An unmanned aerial vehicle route planning method and system based on energy consumption

Abstract

The application belongs to the technical field of unmanned aerial vehicle task planning, and particularly relates to an unmanned aerial vehicle path planning method and system based on energy consumption. The second cost function value of a subnode via a parent node to an end point is calculated based on energy consumption, a battery state penalty factor, a path smoothness factor, an obstacle density factor, an assigned distance weight value and an energy weight value. The total cost function value of each subnode is calculated from the second cost function value and the first cost function value of the subnode via the parent node to a start point. The subnode with the minimum total cost function value is selected as a new parent node, and the process is repeated until the end point information is contained in the searched subnode information. The path formed by connecting all parent nodes is used as the optimal global path for unmanned aerial vehicle flight. The technical scheme of the application can effectively improve the safety of unmanned aerial vehicle flight and reduce the energy consumption of unmanned aerial vehicle flight.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) mission planning technology, specifically relating to an energy-consumption-based UAV trajectory planning method and system. Background Technology

[0002] In the military field, drones are widely used in intelligence reconnaissance, precision strikes, combat support, electronic warfare, and many other areas, enabling them to play an irreplaceable key role in modern warfare and driving continuous innovation and development in military technology.

[0003] Mission planning is a complex functional system encompassing battlefield analysis, mission allocation and trajectory planning, command and control, and many other aspects. It involves the top-level planning of the entire weapon system and related hardware and software design. Trajectory planning, based on existing start and end points and information about obstacles in between, plans a safe, efficient, and energy-saving flight route, including local path planning and global path planning.

[0004] Flight path planning is the foundation of mission allocation. In modern warfare, where air defense technology is becoming increasingly advanced and air defense systems are becoming increasingly sophisticated, flight path planning can effectively improve the safety, efficiency, and performance of UAVs, enabling autonomous navigation and efficient mission completion. It is an effective means to enhance the combat effectiveness of UAVs and to implement integrated reconnaissance and strike operations and long-range precision strikes.

[0005] In recent years, many scholars have proposed many feasible methods for UAV trajectory planning. For example, in the Chinese invention patent application publication CN121252821A, a three-dimensional path planning model and artificial bee colony algorithm are used to construct the optimal flight path of the UAV.

[0006] For example, in the Chinese invention patent application publication CN121209396A, a bidirectional fast search tree algorithm is used, and the dynamic association of target bias and path cost is adopted to obtain the optimal search path.

[0007] Each of the above methods has its advantages and disadvantages. While they can effectively plan the flight path of a drone, drones are aircraft with limited range. If the planned route is too long and the battery capacity is limited, the drone may encounter difficulties returning to its home location. Furthermore, these methods are not very effective at avoiding dangerous areas and obstacles encountered during flight. If avoidance is not timely, it can easily lead to flight accidents.

[0008] Therefore, further research and improvement of UAV trajectory planning methods are needed to enhance UAV trajectory planning capabilities and safety performance. Summary of the Invention

[0009] The purpose of this invention is to provide a method and system for planning the flight path of unmanned aerial vehicles (UAVs) based on energy consumption.

[0010] To achieve the above objectives, the technical solution of the present invention is as follows: An energy-consumption-based UAV trajectory planning method includes the following steps: S1: Rasterize the internal map of the drone, marking the start point, end point, the location of obstacles, and the size of the obstacle area; S2: Starting from the origin in the internal map, traverse several neighboring child nodes, calculate the total cost function value of each child node, and select the child node with the smallest total cost function value as the new parent node. The total cost function value is calculated based on the first cost function value from the child node to the starting point via the parent node and the second cost function value from the child node to the ending point via the parent node. The second cost function value from the child node to the ending point is calculated based on the straight-line distance from the child node to the parent node, energy consumption, battery state penalty factor, path smoothness factor, obstacle density factor, assigned distance weight value, and energy weight value. Energy consumption is calculated based on the straight-line distance, vertical movement distance, energy consumption value, number of turns, and turning energy consumption. The battery state penalty factor is calculated based on the battery consumption ratio factor, which is calculated based on the total flight consumption and current battery level. The path smoothness factor is calculated based on the vertical and horizontal distances. The obstacle density factor is calculated based on the obstacle density value at the location of the child node. S3: Iterate through the child nodes of the new parent node again, calculate the total cost function value of the child nodes, and select the child node with the smallest total cost function value (excluding the child nodes that have been searched) as the new parent node. S4: Repeat step S3 until there is endpoint information in the surrounding nearest child nodes, complete one global path planning, and connect all parent nodes in sequence to form the optimal global path for the drone flight.

[0011] Furthermore, the value of the second cost function in S2 The calculation is performed using the following formula: in This represents the straight-line distance from child node n to parent node. This represents the energy consumption from child node n to parent node. Represents the obstacle density factor. Indicates the battery state penalty factor. Represents the path smoothing factor. This represents the distance weight value, which is 0.4. This represents the energy weight value, which is 0.6, and n is a natural number greater than 0.

[0012] Energy consumption in S2 The calculation is performed using the following formula: in Indicates straight-line distance. Indicates the vertical movement distance. Indicates the number of turns. This represents the energy consumption over a straight-line distance, with a value of 1.0. This represents the energy consumption value for vertical movement distance, with a value of 1.5. This represents the energy consumption value during turning, with a value of 0.5.

[0013] Obstacle density factor in S2 The calculation is performed using the following formula: in This represents the obstacle density value at the location of child node n, with a value range of [0, 1].

[0014] Battery consumption ratio factor in S2 The calculation is performed using the following formula: in Indicates the total flight consumption. This indicates the current battery level.

[0015] Battery state penalty factor in S2 The calculation is performed using the following formula: When >0.9, When the value is in the range [0.7, 0.9], When the value is in the range [0.5, 0.7], When ≤0.4, If the energy consumption is greater than 1, the flight energy consumption exceeds the current battery power, and the drone will not be able to return after completing the mission.

[0016] Vertical distance from child node to parent node in S2 and horizontal distance The calculation is performed using the following formula: in , , This represents the three-dimensional coordinates of child node n. , , Represents the three-dimensional coordinates of the parent node.

[0017] Path smoothness factor The calculation is performed using the following formula: < *20% =0.95 In other cases =1 The first cost function value in S2 The calculation is performed using the following formula: in, The actual cost from the starting point to the parent node is represented by the following formula: in, , , Represents the three-dimensional coordinates of the parent node. The single-step cost of moving from the parent node to the child node n is expressed by the following formula: in, , , This represents the three-dimensional coordinates of child node n. , , This represents the three-dimensional coordinates of the parent node.

[0018] An energy-consumption-based UAV trajectory planning system includes the following modules: Map annotation module: Rasterizes the internal map of the drone, and annotates the start point, end point, location of obstacles, and size of obstacle areas.

[0019] Cost function value calculation module: Starting from the origin in the internal map, it traverses several neighboring child nodes, calculates the total cost function value of each child node, and selects the child node with the smallest total cost function value as the new parent node. The total cost function value is calculated based on the first cost function value from the child node to the starting point via the parent node and the second cost function value from the child node to the ending point via the parent node. The second cost function value from the child node to the ending point is calculated based on the straight-line distance from the child node to the parent node, energy consumption, battery state penalty factor, path smoothness factor, obstacle density factor, assigned distance weight value, and energy weight value. Energy consumption is calculated based on the straight-line distance, vertical movement distance, energy consumption value, number of turns, and turning energy consumption. The battery state penalty factor is calculated based on the battery consumption ratio factor, which is calculated based on the total flight consumption and the current battery level. The path smoothness factor is calculated based on the vertical distance and horizontal distance. The obstacle density factor is calculated based on the obstacle density value at the location of the child node.

[0020] Parent node selection module: Iterate through the nearest child nodes of the new parent node again, calculate the total cost function value of the child nodes, and select the child node with the smallest total cost function value (excluding the child nodes that have already been searched) as the new parent node.

[0021] Path decision module: Repeat step S3 until there is destination information in the surrounding nearest child nodes, complete a global path planning, and connect all parent nodes in sequence to form the optimal global path for the UAV flight.

[0022] Compared with the prior art, the technical solution provided by this invention has the following advantages: Energy consumption calculations, battery state penalty factors, path smoothness factors, and obstacle density factors were introduced. In the energy consumption calculations, the number of turns and turn energy consumption were incorporated to reduce the number of turns during drone flight and select more energy-efficient flight paths.

[0023] Given the high inertia of drones, excessive turning is inconsistent with the kinematic characteristics and inertia of drone flight. Introducing energy consumption can effectively improve the safety of drone flight.

[0024] Introducing a battery state penalty factor can ensure that the UAV completes its mission and returns safely, while introducing a path smoothness factor can reduce energy consumption and improve flight stability. Attached Figure Description

[0025] Figure 1 This is a three-dimensional path planning diagram of the present invention; Figure 2 This is a 3D path planning graph based solely on the straight-line distances between nodes; Figure 3 This is a 3D path planning diagram for the RRT algorithm. Detailed Implementation

[0026] To further understand the content of this invention, the invention will be described in detail with reference to the embodiments. Example 1 S1: Rasterize the internal map of the drone, marking the start point, end point, the location of obstacles, and the size of the obstacle area; S2: Starting from the origin in the internal map, traverse several neighboring child nodes, calculate the total cost function value of each child node, and select the child node with the smallest total cost function value as the new parent node. The total cost function value is calculated based on the first cost function value from the child node to the starting point via the parent node and the second cost function value from the child node to the ending point via the parent node. The second cost function value from the child node to the ending point is calculated based on the straight-line distance from the child node to the parent node, energy consumption, battery state penalty factor, path smoothness factor, obstacle density factor, assigned distance weight value, and energy weight value. Energy consumption is calculated based on the straight-line distance, vertical movement distance, energy consumption value, number of turns, and turning energy consumption.

[0027] The battery state penalty factor is calculated based on the battery consumption ratio factor, which is calculated based on the total flight consumption and the current battery level.

[0028] The path smoothness factor is calculated based on the vertical and horizontal distances.

[0029] The obstacle density factor is calculated based on the obstacle density value at the location of the child node.

[0030] Create an OPEN table and add the starting point information to it. Create a Childnode table and an OpenList table to store child node information and the total cost function value for each child node, respectively. The Childnode table is a three-dimensional array used to store the three-dimensional coordinates of the child nodes, and the OpenList table is also a three-dimensional array, storing the first cost function. Second cost function Sum of total cost functions The value of .

[0031] The traditional definition of the total cost function value is: ,in This represents the first cost function value from the child node to the starting point via the parent node. The second cost function value represents the distance from the child node to the destination node via the parent node; the improved second cost function value. The calculation is performed using the following formula: in This represents the straight-line distance from child node n to parent node. This represents the energy consumption from child node n to parent node. Represents the obstacle density factor. Indicates the battery state penalty factor. Represents the path smoothing factor. This represents the distance weight value, which is 0.4. This represents the energy weight value, which is 0.6, and n is a natural number greater than 0.

[0032] Energy consumption The calculation is performed using the following formula: in Indicates straight-line distance. Indicates the vertical movement distance. Indicates the number of turns. This represents the energy consumption over a straight-line distance, with a value of 1.0. This represents the energy consumption value for vertical movement distance, with a value of 1.5. This represents the energy consumption value during turning, with a value of 0.5.

[0033] Obstacle density factor The calculation is performed using the following formula: in This represents the obstacle density value at the location of child node n. The calculation method is as follows: First, a three-dimensional density map is established, and the proportion of obstacles within a fixed radius (set to 2 grids in this paper) around each grid point is calculated. =Number of obstacles / Total number of grid cells checked. The obstacle density value ranges from 0 to 1, where 0 indicates no obstacles nearby, and 1 indicates the area is completely surrounded by obstacles. The final obstacle density factor is... The value ranges from 1 to 3. In a completely open region, This means that the value of the second cost function is not increased when obstacles are dense. This indicates a significant increase in the value of the second cost function. The advantage of this design is that the drone can choose to fly in areas with low obstacle density, while not completely excluding the possibility of passing through areas with more obstacles, but simply increasing its cost assessment.

[0034] The battery state penalty factor is a key adjustment factor, crucial for ensuring the drone can complete its mission and return home. This paper defines a battery consumption ratio factor. , in, Indicates the total flight consumption. Indicates the current battery level. When this happens, it means the flight energy consumption exceeds the current battery charge, and the drone will not be able to return after completing its mission; when When the value is close to 1, it means that the energy consumption of flight is close to the current battery level, and the drone may not be able to return. Battery state penalty factor. The calculation is divided into the following cases, when When the value is greater than 0.9, the drone's energy consumption is almost equal to the current battery capacity, making it difficult for the drone to return. The possible values ​​are: , when When the value is between 0.7 and 0.9, it indicates that the battery power is barely enough for the drone to complete its mission and return. The value can be: , when A value between 0.5 and 0.7 indicates that the drone has enough remaining battery power to return. The value can be: , when When the value is below 0.4, then... The value can be: .

[0035] Path smoothness factor The design aims to encourage algorithms to generate smoother flight paths, reducing unnecessary high-frequency turns and frequent altitude changes, thereby reducing energy consumption and improving flight stability. The path smoothness factor... Primarily based on vertical distance and horizontal distance The calculation shows that its and The calculation formula is: in , , This represents the three-dimensional coordinates of child node n. , , Represents the three-dimensional coordinates of the parent node. Path smoothness factor. The calculation formula is: when the vertical distance Less than horizontal distance At 20%, In other cases The advantage of this design is that when the vertical distance is less than 20% of the horizontal distance, A smoothness factor of 0.95 means the total cost function will tend to select a smoother horizontal flight path, avoiding unnecessary vertical fluctuations. Since our previously designed energy consumption value for vertical movement is 1.5, a smooth flight path typically consumes less energy, reducing the frequency of acceleration, deceleration, and changes in direction, thus saving battery power. This is especially important for drones with limited batteries, improving the overall safety and reliability of the flight mission. This smoothness factor differs from other factors (especially the battery penalty factor). They work together to guide the algorithm to select more energy-efficient and more stable flight paths while meeting mission requirements.

[0036] S3: Iterate through the nearest child nodes of the new parent node again, calculate the total cost function value of the child nodes, and select the child node with the smallest total cost function value (excluding the child nodes that have already been searched) as the new parent node.

[0037] Starting from the origin, the drone traverses its six nearest neighboring nodes (up, down, left, right, front, and back), adding the 3D coordinates of each child node to the Childnode table. It then sequentially evaluates the total cost function value of each of the six surrounding child nodes and assigns the corresponding coordinates to each child node. , and The value is stored in the OpenList table. The child node with the smallest total cost function value is selected as the new parent node. This new parent node is the next position that the drone will reach. The calculated new parent node is added to the OPEN table.

[0038] S4: Repeat step S3 until there is endpoint information in the surrounding nearest child nodes, complete one global path planning, and connect all parent nodes in sequence to form the optimal global path for the drone flight.

[0039] Starting with the new parent node, the information in the previous Childnode and OpenList tables is cleared. Then, the system continues to traverse the six nearest child nodes surrounding the new parent node, recalculating their total cost function value. The position information of the new child node is added to the Childnode table, and the cost function value is placed in the OpenList table. This process continues until the drone finds its destination position among the child nodes. This design improves search efficiency and reduces energy consumption. Finally, the position information of all found parent nodes is added to the OPEN table, which stores the node positions along all the drone's flight paths. The path formed by connecting these node positions serves as the optimal global path for the drone. Furthermore, during the search for surrounding child nodes of the new parent node, if a child node from the previous step exists, it is not searched again in the current parent node search. This design reduces unnecessary computation and effectively improves the efficiency of drone flight path planning. The location information stored in the OPEN table represents the inflection points during the drone's flight. After introducing the path smoothing factor constraint, the planned flight path of the drone is a smooth path, which is more in line with the kinematic and dynamic characteristics of drone flight and also more in line with the inertia of drone flight. Therefore, the drone no longer needs to use more energy to counteract its own inertia and nonlinear kinematic characteristics, allowing more of the drone's energy consumption to be used in path flight.

[0040] Example 2 Simulation and analysis were performed using MATLAB 2024a. The computer's processor was an Intel(R) Core(TM) i5-10300H CPU @ 2.50GHz, and the memory was 8.0GB. Environmental modeling used a 3D raster map, with the raster map size set to 1.

[0041] In the simulation of this invention, since the algorithm's running time is related to factors such as the computer's CPU frequency and memory usage, the average of 50 program runs is taken as the algorithm's running time. The distance metric is measured in grid size. The simulation experiment was conducted on a 3D map with dimensions (20, 20, 10), a starting point (1, 20, 1), and an ending point (20, 1, 8). The obstacles were set as tall buildings, and their setting format was […]. , , , , , The obstacles are: obstacle 1 [9, 11, 1, 8, 1, 8], obstacle 2 [3, 5, 3, 6, 1, 3], obstacle 3 [16, 18, 16, 18, 1, 5], and obstacle 4 [16, 18, 3, 5, 1, 4]. The initial battery level of the drone is set to 90% of its maximum design battery level.

[0042] This experiment was conducted in three groups. The first group used the method described in this application, and the simulation results are as follows: Figure 1 The second method, based on the method of this application, calculates only the straight-line distance between the child node and the parent node for the second cost function value. The simulation results are as follows: Figure 2 Group 3 uses the commonly used RRT algorithm, and the simulation results are as follows. Figure 3 .

[0043] Figure 1 and Figure 2 In the diagram, green dots represent the starting point of the UAV's planned path, red dots represent the ending point, yellow dots represent all parent nodes in the path planning, solid blue lines represent the planned UAV flight path, and dashed green lines represent the straight-line distance from the starting point to the ending point. Simulation results show that this invention effectively improves the safety and smoothness of UAV flight and significantly reduces energy consumption.

Claims

1. A method for energy-based UAV path planning, characterized in that: Includes the following steps: S1: Rasterize the internal map of the drone, marking the start point, end point, the location of obstacles, and the size of the obstacle area; S2: Starting from the origin in the internal map, traverse several neighboring child nodes, calculate the total cost function value of each child node, and select the child node with the smallest total cost function value as the new parent node. The total cost function value is calculated based on the first cost function value from the child node to the starting point via the parent node and the second cost function value from the child node to the ending point via the parent node. The second cost function value from the child node to the ending point is calculated based on the straight-line distance from the child node to the parent node, energy consumption, battery state penalty factor, path smoothness factor, obstacle density factor, assigned distance weight value, and energy weight value. Energy consumption is calculated based on the straight-line distance, vertical movement distance, distance energy consumption value, number of turns, and turn energy consumption value. The battery state penalty factor is calculated based on the battery consumption ratio factor, which is calculated based on the total flight consumption and the current battery level. The path smoothness factor is calculated based on the vertical and horizontal distances. The obstacle density factor is calculated based on the obstacle density value at the location of the child node; S3: Iterate through the child nodes of the new parent node again, calculate the total cost function value of the child nodes, and select the child node with the smallest total cost function value (excluding the child nodes that have been searched) as the new parent node. S4: Repeat step S3 until there is endpoint information in the surrounding nearest child nodes, complete one global path planning, and connect all parent nodes in sequence to form the optimal global path for the drone flight.

2. The UAV trajectory planning method according to claim 1, characterized in that: S2 in the second cost function value The calculation of the second cost function value S2 in S2 is realized by the following formula: wherein represents a straight-line distance from the sub-node n to the parent node, represents an energy consumption from the sub-node n to the parent node, represents an obstacle density factor, represents a battery state penalty factor, represents a path smoothness factor, represents a distance weight value, which is 0.4, represents an energy weight value, which is 0.6, and n is a natural number greater than 0.

3. The UAV trajectory planning method according to claim 1, characterized in that: Energy consumption in S2 The calculation is achieved by the following formula: wherein represents a straight line distance, represents a vertical movement distance, represents a number of turns, represents a straight line distance energy consumption value, which is 1.0, represents a vertical movement distance energy consumption value, which is 1.5, represents a turn energy consumption value, which is 0.

5.

4. The UAV trajectory planning method according to claim 1, characterized in that: S2 obstacle density factor The calculation is achieved by the following formula: wherein represents the obstacle density value at the location of the child node n, taking the value range [0, 1].

5. The UAV trajectory planning method according to claim 1, characterized in that: S2 battery consumption proportionality factor The calculation of the S2 battery consumption proportionality factor is achieved by the following formula: in Indicates the total flight consumption. This indicates the current battery level.

6. The UAV trajectory planning method according to claim 5, characterized in that: Battery state penalty factor in S2 The calculation is performed using the following formula: When >0.9, When the value is in the range [0.7, 0.9], When the value is in the range [0.5, 0.7], When ≤0.4, If the energy consumption is greater than 1, the flight energy consumption exceeds the current battery power, and the drone will not be able to return after completing the mission.

7. The UAV trajectory planning method according to claim 1, characterized in that: Vertical distance in S2 and horizontal distance The calculation is performed using the following formula: in , , This represents the three-dimensional coordinates of child node n. , , Represents the three-dimensional coordinates of the parent node.

8. The UAV trajectory planning method according to claim 7, characterized in that: Path smoothness factor The calculation is performed using the following formula: < *20% =0.95 In other cases =1.

9. The UAV trajectory planning method according to claim 1, characterized in that: The first cost function value in S2 The calculation is performed using the following formula: in, The actual cost from the starting point to the parent node is represented by the following formula: in, , , This represents the three-dimensional coordinates of the parent node. The single-step cost of moving from the parent node to the child node n is expressed by the following formula: in, , , This represents the three-dimensional coordinates of child node n. , , Represents the three-dimensional coordinates of the parent node.

10. An energy-consumption-based unmanned aerial vehicle (UAV) trajectory planning system, characterized in that: Includes the following modules: Map annotation module: Rasterizes the internal map of the drone, annotating the start point, end point, location of obstacles, and size of obstacle areas; Cost function value calculation module: Starting from the origin in the internal map, it traverses several neighboring child nodes, calculates the total cost function value of each child node, and selects the child node with the smallest total cost function value as the new parent node. The total cost function value is calculated based on the first cost function value from the child node to the starting point via the parent node and the second cost function value from the child node to the ending point via the parent node. The second cost function value from the child node to the ending point is calculated based on the straight-line distance from the child node to the parent node, energy consumption, battery state penalty factor, path smoothness factor, obstacle density factor, assigned distance weight value, and energy weight value. Energy consumption is calculated based on the straight-line distance, vertical movement distance, energy consumption value, number of turns, and turning energy consumption. The battery state penalty factor is calculated based on the battery consumption ratio factor, which is calculated based on the total flight consumption and the current battery level. The path smoothness factor is calculated based on the vertical and horizontal distances. The obstacle density factor is calculated based on the obstacle density value at the location of the child node; Parent node selection module: Iterate through the nearest child nodes of the new parent node again, calculate the total cost function value of the child nodes, and select the child node with the smallest total cost function value (excluding the child nodes that have already been searched) as the new parent node. Path decision module: Repeat step S3 until there is destination information in the surrounding nearest child nodes, complete a global path planning, and connect all parent nodes in sequence to form the optimal global path for the UAV flight.

Images