Guiding search method

Abstract

The invention discloses a guided search method, which comprises the following steps of: S1, generating a target particle swarm by using a stochastic simulation method according to a preset target position, a preset target speed, a preset target course, a preset target information error level, preset notification information delay time and a preset target particle number at a notification moment; s2, sorting the target particles in the target particle swarm to obtain a target particle sequence, determining the course and time of different legs of a searcher according to the target particle sequence, and counting the number of the found target particles; s3, reordering the target particles in the target particle swarm to obtain a new target particle sequence, repeating the step S2, and counting the number of the found target particles; s4, determining a target particle sequence to be reserved by adopting a simulated annealing algorithm; and S5, repeating the steps S3 to S4 until the annealing temperature reaches a preset end value. According to the method, the target discovery probability is maximized under the limited search capability.

Description

Technical Field

[0001] The present invention belongs to the technical field of mobile target searching and tracking, and in particular relates to a guided searching method. Background Art

[0002] In the mobile target search task, traditional methods (such as extended Kalman filtering, particle filtering, and grid search) are usually based on the following assumptions: (1) The target state information is relatively accurate: they rely on high-precision sensors or communication links to continuously update the target parameters. However, in actual applications, the target position, speed, and heading are easily affected by environmental interference (such as ocean currents and wind) or human interference (such as active target avoidance), resulting in significant errors in the prior information; (2) The searcher is sufficiently maneuverable: existing path planning algorithms (such as spiral search and parallel line scanning) usually assume that the searcher's speed is much higher than the target and can quickly cover the suspicious area. However, when the searcher's speed is limited (such as the endurance limit of small drones or the low speed characteristics of underwater robots), it is easy to miss the detection due to path overlap or blind spots; (3) The detection equipment has a wide coverage range: traditional methods default to using wide-field-of-view sensors (such as wide-angle cameras), while the detection equipment in some application scenarios (such as passive detection sonar of underwater vehicles) has a limited detection range and requires a more sophisticated search strategy.

[0003] The defects of the existing technology are as follows: (1) The static probability model has poor adaptability: the search algorithm based on the initial error Gaussian distribution (such as covariance matrix prediction) does not take into account the time-varying characteristics of the error caused by the dynamic motion of the target, and it is difficult to correct the search area in real time; (2) Path planning based on special search patterns relies on human experience: the existing search scheme optimization method usually performs path planning based on special search patterns, which is highly dependent on human experience and has a limited scheme search space. The scheme optimization effect using swarm intelligence optimization algorithm is not ideal. Summary of the Invention

[0004] The technical problem solved by the present invention is to overcome the deficiencies of the prior art and provide a guided search method to maximize the target discovery probability under limited search capabilities.

[0005] The object of the present invention is achieved through the following technical solutions: a guided search method, comprising: step S1: generating a target particle swarm using a random simulation method according to a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset number of target particles at the notification time; step S2: sorting the target particles in the target particle swarm to obtain a target particle sequence, determining the heading and time of different flight segments of the searcher according to the target particle sequence, and counting the number of target particles found; step S3: re-sorting the target particles in the target particle swarm to obtain a new target particle sequence, repeating step S2, and counting the number of target particles found; step S4: determining the target particle sequence to be retained using a simulated annealing algorithm according to the number of target particles found in step S2 and the number of target particles found in step S3; and step S5: repeating steps S3 to S4 until the annealing temperature reaches a preset end value.

[0006] In the above-mentioned guided search method, determining the heading and time of different flight segments of the searcher based on the target particle sequence includes: step S21: using the method of determining the shortest time and heading for discovering a single target particle to determine the heading and length of the first flight segment for the first target particle in the target particle sequence; step S22: determining the target particles that can be discovered in the first flight segment; step S23: for the first undiscovered target particle in the target particle sequence, repeating steps S21 to S22 until all target particles are discovered or the preset distance to the first undiscovered target particle in the target particle sequence cannot be reached, and counting the number of discovered target particles.

[0007] In the above-mentioned guided search method, the method for determining the shortest time and heading for finding a single target particle includes: step S211: establishing the motion equation of a single target particle and the motion equation of a searcher, and obtaining a set of equations based on the motion equation of the single target particle and the motion equation of the searcher, so that the distance between the searcher and the single target particle reaches the detection distance of the searcher to the target; step S212: solving the set of equations in step S211 to obtain the minimum positive solution, that is, the shortest time for finding a single target particle, and multiplying the shortest time for finding a single target particle by the speed of the searcher to obtain the length of the first flight segment; step S213: substituting the shortest time for finding a single target particle into the motion equation of the searcher to obtain the heading of the first flight segment.

[0008] In the above guided search method, the distance between the searcher and a single target particle reaches the searcher's detection distance d s The equations are obtained by the following formula:

[0009]

[0010] Where D is the distance from the searcher to the target at time t, H mis the target heading at the time of notification, V m is the target speed at the time of notification, t is the time, V w is the searcher speed, d s is the detection range of the searcher to the target, H is the heading of the first segment, and X is the target side angle.

[0011] In the guided search method above, the heading of the first leg is obtained by the following formula:

[0012] H=arctan2(V m Tsin(H m )-Dsin(H m +X),V m Tcos(H m )-Dcos(H m +X));

[0013] Among them, V m is the target speed at the time of notification, H m is the target heading at the time of notification, D is the distance from the searcher to the target at time t, T is the time required for approach, and X is the target side angle.

[0014] In the above-mentioned guided search method, determining the target particles that can be found in the first flight segment includes: obtaining a set of equations for the distance between the searcher and other target particles to reach the searcher's detection distance of the target based on the motion equations of other target particles in the target particle sequence and the searcher's motion equation; solving the set of equations to obtain the time for reaching the searcher's detection distance of the target, and if the time is included in the time range of the first flight segment, it is determined that the target particle can be found in the first flight segment.

[0015] In the guided search method described above, the first leg is the line connecting the searcher and the position of the first target particle found.

[0016] A guided search system comprises: a first module for generating a target particle swarm using a random simulation method based on a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset number of target particles at a notification moment; a second module for sorting target particles in the target particle swarm to obtain a target particle sequence, determining the heading and time of different flight segments of a searcher based on the target particle sequence, and counting the number of target particles found; a third module for re-sorting target particles in the target particle swarm to obtain a new target particle sequence, and counting the number of target particles found; and a fourth module for determining a target particle sequence to be retained using a simulated annealing algorithm based on the number of target particles found in the second module and the number of target particles found in the third module.

[0017] In the above-mentioned guided search system, determining the heading and time of different flight segments of the searcher based on the target particle sequence includes: step S21: using the method of determining the shortest time and heading for discovering a single target particle to determine the heading and length of the first flight segment for the first target particle in the target particle sequence; step S22: determining the target particles that can be discovered in the first flight segment; step S23: repeating steps S21 to S22 for the first undiscovered target particle in the target particle sequence until all target particles are discovered or the preset distance to the first undiscovered target particle in the target particle sequence cannot be reached, and counting the number of discovered target particles.

[0018] In the above-mentioned guided search system, the method for determining the shortest time and heading for discovering a single target particle includes: step S211: establishing the motion equation of a single target particle and the motion equation of a searcher, and obtaining a set of equations for the distance between the searcher and the single target particle to reach the detection distance of the searcher to the target based on the motion equation of the single target particle and the motion equation of the searcher; step S212: solving the set of equations in step S211 to obtain the minimum positive solution, that is, the shortest time for discovering a single target particle, and multiplying the shortest time for discovering a single target particle by the speed of the searcher to obtain the length of the first flight segment; step S213: substituting the shortest time for discovering a single target particle into the motion equation of the searcher to obtain the heading of the first flight segment.

[0019] Compared with the prior art, the present invention has the following beneficial effects:

[0020] The present invention maximizes the target discovery probability under limited search capabilities through dynamic probability modeling, search path adaptive optimization and resource allocation strategy. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Various other advantages and benefits will become apparent to those skilled in the art upon reading the detailed description of the preferred embodiment below. The accompanying drawings are for illustration purposes only and are not to be considered as limiting the present invention. The same reference symbols are used throughout the drawings to represent the same components. In the drawings:

[0022] Figure 1 Schematic diagram of a particle capture model for optimizing a guided search solution provided by an embodiment of the present invention;

[0023] Figure 2 Schematic diagram of the principle of determining the shortest approaching time solution provided by an embodiment of the present invention;

[0024] Figure 3 1 is a schematic diagram of a dynamic generation process of a guided search scheme based on a target particle sequence according to an embodiment of the present invention;

[0025] Figure 4Schematic diagram of a discovery probability change curve of a guided search simulated annealing algorithm provided in an embodiment of the present invention;

[0026] Figure 5 It is a schematic diagram of the process of guiding the search for the optimal solution provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that, unless there is a conflict, the embodiments of the present disclosure and the features described in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0028] This embodiment provides a guided search method, which includes: step S1: generating a target particle swarm using a random simulation method according to a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset number of target particles at the notification time; step S2: sorting the target particles in the target particle swarm to obtain a target particle sequence, determining the heading and time of different flight segments of the searcher according to the target particle sequence, and counting the number of target particles found; step S3: re-sorting the target particles in the target particle swarm to obtain a new target particle sequence, repeating step S2, and counting the number of target particles found; step S4: determining a target particle sequence to be retained using a simulated annealing algorithm based on the number of target particles found in step S2 and the number of target particles found in step S3; and step S5: repeating steps S3 to S4 until the annealing temperature reaches a preset end value.

[0029] This example uses Monte Carlo simulation to transform the guided search solution optimization problem into a target particle capture problem. This method proposes a search solution generation method based on the target particle ranking. Specifically, for each particle in the target particle sequence, the shortest maneuvering solution is calculated, starting from the front and ending at the back, to reach a predetermined distance from an undiscovered particle. The particles that can be discovered during this period are then determined, until no particle can be approached to the predetermined distance. Based on this, a simulated annealing algorithm is then used to optimize the search solution.

[0030] Determining the headings and times of different flight segments of the searcher based on the target particle sequence includes: step S21: determining the heading and length of the first flight segment for the first target particle in the target particle sequence using the method for determining the shortest time and heading for discovering a single target particle; wherein the first flight segment is a line connecting the searcher and the position where the first target particle is discovered; step S22: determining the target particles that can be discovered in the first flight segment; step S23: repeating steps S21 to S22 for the first undiscovered target particle in the target particle sequence until all target particles are discovered or the searcher can no longer reach a preset distance to the first undiscovered target particle in the target particle sequence, and counting the number of discovered target particles.

[0031] The method for determining the shortest time and heading for discovering a single target particle includes: step S211: establishing a motion equation for a single target particle and a motion equation for a searcher, and obtaining a set of equations based on the motion equation for the single target particle and the motion equation for the searcher, so that the distance between the searcher and the single target particle reaches the detection distance of the searcher to the target; step S212: solving the set of equations in step S211 to obtain the minimum positive solution, i.e., the shortest time for discovering a single target particle, and multiplying the shortest time for discovering a single target particle by the speed of the searcher to obtain the length of the first flight segment; step S213: substituting the shortest time for discovering a single target particle into the motion equation for the searcher to obtain the heading of the first flight segment.

[0032] The distance between the searcher and a single target particle reaches the searcher's detection distance d s The equations are obtained by the following formula:

[0033]

[0034] Where D is the distance from the searcher to the target at time t, H m is the target heading at the time of notification, V m is the target speed at the time of notification, t is the time, V w is the searcher speed, d s is the detection range of the searcher to the target, H is the heading of the first segment, and X is the target side angle.

[0035] The heading of the first leg is obtained by the following formula:

[0036] H=arctan2(V m Tsin(H m )-Dsin(H m +X),V m Tcos(H m )-Dcos(H m +X));

[0037] Among them, V mis the target speed at the time of notification, H m is the target heading at the time of notification, D is the distance from the searcher to the target at time t, T is the time required for approach, and X is the target side angle.

[0038] Determining the target particles that can be found in the first flight segment includes: obtaining a set of equations for the distance between the searcher and other target particles to reach the searcher's detection distance for the target based on the motion equations of other target particles in the target particle sequence and the searcher's motion equation; solving the set of equations to obtain the time for reaching the searcher's detection distance for the target, and if the time is included in the time range of the first flight segment, it is determined that the target particle can be found in the first flight segment.

[0039] The target position (x m0 ,y m0 ), target speed V m 、Target heading H m , target information error level σ xm ,σ ym ,σ Vm ,σ Hm , notification information delay time t y , and the searcher speed V w , the searcher's detection distance d to the target s , the distance D from the searcher to the target at the current moment, the target side angle X, determine the searcher's optimal maneuvering heading function H w (t), so that the probability of the searcher finding the target is maximized, that is: maxP(H w (t)).

[0040] Based on Monte Carlo simulation theory, the guided search solution optimization problem can be transformed into a particle capture problem, such as Figure 1 As shown in the figure, “.” represents a target sample generated using a random simulation method based on the distribution of target position, maneuvering speed, and maneuvering heading. It is called a target particle. “*” marks the position of the search platform at time 0, and the sector with “*” as the center is the detection sector of the search platform.

[0041] Generate a sequence of all target particles manually or randomly, generate a guided search scheme based on the given target particle sequence and calculate the discovery probability.

[0042] It is carried out in three steps: first, determine the earliest discovery plan; second, evaluate the search effect of a single flight segment; and third, realize dynamic search plan generation.

[0043] (1) Determine the shortest approach time for a single particle

[0044] The nature of the shortest time to engage the enemy is proved: In fact, if the searcher adopts a fixed heading movement, it can be seen from the searcher's maneuvering range circle that if the searcher maneuvers in any other direction, the searcher cannot approach the target to the predetermined distance before time T or T.

[0045] The following proves that if the searcher uses a change of direction maneuver, it cannot approach the target within the predetermined distance on or before time T. Since the sum of the lengths of two sides of a triangle must be greater than the third side, if the headings in the two phases are different, their positions must be within the circle of possible searcher positions shown in the figure, and therefore they cannot approach the target within the predetermined distance.

[0046] like Figure 2 As shown in the figure, M and W are the positions of the target and search platform at time 0, respectively; MK is the target's maneuvering direction; WO is the reverse vector of the target's displacement at time t; D and X are the distance from the search platform to the target and the target's beam angle at time 0, respectively. The circle with O as its center is the search platform's maneuvering range circle.

[0047] (1) Determine the shortest approach time

[0048] List the system of equations:

[0049]

[0050] Square both sides of the above and below equations and add them together to get a quadratic equation:

[0051]

[0052] Right now

[0053]

[0054] The shortest time approach solution is the minimum positive solution t of the quadratic equation.

[0055] (2) Determine the minimum time to approach the course

[0056] The coordinates of point O are calculated from the coordinates of point W, so the longest tracking heading is the direction of vector OM.

[0057] H=arctan2(V m Tsin(H m )-Dsin(H m +X),V m Tcos(H m )-Dcos(H m +X))

[0058] (2) Target detection and time calculation in the direct search phase

[0059] The direct search segment finds the target judgment and time calculation, which can be approached to the predetermined distance judgment and time calculation problem. Given the initial position of the target (x m ,y m ), target speed V m , target heading H m , the searcher's initial position (x w ,y w ), searcher speed V w , searcher heading H w , and the distance D′ that the searcher is expected to approach the target. Determine whether the searcher can approach the target to the predetermined distance, if the time required for approaching can be given.

[0060] Obviously, if the distance D between the searcher and the target at the initial moment is less than or equal to the pre-approach distance D′, the result of the approach to the predetermined distance is 1, and the time required for approach is 0.

[0061] Assuming that the distance D between the searcher and the target at the initial moment is greater than the pre-approach distance D′, a method for determining the approachable predetermined distance and calculating the time is given.

[0062] make

[0063] D ′2 =(x w +V w cosH w tx m -V m cosH m t) 2 +(y w +V w sinH w ty m -V m sinH m t) 2

[0064] Sort out

[0065]

[0066] It is called the close to predetermined distance equation, where

[0067]

[0068] remember

[0069]

[0070] Solving Quadratic Equations

[0071] at 2 +bt+c=0

[0072] If t has a non-negative solution, it can be approached, and the time required for approaching is the smaller non-negative solution; otherwise it cannot be approached.

[0073] (3) Dynamically generate the search plan based on the remaining particles in the particle queue, such as Figure 3 shown.

[0074] M1: Let the current position of the search platform be its initial position, and the discovered state of all particles in the target particle queue be undiscovered, and the position be their respective initial positions;

[0075] M2: Repeat the operations from M3 to M6 until the maximum number of times Max_Steps is reached;

[0076] M3: Set the end flag bend to 1;

[0077] M4: Perform M5 operation on each particle g in the target particle from front to back;

[0078] M5: If particle g has not been discovered, determine the shortest time t and approach plan for the search platform to approach the target particle g from its current position to the discovery distance ds. If it can be approached, set bend to 0, add the approach time and approach heading as new segment data to the search platform maneuver plan, and for all particles that have not been discovered, determine the particles that can be discovered during the period and set their discovered status to 1, and calculate the status of the particles that have not been discovered after time t, and then go to M6;

[0079] M6: If bend is 1, end the search solution generation, otherwise go to M4.

[0080] The optimized goal is to capture the maximum number of target particles, and the simulated annealing algorithm is used to optimize the particle order. New solutions are randomly generated by using the exchange method, displacement method, and inversion method. The acceptance of the new solution is determined based on the trend of the objective function value and the instantaneous temperature. Figure 4 and Figure 5 shown.

[0081] This embodiment also provides a guided search system, which includes: a first module, used to generate a target particle swarm using a random simulation method based on a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset number of target particles at the notification time; a second module, used to sort the target particles in the target particle swarm to obtain a target particle sequence, determine the heading and time of different flight segments of the searcher based on the target particle sequence, and count the number of target particles found; a third module, used to re-sort the target particles in the target particle swarm to obtain a new target particle sequence, and count the number of target particles found; and a fourth module, used to determine the target particle sequence to be retained using a simulated annealing algorithm based on the number of target particles found in the second module and the number of target particles found in the third module.

[0082] This embodiment optimizes the search method for dynamic targets (such as drifting objects at sea, low-altitude aircraft, and ground-moving targets). It is particularly suitable for scenarios where the searcher (such as a drone, unmanned boat, or robot) has low speed, limited sensor coverage, and significant uncertainty in the target's position, speed, and heading. This embodiment maximizes the probability of target discovery within limited search capabilities through dynamic probability modeling, adaptive optimization of search paths, and resource allocation strategies.

[0083] Although the present invention has been disclosed above in terms of preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art may make possible changes and modifications to the technical solutions of the present invention by using the methods and technical contents disclosed above without departing from the spirit and scope of the present invention. Therefore, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solutions of the present invention shall fall within the scope of protection of the technical solutions of the present invention.

Claims

1. A guided search method, characterized in that include: Step S1: Generate a target particle swarm using a random simulation method according to a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset target particle number at the notification time; Step S2: sort the target particles in the target particle group to obtain a target particle sequence, determine the heading and time of the searcher's different flight segments according to the target particle sequence, and count the number of target particles found; Step S3: reorder the target particles in the target particle group to obtain a new target particle sequence, repeat step S2, and count the number of target particles found; Step S4: According to the number of target particles found in step S2 and the number of target particles found in step S3, a simulated annealing algorithm is used to determine the target particle sequence to be retained; Step S5: Repeat steps S3 to S4 until the annealing temperature reaches a preset end value.

2. The guided search method according to claim 1, wherein: Determining the heading and time of different segments of the searcher based on the target particle sequence includes: Step S21: determining the heading and length of the first flight segment for the first target particle in the target particle sequence using a method for determining the shortest time and heading for finding a single target particle; Step S22: determining the target particles that can be found in the first flight segment; Step S23: Repeat steps S21 to S22 for the first undiscovered target particle in the target particle sequence until all target particles are discovered or the target particle cannot reach the preset distance from the first undiscovered target particle in the target particle sequence, and count the number of discovered target particles.

3. The guided search method according to claim 2, wherein: Methods for determining the minimum time and heading to find a single target particle include: Step S211: establishing a single target particle motion equation and a searcher motion equation, and obtaining a set of equations for the distance between the searcher and the single target particle to achieve the searcher's detection distance of the target based on the single target particle motion equation and the searcher motion equation; Step S212: Solve the equations in step S211 to obtain the minimum positive solution, i.e., the shortest time to find a single target particle. Multiply the shortest time to find a single target particle by the searcher's speed to obtain the length of the first flight segment. Step S213: Substitute the shortest time for finding a single target particle into the searcher's motion equation to obtain the heading of the first flight segment.

4. The guided search method according to claim 3, wherein: The distance between the searcher and a single target particle reaches the searcher's detection distance d s The equations are obtained by the following formula: Where D is the distance from the searcher to the target at time t, H m is the target heading at the time of notification, V m is the target speed at the time of notification, t is the time, V w is the searcher speed, d s is the detection range of the searcher to the target, H is the heading of the first segment, and X is the target side angle.

5. The guided search method according to claim 3, wherein: The heading of the first leg is obtained by the following formula: H =arctan2(V m Csin(H m )-Din(H m +X),V m Tcos(H m )-Dcos(H m +X)); Among them, V m is the target speed at the time of notification, H m is the target heading at the time of notification, D is the distance from the searcher to the target at time t, T is the time required for approach, and X is the target side angle.

6. The guided search method according to claim 2, wherein: The target particles that can be found in the first segment are determined to include: According to the motion equations of other target particles in the target particle sequence and the searcher's motion equation, a set of equations is obtained for the distance between the searcher and other target particles to reach the searcher's detection distance of the target; the time to reach the searcher's detection distance of the target is obtained by solving the set of equations. If the time is included in the time range of the first flight segment, it is determined that the target particle can be found in the first flight segment.

7. The guided search method according to claim 2, wherein: The first leg is the line connecting the searcher and the location where the first target particle is found.

8. A guided search system, characterized in that include: The first module is used to generate a target particle swarm using a random simulation method according to a preset target position, a preset target speed, a preset target heading, a preset target information error level, a preset notification information delay time, and a preset target particle number at the notification time; The second module is used to sort the target particles in the target particle group to obtain a target particle sequence, determine the heading and time of the searcher's different flight segments according to the target particle sequence, and count the number of target particles found; The third module is used to reorder the target particles in the target particle group to obtain a new target particle sequence and count the number of target particles found; The fourth module is used to determine the target particle sequence to be retained by using a simulated annealing algorithm according to the number of target particles found in the second module and the number of target particles found in the third module.

9. The guided search system according to claim 8, characterized in that: Determining the heading and time of different segments of the searcher based on the target particle sequence includes: Step S21: determining the heading and length of the first flight segment for the first target particle in the target particle sequence using a method for determining the shortest time and heading for finding a single target particle; Step S22: determining the target particles that can be found in the first flight segment; Step S23: Repeat steps S21 to S22 for the first undiscovered target particle in the target particle sequence until all target particles are discovered or the target particle cannot reach the preset distance from the first undiscovered target particle in the target particle sequence, and count the number of discovered target particles.

10. The guided search system according to claim 9, characterized in that: Methods for determining the minimum time and heading to find a single target particle include: Step S211: establishing a single target particle motion equation and a searcher motion equation, and obtaining a set of equations for the distance between the searcher and the single target particle to achieve the searcher's detection distance of the target based on the single target particle motion equation and the searcher motion equation; Step S212: Solve the equations in step S211 to obtain the minimum positive solution, i.e., the shortest time to find a single target particle. Multiply the shortest time to find a single target particle by the searcher's speed to obtain the length of the first flight segment. Step S213: Substitute the shortest time for finding a single target particle into the searcher's motion equation to obtain the heading of the first flight segment.

Images