JP6829858B2 - Optimal flight network generation method and system - Google Patents

Description

The present invention relates to a method and a system for generating an optimum flight network.

Japanese Unexamined Patent Publication No. 6-149376 (Patent Document 1) discloses a route generation technique for searching for a route that minimizes the route cost based on three-dimensional topographical information. Further, in Japanese Patent No. 3557443 (Patent Document 2) and Japanese Patent No. 3302735 (Patent Document 3), a plurality of way points serving as reference points of a route database are set for three-dimensional map data, and an aircraft is set. A technique for generating a flight path for generating a flight path suitable for flying at a low altitude by connecting a plurality of way points is disclosed.

Japanese Unexamined Patent Publication No. 6-149376 Japanese Patent No. 3557443 Japanese Patent No. 3302735

In the techniques described in Patent Documents 2 and 3, an appropriate flight path is set between waypoints with the set waypoints fixed. However, unmanned aerial vehicles such as drones have to be powered by batteries and cannot fly for a long time, and the flight altitude is limited. Under such conditions, it is necessary to select a route that makes the flight time as long as possible (in other words, consumes as little energy as possible corresponding to fuel consumption). However, the conventional technique cannot sufficiently meet such a demand because the waypoint is fixed.

An object of the present invention is to provide a method and a system for generating an optimum flight network capable of determining a flight path capable of flying at the highest possible speed and consuming less energy.

The method of the present invention sets a plurality of waypoints as reference points of a route database for three-dimensional map data, and is suitable for flying an unmanned vehicle having an upper limit flight altitude with low energy consumption. The target is a method of generating an optimum flight network that generates a route between each waypoint. In the present invention, a plurality of ways set in the three-dimensional map data by executing the first step and the second step for each of a plurality of planned flight areas defined to cover the three-dimensional map data. Determine the proper position of the point.

In the first step, one planned flight area is determined in the three-dimensional map data, and a plurality of waypoints surrounding the planned flight area are defined as a plurality of fixed waypoints that do not move.
In the second step, an unmanned vehicle moves between one or more waypoints and multiple fixed waypoints while repositioning one or more waypoints within the planned flight area surrounded by multiple fixed waypoints. Finds the path that minimizes energy consumption while moving, and obtains the path that minimizes the average energy consumption of one or more waypoints and multiple fixed waypoints. The position is determined as the provisional position of the one or more waypoints. Then, by executing the first step and the second step for each of the plurality of planned flight areas defined to cover the three-dimensional map data, the plurality of waypoints set in the three-dimensional map data can be obtained. Determine the proper position.

In the present invention, the average energy consumption of energy consumed between one or more waypoints and a plurality of fixed waypoints while appropriately changing the position of one or more waypoints in the planned flight area surrounded by fixed waypoints. The position of one or more waypoints from which the path that minimizes is obtained is determined as the provisional position of the one or more waypoints. Therefore, it is possible to reduce the energy consumption required from the starting point to the reaching point as compared with the conventional technique in which the optimum flight network is determined with the waypoints fixed.

When a plurality of planned flight areas are set so as not to overlap, the provisional positions of one or more waypoints become appropriate positions as they are. When a plurality of planned flight areas are set in a partially overlapped state, the provisional positions of one or more waypoints at the overlapping positions are changed to new provisional positions by performing the second step again. There is. The provisional position becomes the proper position when the determination operation of one or more waypoints for all the planned flight areas is completed.

When one or more waypoints are one waypoint, it is the simplest and the time for determining the appropriate position is shortened.

Also, when setting multiple scheduled flight areas, if the next scheduled flight area is set so that the waypoints at the provisional position in the previously set scheduled flight area are adjacent to each other as fixed waypoints. Although the time required for positioning is shortened, it goes without saying that the next scheduled flight area may be set randomly.

It is preferable that a plurality of waypoints located on the outer peripheral portion of the three-dimensional map data are always fixed waypoints. In this way, it is not necessary to prepare map data for determining the appropriate positions of a plurality of waypoints located on the outer peripheral portion.

In the second step, optimization with energy consumption as the objective function can be performed.

In the second step, ant colony optimization or genetic algorithm is used when determining the route, and particle swarm optimization or grid search is used when moving one or more waypoints to determine the route and one or more. Waypoints can be easily determined.

In the ant colony optimization, the function of f = ax + by + cz can be used as a cost function corresponding to the energy consumption and the calculation can be performed. In this case, x is the horizontal movement distance of the flight path, y is the vertical movement distance of the flight path, z is the mean curvature of the flight path, and a, b and c are coefficients determined by the performance of the unmanned vehicle. Is. By using such a cost function, the route can be determined by a simple operation. The cost function is not limited to the above function.

It is a block diagram which shows one Embodiment of the generation system of the optimum flight network of this invention. It is a figure showing the flight schedule area on the 3D map determined by the flight schedule area determination part in the embodiment of FIG. It is the same figure as FIG. 2 which shows how the ant colony optimization part determines the optimum flight path from a moving waypoint in a planned flight area to each fixed waypoint. It is the same figure as FIG. 2 which shows the state which the ant colony optimization part moved the moving waypoint after the processing about the first moving waypoint of FIG. 3 was completed. It is the same figure as FIG. 2 which shows the state which the particle swarm optimization part determined the provisional position of a waypoint. It is the same figure as FIG. 2 which shows each waypoint of the next planned flight area after one provisional position is decided. It is a flowchart which shows the algorithm of the software when the generation system of the optimum flight network shown in FIG. 1 is carried out by using a computer. It is a flowchart which shows the detail of step 2 of FIG. It is the figure which represented the optimum flight network including three planned flight areas on 3D map data.

Hereinafter, an example of the method of generating the optimum flight network of the present invention and the embodiment of the system will be described with reference to the drawings.

In FIG. 1, the optimum flight network generation system of the present embodiment sets a plurality of way points as reference points of a route database for three-dimensional map data, and sets an unmanned flight object having an upper limit flight altitude. It is a system that creates a flight path between each waypoint that is suitable for flying with low energy consumption. Therefore, the optimum flight network generation system of the present embodiment includes a map data storage unit 10 and an appropriate position determination unit 12.

The map data storage unit 10 stores three-dimensional map data in which a plurality of waypoints are set. The three-dimensional map data includes elevation data in addition to the normal two-dimensional map data. As the altitude data, for example, the one generated by the land observation technology satellite can be used.

The appropriate position determination unit 12 determines appropriate positions of a plurality of waypoints and stores them in the map data storage unit 10.

The appropriate position determining unit 12 includes a planned flight area determining unit 14, a provisional position determining unit 16, and a waypoint setting unit 18. The scheduled flight area determination unit 14 determines one planned flight area in the three-dimensional map data read from the map data storage unit 10, and six waypoints surrounding the planned flight area are six fixed waypoints that do not move. To be determined. FIG. 2 shows a planned flight area on a three-dimensional map determined by the planned flight area determination unit 14. In FIG. 2, the flight path covers the entire three-dimensional map in a triangular mesh shape, and each intersection of the flight paths is a waypoint. Of the large number of waypoints, the area surrounded by the six fixed waypoints WPf1 to WPf6 is designated as the scheduled flight area, and one waypoint included in the planned flight area is designated as the moving waypoint WPm.

In the present embodiment, a plurality of waypoints located on the outer peripheral portion of the three-dimensional map data are always set as fixed waypoints. In the three-dimensional map data shown in FIG. 2, the fixed waypoints WPf1 to WPf6 are located on the outer peripheral portion of the planned flight area, and are always fixed weight points until the appropriate position of the moving waypoint WPm is determined. Therefore, it does not become a movement weight point. In this way, it is not necessary to prepare data for determining the appropriate position of the waypoints located on the outer peripheral portion. These fixed waypoints WPf1 to WPf6 are also moved when they become sequentially moving waypoints WPm and determine an appropriate position.

The provisional position determination unit 16 receives the flight schedule area set by the flight schedule area determination unit 14 and the three-dimensional map data, determines the optimum position of the movement waypoint WPm in the flight schedule area, and determines the position of the optimum movement waypoint WPm, and the provisional waypoint WPp. And. In the present embodiment, one provisional waypoint WPp is determined in one process from the viewpoint of saving resources and shortening the determination time. However, depending on the conditions, it is possible to provide a plurality of moving waypoints in one planned flight area and obtain a plurality of provisional waypoints WPp at the same time. In order to determine the provisional waypoint WPp, the provisional position determination unit 16 performs a two-step determination process in the ant colony optimization unit 20 and the particle swarm optimization unit 22.

Ant colony optimization is a probabilistic model of how ants efficiently move from their nests to their feeding grounds using pheromones. Ants go out to get food from their nest colonies, but when they find food, they bring back some of it, but at that time they return with pheromones. Then, when the ants who go looking for food find the pheromone, they move along the route in search of the food ahead and strengthen the pheromone on the way back. Pheromones evaporate over time, so a more suitable route is selected.

As shown in FIG. 3, the ant colony optimization unit 20 that executes the first step optimally flies the route that consumes the least energy in the flight route from the moving waypoint WPm to each fixed waypoint WPf1 to WPf6. Set as a route. Therefore, a total of six optimum flight paths are generated, one for each of the moving waypoints WPm and each fixed waypoint WPf1 to WPf6. That is, a route that minimizes energy consumption from the moving waypoint WPm corresponding to the colony to each fixed waypoint WPf1 to 6 is obtained by using the above-mentioned ant colony optimization method. Incidentally, the first selected moving waypoint WPm in FIG. 3 is on the intersection of the triangular mesh-shaped flight paths. In the present embodiment, which path has the minimum energy consumption is determined by an optimization operation with the energy consumption as the objective function when the unmanned aerial vehicle flies. That is, the route that minimizes the following cost function is set as the optimum flight route.

f = ax + by + cz (1)
In equation (1) representing the cost function,
x: Horizontal movement distance of the flight path y: Vertical movement distance of the flight path (integration)
z: Mean curvature of the flight path.

The longer the horizontal movement distance of the flight path, the more energy is consumed, the integrated horizontal movement distance (not the altitude difference between the start point and the end point) is directly reflected in the energy consumption, and the mean curvature is that the unmanned air vehicle bends. This is because energy is used for acceleration / deceleration. a, b and c are constants determined by the performance of the unmanned aerial vehicle, and are weight constants of numerical values such as 0.08, 0.8 and 0.01. The constant is determined for each model of unmanned aerial vehicle, and data can be acquired by actually flying the unmanned aerial vehicle, and it can also be derived from the specifications provided by the manufacturer. In addition, there is a considerable difference in the constant depending on whether the unmanned aerial vehicle is actually loaded or empty, but if the unmanned aerial vehicle goes to the destination with the actual load and returns to the starting point with the empty load, it is about half of the maximum load weight. Of course, the constant may be calculated with a heavy load loaded. By using such a cost function, the route can be determined by a simple operation.

The function storage unit 24 stores the above cost function, stores each constant input from the outside, puts each constant in the cost function, and outputs the cost function to the ant colony optimization unit 20. Note that the cost function is not limited to the above, but includes multiplication, adds other factors, or considers evaluation criteria other than energy consumption (for example, the time required to reach the destination). You may be.

As shown in FIG. 3, the ant colony optimization unit 20 first moves the moving waypoint WPm as shown in FIG. 4 when the optimum flight path for one moving waypoint WPm is obtained, and then moves as shown in FIG. Then, for each moving waypoint WPm, six optimum flight paths are obtained in the same manner as above. The position and number of moving waypoints WPm for which the optimum flight path is required in one process are not limited, but it is easiest to assume one moving waypoint WPm in one planned flight area. It is considered efficient to concentrate the moving points of the moving waypoint WPm near the center of the planned flight area without dispersing the initial positions throughout. Theoretically, it is adjusted according to the size of the planned flight area, resources, processing time, etc., but it is, for example, 100 locations.

Next, the particle swarm optimization unit 22 that executes the second step minimizes the average energy consumption of the energy consumed between the moving waypoint WPm and the six fixed waypoints WPf by the particle swarm optimization process. The position of one moving waypoint WPm from which the above route is obtained is discovered, and the position is determined as the provisional position of the provisional waypoint WPp.

Particle swarm optimization is a model of how one fish finds a good pathway in a school and the rest follow it. In a multidimensional space, a group of particles with position and velocity move around when trying to find an optimal solution. At this time, the position of the particle having the best value is notified to the whole, and the particle at the local best position is notified to the vicinity.

The particle swarm optimization unit 22 receives the positions of the plurality of moving way points WPm and the average energy consumption of the six optimum flight paths from the ant colony optimization unit 20, and performs particle swarm optimization processing based on these. That is, the position of the provisional waypoint WPp (= particle) that minimizes the average energy consumption is obtained, and the position (= particle) of the moving waypoint WPm at the local best position is notified to the moving waypoint that is in the vicinity in calculation. Each travel waypoint remembers the best position it has ever found. Then, the current position is compared with the best position, and if the best position is better, the moving waypoint adjusts the speed and returns to the best position. Based on these notifications, the position and speed of each moving waypoint WPm (= particle) are updated.

The above "average" in the present embodiment is a simple arithmetic mean, but it may be appropriate to use another average such as a geometric mean. For example, if the planned flight area is near the planned flight area, the unmanned aircraft should fly in the planned flight area in order to deliver the cargo carried in by other means of transportation such as trucks from the accumulated area to individual destinations. This is the case. Unmanned aerial vehicles heading away from such agglomerations are mostly unloaded, and unmanned aerial vehicles heading closer to the agglomerations are mostly empty. In this way, when it is expected that an unmanned aerial vehicle with an actual load and an empty load will fly in the same flight path in an extremely biased direction, the constant of the cost function is changed between the actual load and the empty load, and the direction is changed. By applying the harmonic mean in consideration, it can be expected to obtain the optimum flight route that matches the actual operation of the unmanned aerial vehicle.

The waypoint setting unit 18 receives the position of each particle whose position and velocity have been updated = the position of the moving waypoint WPm from the particle swarm optimization unit 22, and if reprocessing is required, the ant colony optimization unit 20 is used. Output, and if reprocessing is not required, replace the provisional position of the provisional waypoint WPp, which had the minimum average energy consumption for one scheduled flight area, with a new fixed waypoint WPf and output it to the planned flight area determination unit 14. To do. FIG. 5 shows a state in which the provisional position of the provisional waypoint WPp is determined. Finally, the provisional position is set as an appropriate position and output to the map data storage unit 10.

When the provisional position of one provisional waypoint WPp is determined, the flight schedule area determination unit 14 that receives this determines the flight schedule area to be processed next. At that time, in the present embodiment, in order to shorten the time required for position determination, the position of the provisional waypoint WPp is determined and updated in advance adjacent to the planned flight area including the updated fixed waypoint WPf. Set the planned flight area. Among them, it is more efficient to set the flight planned area surrounded by the previously updated fixed waypoints. As shown in FIG. 6, the provisional waypoint WPp of FIG. 5 is set as a new fixed waypoint WPf7, and a new planned flight area surrounded by the fixed waypoints WPf2, 4, 8 to 10 is set and included in the area. The fixed waypoint WPf3 that had been used is used as a new moving waypoint WPm, and the process of determining the provisional position is being continued. Needless to say, the planned flight area to be processed next may be randomly set or selected according to other rules as long as the processing time and the burden on resources are not a problem.

In other embodiments, other techniques such as genetic algorithms can be used instead of ant colony optimization when determining the pathway. Further, when moving one or more waypoints, other methods such as grid search can be used instead of particle swarm optimization.

Such an optimum flight network generation system can be configured by software including a personal computer control device, an arithmetic unit, a main storage device, peripheral devices such as an auxiliary storage device, and an OS. For ant colony optimization and particle swarm optimization, dedicated package software is on sale and available.

Next, the software algorithm used when the optimum flight network generation system according to the present embodiment is realized by using a computer will be described with reference to FIGS. 7 and 8. First, the planned flight area determination unit 14 of the appropriate position determination unit 12 reads out the three-dimensional map data in which a plurality of way points, which are initially the intersections of the triangular mesh-shaped flight paths, are set from the map data storage unit 10 (step). S1), as shown in FIG. 2, as a first step, a planned flight area surrounded by six fixed way points WPf1 to WPf6 that do not move is determined (step S2).

Next, as shown in FIG. 3 and below, the provisional position determination unit 16 has six fixed waypoints WPf1 surrounding the planned flight area within the planned flight area determined by the planned flight area determination unit 14 as a second step. The position of one moving waypoint WPm that consumes the least amount of energy for the unmanned aircraft to move between and WPf6 is determined as the provisional position of one provisional waypoint WPp (step S3).

In the provisional position determination in step S3, as shown in the flowchart of FIG. 7, the energy consumption from each of the plurality of movement points WPm (the position is determined in step S31) in the planned flight area to the fixed way points WPf1 to WPf6 is minimized. Step 32 of ant colony optimization to obtain each optimum path and the position of the moving way point WPm (= particle) that minimizes the average energy consumption are obtained, and this is used as the provisional way point WPp, and the position is the whole particle ( The particle swarm optimization step S34 which notifies the particle (way point) and the particle (way point) at the local best position to the neighboring particles, and updates the position and velocity of each particle based on these notifications. The provisional position is determined through two steps of optimization.

Next, it is determined whether the position of the provisional waypoint WPp may be determined (step S36). This determination is based on the number of times steps S32 and S34 are repeated in this embodiment. It is clear that the more the number of times the iterative process is performed, the closer the result is to the optimum, but it is preferable that the particle swarm optimization is completed at an appropriate number of times due to the problems of resources and processing time. In the present embodiment, when steps S32 and 34 are repeated 5 times for the same scheduled flight area, the position of the provisional waypoint WPp this time is the optimum waypoint position in the planned flight area at present. Judging that, a provisional position is decided. If the number of repetitions is less than 5, the position of each moved waypoint WPm moved in the particle swarm optimization in step S34 is updated (step S38) to the position of each updated moving waypoint WPm. The process of ant colony optimization (step 32) is repeated.

In other embodiments, the determination of whether or not the position of the provisional waypoint WPp may be determined is, for example, by comparing the previous provisional position with the current provisional position and making a slight difference such that they are the same or can be ignored. If this is not the case, it may be done based on the presumption that this provisional position is optimal.

When the provisional position determination in step S3 for one planned flight area is completed, it is determined whether or not there is an unprocessed planned flight area in the three-dimensional map data (step S4), and if there is, it is stored in the map data storage unit 10. The moving waypoint at the provisional position is changed to a fixed waypoint on the three-dimensional map data (step S5), the process returns to step S1 and steps S2 and subsequent steps are repeated for the new planned flight area.

When it is determined in step S4 that the processing for the entire planned flight area is completed, that is, the fixed waypoints other than the fixed waypoints located on the outermost periphery of the triangular mesh set in the 3D map data are moved waypoints. After determining the provisional position, it is determined whether or not the provisional position of the fixed waypoint determined so far may be set as an appropriate position (step S6). In the present embodiment, this determination is repeated three times. If the number of times is 2 or less, the process of determining the appropriate position again is executed for each planned flight area. Therefore, the positions of the provisional waypoints WPp differ between the first and second times, and the second and third times, but it is clear that they converge as the number of times increases, and in the present embodiment, repeat three times. It is presumed that the optimum result was reached. In other embodiments, the number of processes may vary depending on the size of the area of the three-dimensional map data and the number of planned flight areas. In another embodiment, the determination in step S6 is based on collation with the previous result.

When the provisional position is determined to be an appropriate position in step S6, the appropriate positions of the plurality of waypoints set in the three-dimensional map data are determined, written in the map data storage unit 10 (step S7), and the processing of the present embodiment is performed. Is finished.

FIG. 9 is an example of the optimum flight network generated by the method for generating the optimum flight network according to the present embodiment for the three planned flight areas. Initially, the three waypoints WP1 to WP1 to 3 which were regularly arranged in a triangular mesh shape were displaced from the initial positions as a result of performing the processes shown in FIGS. 7 and 8 using the system of FIG. The path to the fixed point on the outer circumference is no longer straight. Until the result of FIG. 9 was obtained, the optimization was first repeated 5 times for the three planned flight areas, and this was repeated 3 times. Therefore, the optimum flight network of FIG. 9 performed the optimization of steps S33 and 34 45 times in total. You have done it.

The flight path with the waypoint at the appropriate position written in the map data storage unit 10 as the intersection is expected to minimize energy consumption by the unmanned aircraft flying along the waypoint. By performing the above processing on the fixed way points located at the intersections of the triangular meshes on the 3D map data (excluding the fixed way points located on the outermost periphery of the correct map data), on the 3D map data. The optimum flight network can be set for all areas.

As described above, in the method of generating the optimum flight network of the present embodiment, since the optimum flight network can be generated by moving the waypoints, the optimum flight network has been determined with the waypoints fixed. Compared with the above technology, the energy consumption required from the starting point to the reaching point can be reduced.

When a plurality of planned flight areas are set so as not to overlap, the provisional positions of one or more waypoints become appropriate positions as they are. When a plurality of planned flight areas are set in a partially overlapped state, the provisional positions of one or more waypoints at the overlapping positions are changed to new provisional positions by performing the second step again. There is. The provisional position becomes an appropriate position when the determination operation of one or more waypoints for all the planned flight areas covering the three-dimensional map data is completed.

As described above, according to the method and system for generating the optimum flight network of the present invention, it is possible to determine the optimum flight path of the unmanned aerial vehicle that consumes less energy.

10 Map data storage unit 12 Appropriate position determination unit 14 Flight schedule area determination unit 16 Temporary position determination unit 18 Waypoint setting unit 20 Ant colony optimization unit 22 Particle swarm optimization unit 24 Cost function storage unit

Claims (12)

Using a computer, set multiple way points that serve as reference points for the route database for 3D map data, and use a computer to create a flight route suitable for flying an unmanned aircraft with a defined upper limit flight altitude with low energy consumption. It is a method of generating the optimum flight network generated between each way point.
The computer
The first step of determining one planned flight area in the three-dimensional map data and defining a plurality of waypoints surrounding the planned flight area as a plurality of fixed waypoints that do not move.
The unmanned air vehicle moves between the one or more waypoints and the plurality of fixed waypoints while changing the position of one or more waypoints in the planned flight area surrounded by the plurality of fixed waypoints. The one or more that finds the path that minimizes the energy consumption to move, and obtains the path that minimizes the average energy consumption of the energy consumption between the one or more waypoints and the plurality of fixed waypoints. The second step of determining the position of the waypoint as the provisional position of the one or more waypoints,
When each of the plurality of planned flight areas determined to cover the three-dimensional map data is executed and the plurality of planned flight areas are set in a non-overlapping state, the one or more waypoints are used. If the provisional position of is determined as the proper position as it is and the plurality of planned flight areas are set in a partially overlapped state, the second step is performed again to determine the provisional position, and the plurality of flight schedules are scheduled. When the operation of determining one or more waypoints with respect to the region is completed, the provisional positions of the one or more waypoints are determined as appropriate positions, so that the appropriateness of the plurality of waypoints set in the three-dimensional map data is appropriate. A method of generating an optimal flight network, which is characterized by determining a position. The method for generating an optimum flight network according to claim 1 , wherein the next scheduled flight area is set adjacent to the planned flight area including the one or more waypoints whose provisional position has been determined earlier. The method for generating an optimum flight network according to claim 1 , wherein a plurality of planned flight areas are randomly set. The method for generating an optimum flight network according to claim 1, wherein in the first step and the second step, optimization with energy consumption as an objective function is performed. The second step is characterized by using an ant colony optimization or genetic algorithm when determining the pathway and using particle swarm optimization or grid search when moving the one or more waypoints. The method for generating an optimum flight network according to claim 1. In the ant colony optimization, an operation is performed by using the function of f = ax + by + cz as a cost function corresponding to energy consumption.
However, x is the horizontal movement distance of the flight path, y is the vertical movement distance of the flight path, z is the mean curvature of the flight path, and a, b, and c are determined by the performance of the unmanned vehicle. The method for generating an optimum flight network according to claim 5 , wherein the flight network is a coefficient. Multiple way points that serve as reference points for the route database are set for the 3D map data, and a flight route suitable for flying an unmanned aircraft with a defined upper limit flight altitude with low energy consumption is set between each way point. It is a generation system of the optimum flight network to be generated in
A map data storage unit that stores the three-dimensional map data in which a plurality of waypoints are set, and a map data storage unit.
It is provided with an appropriate position determining unit that determines appropriate positions of the plurality of waypoints and stores them in the map data storage unit.
The proper position-fixing unit uses a computer
The first step of determining one planned flight area in the three-dimensional map data and defining a plurality of waypoints surrounding the planned flight area as a plurality of fixed waypoints that do not move.
The unmanned air vehicle moves between the one or more waypoints and the plurality of fixed waypoints while changing the position of one or more waypoints in the planned flight area surrounded by the plurality of fixed waypoints. The one or more that finds the path that minimizes the energy consumption to move, and obtains the path that minimizes the average energy consumption of the energy consumption between the one or more waypoints and the plurality of fixed waypoints. The second step of determining the position of the waypoint as the provisional position of the one or more waypoints,
When each of the plurality of scheduled flight areas defined to cover the three-dimensional map data is executed and the plurality of scheduled flight areas are set in a non-overlapping state, the one or more waypoints are used. If the provisional position of is determined as the proper position as it is and the plurality of scheduled flight areas are set in a partially overlapped state, the second step is performed again to determine the provisional position, and the plurality of scheduled flights are scheduled. When the operation of determining one or more waypoints with respect to the region is completed, the provisional positions of the one or more waypoints are determined as appropriate positions, so that the appropriateness of the plurality of waypoints set in the three-dimensional map data is appropriate. An optimal flight network generation system characterized by being configured to determine position. The optimal flight network generation system according to claim 7 , wherein the next scheduled flight area is set adjacent to the planned flight area including the one or more waypoints whose provisional position has been determined earlier. The optimal flight network generation system according to claim 7 , wherein the plurality of scheduled flight areas are randomly set. The optimal flight network generation system according to claim 7 , wherein in the first step and the second step, optimization with energy consumption as an objective function is performed. The second step is characterized by using an ant colony optimization or genetic algorithm when determining the pathway and using particle swarm optimization or grid search when moving the one or more waypoints. The optimal flight network generation system according to claim 7 . In the ant colony optimization, an operation is performed by using the function of f = ax + by + cz as a cost function corresponding to energy consumption.
However, x is the horizontal movement distance of the flight path, y is the vertical movement distance of the flight path, z is the mean curvature of the flight path, and a, b and c are determined by the performance of the unmanned vehicle. The optimal flight network generation system according to claim 11 , characterized in that it is a coefficient.