Unmanned aerial vehicle real-time communication and control method based on 5.5 G network

Abstract

The invention discloses an unmanned aerial vehicle real-time communication and control method based on a 5.5 G network, relates to the technical field of unmanned aerial vehicle communication, and dynamically constructs a task planning distribution and path optimization model by collecting flight path information and environment data of an unmanned aerial vehicle in real time. The method comprises the following steps: firstly, constructing a distribution fitness function by collecting unmanned aerial vehicle and task information in advance, and iteratively solving a task distribution scheme of each unmanned aerial vehicle by using a genetic algorithm and a reinforcement learning optimization model; and then, a task execution sequence is sent to each unmanned aerial vehicle, path information collected in the flight process is received in real time through a 5.5 G network, a path fitness function is constructed based on the information, and the flight path is iteratively adjusted by using a genetic algorithm, so that efficient and safe execution of the task is ensured.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicle communication technology, and in particular to a real-time communication and control method for unmanned aerial vehicles based on a 5.5G network. Background Art

[0002] With the rapid development of drone technology, they have shown great application potential in logistics and transportation, emergency rescue, patrol monitoring and other fields. However, to give full play to the advantages of drones in these scenarios, efficient and reliable real-time communication and control become essential. Traditional communication networks are difficult to meet the needs of real-time control and data transmission of drones due to limited bandwidth and high latency. Therefore, 5.5G networks have become an ideal choice for real-time communication and control of drones due to their large bandwidth, low latency and high reliability.

[0003] Existing drone communication and control technologies mainly rely on 4G networks and Wi-Fi, which have limited bandwidth and coverage, unstable communications and high latency, and cannot meet the needs of multi-drone collaborative tasks and real-time control in complex environments. In addition, existing technologies also have many shortcomings in task planning and path optimization. For example, task allocation and path planning often rely on preset data and cannot be adjusted dynamically in real time, resulting in slow response speed and inaccurate path planning when facing emergencies or complex environments, which in turn affects the success rate and safety of the mission. Therefore, there is an urgent need for a technical solution that can realize real-time task planning allocation and path optimization to improve the operational efficiency and safety of drones.

[0004] The Chinese patent with the authorization announcement number CN117320106B discloses a Beidou-based forestry UAV intelligent communication system and terminal, including a path planning system, which is used to generate the initial planning path for the UAV to perform the task through the operation task, regional characteristics and UAV parameters; an autonomous obstacle avoidance system, when the distance between the UAV and the obstacle is the shortest distance, adjusts the UAV flight heading angle so that the UAV can bypass the obstacle along the boundary of the obstacle area; after bypassing the obstacle, it returns to the initial path and continues to fly to the preset position. However, the solution failed to design an autonomous path planning system, and only returned to the initial path after bypassing the obstacle, which lacks intelligence;

[0005] To this end, the present invention proposes a real-time communication and control method for UAV based on 5.5G network. Summary of the invention

[0006] The present invention aims to solve at least one of the technical problems existing in the prior art. To this end, the present invention proposes a real-time communication and control method for UAVs based on a 5.5G network, which ensures efficient and safe execution of tasks.

[0007] To achieve the above purpose, a real-time communication and control method for UAV based on 5.5G network is proposed, which includes the following steps:

[0008] Step 1: Collect drone information set and mission information set in advance;

[0009] Step 2: Construct an allocation fitness function based on the drone information set and the task information set;

[0010] Step 3: Based on the allocation fitness function, use the genetic algorithm and reinforcement learning optimization model to iteratively solve and output the task allocation results of each drone;

[0011] Step 4: Send the task execution order corresponding to the task allocation result to each drone, and receive the flight path information collected by each drone during flight in real time through the 5.5G network;

[0012] Step 5: Construct a path fitness function based on the flight path information and mission information set;

[0013] Step 6: Based on the path fitness function, use the genetic algorithm to iteratively adjust the flight path;

[0014] The drone information set is collected in the following manner:

[0015] Collect the drone numbers, initial positions, maximum load weights, and speeds of all drones that need to perform tasks in advance to form a drone information set;

[0016] The task information set is collected in the following manner:

[0017] Collect the list of tasks to be executed, including the number of each task, the three-dimensional spatial coordinates of the destination location of the task, the priority of each task, the load of each task, and the estimated completion time of each task, to form a task information set;

[0018] The method of constructing the allocation fitness function based on the drone information set and the task information set is:

[0019] Determine the scheduling efficiency item of the allocation fitness function based on the speed, initial position, destination position and expected completion time of each task of the UAV;

[0020] According to the priority of each task, determine the emergency degree item of the allocation fitness function;

[0021] According to the load of each task, determine the load balancing item of the allocation fitness function;

[0022] According to the dispatch efficiency item, emergency degree item and load balance item, the allocation fitness function is constructed;

[0023] The method of iteratively solving and outputting the task allocation results of each UAV based on the allocation fitness function using the genetic algorithm and the reinforcement learning optimization model includes the following steps:

[0024] Step 31: Represent the task allocation scheme in the form of first chromosomes, each first chromosome includes the arrangement order of a number of task nodes, and each task node represents a task allocation;

[0025] Step 32: randomly generate L1 initial task allocation schemes to form an initial population; L1 is a preset quantity parameter;

[0026] Step 33: Use the roulette wheel selection method or the tournament selection method to select L2 first chromosomes with the highest allocation fitness from the initial population as parents for breeding the first chromosomes of the new generation; L2 is a preset quantity parameter;

[0027] Step 34: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent generation first chromosomes to generate a new first chromosome;

[0028] Step 35: performing task exchange or task insertion mutation on the newly generated first chromosome at a preset mutation rate;

[0029] Step 36: Optimize the task allocation strategy by using the reinforcement learning optimization model;

[0030] The method of optimizing the task allocation strategy by using the reinforcement learning optimization model is as follows:

[0031] Represent the state of the reinforcement learning optimization model as the current task assignment and drone state;

[0032] Represent the actions of the reinforcement learning optimization model as task allocation decisions to move from the current state to the next state;

[0033] The reward function of the reinforcement learning optimization model is set to evaluate the change in the distribution fitness of each action, usually the difference between the distribution fitness before and after the action is performed;

[0034] Use the Q-learning algorithm to update the reward value Q of the reinforcement learning optimization model and optimize the task allocation strategy;

[0035] Step 37: Iterate steps 33 to 36, repeat the selection, crossover, mutation and allocation fitness evaluation process until the preset number of iterations is reached or the allocation fitness value converges, and then output the final task allocation plan;

[0036] The method of receiving the flight path information collected by each drone during flight in real time through the 5.5G network is as follows:

[0037] During the flight, each drone collects the location information of the drone in real time through the GPS positioning system, detects the obstacle information in the flight environment through the LIDAR system in real time, and collects the meteorological data in the flight environment through the meteorological sensor in real time; the location information, obstacle information and meteorological data constitute the flight path information;

[0038] The flight path information collected in real time is transmitted to the control center via the 5.5G network;

[0039] The method of constructing the path fitness function based on the flight path information and the mission information set is:

[0040] Based on the position information in the flight path information and the destination position of each task in the task information set, a path efficiency term of the path fitness function is constructed;

[0041] Based on the obstacle information and meteorological data in the flight path information, the path safety term of the path fitness function is constructed;

[0042] The path fitness function is a weighted sum of a path efficiency term and a path safety term, and the weights of the path efficiency term and the path safety term are both preset proportional coefficients;

[0043] The method of iteratively adjusting the flight path using a genetic algorithm based on the path fitness function comprises the following steps:

[0044] Step 61: Represent the path as a second chromosome, each second chromosome comprising a number of path nodes;

[0045] Step 62: randomly generate L3 initial paths to form an initial path population; L3 is a preset quantity parameter;

[0046] Step 63: Use the roulette wheel selection method or the tournament selection method to select L4 second chromosomes with the highest path fitness from the population as parents for breeding a new generation of second chromosomes; L4 is a preset quantity parameter;

[0047] Step 64: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent second chromosomes to generate a new second chromosome;

[0048] Step 65: performing a mutation of waypoint exchange or waypoint insertion on the newly generated second chromosome at a preset mutation rate;

[0049] Step 66: Calculate the path fitness value of the second chromosome of the new generation, and select the chromosome whose path fitness is greater than the preset fitness threshold to be retained in the next generation;

[0050] Step 67: Repeat the selection, crossover, mutation and path fitness evaluation process until the preset number of iterations is reached or the path fitness value converges, and output the second chromosome with the highest fitness. The path represented by the chromosome is the flight path that the UAV will subsequently execute.

[0051] An electronic device is proposed, comprising: a processor and a memory, wherein the memory stores a computer program that can be called by the processor;

[0052] The processor executes the above-mentioned real-time communication and control method of a drone based on a 5.5G network by calling the computer program stored in the memory.

[0053] A computer-readable storage medium is provided, on which a rewritable computer program is stored;

[0054] When the computer program runs on a computer device, the computer device executes the above-mentioned real-time communication and control method for a drone based on a 5.5G network.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention dynamically constructs a task planning allocation and path optimization model by collecting the flight path information and environmental data of the drone in real time. First, by collecting drone and task information in advance, constructing an allocation fitness function, and using a genetic algorithm and a reinforcement learning optimization model, iteratively solves the task allocation plan for each drone. Then, the task execution order is sent to each drone, and the path information collected during the flight is received in real time through the 5.5G network. Based on this information, a path fitness function is constructed, and a genetic algorithm is used to iteratively adjust the flight path; this ensures efficient and safe execution of tasks, significantly improves the performance of drones in application scenarios such as logistics transportation, emergency rescue, patrol monitoring, etc., gives full play to the potential of drones, and promotes the development of related industries. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a flow chart of a method for real-time communication and control of a drone based on a 5.5G network in Embodiment 1 of the present invention;

[0058] Figure 2 This is a flowchart of the iteration of the task allocation results in Example 1 of the present invention. DETAILED DESCRIPTION

[0059] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0060] Example 1

[0061] like Figure 1 As shown, a real-time communication and control method for a UAV based on a 5.5G network includes the following steps:

[0062] Step 1: Collect drone information set and mission information set in advance;

[0063] Step 2: Construct an allocation fitness function based on the drone information set and the task information set;

[0064] Step 3: Based on the allocation fitness function, use the genetic algorithm and reinforcement learning optimization model to iteratively solve and output the task allocation results of each drone;

[0065] Step 4: Send the task execution order corresponding to the task allocation result to each drone, and receive the flight path information collected by each drone during flight in real time through the 5.5G network;

[0066] Step 5: Construct a path fitness function based on the flight path information and mission information set;

[0067] Step 6: Based on the path fitness function, use the genetic algorithm to iteratively adjust the flight path.

[0068] The drone information set is collected in the following manner:

[0069] Collect the drone numbers, initial positions, maximum load weights, and speeds of all drones that need to perform tasks in advance to form a drone information set;

[0070] Among them, the initial position is the three-dimensional spatial coordinates of the location of the drone before it performs the mission, and the three-dimensional spatial coordinates can be obtained through the global three-dimensional spatial coordinate system constructed by the satellite positioning system; the maximum load weight is the maximum load weight that the drone can carry, and the speed is the flight speed of the drone.

[0071] The task information set is collected in the following manner:

[0072] Collect the list of tasks to be executed, including the number of each task, the three-dimensional spatial coordinates of the destination location of the task, the priority of each task, the load of each task, and the estimated completion time of each task, to form a task information set;

[0073] Specifically, the priority refers to the importance and urgency of the task, which is manually determined in advance based on the actual situation of the task. The load capacity is the weight of items that need to be carried or transported after performing the task, and the load capacity is particularly important in transportation tasks. The estimated completion time is the time required for the drone to perform the corresponding task after arriving at the destination, which is generally determined based on actual experience.

[0074] Furthermore, the method of constructing the allocation fitness function based on the drone information set and the task information set is:

[0075] Determine the scheduling efficiency item of the allocation fitness function based on the speed, initial position, destination position and expected completion time of each task of the UAV;

[0076] According to the priority of each task, determine the emergency degree item of the allocation fitness function;

[0077] According to the load of each task, determine the load balancing item of the allocation fitness function;

[0078] According to the dispatching efficiency term, emergency degree term and load balancing term, the allocation fitness function is constructed.

[0079] Preferably, the scheduling efficiency item is determined as follows:

[0080] Mark each drone with a number j;

[0081] Mark each task with an i;

[0082] The three-dimensional spatial coordinates of the target position of the i-th task are marked as Pi;

[0083] The estimated completion time of the i-th task is marked as TCi;

[0084] The speed of the jth UAV is denoted as vj;

[0085] The task sequence assigned to the j-th UAV is marked as Xj, where the task sequence stores the numbers of several tasks in order, indicating that the j-th UAV needs to perform the tasks in the sequence in order according to the task sequence;

[0086] The tasks in the task sequence Xj are numbered as xj1, xj2, ..., xjk, xj(k+1), ..., xjK in sequence; where K is the number of all tasks in the task sequence Xj, and k is any one of 1, 2, ..., K.

[0087] It can be understood that each xjk corresponds to a task number i; for example, in the example with the following task and drone information:

[0088] Mission Information:

[0089] Task 1: Destination (10,20,5), priority 3, estimated completion time 60 minutes, payload 2kg;

[0090] Task 2: Destination (15,25,3), priority 2, estimated completion time 45 minutes, payload 3kg;

[0091] Task 3: Destination (20,30,2), priority 1, estimated completion time 30 minutes, load 1kg;

[0092] Drone Information:

[0093] Drone 1: initial position (5,5,0), maximum load weight 5kg;

[0094] Drone 2: initial position (10,10,0), maximum load weight 4kg;

[0095] The resulting allocation is:

[0096] Drone 1 performs missions 1 and 3;

[0097] Drone 2 performs mission 2;

[0098] Then the number sequences x11 and x13 of drone 1 correspond to tasks 1 and 3 respectively; the number x22 of drone 2 corresponds to task 2.

[0099] For any task xjk, calculate the total completion time TTxjk of the task, where the total completion time is the sum of the UAV flight time TFxjk and the estimated completion time Tci required to perform the task;

[0100] in, Among them, d(Pxjk, Pxj(k-1)) represents the spatial distance between positions Pxjk and Pxj(k-1). It can be understood that Pxj(k-1) is the execution location of the previous task of task xjk, and Pxjk is the execution location of task xjk, that is, d(Pxjk, Pxj(k-1)) represents the distance that the drone flies from the previous task execution location to the task execution location. The distance divided by the drone's flight speed is the required flight time;

[0101] Calculate the total completion time of all tasks and retrieve the task with the longest total completion time, marking it as Tmax;

[0102] Then the calculation formula of the scheduling efficiency term is: Among them, n is the number of all tasks; it can be understood that when Tmax is too large, that is, when there is a task sequence with a total completion time that is too long, the value of the scheduling efficiency item will continue to accumulate larger values, thereby causing the scheduling efficiency item to be too large. Therefore, the scheduling efficiency item reflects the time efficiency of task allocation.

[0103] Furthermore, the emergency level item is determined in the following manner:

[0104] Mark the priority of task xjk as pxjk;

[0105] The calculation formula of the emergency degree item is: It can be understood that pxjk-pxj(k-1) measures the priority difference between two adjacent tasks. Obviously, when the low-priority task is executed first, the value will be greater than 0, and the greater the priority difference, the greater the value. Therefore, the emergency level reflects the priority of executing emergency tasks.

[0106] Furthermore, the load balancing item is determined as follows:

[0107] Mark the load of the xjkth task as wxjk;

[0108] The maximum load weight of the j-th UAV is marked as WMj;

[0109] For the jth UAV, the calculation formula for its total load WTxjk when performing the xjkth task is:

[0110] Then the calculation formula of the load balancing item is: Among them, |WMj-WTxjK| reflects the utilization rate of the j-th UAV load. The closer it is to the maximum load weight, the closer |WMj-WTxjK| is to 0, indicating that its utilization rate is higher; is the average value of all WTxjK, It reflects the balance of the load of each drone to avoid uneven distribution; α1 and α2 are both preset proportional coefficients;

[0111] Thus, the method of constructing the allocation fitness function is:

[0112]

[0113] Among them, β1, β2 and β3 are all preset proportional coefficients.

[0114] like Figure 2As shown, the iterative solution and output of the task allocation results of each UAV based on the allocation fitness function using the genetic algorithm and the reinforcement learning optimization model includes the following steps:

[0115] Step 31: Represent the task allocation scheme in the form of first chromosomes, each first chromosome includes the arrangement order of a number of task nodes, and each task node represents a task allocation;

[0116] Step 32: randomly generate L1 initial task allocation schemes to form an initial population; L1 is a preset quantity parameter;

[0117] Step 33: Use the roulette wheel selection method or the tournament selection method to select L2 first chromosomes with the highest allocation fitness from the initial population as parents for breeding the first chromosomes of the new generation; L2 is a preset quantity parameter; generally, L2 is set to

[0118] Step 34: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent generation first chromosomes to generate a new first chromosome;

[0119] Step 35: performing task exchange or task insertion mutation on the newly generated first chromosome at a preset mutation rate;

[0120] Step 36: Optimize the task allocation strategy by using the reinforcement learning optimization model.

[0121] The method of optimizing the task allocation strategy by using the reinforcement learning optimization model is as follows:

[0122] Represent the state of the reinforcement learning optimization model as the current task assignment and drone state;

[0123] Represent the actions of the reinforcement learning optimization model as task allocation decisions to move from the current state to the next state;

[0124] The reward function of the reinforcement learning optimization model is set to evaluate the change in the distribution fitness of each action, usually the difference between the distribution fitness before and after the action is performed;

[0125] Use the Q-learning algorithm to update the reward value Q of the reinforcement learning optimization model and optimize the task allocation strategy;

[0126] Step 37: Iterate steps 33 to 36, repeat the selection, crossover, mutation and allocation fitness evaluation process until the preset number of iterations is reached or the allocation fitness value converges, and then output the final task allocation plan.

[0127] The method of receiving the flight path information collected by each drone during flight in real time through the 5.5G network is as follows:

[0128] During the flight, each drone collects the location information of the drone in real time through the GPS positioning system, detects the obstacle information in the flight environment through the LIDAR system in real time, and collects the meteorological data in the flight environment through the meteorological sensor in real time; the location information, obstacle information and meteorological data constitute the flight path information;

[0129] Specifically, the location information includes but is not limited to the current location, flight altitude, flight speed and flight direction of the drone; the obstacle information includes but is not limited to the location information and size of the obstacle; the meteorological data includes but is not limited to wind speed, wind direction, rainfall and visibility;

[0130] The flight path information collected in real time is transmitted to the control center via the 5.5G network.

[0131] The method of constructing the path fitness function based on the flight path information and the mission information set is:

[0132] Based on the position information in the flight path information and the destination position of each task in the task information set, a path efficiency term of the path fitness function is constructed;

[0133] Based on the obstacle information and meteorological data in the flight path information, the path safety term of the path fitness function is constructed;

[0134] The path fitness function is a weighted sum of a path efficiency term and a path safety term, and the weights of the path efficiency term and the path safety term are both preset proportional coefficients.

[0135] Specifically, the path efficiency term is constructed as follows:

[0136] The flight path of the UAV is divided into several path nodes at equal distances, each path node corresponds to the real-time spatial position of the UAV;

[0137] Calculate the Euclidean distance between every two adjacent path nodes;

[0138] The sum of the Euclidean distances calculated for all path nodes is taken as the path efficiency term; it can be understood that the path efficiency term expresses the total distance flown by the UAV, and the greater the distance, the lower its flight efficiency.

[0139] Furthermore, the path security item is constructed as follows:

[0140] For each path node, calculate its obstacle risk score SDf and weather risk score SMf, where f is the number of the path node in the flight path;

[0141] The obstacle risk score SDf is calculated as follows:

[0142] Wherein, e is a natural constant, DOf is the Euclidean distance between path node f and the nearest obstacle, Dth is the preset safety distance threshold; γ1 is the preset proportional coefficient;

[0143] The weather risk score SMf is calculated as follows:

[0144] Wherein, SWf is the wind speed at path node f, SRf is the rainfall at path node f, SSf is the visibility at path node f; γ2, γ3 and γ4 are all preset proportional coefficients;

[0145] Calculate a comprehensive risk value for each path node, where the comprehensive risk value is a weighted sum of an obstacle risk score SDf and a weather risk score SMf, where the weights of the obstacle risk score SDf and the weather risk score SMf are set in advance based on experience;

[0146] The sum of the comprehensive risk values ​​of all path nodes is the path safety item.

[0147] The method of iteratively adjusting the flight path using a genetic algorithm based on the path fitness function comprises the following steps:

[0148] Step 61: Represent the path as a second chromosome, each second chromosome comprising a number of path nodes;

[0149] Step 62: randomly generate L3 initial paths to form an initial path population; L3 is a preset quantity parameter;

[0150] Step 63: Use the roulette wheel selection method or the tournament selection method to select L4 second chromosomes with the highest path fitness from the population as parents for breeding the second chromosomes of the new generation; L4 is a preset quantity parameter; generally, L4 is set to

[0151] Step 64: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent second chromosomes to generate a new second chromosome;

[0152] Step 65: performing a mutation of waypoint exchange or waypoint insertion on the newly generated second chromosome at a preset mutation rate;

[0153] Step 66: Calculate the path fitness value of the second chromosome of the new generation, and select the chromosome whose path fitness is greater than the preset fitness threshold to be retained in the next generation;

[0154] Step 67: Repeat the selection, crossover, mutation and path fitness evaluation process until the preset number of iterations is reached or the path fitness value converges, and output the second chromosome with the highest fitness. The path represented by the chromosome is the flight path that the UAV will subsequently execute.

[0155] Example 2

[0156] The present application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable codes, and when the computer-readable codes are executed by one or more processors, the real-time communication and control method of a drone based on a 5.5G network as described above may be executed.

[0157] The method or device according to the embodiment of the present application can also be implemented with the help of the architecture of the electronic device described below. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, an input / output component, a hard disk, etc. A storage device in the electronic device, such as a ROM or a hard disk, can store the real-time communication and control method of a drone based on a 5.5G network provided in the present application.

[0158] Furthermore, the electronic device may also include a user interface. Of course, the above architecture is only exemplary, and when implementing different devices, one or more components in the above electronic device may be omitted according to actual needs.

[0159] Example 3

[0160] The present application discloses a computer-readable storage medium. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by the processor, the real-time communication and control method of a drone based on a 5.5G network according to the embodiment of the present application described with reference to the above figures can be executed. The computer-readable storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache), etc. Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0161] In addition, according to the implementation of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, the non-transitory machine-readable storage medium stores machine-readable instructions, the machine-readable instructions can be run by a processor to execute instructions corresponding to the method steps provided by the present application, and when the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0162] The methods, apparatuses, and devices of the present application may be implemented in many ways. For example, the methods, apparatuses, and devices of the present application may be implemented by software, hardware, firmware, or any combination of software, hardware, and firmware. The above order of steps for the method is for illustration only, and the steps of the method of the present application are not limited to the order specifically described above unless otherwise specified. In addition, in some embodiments, the present application may also be implemented as programs recorded in a recording medium, which include machine-readable instructions for implementing the method according to the present application. Therefore, the present application also covers recording media storing programs for executing the method according to the present application.

[0163] In addition, the parts of the above-mentioned technical solutions provided in the embodiments of the present application that are consistent with the implementation principles of the corresponding technical solutions in the prior art are not described in detail to avoid excessive redundancy.

[0164] The specific implementation modes as described above further describe the purpose, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation mode of the present invention and is not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

[0165] The above preset parameters or preset thresholds are all set by technicians in this field according to actual conditions or obtained through large-scale data simulation.

[0166] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A real-time communication and control method for unmanned aerial vehicles based on 5.5G network, characterized in that: The following steps are involved: Step 1: Collect drone information set and mission information set in advance; Step 2: Construct an allocation fitness function based on the drone information set and the task information set; Step 3: Based on the allocation fitness function, use the genetic algorithm and reinforcement learning optimization model to iteratively solve and output the task allocation results of each drone; Step 4: Send the task execution order corresponding to the task allocation result to each drone, and receive the flight path information collected by each drone during flight in real time through the 5.5G network; Step 5: Construct a path fitness function based on the flight path information and mission information set; Step 6: Based on the path fitness function, use the genetic algorithm to iteratively adjust the flight path.

2. The method for real-time communication and control of a drone based on a 5.5G network according to claim 1, characterized in that: The task information set is collected in the following manner: Collect the list of tasks to be executed, the number of each task, the three-dimensional spatial coordinates of the destination location of the task, the priority of each task, the load of each task, and the estimated completion time of each task to form a task information set.

3. The method for real-time communication and control of a drone based on a 5.5G network according to claim 2, characterized in that: The method of constructing the allocation fitness function based on the drone information set and the task information set is: Determine the scheduling efficiency item of the allocation fitness function based on the speed, initial position, destination position and expected completion time of each task of the UAV; According to the priority of each task, determine the emergency degree item of the allocation fitness function; According to the load of each task, determine the load balancing item of the allocation fitness function; According to the dispatching efficiency term, emergency degree term and load balancing term, the allocation fitness function is constructed.

4. The method for real-time communication and control of a drone based on a 5.5G network according to claim 3 is characterized in that: The method of iteratively solving and outputting the task allocation results of each UAV based on the allocation fitness function using the genetic algorithm and the reinforcement learning optimization model includes the following steps: Step 31: Represent the task allocation scheme in the form of first chromosomes, each first chromosome includes the arrangement order of a number of task nodes, and each task node represents a task allocation; Step 32: randomly generate L1 initial task allocation schemes to form an initial population; L1 is a preset quantity parameter; Step 33: Use the roulette wheel selection method or the tournament selection method to select L2 first chromosomes with the highest allocation fitness from the initial population as parents for breeding the first chromosomes of the new generation; L2 is a preset quantity parameter; Step 34: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent generation first chromosomes to generate a new first chromosome; Step 35: performing task exchange or task insertion mutation on the newly generated first chromosome at a preset mutation rate; Step 36: Optimize the task allocation strategy by using the reinforcement learning optimization model; Step 37: Iterate steps 33 to 36, repeat the selection, crossover, mutation and allocation fitness evaluation process until the preset number of iterations is reached or the allocation fitness value converges, and then output the final task allocation plan.

5. The method for real-time communication and control of a drone based on a 5.5G network according to claim 4 is characterized in that: The method of optimizing the task allocation strategy by using the reinforcement learning optimization model is as follows: Represent the state of the reinforcement learning optimization model as the current task assignment and drone state; Represent the actions of the reinforcement learning optimization model as task allocation decisions to move from the current state to the next state; The reward function of the reinforcement learning optimization model is set to evaluate the change in the fitness of the distribution for each action; Use the Q-learning algorithm to update the reward value Q of the reinforcement learning optimization model and optimize the task allocation strategy.

6. The method for real-time communication and control of a UAV based on a 5.5G network according to claim 5, characterized in that: The method of receiving the flight path information collected by each drone during flight in real time through the 5.5G network is as follows: During the flight, each drone collects the location information of the drone in real time through the GPS positioning system, detects the obstacle information in the flight environment through the LIDAR system in real time, and collects the meteorological data in the flight environment through the meteorological sensor in real time; the location information, obstacle information and meteorological data constitute the flight path information; The flight path information collected in real time is transmitted to the control center via the 5.5G network.

7. The method for real-time communication and control of a UAV based on a 5.5G network according to claim 6, characterized in that: The method of constructing the path fitness function based on the flight path information and the mission information set is: Based on the position information in the flight path information and the destination position of each task in the task information set, a path efficiency term of the path fitness function is constructed; Based on the obstacle information and meteorological data in the flight path information, the path safety term of the path fitness function is constructed; The path fitness function is a weighted sum of a path efficiency term and a path safety term, and the weights of the path efficiency term and the path safety term are both preset proportional coefficients.

8. The method for real-time communication and control of a drone based on a 5.5G network according to claim 7, characterized in that: The method of iteratively adjusting the flight path using a genetic algorithm based on the path fitness function comprises the following steps: Step 61: Represent the path as a second chromosome, each second chromosome comprising a number of path nodes; Step 62: randomly generate L3 initial paths to form an initial path population; L3 is a preset quantity parameter; Step 63: Use the roulette wheel selection method or the tournament selection method to select L4 second chromosomes with the highest path fitness from the population as parents for breeding a new generation of second chromosomes; L4 is a preset quantity parameter; Step 64: performing a crossover operation of a single-point crossover or a multi-point crossover to combine the parent second chromosomes to generate a new second chromosome; Step 65: performing a mutation of waypoint exchange or waypoint insertion on the newly generated second chromosome at a preset mutation rate; Step 66: Calculate the path fitness value of the second chromosome of the new generation, and select the chromosome whose path fitness is greater than the preset fitness threshold to be retained in the next generation; Step 67: Repeat the selection, crossover, mutation and path fitness evaluation process until the preset number of iterations is reached or the path fitness value converges, and output the second chromosome with the highest fitness. The path represented by the chromosome is the flight path that the UAV will subsequently execute.

9. An electronic device, characterized in that: include: A processor and a memory, wherein: The memory stores a computer program that can be called by the processor; The processor executes the real-time communication and control method for a drone based on a 5.5G network as described in any one of claims 1 to 8 in the background by calling the computer program stored in the memory.

10. A computer-readable storage medium, characterized in that: A rewritable computer program is stored thereon; When the computer program runs on a computer device, the computer device executes a 5.5G network-based drone real-time communication and control method as described in any one of claims 1 to 8 in the background.

Images