Unmanned aerial vehicle path planning system and method for emergency rescue

Abstract

The invention discloses an unmanned aerial vehicle path planning system and method for emergency rescue, and belongs to the field of unmanned aerial vehicle auxiliary emergency rescue, and the method comprises the steps: S1, constructing an unmanned aerial vehicle path planning scene graph, and defining the space and time constraints of a flight task; s2, defining and reconstructing a fitness function to optimize a path planning target; s3, considering sensor access limitation and path closing requirements, and establishing related constraint conditions; s4, establishing a mathematical model of a path planning problem according to system requirements; and S5, solving a path planning problem by adopting a genetic algorithm according to the established mathematical model. The problem that in the prior art, when an unmanned aerial vehicle executes a task, the flight duration and the flight distance are limited due to energy limitation is solved. By optimizing the path planning, the flight efficiency and the quality of the path planning are improved, and effective technical support is provided for the application of the unmanned aerial vehicle in emergency rescue actions.

Description

Technical Field

[0001] The present invention relates to the field of drone-assisted emergency rescue, and more specifically, to a drone path planning system and method for emergency rescue. Background Art

[0002] In today's society, in the field of emergency rescue, drones are gradually becoming an important rescue force. Even though drones now have broad development prospects in data collection applications in emergency rescue, the drone is always limited by its advantage (small size), which restricts the fuel it can carry, resulting in a limited maximum flight distance of the drone. Therefore, it is necessary to optimize the flight path of the drone during data collection in emergency situations. As described above, in the ground-air collaborative rescue scenario of the drone for emergency rescue information collection, optimizing the data collection path of the drone will greatly improve the data collection efficiency of the drone and enable the drone to have better coordination with ground workers.

[0003] Due to the variability of the climate and the influence of clouds and fog, the image clarity obtained by traditional satellite and aerial photography technologies is often unsatisfactory. In this context, drone technology, with its advantages of low-altitude flight, less influence of clouds and fog, and high image resolution, demonstrates its important role in disaster monitoring and emergency rescue. Drones can not only fly for a long time but also carry out large-scale maneuvering operations, providing flexible monitoring capabilities, quickly obtaining a wide range of situations after a disaster, such as the degree of disaster, traffic conditions, and potential risks of secondary disasters, while ensuring the safety of the personnel performing the tasks and effectively preventing casualties.

[0004] Drone remote sensing technology, as the core of drone applications, is an advanced technology integrating communication, positioning, sensing, and applications. This technology realizes the precise control of drones through ground and airborne control systems, using technologies such as positioning, communication, and sensors. The ground control center sends flight instructions to the drone through wireless communication, collects data, and performs processing, modeling, and analysis of remote sensing data to achieve the automation and intelligent operation of the drone. Compared with satellite and manned aerial remote sensing, drone remote sensing has obvious advantages in temporal and spatial resolution and can respond to diverse rescue needs in real time. In addition, the flexibility of the drone platform allows it to carry a variety of sensor devices, such as high-definition cameras, thermal imagers, infrared night vision devices, and emergency communication devices, quickly replace the devices according to actual needs, obtain high-quality image data, and transmit the data with high efficiency, greatly improving the efficiency and success rate of rescue operations.

[0005] The applications of these drones have greatly improved the rescue efficiency and ensured that rescue supplies can be quickly and accurately delivered to people in need.

[0006] The prior art has the following disadvantages: Due to the damage of buildings and the impassability of roads in the disaster area, the sensors deployed on the ground cannot complete information collection in the first time, thus affecting the rescue efficiency. In addition, due to the limited energy of the drone itself, the flight duration and flight distance are restricted, so path planning has become a key issue to solve the problem. How to achieve the maximum information collection efficiency with the shortest flight path is a technical problem to be solved urgently. Summary of the Invention

[0007] The purpose of the present invention is to overcome the deficiencies of the prior art, and provide a drone path planning system and method for emergency rescue, which solves the problem that the prior art fails to well combine optimization methods with data processing efficiency.

[0008] The purpose of the present invention is achieved through the following solutions:

[0009] A drone path planning system for emergency rescue, comprising:

[0010] A drone data collection network composed of drones and ground sensor nodes, wherein the drones are equipped with sensors and have data processing and storage modules;

[0011] The drone is used for data collection, and is equipped with corresponding sensor devices to achieve precise monitoring and data collection of various sensors deployed on the ground;

[0012] A connection is established between the drone and the sensors deployed on the ground through the data processing and storage module, for receiving, processing and storing the data transmitted from the ground nodes.

[0013] Furthermore, the sensors deployed on the ground are randomly distributed within a set area and are all on the same horizontal plane; the drone flies directly above the target area to complete comprehensive data collection of all sensors.

[0014] A drone path planning method for emergency rescue, based on the drone path planning system for emergency rescue described in any one of the above, comprises the following steps:

[0015] Step S1: Construct a drone path planning scenario graph, and define the spatial and temporal constraints of the flight mission;

[0016] Step S2: Define and reconstruct the fitness function to optimize the path planning objective;

[0017] Step S3: Consider the sensor access restrictions and the path closure requirements, and establish relevant constraint conditions;

[0018] Step S4: According to the system requirements, establish a mathematical model for the path planning problem;

[0019] Step S5: Solve the path planning problem using the genetic algorithm according to the established mathematical model.

[0020] Further, in step S1, when constructing the UAV path planning scenario graph and defining the spatial and temporal constraints of the flight mission, it specifically includes the following sub-steps:

[0021] Considering a UAV data collection network composed of a UAV for emergency rescue and X ground sensor nodes, assuming that there are n randomly distributed data collection points within an interval, and they are all randomly distributed on this horizontal plane, where X i ={x i ,y i} represents the two-dimensional coordinates of the i-th sensor, x=(x 1 ,x 2 ,...,x n ) represents the order in which the UAV passes through each sensor; d(x i ,x i+1 ) is the distance between sensor x i and sensor x i+1 ; n is the number of randomly distributed data collection points; d(x n ,x 1 ) is the distance from the position of the last sensor to be collected to the starting point to ensure that the UAV can successfully fly back to the starting point;

[0022] In the information collection of the UAV for emergency rescue, assuming that the UAV always flies horizontally at a constant speed and the UAV needs to fly directly above the target to perform complete information collection, then when the flight path of the UAV is the shortest, its time consumption is also the shortest; by default, the starting point is also regarded as a sensor that needs to collect information, that is, the starting point is x 1 .

[0023] Further, in step S2, when defining and reconstructing the fitness function to optimize the path planning objective, it specifically includes the following sub-steps:

[0024] Let the path planning objective be to minimize the total path distance, and the fitness function is associated with the objective function;

[0025] The fitness function F is defined based on the objective function f, where f is the total path distance of the sensor sequence; assume that the fitness value increases as the quality of the solution improves, and define the fitness function as a function of the reciprocal of the objective function;

[0026] Assume that the objective function f(x) is the total travel distance of the individual x, and the fitness function F is expressed as:

[0027]

[0028] Add a small constant ∈ to ensure that the denominator is not zero:

[0029]

[0030] where ∈ is a constant close to zero.

[0031] Furthermore, in step S3, the establishment of the relevant constraint conditions specifically includes the sub-step: The sensor sequence cannot have duplicate sensors.

[0032] Furthermore, in step S3, the establishment of the relevant constraint conditions specifically includes the sub-step: The path starts and ends at the same sensor.

[0033] Furthermore, in step S4, the establishment of the mathematical model of the path planning problem specifically includes the following sub-steps:

[0034] Set the minimization objective function f(x), that is, the total length of the path, expressed as:

[0035]

[0036] Here, d(x i ,x i+1 ) is the distance between sensor x i and sensor x i+1 , d(x n ,x 1 ) represents the distance from the last sensor to the first sensor, ensuring that the UAV can fly back to the starting point; introduce the binary decision variable x ij to represent whether there is a path from sensor i to sensor j; where, x ij = 1 means there is a path from sensor i to sensor j, and x ij = 0 means there is no path; constraint C1 means that each sensor is visited only once; constraint C2 means that the path must start from the starting point, visit all other sensors and return to the starting point; constraint C3 ensures that the path starts from the starting point and finally returns to the starting point.

[0037] Furthermore, in step S5, the solution of the path planning problem by using the genetic algorithm according to the established mathematical model specifically includes the following sub-steps:

[0038] Step S51: Initialization, randomly generate the initial population, and each individual represents a possible sequence of sensors;

[0039] Step S52: Evaluate and calculate the fitness of each individual;

[0040] Step S53: Select individuals for reproduction according to the fitness;

[0041] Step S54: Generate new individuals through crossover operation, and exchange partial sequences of two individuals;

[0042] Step S55: Randomly change some parts of the individuals to introduce new mutations;

[0043] Step S56: Repeat the processes of evaluation, selection, crossover, and mutation until the stopping condition is met.

[0044] Furthermore, in step S53, individuals with higher fitness have a higher probability of being selected.

[0045] The beneficial effects of the present invention include:

[0046] The present invention solves the problem that in the prior art, the flight duration and flight distance of drones are limited due to energy constraints during mission execution. By optimizing path planning, the flight efficiency and the quality of path planning are improved, providing effective technical support for the application of drones in emergency rescue operations.

[0047] The present invention combines drone data sensing and path planning with an optimization algorithm (genetic algorithm), achieving the goal of maximizing geographical fairness and data volume weight while ensuring the quality of data collection.

[0048] During the search process, the present invention can effectively avoid local optimal solutions and quickly converge to the global optimal solution, thus providing an economical and efficient flight path for the drone in a relatively short time. Description of the Drawings

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0050] Figure 1 It is a scenario diagram of the drone path planning for emergency rescue in the embodiment of the present invention;

[0051] Figure 2 It is a sensor distribution diagram in the embodiment of the present invention;

[0052] Figure 3 It is the 16th iteration diagram in the embodiment of the present invention;

[0053] Figure 4 It is the 32nd iteration diagram in the embodiment of the present invention;

[0054] Figure 5 It is the 64th iteration diagram in the embodiment of the present invention;

[0055] Figure 6 This is the 154th iteration graph in the embodiment of the present invention;

[0056] Figure 7 This is the 197th iteration graph in the embodiment of the present invention;

[0057] Figure 8 This is the 266th iteration graph in the embodiment of the present invention;

[0058] Figure 9 This is the shortest path convergence graph in the embodiment of the present invention;

[0059] Figure 10 This is the 583rd iteration graph with 10,000 iterations in the embodiment of the present invention;

[0060] Figure 11 This is the 5901st iteration graph with 10,000 iterations in the embodiment of the present invention. Detailed implementation manners

[0061] All features disclosed in all embodiments in this specification, or all steps in any disclosed methods or processes, except for mutually exclusive features and / or steps, can be combined and / or extended, replaced in any manner.

[0062] In a preferred embodiment, the present invention specifically provides a UAV path planning system for emergency rescue, including: a UAV data acquisition network composed of UAVs and ground sensor nodes, where the UAVs are equipped with sensors and have data processing and storage modules.

[0063] The UAVs in the solution of this embodiment are used for data acquisition, carrying corresponding sensor devices to achieve precise monitoring and data acquisition of various sensors deployed on the ground.

[0064] The UAVs in the solution of this embodiment establish connections with the sensors deployed on the ground through data processing and storage modules, and are used to receive, process, and store the data transmitted from the ground nodes.

[0065] The ground sensors in the solution of this embodiment are randomly distributed within a certain area and are all on the same horizontal plane. The UAV needs to fly directly above the target area to complete the comprehensive data acquisition of all sensors.

[0066] In other preferred embodiments of the present invention, based on the above system, a UAV path planning method for emergency rescue is further provided, including the following steps:

[0067] Step S1: Construct a UAV path planning scenario graph and define the spatial and temporal constraints of the flight mission;

[0068] Step S2: Define and reconstruct the fitness function to optimize the path planning objective;

[0069] Step S3: Consider the sensor access limitations and the path closure requirement, and establish relevant constraint conditions;

[0070] Step S4: Establish a mathematical model for the path planning problem according to the system requirements;

[0071] Step S5: Solve the path planning problem using a genetic algorithm according to the established model.

[0072] In a further implementation manner of the above method embodiment, in step S1, considering a UAV data acquisition network composed of a UAV for emergency rescue and X ground sensor nodes, assuming that there are n randomly distributed data acquisition points within the interval, and they are all randomly distributed on this horizontal plane, where X i ={x i , y i} represents the two-dimensional coordinates of the i-th sensor, x=(x 1 , x 2 ,..., x n ) represents the sequential arrangement (i.e., individual or solution) of the UAV passing through each sensor; d(x i , x i+1 ) is the distance between sensor x i and sensor x i+1 ; n is the number of randomly distributed data acquisition points; d(x n , x 1 ) is the position from the last sensor position to be collected to the starting point, so as to ensure that the UAV can successfully fly back to the starting point. In the information collection of the UAV for emergency rescue, it is often required to collect the information of all target sensors at the fastest speed. Assuming that the UAV always flies horizontally at a constant speed and the UAV needs to fly directly above the target to perform complete information collection, then when the UAV flight path is the shortest, its time consumption is also the shortest. By default, the starting point is also regarded as a sensor that needs to collect information, that is, the starting point is x 1 .

[0073] In a further implementation manner of the above method embodiment, in step S2, the path planning objective is to minimize the total path distance. Therefore, the fitness function is usually associated with the objective function. The fitness function F can be defined based on the objective function f, where f is the total path distance of the individual (sensor sequence). To make the fitness function more suitable for the genetic algorithm, assuming that the fitness value increases as the quality of the solution improves, the fitness function is defined as a function of the reciprocal (or negative value) of the objective function.

[0074] Suppose the objective function f(x) is the total travel distance of individual x, and the fitness function F can be expressed as:

[0075]

[0076] To avoid the situation of division by zero (i.e., a perfect solution with zero travel distance), a small constant ∈ is added to ensure that the denominator is not zero:

[0077]

[0078] where ∈ is a constant close to zero.

[0079] In a further implementation of the above method embodiment, in step S3, each point where the sensor is located can only be visited once, which means that the sensor sequence in the solution cannot have duplicate sensors. For example, if the sequence is (1, 2, 1, 3, 2), then it does not meet this constraint because sensor 1 and sensor 2 are visited twice.

[0080] In step S3, the path must start and end at the same sensor (starting point / ending point), which is a requirement to ensure that the path is closed. It must start from the starting point, visit all other sensors, and then return to the starting point. This means that the first and last sensors in the sequence must be the same.

[0081] In a further implementation of the above method embodiment, in step S4, let the objective function f(x) be minimized, that is, the total length of the path, which can be expressed as:

[0082]

[0083] Here, d(x i , x i+1 ) is the distance between sensor x i and sensor x i+1 , and d(x n , x 1 ) represents the distance from the last sensor to the first sensor (starting point), ensuring that the drone can fly back to the starting point. The binary decision variable x ij is introduced to represent whether there is a path from sensor i to sensor j. Specifically, x ij = 1 indicates that there is a path from sensor i to sensor j, and x ij = 0 indicates that there is no path. Constraint C1 means that each sensor is only visited once; constraint C2 means that the path must start from the starting point, visit all other sensors and return to the starting point; constraint C3 ensures that the path starts from the starting point and finally returns to the starting point.

[0084] In a further implementation of the above method embodiment, step S5 specifically includes the following sub-steps:

[0085] Sub-step S51: Initialization. Randomly generate an initial population, where each individual represents a possible solution (a sequence of sensors).

[0086] Sub-step S52: Evaluate and calculate the fitness of each individual.

[0087] Sub-step S53: Select individuals for reproduction according to their fitness. Individuals with higher fitness have a higher probability of being selected.

[0088] Sub-step S54: Generate new individuals through crossover operations, which involve swapping parts of the sequences of two individuals.

[0089] Sub-step S55: Randomly change some parts of the individuals to introduce new mutations.

[0090] Sub-step S56: Repeat the processes of evaluation, selection, crossover, and mutation until the stopping condition is met.

[0091] In other embodiments of the present invention, a UAV path planning solution for emergency rescue is provided. In specific implementation, based on the system and method in the above preferred embodiment, specifically considering a UAV data acquisition network composed of a UAV for emergency rescue and X ground sensor nodes, assuming that there are n randomly distributed data acquisition points within an interval, and they are all randomly distributed on this horizontal plane, where X i ={x i ,y i} represents the two-dimensional coordinates of the i-th sensor, x=(x 1 ,x 2 ,...,x n ) represents the sequential arrangement of the UAV passing through each sensor (i.e., an individual or a solution); d(x i ,x i+1 ) is the distance between sensor x i and sensor x i+1 ; n is the number of randomly distributed data acquisition points; d(x n ,x 1 ) is the position from the last sensor to be collected to the starting point, to ensure that the UAV can successfully fly back to the starting point. In the information collection of UAVs for emergency rescue, it is often required to collect the information of all target sensors at the fastest speed. Assuming that the UAV always flies horizontally at a constant speed and the UAV needs to fly directly above the target to perform complete information collection, then when the UAV flight path is the shortest, the time consumed is also the shortest. By default, the starting point is also regarded as a sensor that needs to collect information, that is, the starting point is x 1 . Figure 1 It is a scenario diagram of UAV path planning for emergency rescue.

[0092] (1) Problem formulation

[0093] In a genetic algorithm, the utility function (commonly known as the fitness function) is used to evaluate the quality of each individual, so that more adaptable individuals can be preferentially selected during the selection process of the genetic algorithm. The goal is to minimize the total path distance, so the fitness function is usually associated with the objective function. The fitness function F can be defined based on the objective function f, where f is the total path distance of an individual (sensor sequence). To make the fitness function more suitable for the genetic algorithm, it is usually desired that the fitness value increases as the quality of the solution improves. Since the goal is to minimize the travel distance, the fitness function can be defined as a function of the reciprocal (or negative value) of the objective function.

[0094] Assuming that the objective function f(x) is the total travel distance of individual x, then the fitness function F can be expressed as:

[0095]

[0096] To avoid division by zero (i.e., a perfect solution with zero travel distance), we can add a small constant ∈ to ensure that the denominator is not zero:

[0097]

[0098] where ∈ is a small constant close to zero.

[0099] Taking the shortest distance for the UAV to fly and collect all sensor information within the area as the objective function, the problem is modeled, and the fitness function for UAV path planning in this scenario has been derived through the above content. Now, the problem is modeled as:

[0100]

[0101] Here, d(x i , x i+1 ) is the distance between sensor x i and sensor x i+1 , and d(x n , x 1 ) represents the distance from the last sensor to the first sensor (starting point), ensuring that the UAV can fly back to the starting point. The binary decision variable x ij is introduced to represent whether there is a path from sensor i to sensor j. Specifically, x ij = 1 indicates that there is a path from sensor i to sensor j, and x ij = 0 indicates that there is no path. Constraint C1 means that each sensor is visited only once; Constraint C2 means that the path must start from the starting point, visit all other sensors and return to the starting point; Constraint C3 ensures that the path starts from the starting point and finally returns to the starting point.

[0102] (2) Problem Solving

[0103] The problem is defined as finding the shortest path such that the drone can visit all sensors and return to the starting point. To solve this problem, a genetic algorithm is used for solution, and the steps of the genetic algorithm are as follows:

[0104] Step 1, Initialization and Configuration.

[0105] Generate sensor coordinates: Use the rand function to generate a 30×2 matrix xy, representing the coordinates of 30 sensors in a two-dimensional space.

[0106] Create a distance matrix: If the input distance matrix dmat is empty, calculate and generate the distance matrix dmat according to the sensor coordinates xy. This is done by calculating the Euclidean distance between each pair of sensors.

[0107] Configure GA parameters: Set parameters such as population size PopSize, number of iterations NumIter, whether to display progress ShowProg, and whether to display results ShowResult.

[0108] Step 2, Population Initialization.

[0109] Create an initial population: Randomly generate sensor sequences in the Pop matrix, and each individual represents a possible path. The first individual (starting point) is arranged in the order of sensor numbers, and the remaining individuals are randomly arranged.

[0110] Step 3, GA Main Loop.

[0111] Fitness evaluation: Calculate the total path of each individual and store it in the TotalDist array, which represents the fitness of each individual.

[0112] Optimal solution search: In each iteration, find the individual with the minimum total path and update the global optimal solution GlobalMin and the optimal path OptRoute.

[0113] Visualize progress: If ShowProg is true, display the current optimal path and total path distance in real time during the iteration.

[0114] Genetic operations: Generate a new population through crossover and mutation operations. Here are custom genetic operations, including Flip, Swap, and Slide mutations.

[0115] Step 4, Result Output and Visualization.

[0116] Result Output: If ShowResult is true, then after the algorithm ends, the final optimal path and the positions of all sensors are drawn, and the history of the best solution is displayed.

[0117] Data Writing: Write the optimal path and the corresponding total path distance into the OptimPath.txt file.

[0118] Step 5, return the result.

[0119] Convert the encoded sensor sequence back to the actual sensor access path, and output the individual with the highest fitness (i.e., the shortest path) as the solution to the problem. Through this modeling process, the process of using the genetic algorithm to solve the UAV path planning problem for rescue can be transformed into a mathematical model, which can be used for further analysis and optimization.

[0120] The technical effect verification of the solution of the embodiment of the present invention is as follows: The method of the present invention can effectively balance the calculation cost and time efficiency within a limited number of iterations, verifying its effectiveness in UAV path planning. This method shows good performance in the UAV path planning problem in the emergency rescue scenario.

[0121] Specifically, Figure 2 Shows the distribution of sensors in the acquisition area. There are 30 sensors randomly distributed in a 1000m * 1000m area in this image.

[0122] The genetic algorithm is adopted, including operations such as population initialization, fitness evaluation, selection, crossover, and mutation. In each iteration, the current optimal solution is recorded and updated. Figure 3 、 Figure 4 、 Figure 5 Show the path diagrams of the 16th, 32nd, and 64th iterations respectively. It can be seen from the figures that the paths of the UAV are relatively messy in the initial stage, but as the number of iterations increases, the paths gradually simplify. Figure 6 、 Figure 7 、 Figure 8 Are the path diagrams of the 154th, 197th, and 266th iterations respectively. At this time, the paths tend to be closed, and the crossed and repeated paths are significantly reduced.

[0123] The convergence diagram is as Figure 9 shown. Where the x-axis represents the number of iterations, and the y-axis represents the total distance of the optimal path. The number of iterations determines the number of times the algorithm searches for the optimal solution in the entire search space. A higher number of iterations means that the algorithm has more opportunities to explore the solution space, thereby increasing the probability of finding the global optimal solution.

[0124] The number of iterations set for this simulation is 1000. After multiple simulations, it is found that in most cases, 1000 iterations are sufficient to find the optimal path (the shortest path). When the number of iterations is increased to 10000, in 50 simulations, only in 1 simulation did the optimal path not appear within the first 1000 iterations. Some iterations of this simulation are as shown in Figure 10 , Figure 11 . In this simulation, the optimal path in the 5901st iteration is 27.4905 meters shorter than the path distance in the 583rd iteration. However, it should be noted that increasing the number of iterations to 10000 significantly increases the computational workload and time, while the change in the optimized distance is very limited. In contrast, when the number of iterations is 1000, the algorithm can find the optimal path with an accuracy of 98%, showing relatively high efficiency and stability.

[0125] The units involved in the embodiments of the present invention can be implemented in software or in hardware, and the described units can also be provided in a processor. Among them, the names of these units do not, in some cases, constitute a limitation on the unit itself.

[0126] According to one aspect of the embodiments of the present invention, there is provided a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the methods provided in the above various optional implementation manners.

[0127] As another aspect, the embodiments of the present invention further provide a computer-readable medium, which may be included in the electronic device described in the above embodiments; or may exist separately without being assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by an electronic device, the electronic device implements the methods described in the above embodiments.

Claims

1. A UAV path planning system for emergency rescue, characterized in that: include: A drone data collection network consisting of drones and ground sensor nodes, where the drones are equipped with sensors and have data processing and storage modules; The drone is used for data collection and is equipped with corresponding sensor equipment to achieve accurate monitoring and data collection of various sensors deployed on the ground; The UAV is connected to the sensors deployed on the ground through the data processing and storage module to receive, process and store data transmitted from the ground nodes.

2. The UAV path planning system for emergency rescue according to claim 1 is characterized in that: The ground-deployed sensors are randomly distributed in a set area and are all on the same horizontal plane; the UAV flies directly above the target area to complete comprehensive data collection of all sensors.

3. A method for UAV path planning for emergency rescue, characterized in that: The UAV path planning system for emergency rescue according to any one of claims 1 or 2 comprises the following steps: Step S1: Construct a UAV path planning scenario graph to define the spatial and temporal constraints of the flight mission; Step S2: define and reconstruct the fitness function to optimize the path planning goal; Step S3: Considering sensor access restrictions and path closure requirements, establish relevant constraints; Step S4: Establish a mathematical model of the path planning problem according to system requirements; Step S5: Based on the established mathematical model, a genetic algorithm is used to solve the path planning problem.

4. The method for emergency rescue UAV path planning according to claim 3, characterized in that: In step S1, the construction of the UAV path planning scene graph defines the spatial and temporal constraints of the flight mission, which specifically includes the following sub-steps: Considering the UAV data collection network composed of emergency rescue UAVs and X ground sensor nodes, it is assumed that there are n randomly distributed data collection points in the interval, and they are all randomly distributed on this horizontal plane, where X i ={x i ,y i } represents the two-dimensional coordinates of the i-th sensor, x=(x1,x2,...,x n ) represents the order in which the drone passes each sensor; d(x i ,x i+1 ) is the sensor x i To sensor x i+1 The distance between them; n is the number of randomly distributed data collection points; d(x n ,x1) is the last sensor position to be collected from the starting point to ensure that the drone can successfully fly back to the starting point; In the drone information collection for emergency rescue, assuming that the drone always flies at a constant speed in the horizontal direction and needs to fly directly above the target to collect complete information, the time consumed is also the shortest when the drone's flight path is the shortest; the default starting point is also used as a sensor that needs to collect information, that is, the starting point is x1.

5. The method for emergency rescue UAV path planning according to claim 4, characterized in that: In step S2, the fitness function is defined and reconstructed to optimize the path planning target, which specifically includes the following sub-steps: Assume that the path planning goal is to minimize the total path distance, and the fitness function is associated with the objective function; The fitness function F is defined based on the objective function f, where f is the total path distance of the sensor sequence; the fitness value is assumed to increase with the improvement of the quality of the solution, and the fitness function is defined as a function of the inverse of the objective function; Assuming that the objective function f(x) is the total travel distance of individual x, the fitness function F is expressed as: Add a small constant ∈ to ensure that the denominator is not zero: Here, ∈ is a constant close to zero.

6. The method for emergency rescue UAV path planning according to claim 5, characterized in that: In step S3, the establishment of relevant constraint conditions specifically includes the sub-steps of: the sensor sequence cannot have repeated sensors.

7. The method for emergency rescue UAV path planning according to claim 5, characterized in that: In step S3, the establishment of relevant constraint conditions specifically includes the sub-steps of: the path starts and ends at the same sensor.

8. The method for emergency rescue UAV path planning according to claim 5, characterized in that: In step S4, the mathematical model of the path planning problem is established, which specifically includes the following sub-steps: Let the minimized objective function f(x), that is, the total length of the path, be expressed as: Here, d(x i ,x i+1 ) is the sensor x i To sensor x i+1 The distance between them, d(x n , x1) represents the distance from the last sensor to the first sensor, ensuring that the drone can fly back to the starting point; introduce the binary decision variable x ij To indicate whether there is a path from sensor i to sensor j; where x ij = 1 means there is a path from sensor i to sensor j, x ij = 0 means there is no path; constraint C1 means each sensor is visited only once; constraint C2 means the path must start from the starting point, visit all other sensors and return to the starting point; constraint C3 ensures that the path starts from the starting point and eventually returns to the starting point.

9. The method for emergency rescue UAV path planning according to claim 8, characterized in that: In step S5, the path planning problem is solved by using a genetic algorithm according to the established mathematical model, which specifically includes the following sub-steps: Step S51: Initialization, randomly generating an initial population, each individual represents a possible sequence of sensors; Step S52: Evaluate and calculate the fitness of each individual; Step S53: select individuals for reproduction according to fitness; Step S54: generate a new individual through a crossover operation, exchanging partial sequences of two individuals; Step S55: randomly changing certain parts of the individual to introduce new mutations; Step S56: Repeat the process of evaluation, selection, crossover and mutation until the stopping condition is met.

10. The method for emergency rescue UAV path planning according to claim 3, characterized in that: In step S53, individuals with higher fitness have a higher probability of being selected.

Images