Real-time online trajectory planning method for UAV for returning spacecraft measurement mission
Through real-time online track planning method, combined with ternary sampling method and sparse processing, the real-time tracking and planning problem of dynamic targets of spacecraft return capsules is solved, and the efficient tracking coverage of the return capsules by drones is achieved, and the search and rescue information network is improved. Measurement and control capabilities.
Patent Information
- Application Number
- CN202210952570.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-09
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2042-08-09
AI Technical Summary
The existing drone track planning technology lacks real-time tracking planning methods that are adapted to dynamic targets of spacecraft return capsules, especially when ballistic updates are frequent, it is difficult to meet the rapid planning needs of return capsules.
A real-time online track planning method for a return spacecraft is adopted, and the ballistic point of the return capsule is predicted, the boundary of the detection range of the optical sensor and the aircraft speed are used as constraints, combined with ternary sampling method and sparse track point processing , calculate the optimal track point and sparse it to ensure that the drone can track the return capsule in real time.
It realizes efficient tracking coverage of the spacecraft return capsule, improves the measurement and control coverage of the search and rescue information network, solves the problem of insufficient measurement performance when relying solely on the land-based measurement and control system, can provide continuous intuitive image information throughout the process, and supports efficient search and rescue. Task.
Smart Images

Figure CN115525062B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the field of trajectory planning and relates to a real-time online trajectory planning method for an unmanned aerial vehicle (UAV) for a returning spacecraft measurement mission. Background Art
[0002] A lot of research has been conducted at home and abroad on UAV trajectory planning. Compared with domestic research, foreign research started earlier. In the 1990s, foreign researchers focused on single UAV static trajectory planning and two-dimensional planning environment modeling. Due to the complex and changeable flight environment in the actual environment, many research results have been produced for environmental modeling optimization and UAV flight stability optimization. For example, Ruz et al. used mixed integer linear programming (MILP) to detect threat areas and obstacle spaces. In order to enable unmanned aircraft to quickly avoid radar detection areas, the MILP method makes the nonlinear radar detection function produce linear characteristics. Hota et al. studied a three-dimensional planning method that makes the aircraft flight more stable. The entire trajectory is constructed by connecting the circle with the minimum turning radius and a straight line, which makes any two adjacent points on the trajectory have the characteristics of smooth transition.
[0003] Compared with foreign countries, the research on trajectory planning in China started later and is in the stage of catching up with the West in terms of trajectory planning technology. However, domestic researchers have also made many outstanding achievements in the technology of unmanned aircraft trajectory planning. Zhang Jialong et al. applied the composite vector artificial potential field to trajectory optimization, used the potential field gravity to track the target, and used the potential field repulsion to stay away from obstacles, effectively solving the three-dimensional obstacle avoidance problem of unmanned aircraft. Zhang et al. used the hierarchical directed dynamic programming (DP) method to improve the adaptability of the DP algorithm in special terrain and the convergence speed of the algorithm, thereby improving the algorithm operation efficiency. In order to solve the problem that the B-spline curve cannot consider the initial and final states of the aircraft during trajectory optimization, Yang Jian et al. first merged the Dubins curve and the B-spline curve and planned an optional trajectory, and then used a recursive method to use the initial trajectory of the combined algorithm as the initial solution of the next optimal control algorithm, thereby establishing a refined trajectory planning model.
[0004] The current UAV trajectory planning methods are all for the reconnaissance, attack and obstacle avoidance of static targets. There is a lack of research on trajectory planning methods for continuous tracking of dynamic targets such as spacecraft return capsules. In addition, the return process of the spacecraft is affected by various uncertain factors such as flight control accuracy and high-altitude wind. The trajectory of the return capsule needs to be constantly updated. Especially in the ballistic return process, the difference between the return capsule trajectory and the theoretical trajectory is very large. Therefore, when using UAVs to carry out real-time scene measurement of the return capsule, it is necessary to quickly plan and generate a flight trajectory that meets the detection requirements based on the latest predicted trajectory, which puts forward high requirements for the real-time performance of trajectory planning. The current UAV trajectory planning technology lacks a fast algorithm that is suitable for the return capsule tracking scenario. Summary of the invention
[0005] Technical issues to be solved
[0006] In order to avoid the shortcomings of the prior art, the present invention proposes a real-time online trajectory planning method for an unmanned aerial vehicle (UAV) for a returning spacecraft measurement mission.
[0007] Technical Solution
[0008] A real-time online trajectory planning method for a UAV for a returning spacecraft measurement mission, characterized by the following steps:
[0009] Step 1. Initial configuration data:
[0010] 1. The predicted trajectory point of the return capsule within the predetermined tracking arc of the UAV is Ψ = {L, B, H}, where L, B, and H represent the longitude, latitude, and altitude of the return capsule in the WGS84 system respectively; the time interval between the trajectory points is t;
[0011] 2. The detection range boundary of the drone optical sensor: including the maximum detection distance ρ max , horizontal detection range (Φmim, Φmax), vertical detection range Θmim, Θmax);
[0012] 3. Convert the longitude and latitude coordinates to XY horizontal and vertical coordinates, and use the aircraft speed v as the fixed value correction formula:
[0013]
[0014] Where b0 and l0 represent the reference points of latitude and longitude respectively, b and l represent latitude and longitude respectively, and x and y represent the horizontal and vertical coordinates respectively;
[0015] Step 2: Calculate the initial track point:
[0016] The starting point of the trajectory of the aircraft is G(0)=(x0,y0), and the end point is G(N)=(x N ,y N), the starting point of the aircraft trajectory planning is:
[0017]
[0018] Where: θ is the relative angle between the perpendicular bisector of the trajectory point line and the abscissa axis;
[0019] Step 3: Calculate the subsequent track points: The starting point of the trajectory of the initial point of the aircraft is G(0) = (x0, y0), and the end point is G(N) = (x N ,y N ), the starting point of the aircraft trajectory planning is:
[0020]
[0021] Where: θ is the relative angle between the perpendicular bisector of the trajectory point line and the horizontal axis; L is the expected distance;
[0022] Step 4: Calculate the subsequent track points: take the initial point as the starting point, the current speed direction as the stepping direction, and take the average flight speed of the drone as the step length to expand forward one step to obtain the second track point. The "ternary" sampling method is used to determine the track point later;
[0023] 1. Determine three points to be selected at three angles of α, 0, and -α in the original flight speed direction;
[0024] 2. Determine the total cost function Cost of the optimal track point:
[0025] Cost = Cost p +Cost e +Cost d
[0026] Calculate the cost function corresponding to each candidate track point, and the track point corresponding to the smallest result is the optimal track point:
[0027]
[0028] Cost d Cost function for no-fly zone constraint track points:
[0029] Where d obs represents the distance between the candidate track point and the obstacle, D0 represents the predicted distance, and u(·) represents the step function; when the distance between the candidate track point and the obstacle is less than the predicted distance D0, the cost function is set to the activated state for track optimization, otherwise the cost function is in the dormant state:
[0030] Cost p Cost function for constraining track points to the azimuth coverage of the optical sensor:
[0031]
[0032] Where: opt(·) is the direction with the maximum gain within the sensor detection range; the azimuth angle between the selected track point and the return capsule is θ;
[0033] Cost e Cost function for constraining track points for the pitch coverage of the optical sensor:
[0034]
[0035] Among them: When there is no roll angle at the selected track point, the pitch angle with the return capsule is
[0036] Step 5: Sparse track points:
[0037] 1. Traverse the complete track and sparse track points:
[0038] If the turning radius of three adjacent points is less than 3 times the minimum turning radius R, then delete one point in the middle of the three points to make the overall track sparse and increase the turning radius of the three adjacent points;
[0039] The next track point is selected at the straight line according to the predetermined straight line thinning rate. If the turning radius formed by the three adjacent points is less than 3 times the minimum turning radius R, the next track point is selected using the curve thinning rate, and this cycle is repeated until the entire track is traversed;
[0040] 2. Sparse obstacle avoidance: From the starting point of the track, use the straight line sparse rate (this value can be set on the interface) to find the track point as the sparse pre-selected point, connect the pre-selected point with the previous point. If the connecting line passes through an obstacle, it means that this pre-selected point does not meet the obstacle avoidance requirements. Then, step forward from this pre-selected point to search until a track point that meets the obstacle avoidance requirements is found. This track point is used as the determined sparse point; then start from the determined sparse point and loop the above process to traverse the entire track. Then all the determined sparse points form a new track. This track is the sparse track to complete the online track planning.
[0041] The maximum detection distance of the optical load max :
[0042]
[0043] Where D0 is the effective aperture of the optical system, ξ is the correction coefficient, τ0 is the transmittance of the optical system, and D * is the detection ratio of the subsystem, a is the length of the photosensitive area of the detector pixel, b is the width of the photosensitive area of the detector pixel, f is the focal length, t int is the detector integration time of each subsystem, F is the F number, τ ais the atmospheric transmittance of each subsystem, J is the total target illumination energy of each subsystem, and SNR is the signal-to-noise ratio of the output signal of the detection system of each subsystem.
[0044] The time interval t of step 1 is 1s.
[0045] The α is determined according to the minimum turning radius R of the aircraft:
[0046]
[0047] Beneficial Effects
[0048] The present invention proposes a real-time online trajectory planning method for UAVs for returning spacecraft measurement missions. Under the condition that the trajectory prediction of the return capsule is known, the flight performance and flight restriction area of the UAV are constrained, and the real-time dynamic tracking of the spacecraft return capsule is taken as the core goal. While fully considering the real-time performance of the trajectory planning method, the flight trajectory of the UAV tracking the return capsule is planned, and the maneuvering operation characteristics of the UAV and the detection payload are organically combined to improve the tracking coverage of the return capsule as much as possible. After applying this method, the UAV equipped with an optical detection payload can effectively meet the requirements of the aerial dynamic tracking of the spacecraft return capsule, realize the coordinated operation with the existing land-based measurement and control system that guarantees the tracking mission of the spacecraft return capsule, and construct an air-ground integrated, mutually complementary, and multi-integrated search and rescue information network, effectively solving the problem of poor low-elevation angle measurement performance and insufficient measurement and control coverage of the landing stage of the spacecraft return capsule generated by the existing search and rescue information network when relying solely on the land-based measurement and control system. Ensure that the search and rescue information network can provide full-process continuous intuitive image information for the accurate judgment of the landing search and rescue situation, which is helpful to efficiently complete the search and rescue recovery mission.
[0049] The present invention adopts a ternary sampling method, integrates the flight dynamics constraints of the UAV into the sampling mechanism of the track step, greatly simplifies the feasible solution space of track planning, and adopts an integrated roll angle iteration method to reduce the cumbersome coordinate conversion. The real-time requirements are fully considered in terms of optimization method route selection and calculation processing. Compared with the existing UAV track planning methods, the present invention has the characteristics of simplicity, strong engineering practicability and high real-time performance, and can effectively adapt to the UAV track planning problems in the measurement and support mission of the spacecraft return capsule. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] Figure 1 :Flowchart of the UAV trajectory planning method for returning spacecraft measurement missions
[0051] Figure 2 :Schematic diagram of the initial point search method
[0052] Figure 3 :Flowchart of the initial point search method
[0053] Figure 4 : Track point sampling diagram
[0054] Figure 5 : Obstacle avoidance diagram
[0055] Figure 6 :Sparse diagram of track points
[0056] Figure 7 : Track point sparse flow chart
[0057] Figure 8 : Track point sparse flow chart
[0058] Fig. 9 : Schematic diagram of the flight path of a drone in the embodiment
[0059] Fig.10 : UAV sensor coverage
[0060] Fig.11 : Method running time DETAILED DESCRIPTION
[0061] The present invention will now be further described with reference to the embodiments and the accompanying drawings:
[0062] The method proposed in this invention includes four steps: initial configuration, initial track point calculation, subsequent track point calculation and track point sparseness. The overall method flow chart is as follows: Figure 1 shown.
[0063] Step 1: Initial Configuration
[0064] The initial configuration phase mainly provides necessary data support for trajectory planning and carries out preliminary processing.
[0065] (1) Data support
[0066] The data provided mainly include the predicted trajectory point of the return capsule within the UAV’s predetermined tracking arc and the detection range boundary of the UAV’s optical sensor.
[0067] The ballistic point is represented by Ψ={L,B,H}, where L, B, and H represent the longitude, latitude, and altitude of the return capsule in the WGS84 system, respectively, and the time interval is 1s.
[0068] The sensor detection range space is determined by the maximum detection distance ρ max , horizontal detection range (Φmim, Φmax), vertical detection range Θmim, Θmax). The maximum detection distance is calculated in the next step of pre-processing according to the target characteristics and background environment.
[0069] (2) Pre-processing
[0070] ①Coordinate conversion
[0071] Assuming that the flight altitude of the drone is a constant, that is, the altitude of the starting point and the altitude of the end point remain consistent, the three-dimensional space planning problem is transformed into a two-dimensional space planning problem, which greatly reduces the amount of calculation.
[0072] The actual planning space has continuously changing values. The computer does not have enough computing power to meet the needs of continuous value calculations, so it is necessary to discretize the planning space.
[0073] Use the following formula to transform the longitude and latitude coordinates into XY horizontal and vertical coordinates
[0074]
[0075] Among them, b0 and l0 represent the reference points of latitude and longitude respectively, b and l represent latitude and longitude respectively, and x and y represent the horizontal and vertical coordinates respectively.
[0076] In order to facilitate the calculation of the distance in the XY horizontal and vertical coordinates, the aircraft speed v is used as a constant correction formula, as shown in the following formula
[0077]
[0078] Then the distance represented by an x is the magnitude of the aircraft's speed v.
[0079] Use the following formula to transform the XY horizontal and vertical coordinates into longitude and latitude coordinates
[0080]
[0081] All of the above information processing uses a fast grid mapping method to achieve the conversion from L, B to rectangular coordinates X, Y, avoiding the tediousness and complexity of traditional coordinate conversion.
[0082] ①Maximum detection distance R of optical payload
[0083]
[0084] In the above formula, D0 is the effective aperture of the optical system, ξ correction coefficient, τ0 is the transmittance of the optical system, and D * is the detection ratio of the subsystem, a is the length of the photosensitive area of the detector pixel, b is the width of the photosensitive area of the detector pixel, f is the focal length, t int is the detector integration time of each subsystem, F is the F number, τ a is the atmospheric transmittance of each subsystem, J is the total target illumination energy of each subsystem, and SNR is the signal-to-noise ratio of the output signal of the detection system of each subsystem.
[0085] Step 2: Calculate the initial track point
[0086] Assuming that the target's speed is greater than that of the aircraft, in order to achieve a better tracking effect, the aircraft's position should be set at a certain lead distance near the theoretical ballistic point. The lead distance has a certain correlation with the relative speed. The greater the relative speed, the greater the lead distance, and vice versa.
[0087] Assuming the target theoretical trajectory is a straight line, Figure 2 As shown, the figure shows the speed of the target, v0 is the speed of the aircraft, point q is the initial point of the aircraft, and point p is the target
[0088] Initial point, initial point search method diagram see Figure 2 , the initial point search method flow chart is shown in Figure 3
[0089] The following formula can be obtained
[0090]
[0091] Where N represents the number of selected trajectory points, m is the point selection parameter, L is the expected distance, γ is the maximum azimuth angle, and v is the target speed.
[0092] Define the trajectory starting point of the aircraft as G(0)=(x0,y0) and the end point as G(N)=(x N ,y N ), we can get the starting point of the aircraft trajectory planning as
[0093]
[0094] Where θ is the relative angle between the perpendicular bisector of the line connecting the ballistic points and the horizontal axis.
[0095] Step 3: Calculate subsequent track points
[0096] When calculating the subsequent track points, the idea of "one step, one move" is adopted. According to the method of benchmarking the predicted ballistic points, one step is taken forward every 1s. Taking into account the constraints of the no-fly zone and the flight performance of the UAV, the optimal UAV flight track point corresponding to each ballistic point is calculated, so that the track points can be updated online in real time at any time according to the latest predicted ballistic points of the return capsule to meet the observation requirements of the space search and rescue information network. Taking the initial point as the starting point, the current speed direction as the step direction, and the average flight speed of the UAV as the step length, one step is extended forward to obtain the second track point. The "ternary" sampling method is used to determine the track points. See Figure 4 , the specific steps are as follows:
[0097] ① In the figure, P1 is the current known track point, P21, P20, and P22 are three candidate track points determined according to the "ternary" sampling method. The angles between these three candidate points and the original flight speed direction are α, 0, and -α, respectively, where α is calculated based on the aircraft's minimum turning radius R. The specific formula is as follows:
[0098]
[0099] The positions of points P21, P20, and P22 are calculated based on the aircraft's flight speed and deflection angle.
[0100] ② Design the cost function of the no-fly zone constraint track point. When the distance between the selected track point and the obstacle is less than a certain predicted distance D0, set the cost function to the active state for track optimization, otherwise the cost function is dormant, that is, the design cost function is
[0101]
[0102] Where d obs represents the distance between the candidate track point and the obstacle, D0 represents the predicted distance, and u(·) represents the step function.
[0103] Obstacle avoidance diagram Figure 5 As shown, the black circle is an obstacle, the outer dotted line is the advance obstacle avoidance distance, and the expansion node cannot be expanded into the dotted line to achieve advance obstacle avoidance. When the expansion node is outside the dotted line, this obstacle avoidance cost function is disabled.
[0104] ⑤ Design the cost function of the optical sensor’s azimuth coverage range to constrain the track point. The azimuth angle between the selected track point and the return capsule is θ, then the cost function is:
[0105]
[0106] where opt(·) is the direction with maximum gain within the sensor detection range.
[0107] ⑥ Design the cost function of the pitch coverage range of the optical sensor to constrain the track points.
[0108] When there is no roll angle at the selected track point, the pitch angle of the return capsule is The corresponding cost function at this time is:
[0109]
[0110] In addition, the roll angle of the drone will affect the cost calculation of the pitch angle in the trajectory planning when turning. The traditional method uses the flight simulation model and the trajectory planning to iterate each other to approximately obtain the matching roll angle. This method will reduce the real-time performance of the trajectory planning algorithm. The present invention proposes an integrated roll angle iteration method, which can significantly improve the real-time performance.
[0111]
[0112] ⑦ Determine the total cost function Cost of the optimal track point.
[0113] Cost = Cost p +Cost e +Cost d (12)
[0114] ⑧ Calculate the cost function corresponding to each candidate track point, and the track point corresponding to the smallest result is the optimal track point.
[0115]
[0116] Step 4: Sparse track points
[0117] Usually, too many track points are set during the flight of the drone, which is not conducive to the control system. Therefore, the above track points need to be sparsely processed to convert them into information that meets the binding requirements of the drone. The specific steps are as follows: ① Track point sparse, track point sparse indication Figure 6
[0118] like Figure 6 As shown in FIG. 1 , if the turning angles of three adjacent points are too large, one point in the middle of the three points is deleted, so that the overall track becomes sparse and the turning radius of the three adjacent points increases.
[0119] The next track point is selected according to the line thinning rate. If the turning radius formed by the three adjacent points is less than a certain value, the next track point is selected according to the curve thinning rate. This cycle is repeated until the entire track is traversed. Figure 7 shown.
[0120] ②Sparse obstacle avoidance
[0121] Since the sparse algorithm will cause the track to change slightly, there is a possibility that the sparse track will pass through obstacles, so obstacle avoidance is required.
[0122] First, use a certain sparse rate to find a track point as a sparse pre-selected point from the starting point of the track. Connect this pre-selected point with the previous point. If the line passes through an obstacle, it means that this pre-selected point does not meet the obstacle avoidance requirements. Then, step forward from this pre-selected point to search until a track point that meets the obstacle avoidance requirements is found. This track point is used as a determined sparse point. Then, start from the determined sparse point and repeat the above process to traverse the entire track. Then all determined sparse points form a new track, which is the sparse track. The sparse obstacle avoidance method process is as follows: Figure 8 shown.
[0123] Specific implementation examples:
[0124] Assume that the spacecraft return capsule flies due east at a speed of 6200m / s and -5deg inclination at an altitude of 100km until landing, and opens the parachute at a speed of 190m / s at an altitude of 10km. It flies for about 10 minutes before landing. The radius of the obstacle avoidance area is 30km, and the dotted line at the outer edge of the buffer zone is 10km away from the obstacle avoidance circle. The effective detection range of the UAV optical sensor is 400km, the pitch angle coverage range is (-40°, 40°), and the azimuth angle coverage range is (-70°.70°). In order to ensure the highest coverage of optical scene tracking of the spacecraft return capsule, the trajectory planning method proposed in the present invention is applied to optimize the tracking flight track of the UAV, such as Fig. 9 As shown, the optical sensor tracking rate can reach 93.96%. Fig.10 As shown in Figure 2, under the premise of effectively avoiding obstacles, the operating time of the drone optical sensor is only 1.19s. Fig.10 As shown, it has high real-time performance and effectiveness, and can effectively meet application scenarios.
Claims
1. A real-time online trajectory planning method for UAVs in returning spacecraft measurement missions. Features Here are the steps: Step 1. Initial configuration data: The predicted trajectory point Ψ of the return capsule within the predetermined tracking arc of the UAV is {L, B, H}, where L, B, and H represent the longitude, latitude, and altitude of the return capsule in the WGS84 system respectively; the time interval between the trajectory points is t; The detection range boundary of the drone optical sensor: including the maximum detection distance ρ max , horizontal detection range (Φmim, Φmax), vertical detection range Θmim, Θmax); Transform the longitude and latitude coordinates to XY horizontal and vertical coordinates, and use the aircraft speed v as a constant to correct the formula: Among them, b 0 and l 0 represents the reference point of latitude and longitude respectively, b and l represent latitude and longitude respectively, x and y represent the horizontal and vertical coordinates respectively; Step 2: Calculate the initial track point: The starting point of the trajectory of the aircraft is G(0)=(x 0 ,y 0 ), the end point is G(N)=(x N ,y N ), the starting point of the aircraft trajectory planning is: Where: θ is the relative angle between the perpendicular bisector of the trajectory point line and the abscissa axis; Step 3: Calculate the subsequent track points: take the initial point as the starting point, the current speed direction as the stepping direction, and take the average flight speed of the drone as the step length to move forward one step to get the second track point. The "ternary" sampling method is then used to determine the track points. Determine three points to be selected at three angles of α, 0, and -α in the original flight speed direction; Determine the total cost function Cost of the optimal track point: Cost=Cost p +Cost e +Cost d Calculate the cost function corresponding to each candidate track point, and the track point corresponding to the smallest result is the optimal track point: Cost d Cost function for no-fly zone constraint track points: Where d obs Indicates the distance between the candidate track point and the obstacle, D 0 represents the predicted distance, u(·) represents the step function; when the distance between the candidate track point and the obstacle is less than the predicted distance D 0 When , set the cost function to active state for trajectory optimization, otherwise the cost function is dormant: Cost p Cost function for constraining track points to the azimuth coverage of the optical sensor: Where: opt(·) is the direction with the maximum gain within the sensor detection range; the azimuth angle between the selected track point and the return capsule is θ; Cost e Cost function for constraining track points for the pitch coverage of the optical sensor: Among them: When there is no roll angle at the selected track point, the pitch angle with the return capsule is Step 4: Sparse track points: Traverse the sparse track points of the entire track: If the turning radius of three adjacent points is less than 3 times the minimum turning radius R, then delete one point in the middle of the three points to make the overall track sparse and increase the turning radius of the three adjacent points; The next track point is selected at the straight line according to the predetermined straight line thinning rate. If the turning radius formed by the three adjacent points is less than 3 times the minimum turning radius R, the next track point is selected using the curve thinning rate, and this cycle is repeated until the entire track is traversed; Sparse obstacle avoidance: From the starting point of the track, use the straight line sparse rate to find the track point as the sparse pre-selected point, connect the pre-selected point with the previous point. If the line passes through an obstacle, it means that this pre-selected point does not meet the obstacle avoidance requirements. Then, step forward from this pre-selected point to search until a track point that meets the obstacle avoidance requirements is found. This track point is used as the determined sparse point; then start from the determined sparse point and loop the above process to traverse the entire track. Then all the determined sparse points form a new track. This track is the sparse track to complete the online track planning.
2. According to the real-time online trajectory planning method for UAV for returning spacecraft measurement missions according to claim 1, Features: Maximum detection distance of optical payload ρ max : Where D 0 is the effective aperture of the optical system, ξ correction coefficient, τ 0 Optical system transmittance, D * is the detection ratio of the subsystem, a is the length of the photosensitive area of the detector pixel, b is the width of the photosensitive area of the detector pixel, f is the focal length, t int is the detector integration time of each subsystem, F is the F number, τ a is the atmospheric transmittance of each subsystem, J is the total target illumination energy of each subsystem, and SNR is the signal-to-noise ratio of the output signal of the detection system of each subsystem.
3. According to the real-time online trajectory planning method for UAV for returning spacecraft measurement missions as described in claim 1, Features: The time interval t of step 1 is 1s.
4. According to the real-time online trajectory planning method for UAV for returning spacecraft measurement missions as described in claim 1, Features: The α is determined according to the minimum turning radius R of the aircraft:
