-044 continued its descent, lowering its altitude. Figure 4As shown in Figure (b), the circling and waiting adjustment command ensured that the two low-altitude aircraft maintained a relative distance of more than 2,500 m, and the spatiotemporal overlap was successfully resolved.
[0230] Figure 4 In Figure (c), two manned aircraft (A-001 and A-013) simultaneously took off from landing field F at 6:55:44.38, resulting in a temporal overlap. Since both are unmanned A-type low-altitude aircraft, according to the priority decision rule, A-013 was delayed while A-001 proceeded with its normal takeoff. Figure 4 As shown in Figure (d), after the delay, A-001 took off first and reached a cruising altitude of 300 m, increasing the relative distance between the two aircraft to 2500 m. Subsequently, A-013 took off, and the relative distance between the two low-altitude aircraft decreased but remained greater than 1059.9 m, confirming that the adjustment command successfully resolved the current spatiotemporal overlap without triggering a new spatiotemporal overlap.
[0231] In this embodiment, the delay command is the primary adjustment command, while the rerouting command is used only sparingly. The reason why the delay command is the primary adjustment command is that altitude, deceleration, and heading adjustments all consume additional power, while the delay command can not only successfully resolve the spatiotemporal overlap but also avoid consuming additional power. Therefore, the reinforcement learning agent prioritizes the delay command.
[0232] In planning conflict-free pre-takeoff flight paths, the mixed operation of manned and unmanned low-altitude aircraft must be fully considered; the performance differences of communication, navigation, and surveillance equipment in urban and suburban areas must also be taken into account. The three-dimensional dynamic safety boundary dimensions of the low-altitude aircraft are calculated, and a discretized three-dimensional mesh array is established based on these dimensions. The preset spatiotemporal path is imported into the discretized three-dimensional mesh array to identify conflicts. A spatiotemporal path optimization algorithm is designed and used to resolve conflicts, resulting in a conflict-free pre-takeoff flight path for the low-altitude aircraft. This invention achieves the detection and resolution of conflicts in the preset pre-takeoff flight path of low-altitude aircraft and can be used to formulate conflict-free preset and actual pre-takeoff flight paths for low-altitude aircraft.
[0233] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A conflict-free flight path planning method for low-altitude aircraft before takeoff, characterized in that, include: S1: Calculate the basic protection interval for low-altitude aircraft based on their performance characteristics. The range of aerodynamic disturbances generated by low-altitude aircraft during flight is solved by taking the roll moment coefficient generated by the wake vortex of the low-altitude aircraft as equal to the standard roll moment coefficient as the critical condition; the three-dimensional dynamic safety boundary of the low-altitude aircraft is determined based on the basic protection interval of the low-altitude aircraft and the range of aerodynamic disturbances generated by the low-altitude aircraft during flight. The horizontal dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the horizontal direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the horizontal direction; the vertical dimension of the three-dimensional dynamic safety boundary of a low-altitude aircraft is equal to the maximum value of the basic protection interval distance of the low-altitude aircraft in the vertical direction and the range of aerodynamic disturbances generated by the low-altitude aircraft in the vertical direction. S2: Discretize the airspace into a regular three-dimensional grid array based on the three-dimensional dynamic safety boundary of low-altitude aircraft; S3: Construct the pre-takeoff spatiotemporal path of each low-altitude aircraft, map the pre-takeoff spatiotemporal path of each low-altitude aircraft onto a three-dimensional grid array, and identify all pre-takeoff spatiotemporal paths with spatiotemporal overlap by analyzing the expected spatiotemporal occupancy of the low-altitude aircraft in the three-dimensional grid array. S4: Construct a spatiotemporal path optimization algorithm and set priority decision rules for various types of low-altitude aircraft; S5: Based on the priority decision rules for various types of low-altitude aircraft and the spatiotemporal path optimization algorithm, generate the spatiotemporal path adjustment command with the minimum comprehensive adjustment cost for the pre-set spatiotemporal path before takeoff where there is spatiotemporal overlap, and integrate all spatiotemporal path adjustment commands to generate a conflict-free flight track before takeoff.
2. The method for conflict-free flight path planning before takeoff for low-altitude aircraft according to claim 1, characterized in that, The basic protection interval for low-altitude aircraft is calculated based on their performance characteristics, including: S111: Determine the reaction time based on the driving mode and the typical communication latency based on the operating environment; S112: Calculate flight technical error based on the maximum vertical speed that the human body can withstand, the maximum vertical acceleration that the human body can withstand, and the passenger's reaction time to the change in acceleration. S113: Calculate the basic horizontal protection interval for low-altitude aircraft based on navigation error, flight technical error, reaction time, typical communication delay, control center response delay, and maneuvering delay. S114: Calculate the vertical foundation protection interval based on navigation error and flight technical error; The basic protection interval includes a horizontal basic protection interval and a vertical basic protection interval.
3. The method for conflict-free flight path planning before takeoff for low-altitude aircraft according to claim 2, characterized in that, In step S111, determining the reaction time based on the driving mode includes: For manned flight mode, the pilot's typical reaction time is used as the reaction time; For unmanned driving mode, the typical response delay of the flight control system is used as the reaction time; Typical communication latency is determined based on the operating environment, including: For urban environments, typical navigation errors in urban environments are used as typical communication delays; For suburban environments, typical navigation errors in suburban environments are used as typical communication delays.
4. The method for conflict-free pre-takeoff trajectory planning for low-altitude aircraft according to claim 1, characterized in that, Step S2 includes: In the horizontal direction, a standard map projection method is used to map the geographic region to a Cartesian coordinate system and construct a regular planar grid array. The planar grid is iteratively subdivided at a fixed ratio until the size of the finest level planar grid unit is equal to the minimum scale of the three-dimensional dynamic safety boundary. A height dimension is added vertically to each planar grid to form a three-dimensional grid; the height of a single three-dimensional grid is planned to be the upper limit of the spatial height. , The height of a single 3D mesh is any integer; the height of a single mesh matches the size of the finest level planar mesh on the horizontal plane.
5. The method for conflict-free flight path planning before takeoff for low-altitude aircraft according to claim 1, characterized in that, Constructing the pre-set spatiotemporal path before takeoff includes: S31: Define the starting takeoff and landing field, the end point of the vertical takeoff phase, the end point of the climb phase, the end point of the cruise phase, and the starting and ending takeoff and landing fields of the vertical landing phase as key flight phase transition points and mark them as target points. S32: Assign key state parameters to each target point for determining spatiotemporal overlap. Key state parameters include the target point's spatial coordinates, instantaneous velocity, and remaining battery power. S33: Connect the target points according to the key status parameters to form a pre-set spatiotemporal path before takeoff.
6. The method for conflict-free pre-takeoff trajectory planning for low-altitude aircraft according to claim 1, characterized in that, In step S3, the expected spatiotemporal occupancy of the 3D mesh by low-altitude aircraft includes the expected occupancy range of the 3D mesh coordinates and the expected effective time window. Based on the interpolation principle and the geographic coordinates of the target point, calculate the current geographic coordinates of the low-altitude aircraft. According to the spatial mapping rules, the current geographic coordinates of the low-altitude aircraft and the three-dimensional dynamic safety boundary of the low-altitude aircraft at the current location are converted into multiple continuous three-dimensional grid coordinate ranges that the low-altitude aircraft is expected to occupy. The expected start time of the effective time window is calculated by integrating a preset motion model from takeoff time to arrival at the current geographical location; The expected end time of the effective time window is calculated based on the start time of the expected effective time window, taking into account the expected motion state of the current geographical location and the projected length of the three-dimensional dynamic safety boundary of the low-altitude aircraft in the motion direction.
7. The method for conflict-free flight path planning before takeoff for low-altitude aircraft according to claim 1, characterized in that, The pre-set spatiotemporal path identification method for aircraft with spatiotemporal overlap is as follows: the expected spatiotemporal occupancy segments of the two low-altitude aircraft have an overlap of effective time windows on at least one three-dimensional grid.
8. The method for pre-takeoff conflict-free trajectory planning for low-altitude aircraft according to claim 1, characterized in that, The spatiotemporal path optimization algorithm includes a policy-value network, where the policy in the policy network represents the logic of the agent in selecting and adjusting instructions, and the value in the value network is the sum of the current reward value obtained by adopting the policy and the expected future reward value. The spatiotemporal path optimization algorithm is trained through a progressive training environment spectrum from easy to difficult. The trained spatiotemporal path optimization algorithm, taking into account the takeoff and landing field capacity, airspace structure, and operational rule constraints, generates spatiotemporal path adjustment instructions with the lowest overall adjustment cost.
9. The method for pre-takeoff conflict-free trajectory planning for low-altitude aircraft according to claim 1, characterized in that, Spatiotemporal path adjustment instructions include takeoff and landing site adjustment instructions and flight route adjustment instructions; Landing site adjustment instructions include delays and circling with a waiting procedure; Route adjustment instructions include acceleration, deceleration, climb, descent, yaw, and delay; The comprehensive adjustment cost is a multi-objective optimization function that comprehensively considers the total change in altitude, total change in heading, total power consumption, total delay time, and total change in speed, and solves for the overall minimum value by weighted summation.