Low-altitude multi-unmanned aerial vehicle cooperative prevention and control scheduling optimization method

Abstract

The invention discloses a low-altitude multi-unmanned aerial vehicle cooperative prevention and control scheduling optimization method, and relates to the technical field of unmanned aerial vehicle cooperative scheduling. Collecting and fusing geographic space, target, environment and unmanned aerial vehicle state multi-source data, and generating a prevention and control situation data set; modeling core prevention and control tasks and dividing priorities; establishing an unmanned aerial vehicle resource constraint model and initially allocating resources; constructing a multi-target collaborative scheduling optimization model and solving; scene changes are monitored in real time, scheduling is dynamically adjusted, and conflicts are resolved; and establishing a prevention and control effect evaluation system, and iteratively optimizing a scheduling strategy. The method adapts to low-altitude prevention and control scene requirements, multi-source data fusion improves scheduling decision accuracy, and dynamic scheduling and conflict resolution guarantee stable operation; the resource allocation and strategy iteration are optimized, the prevention and control task completion quality is improved, the resource utilization rate and scene adaptability are improved, and the prevention and control efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) collaborative scheduling technology, and in particular to an optimization method for low-altitude multi-UAV collaborative prevention and control scheduling. Background Technology

[0002] With the rapid development of the low-altitude economy, drones are increasingly used in security monitoring, border control, and emergency response. Multi-drone collaborative control has become a core means to improve control efficiency and coverage. However, the complex and ever-changing low-altitude environment, with its uncertain target trajectories, numerous environmental interference factors, and strict constraints on drone resources, places extremely high demands on the accuracy, real-time performance, and stability of multi-drone collaborative scheduling. Existing technologies mostly focus on task scheduling and resource allocation for general scenarios, making it difficult to adapt to the specific needs of control scenarios and exhibiting significant shortcomings.

[0003] The patent application CN117880846A, entitled "A Task Scheduling and Resource Allocation Method for a Multi-UAV Collaborative Computing Network," mentions establishing a multi-UAV collaborative computing network system model. It describes using Lyapunov optimization and Lagrange duality to jointly optimize task scheduling and resource allocation, minimizing the system's total energy consumption. However, this technology primarily targets the offloading of computing tasks from the computing network, with the core objective of reducing energy consumption. It fails to consider the timeliness and collaborative requirements of special tasks such as target tracking and emergency interception in air defense scenarios. It lacks integration of key air defense information such as geospatial data and target threat levels, and it does not address multi-UAV path conflict resolution and dynamic scheduling mechanisms, making it difficult to directly apply to low-altitude air defense scenarios.

[0004] The patent application document with publication number CN115243321B and title "A Multi-UAV Collaborative Communication and Computation Task Scheduling Method and System" mentions establishing a multi-UAV aerial computing platform and determining the computing UAVs for sub-tasks based on a two-layer game theory method to minimize the total processing latency and achieve load balancing. However, this technology focuses on the uploading and unloading scheduling of general computing tasks and does not design collaborative rules for the characteristics of tasks such as area patrol and target acquisition in prevention and control scenarios. It also does not integrate multi-source data such as environmental parameters (e.g., wind speed, visibility) and UAV status (e.g., remaining battery power, sensor operating status), resulting in limited dynamic adjustment capabilities. When the target trajectory changes abruptly or a new task is inserted in the prevention and control scenario, it is difficult to quickly respond and optimize the scheduling scheme.

[0005] The patent application document with publication number CN117236520A and titled "A Distributed Multi-UAV Cluster Collaborative Scheduling System and Method Thereof" mentions using deep learning to extract image features of inspection paths and fault points, assess the disaster situation, and generate an optimal scheduling scheme, which is applicable to power distribution network fault repair scenarios. However, this technology focuses on inspection and fault assessment, lacking the multi-task priority division and UAV collaborative operation rule design required for prevention and control scenarios. It does not consider key prevention and control indicators such as target threat level and emergency response timeliness. Resource allocation is based solely on image assessment results, without considering constraints such as UAV computing power and endurance, and therefore cannot meet the efficient scheduling requirements of multi-UAV collaborative prevention and control.

[0006] In summary, existing technologies suffer from three major limitations: First, insufficient scenario adaptability, failing to design dedicated scheduling schemes for the dynamic nature of targets, the timeliness of tasks, and the complexity of coordination in prevention and control scenarios; second, insufficient integration of multi-source information, failing to effectively integrate key prevention and control data such as geospatial data, environmental parameters, target information, and UAV status, resulting in insufficient accuracy in scheduling decisions; and third, weak dynamic scheduling and conflict resolution capabilities, making it difficult to cope with emergencies such as changes in the low-altitude environment, target movement, and equipment failure, and lacking a quantitative task priority assessment and scientific resource allocation mechanism. These problems lead to low task completion rates, high response latency, and insufficient resource utilization in multi-UAV collaborative prevention and control, failing to meet the practical needs of low-altitude prevention and control. Therefore, a targeted optimization method for scheduling multi-UAV collaborative prevention and control in low-altitude environments is urgently needed to overcome the bottlenecks of existing technologies. Summary of the Invention

[0007] The present invention proposes a low-altitude multi-UAV collaborative prevention and control scheduling optimization method to solve the problems mentioned in the prior art.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: a low-altitude multi-UAV collaborative prevention and control scheduling optimization method, comprising: Steps for collecting and fusing multi-source information in the prevention and control scenario: Collect geospatial data, target information, environmental parameters, and UAV status data in the low-altitude area; denoise the data using the Kalman filter algorithm; integrate the data using a weighted fusion algorithm to generate a unified prevention and control scenario situation dataset. Prevention and control task modeling and priority division steps: Based on the situation dataset, a prevention and control task model is constructed. According to the target threat level, task urgency, impact range, and resource demand difficulty, a task priority evaluation system is constructed, and three priority levels are divided: Level 1 urgent threat, Level 2 important prevention and control, and Level 3 routine patrol. The steps for modeling and initial allocation of resource constraints among drones are as follows: Establish a drone resource constraint model, including the flight time, type and number of carried equipment, communication coverage radius, data processing capability, and maximum flight speed of each drone; record the current position, remaining battery power, and assigned tasks of each drone; calculate the matching score based on the matching degree between task requirements and drone resources; use a greedy algorithm to perform initial resource allocation; and assign a suitable drone cluster to each task. The steps for constructing and solving the collaborative scheduling optimization model are as follows: With the goals of maximizing the completion rate of prevention and control tasks, minimizing task execution delay, and minimizing the total energy consumption of the system, a collaborative scheduling optimization model is constructed. Task dependency constraints, UAV collaborative operation rules, regional conflict avoidance constraints, and battery life constraints are introduced. The model is solved by combining the particle swarm optimization algorithm with the Lagrange dual method. The particle swarm optimization algorithm improves the convergence speed by dynamically adjusting the inertia weight and learning factor. Dynamic scheduling and conflict resolution steps: Real-time monitoring of changes in the situation of the prevention and control scenario, drone status updates, and task execution progress. When there are changes in the target movement trajectory, drone malfunctions, new task insertions, or sudden changes in environmental parameters, the dynamic scheduling mechanism is triggered. Based on the rolling time domain optimization idea, time windows are divided and the scheduling model is re-solved in each window. Conflict detection algorithms are used to identify drone path conflicts and resource competition conflicts. Conflicts are resolved through path replanning, dynamic adjustment of task priorities, activation of backup drones, and task splitting. Prevention and control effectiveness evaluation and strategy iteration: Establish a prevention and control effectiveness evaluation index system, calculate the values ​​of various indicators through real-time data collection and analysis, and use the random forest algorithm to iteratively optimize the scheduling strategy based on the evaluation results, adjusting task priority weights, resource allocation rules, collaborative scheduling parameters, and algorithm solution coefficients.

[0009] Furthermore, it also includes: Task priority calculation steps: Quantify task priority through indicators. The formula is: Pt=ω1×Tl+ω2×Ed+ω3×Sr+ω4×Rd, where Pt is the priority coefficient of the t-th task, ω1 is the target threat level weight, ω2 is the task urgency weight, ω3 is the impact range weight, ω4 is the resource requirement weight, and ω1+ω2+ω3+ω4=1, Tl is the target threat level, Ed is the task urgency, Sr is the impact range value, and Rd is the resource requirement difficulty. The steps for evaluating the efficiency of collaborative scheduling are as follows: Quantify the collaborative execution effect among UAVs, using the following formula: ,in The efficiency coefficient for coordinated scheduling. The actual number of tasks completed. Average task execution efficiency refers to the number of tasks completed per unit of time. For resource utilization, This represents the theoretical maximum number of tasks that can be completed. This represents the optimal task execution efficiency threshold. For resource utilization rate.

[0010] Furthermore, in the step of information collection and fusion for the prevention and control scenario, geospatial data is acquired through satellite remote sensing and lidar scanning technology; target information is collected through infrared sensors, visible light cameras, and radar equipment carried by UAVs; environmental parameters are collected collaboratively through ground sensors deployed in the prevention and control area and UAV-borne sensors; UAV status data is uploaded in real time through IoT communication protocols; the process noise covariance and observation noise covariance of the Kalman filter algorithm are dynamically adjusted according to scenario characteristics; and the weights of the weighted fusion algorithm are dynamically allocated based on data credibility.

[0011] Furthermore, in the steps of modeling and prioritizing the prevention and control tasks, the task execution requirements clearly define the type of drone, sensor configuration, and execution time range required for the task. The time window is set according to the prevention and control needs. The target threat level is determined by expert evaluation based on the target's behavioral characteristics and the degree of harm. The task urgency is calculated based on the time difference between the task deadline and the current time. The impact range is determined by the area defined by the geographic information system. The difficulty of resource requirements is based on the number of drones, equipment types, and computational resources required to complete the task.

[0012] Furthermore, in the resource constraint modeling and initial allocation steps among UAVs, the maximum flight time of the UAVs is calculated using battery capacity and energy consumption models. The types of equipment carried include infrared cameras, high-definition cameras, radar, and communication relay equipment. The communication coverage is calculated based on the UAV's flight altitude and the power of the communication equipment. The data processing capability is determined by processor performance and algorithm efficiency. The matching degree between task requirements and UAV resources is calculated from four dimensions: equipment adaptability, flight time matching, communication coverage matching, and processing capability matching. The greedy algorithm prioritizes allocating the UAV cluster with the best resources to high-priority tasks.

[0013] Furthermore, in the steps of constructing and solving the collaborative scheduling optimization model, the task dependency constraint sets the task logic with a sequential execution order, the UAV collaboration rule specifies the communication collaboration method and action coordination requirements of the UAV cluster for the same task, the regional conflict avoidance constraint ensures that the UAV flight paths do not overlap and the flight altitudes do not conflict, the inertia weight of the particle swarm optimization algorithm adopts a linear decreasing strategy, the learning factor adopts an adaptive adjustment strategy, and the Lagrange dual method is used to handle the inequality constraints in the model, transforming the multi-objective optimization problem into a single-objective optimization problem to be solved.

[0014] Furthermore, in the dynamic scheduling adjustment and conflict resolution steps, the system dynamically adjusts according to the complexity of the scenario. The conflict detection algorithm adopts a combination of spatial conflict detection based on grid partitioning and temporal conflict detection based on time slices. The path replanning adopts an algorithm combined with the artificial potential field method. The standby drone activation strategy sets the standby state, activation conditions, and replacement process of the standby drone.

[0015] Furthermore, in the steps of evaluating the prevention and control effect and iterating the strategy, the task completion time is the ratio of the actual task completion time to the preset completion time, the target capture rate is the ratio of the number of successfully captured targets to the total number of targets, the area coverage integrity is the ratio of the actual coverage area to the total area of ​​the prevention and control area, the resource utilization rate includes the power utilization rate, equipment utilization rate, and drone utilization rate, the system energy consumption level is the sum of the flight energy consumption, equipment energy consumption, and communication energy consumption of all drones, and the training data of the random forest algorithm includes historical scheduling data, scene parameters, and evaluation results.

[0016] Furthermore, it also includes: Emergency response enhancement steps: For first-priority tasks, establish a rapid response mechanism, pre-plan emergency flight paths, reserve emergency drone resources, and simplify the scheduling decision-making process. When a first-priority task is triggered, the emergency drone cluster will start execution within 30 seconds. The real-time and security of data transmission will be ensured through a dedicated communication link. A priority preemption mechanism will be adopted to ensure that emergency tasks obtain the best resource allocation, while recording emergency scheduling process data.

[0017] Compared with existing technologies, the beneficial effects of this invention are: First, the adaptability to different scenarios is significantly enhanced, accurately matching core prevention and control needs. This invention designs dedicated scheduling logic and coordination rules for four core tasks in low-altitude prevention and control: area patrol, target tracking, emergency interception, and information transmission. A task priority evaluation system quantifies the target threat level and task urgency, ensuring that high-priority tasks receive priority resource support. Compared to existing general-purpose scheduling technologies, this invention deeply aligns with the practical needs of prevention and control scenarios, flexibly adapting to diverse application modes such as single-person patrols, multi-team collaboration, and emergency response. This effectively solves the problem of poor scenario adaptability in existing technologies and significantly improves the quality of prevention and control task completion.

[0018] Secondly, deep integration of multi-source information enhances the accuracy of scheduling decisions. This invention integrates multiple types of data, including geospatial data, target information, environmental parameters, and UAV status, and generates a unified situational awareness dataset through Kalman filtering denoising and weighted fusion algorithms, providing comprehensive support for scheduling decisions. Compared to the single-dimensional data dependence of existing technologies, the multi-source data fusion mechanism of this invention can accurately depict the low-altitude air defense environment and UAV operational status, making task allocation, path planning, and resource scheduling more aligned with actual scenarios, avoiding scheduling deviations caused by incomplete information, and providing a reliable data foundation for collaborative prevention and control.

[0019] Furthermore, the dynamic scheduling and conflict resolution mechanism ensures the stability and continuity of scheduling. Based on the concept of rolling time-domain optimization, this invention monitors scene changes and UAV status in real time. When situations such as target trajectory changes, equipment failures, or new task insertions occur, scheduling adjustments are quickly triggered. Conflicts are resolved through path replanning, activation of backup UAVs, and dynamic adjustment of task priorities. This dynamic response mechanism effectively solves the problems of rigid scheduling and difficulty in responding to emergencies in existing technologies, ensuring the continuous and stable operation of multi-UAV collaborative prevention and control in complex and ever-changing low-altitude environments, and improving the robustness of scheduling.

[0020] Finally, scientifically optimizing resource allocation improves prevention and control efficiency and resource utilization. This invention, through a multi-objective optimization model and an improved particle swarm optimization algorithm, achieves optimal resource allocation while satisfying multiple objective constraints such as task completion rate, latency, and energy consumption. Combined with quantified task prioritization and collaborative efficiency evaluation, it avoids resource waste and load imbalance. Simultaneously, through prevention and control effect evaluation and strategy iteration mechanisms, scheduling parameters and algorithm coefficients are continuously optimized, enabling the system to adapt to different prevention and control scenarios and task requirements. This improves prevention and control efficiency while reducing system energy consumption, achieving dual optimization of prevention and control effectiveness and resource utilization, with very broad application prospects. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of a low-altitude multi-UAV collaborative prevention and control scheduling optimization method proposed in this invention; Figure 2 This is a schematic diagram comparing the task completion rate under different scheduling methods of the low-altitude multi-UAV collaborative prevention and control scheduling optimization method proposed in this invention. Figure 3 This is a schematic diagram comparing the resource utilization rate of different numbers of drones in the low-altitude multi-UAV collaborative prevention and control scheduling optimization method proposed in this invention. Figure 4 This diagram illustrates the comparison of emergency response delays under different scheduling methods for the low-altitude multi-UAV collaborative prevention and control scheduling optimization method proposed in this invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0024] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0025] Reference Figures 1 to 4 A method for optimizing the collaborative prevention and control scheduling of multiple low-altitude unmanned aerial vehicles (UAVs) includes the following steps: The system collects and integrates multi-source information for epidemic prevention and control scenarios. It collects geospatial data, target information, environmental parameters, and UAV status data in low-altitude areas. The geospatial data includes topography, obstacle distribution, boundaries and coordinate range of the prevention and control area. The target information includes target type, real-time location, movement speed, and threat level. The environmental parameters include wind speed, visibility, precipitation intensity, and electromagnetic interference intensity. The UAV status data includes remaining battery power, flight time, payload equipment type, sensor operating status, and communication link quality. The system uses Kalman filtering algorithm to denoise the multi-source data and a weighted fusion algorithm to integrate the data and generate a unified epidemic prevention and control scenario situation dataset. Prevention and control task modeling and prioritization: Based on the situational dataset, a prevention and control task model is constructed, which includes four core tasks: regional patrol, target tracking, emergency interception, and information transmission. The execution requirements, time windows, resource requirements, and expected goals of each task are clearly defined. A task priority evaluation system is constructed based on the target threat level, task urgency, scope of impact, and difficulty of resource requirements, and three priority levels are divided into Level 1 urgent threat, Level 2 important prevention and control, and Level 3 routine patrol, to ensure that high-priority tasks are given priority in obtaining resources and execution authority. Multi-UAV resource constraint modeling and initial allocation: Establish a UAV resource constraint model to clarify the maximum flight time, type and number of portable devices, communication coverage radius, data processing capability, and maximum flight speed of each UAV. Record the current position, remaining battery power, and assigned tasks of each UAV. Calculate the matching score based on the matching degree between task requirements and UAV resources. Use a greedy algorithm for initial resource allocation to assign a suitable UAV cluster to each task, ensuring the basic feasibility of task execution. The collaborative scheduling optimization model is constructed and solved with the objectives of maximizing the completion rate of prevention and control tasks, minimizing task execution delay, and minimizing total system energy consumption. A multi-objective collaborative scheduling optimization model is constructed, which introduces task dependency constraints, UAV collaborative operation rules, regional conflict avoidance constraints, and battery life constraints. An improved particle swarm optimization algorithm combined with the Lagrange dual method is used to solve the model. The improved particle swarm optimization algorithm improves the convergence speed by dynamically adjusting the inertia weight and learning factor, and determines the optimal flight path, task execution order, and resource scheduling strategy of UAVs to achieve efficient collaboration of multiple UAVs. Dynamic scheduling and conflict resolution are implemented. The system monitors changes in the situation of the prevention and control scenario, updates the status of drones, and the progress of task execution in real time. When situations such as changes in target movement trajectory, drone malfunctions, insertion of new tasks, or sudden changes in environmental parameters occur, a dynamic scheduling mechanism is triggered. Based on the idea of ​​rolling time-domain optimization, time windows are divided and the scheduling model is resolved in each window. A conflict detection algorithm is used to identify drone path conflicts and resource competition conflicts. Conflicts are resolved through path replanning, dynamic adjustment of task priorities, activation of backup drones, and task splitting to ensure the continuity and stability of scheduling. The effectiveness of prevention and control measures is evaluated and strategies are iterated. An evaluation index system for prevention and control effectiveness is established, including task completion timeliness, target capture rate, regional coverage integrity, resource utilization rate, system energy consumption level, and conflict occurrence rate. The values ​​of each index are calculated through real-time data collection and analysis. Based on the evaluation results, the scheduling strategy is iteratively optimized using the random forest algorithm. Task priority weights, resource allocation rules, collaborative scheduling parameters, and algorithm solution coefficients are adjusted to continuously improve the adaptability, accuracy, and efficiency of prevention and control scheduling.

[0026] This invention also includes: Accurate task priority calculation steps: Task priority is quantified through multi-dimensional indicators. The specific formula is: Pt=ω1×Tl+ω2×Ed+ω3×Sr+ω4×Rd, where Pt is the priority coefficient of the t-th task, the larger the value, the higher the priority; ω1 is the target threat level weight; ω2 is the task urgency weight; ω3 is the impact range weight; ω4 is the resource requirement weight, and ω1+ω2+ω3+ω4=1; Tl corresponds to the target threat level, with values ​​from 1 to 5; level 1 corresponds to the lowest threat, with a value of 1; level 5 corresponds to the highest threat, with a value of 5; Ed is the task urgency, with a value from 0 to 1, the higher the urgency, the larger the value; Sr is the impact range, in square kilometers; Rd is the resource requirement difficulty, with a value from 0 to 1, the more complex the requirement, the larger the value. Through the comprehensive calculation of multi-dimensional indicators, the scientific quantification of task priority is achieved, providing an accurate basis for scheduling decisions and avoiding subjective bias in priority determination. The steps for evaluating the efficiency of collaborative scheduling are as follows: Quantify the effect of multi-UAV collaborative execution, using the following formula: ,in This is the collaborative scheduling efficiency coefficient, ranging from 0 to 1. A larger value indicates higher collaborative efficiency. The actual number of tasks completed. Average task execution efficiency refers to the number of tasks completed per unit of time. Resource utilization rate is the ratio of the actual amount of resources used to the total amount of available resources. This represents the theoretical maximum number of tasks that can be completed. This represents the optimal task execution efficiency threshold. The maximum resource utilization rate is 100%. This calculation intuitively reflects the overall effect of collaborative scheduling, provides data support for strategy optimization, and identifies weak links in the scheduling process.

[0027] In this invention, during the multi-source information collection and fusion step in the prevention and control scenario, geospatial data is acquired through satellite remote sensing and lidar scanning technology with meter-level accuracy. Target information is collected through infrared sensors, visible light cameras, and radar equipment mounted on UAVs at a frequency of 10Hz. Environmental parameters are collected collaboratively by ground sensors deployed in the prevention and control area and UAV-borne sensors. UAV status data is uploaded in real time through IoT communication protocols. The process noise covariance and observation noise covariance of the Kalman filter algorithm are dynamically adjusted according to the scenario characteristics. The weights of the weighted fusion algorithm are dynamically allocated based on data credibility to ensure the accuracy and timeliness of the fused dataset.

[0028] In this invention, in the steps of modeling and prioritizing prevention and control tasks, the task execution requirements clearly define the type of drone, sensor configuration, and execution time range required for the task. The time window is set according to the prevention and control needs, with a minimum of 5 minutes and a maximum of the drone's maximum endurance. The target threat level is determined by expert evaluation based on the target's behavioral characteristics and potential harm level. The task urgency is calculated based on the time difference between the task deadline and the current time. The impact range is determined by the area defined by the geographic information system. The resource requirement difficulty is based on the number of drones, equipment types, and computational resources required to complete the task.

[0029] In this invention, during the modeling and initial allocation of resource constraints for multiple UAVs, the maximum flight time of the UAVs is calculated using battery capacity and energy consumption models. Portable equipment types include infrared cameras, high-definition cameras, radar, and communication relay equipment. Communication coverage is calculated based on UAV flight altitude and communication equipment power. Data processing capability is determined by processor performance and algorithm efficiency. The matching degree between task requirements and UAV resources is calculated from four dimensions: equipment compatibility, flight time matching, communication coverage matching, and processing capability matching. A greedy algorithm prioritizes allocating the UAV cluster with the best resources to high-priority tasks, ensuring the priority execution of core tasks.

[0030] In this invention, during the construction and solution steps of the collaborative scheduling optimization model, the task dependency constraints clearly define the task logic with a sequential execution order. The UAV collaboration rules specify the communication collaboration methods and action coordination requirements of UAV clusters with the same task. The regional conflict avoidance constraints ensure that UAV flight paths do not overlap and flight altitudes do not conflict. The inertia weight of the improved particle swarm optimization algorithm adopts a linear decreasing strategy, the learning factor adopts an adaptive adjustment strategy, and the Lagrange dual method is used to handle the inequality constraints in the model, transforming the multi-objective optimization problem into a single-objective optimization problem, thereby improving the solution efficiency and the optimality of the solution.

[0031] In this invention, during the dynamic scheduling and conflict resolution steps, the real-time monitoring frequency is 5Hz, and the time window length for rolling time domain optimization is set to 3-10 minutes, which can be dynamically adjusted according to the complexity of the scenario. The conflict detection algorithm adopts a combination of spatial conflict detection based on grid partitioning and temporal conflict detection based on time slices. The path replanning adopts the A* algorithm combined with the artificial potential field method to ensure the optimality and safety of the path. The backup drone activation strategy clearly defines the standby status, activation conditions, and replacement process of the backup drone to ensure the continuity of the mission in case of failure.

[0032] In this invention, in the steps of prevention and control effect evaluation and strategy iteration, the task completion time is the ratio of the actual task completion time to the preset completion time; the target capture rate is the ratio of the number of successfully captured targets to the total number of targets; the area coverage integrity is the ratio of the actual coverage area to the total area of ​​the prevention and control area; the resource utilization rate includes power utilization rate, equipment utilization rate, and drone utilization rate; the system energy consumption level is the sum of the flight energy consumption, equipment energy consumption, and communication energy consumption of all drones; and the training data of the random forest algorithm includes historical scheduling data, scene parameters, and evaluation results. The model parameters are optimized through iterative training to improve the accuracy of strategy adjustment.

[0033] This invention also includes: Emergency response enhancement steps: Establish a rapid response mechanism for first-priority tasks, pre-plan emergency flight paths, reserve emergency drone resources, and simplify scheduling decision-making processes. When a first-priority task is triggered, the emergency drone cluster can start execution within 30 seconds. Real-time and secure data transmission is ensured through a dedicated communication link. A priority preemption mechanism is adopted to ensure that emergency tasks receive optimal resource allocation. At the same time, emergency scheduling process data is recorded to provide case support for subsequent emergency strategy optimization.

[0034] Example 1: Optimized Implementation of Collaborative Low-Altitude Prevention and Control Scheduling at Borders This embodiment is applied to a low-altitude air defense scenario in a border area. The area covers 500 square kilometers and includes complex terrain such as mountains, hills, and grasslands. There are risks of illegal border crossing and transportation of contraband. Ten drones are deployed to form a collaborative defense cluster, including six regular patrol drones, three target tracking drones, and one emergency interception drone. It is necessary to achieve 24-hour uninterrupted defense and take into account both coverage integrity and emergency response speed.

[0035] I. Core Implementation Details Multi-source information collection and fusion for epidemic prevention and control scenarios: Geospatial data is acquired through satellite remote sensing and lidar scanning to accurately map terrain, obstacle distribution, and boundary coordinates of the prevention and control area with meter-level accuracy; Target information is collected collaboratively by infrared sensors and radar equipment on UAVs and ground monitoring stations. Infrared sensors capture thermal images of targets at night, while radar equipment detects the position and speed of moving targets at a frequency of 10Hz, clearly identifying target type, real-time location, movement speed, and threat level; Environmental parameters are collected through ground sensors deployed along the border and UAV-borne sensors, including wind speed, visibility, precipitation intensity, and electromagnetic interference intensity. The wind speed sensor collects data every 5 seconds to address the impact of sandstorms in border areas; UAV status data includes remaining battery power, flight time, payload equipment operating status, and communication link quality, which are uploaded to the dispatch center in real time via IoT communication protocols. The Kalman filter algorithm is used to denoise the multi-source data, filtering out radar signal interference caused by wind and sand and communication data distortion caused by electromagnetic interference. Then, the data is integrated by a weighted fusion algorithm, and the weights are dynamically assigned according to the data credibility: infrared data weight 0.3, radar data weight 0.4, environmental data weight 0.15, and UAV status data weight 0.15, to generate a unified prevention and control scenario situation dataset.

[0036] Prevention and control task modeling and prioritization: Based on the situational awareness dataset, three core task categories are constructed: regional patrol tasks, covering the entire border and completing a full patrol every 2 hours; target tracking tasks, continuously tracking and uploading the trajectory of detected moving targets; and emergency interception tasks, targeting confirmed illegal border crossing targets, implementing close-range expulsion or interception. The execution requirements for each task are clearly defined: patrol tasks must cover all key areas, tracking task positioning error must not exceed 10 meters, and interception tasks must respond before the target enters the core area. Priorities are assigned based on target threat level, task urgency, impact range, and resource requirements. Level 1 urgent threats correspond to illegal border crossing personnel or vehicles, with the highest priority coefficient; Level 2 important prevention and control corresponds to suspicious moving targets requiring further confirmation; and Level 3 routine patrols are daily prevention and control tasks, with the lowest priority.

[0037] Multi-UAV Resource Constraint Modeling and Initial Allocation: A UAV resource constraint model is established. A standard patrol UAV has a maximum endurance of 8 hours, can carry an infrared camera and lightweight radar, and has a communication coverage radius of 50 kilometers. A target tracking UAV has a maximum endurance of 6 hours, is equipped with a high-definition infrared camera and a high-precision positioning module, and has stronger data processing capabilities. An emergency interception UAV has a maximum endurance of 4 hours, is equipped with acoustic and optical deterrent devices and communication jamming devices, and has a higher maximum flight speed. The current location, remaining battery power, and assigned tasks of each UAV are recorded. A task-UAV resource matching score is calculated based on four dimensions: equipment compatibility, endurance matching, communication coverage matching, and processing capacity matching. A greedy algorithm is used for initial allocation, prioritizing the allocation of primary tasks to emergency interception and target tracking UAVs, secondary tasks to the remaining tracking UAVs and some patrol UAVs, and tertiary patrol tasks to the remaining patrol UAVs, ensuring that core tasks receive priority access to suitable resources.

[0038] Construction and Solution of the Cooperative Scheduling Optimization Model: With the objectives of maximizing the completion rate of prevention and control tasks, minimizing task execution latency, and minimizing total system energy consumption, a multi-objective cooperative scheduling optimization model is constructed. This model introduces constraints such as task dependency (tracking tasks must start after patrol tasks detect the target), UAV cooperative operation rules (interception UAVs must maintain communication and coordination with tracking UAVs), regional conflict avoidance constraints (UAV flight path spacing must be no less than 500 meters), and battery life constraints (remaining battery power must meet return trip requirements). An improved particle swarm optimization algorithm combined with the Lagrange dual method is used to solve the model. The improved particle swarm optimization algorithm uses a linear decreasing strategy for the inertia weight, reducing it from 0.9 to 0.4. The learning factor is adaptively adjusted, with the cognitive factor decreasing from 2.0 to 1.5 and the social factor increasing from 1.5 to 2.0, improving the convergence speed. The Lagrange dual method handles inequality constraints, transforming the multi-objective optimization problem into a single-objective optimization problem. This determines the optimal flight path for UAVs, the task execution order, and resource scheduling strategies. For example, patrol UAVs use a zoned rotation route, tracking UAVs dynamically adjust their paths based on the target trajectory, and interception UAVs have pre-set emergency response routes.

[0039] Dynamic scheduling and conflict resolution: Real-time monitoring of changes in the situation of the prevention and control scenario triggers a dynamic scheduling mechanism when there are sudden changes in the target's movement trajectory, wind speeds exceeding 15 m / s, drone battery levels below 20%, or new tasks are inserted. Based on the rolling time-domain optimization concept, a 5-minute time window is set, and the scheduling model is resolved within each window. A combination of grid-based spatial conflict detection and time-slice-based temporal conflict detection is used to identify drone path conflicts and resource competition conflicts. When the paths of two drones are about to overlap, path replanning is performed using the A* algorithm combined with the artificial potential field method; when the tracking drone's battery is low, a backup tracking drone is activated to take over, and the original drone returns to recharge; when a new primary task is inserted, the execution order of low-priority tasks is adjusted to free up resources for emergency tasks.

[0040] Prevention and Control Effectiveness Evaluation and Strategy Iteration: An evaluation index system for prevention and control effectiveness is established, including task completion timeliness, target capture rate, regional coverage integrity, resource utilization rate, system energy consumption level, and conflict occurrence rate. Task completion timeliness is the ratio of actual completion time to the preset completion time; the target capture rate is the ratio of the number of successfully tracked and dealt with targets to the total number of detected targets; and regional coverage integrity is the ratio of the actual covered area to the total area of ​​the prevention and control zone. Through real-time data collection and analysis, the values ​​of each index are calculated. Based on the evaluation results, a random forest algorithm is used to iteratively optimize the scheduling strategy, adjusting task priority weights, resource allocation rules, and collaborative scheduling parameters. For example, environmental parameter weights are adjusted based on the frequency of seasonal sandstorms, and patrol routes are optimized based on target activity patterns.

[0041] Enhanced Emergency Response: A rapid response mechanism is established for top-priority tasks, with three pre-planned emergency flight paths and two backup drones on standby, simplifying the dispatch and decision-making process. Upon detection of an illegal border-crossing target, the emergency drone swarm initiates execution within 30 seconds, ensuring data transmission via a dedicated communication link. A priority preemption mechanism is employed, suspending lower-priority patrol tasks and allocating optimal resources to emergency tasks. Emergency dispatch process data, including target trajectory, drone response time, and handling results, is recorded to provide case support for subsequent emergency strategy optimization.

[0042] Table 1 shows a comparison of the effectiveness of low-altitude air defense and control dispatching at the border: Table 1

[0043] Table 1 clearly demonstrates the significant advantages of this invention in border low-altitude defense scenarios. Traditional scheduling methods, employing fixed patrol routes and static resource allocation, struggle to handle complex terrain and dynamic targets, resulting in low target acquisition rates, delayed responses, blind spots in area coverage, resource waste, and frequent conflicts. This application achieves precise situational awareness through deep fusion of multi-source information, dynamically allocates resources based on priorities, implements an improved optimization algorithm for optimal path and task scheduling, rapidly resolves conflicts through a dynamic adjustment mechanism, and significantly shortens response time through an emergency response mechanism. High target acquisition rates and coverage integrity ensure effective border defense, while high resource utilization and low conflict rates reduce operating costs, fully meeting the practical needs of border defense.

[0044] Example 2: Optimized Implementation of Collaborative Prevention and Control Scheduling for Urban Low-Altitude Security This embodiment is applied to a low-altitude security scenario in the core area of ​​a city. The area covers 50 square kilometers, with dense high-rise buildings and a large population flow. There are risks of suspicious personnel gathering and illegal flight of drones. Eight drones are deployed to form a collaborative security cluster, including three fixed-point monitoring drones, four mobile patrol drones, and one emergency response drone. The goal is to achieve full coverage of key areas and rapid response to emergencies, while taking into account both the effectiveness of prevention and control and adaptability to the urban environment.

[0045] I. Core Implementation Details Multi-source information collection and fusion for epidemic prevention and control scenarios: Geospatial data is acquired through urban GIS and LiDAR scanning to accurately map building distribution, road networks, and boundaries of prevention and control areas, and to mark the coordinates of obstacles such as high-rise buildings and bridges; Target information is collected through high-definition cameras, acoustic sensors, and miniature radars mounted on UAVs. High-definition cameras capture the characteristics of people and vehicles on the ground, acoustic sensors identify abnormal sounds, and miniature radars detect low-altitude flying targets at a frequency of 15Hz, clarifying target type, real-time location, behavioral characteristics, and threat level; Environmental parameters are collected through urban environmental monitoring stations and UAV-borne sensors, including temperature, humidity, visibility, building obstruction, and electromagnetic interference intensity, with a focus on monitoring signal obstruction and electromagnetic interference in densely populated high-rise areas; UAV status data includes remaining battery power, flight time, operational status of payload equipment, and communication link quality, uploaded in real time via IoT communication protocols. A Kalman filter algorithm is used to remove signal noise caused by building reflections, and a weighted fusion algorithm integrates multi-source data, with camera data weighted at 0.4, radar data at 0.3, acoustic data at 0.1, and environmental and status data at 0.2, generating a unified epidemic prevention and control scenario situation dataset.

[0046] Prevention and control task modeling and prioritization: Based on the situational dataset, three core task categories are constructed: emergency response tasks address sudden incidents and unauthorized drone flights; key monitoring tasks address suspicious gatherings and illegal road occupation; and routine patrol tasks address daily prevention and control of urban roads and key areas. The execution requirements for each task are clearly defined: emergency response tasks require arrival at the scene within 5 minutes; key monitoring tasks require continuous tracking of target dynamics; and routine patrol tasks require coverage of key areas once per hour. Priorities are assigned based on target threat level, task urgency, impact range, and resource requirements: Level 1 emergency threat corresponds to sudden incidents and unauthorized drone flights; Level 2 important prevention and control corresponds to suspicious gatherings and violations; and Level 3 routine patrol is the daily inspection task.

[0047] Multi-UAV Resource Constraint Modeling and Initial Allocation: A UAV resource constraint model is established. Fixed-point monitoring UAVs have a maximum endurance of 10 hours, are equipped with high-definition cameras and infrared sensors, and are fixedly deployed on the top of high-rise buildings, with a communication coverage radius of 30 kilometers. Mobile patrol UAVs have a maximum endurance of 6 hours, are equipped with high-definition cameras, acoustic sensors, and lightweight radar, and can flexibly adjust patrol routes. Emergency response UAVs have a maximum endurance of 5 hours, and are equipped with audible and visual alarm devices, communication relay modules, and non-lethal jamming devices. The current location, remaining battery power, and assigned tasks of each UAV are recorded. A matching score is calculated based on four dimensions: equipment compatibility, endurance matching, communication coverage matching, and urban environment adaptability. A greedy algorithm is used for initial allocation. First-level tasks are prioritized for emergency response UAVs, second-level tasks are assigned to mobile patrol UAVs, and third-level tasks are undertaken by fixed-point monitoring UAVs and the remaining mobile patrol UAVs, ensuring accurate matching of tasks and UAV resources.

[0048] Construction and Solution of Collaborative Scheduling Optimization Model: With the objectives of maximizing the completion rate of prevention and control tasks, minimizing task execution delay, and minimizing total system energy consumption, a multi-objective collaborative scheduling optimization model is constructed. This model introduces constraints such as task dependency (key monitoring tasks must be initiated after targets are detected during routine patrols), UAV collaborative operation rules (emergency response UAVs must share target data with fixed-point monitoring UAVs), regional conflict avoidance constraints (UAV flight altitude stratification: 100 meters for low altitude, 200 meters for medium altitude, and 300 meters for high altitude), and battery life constraints (remaining battery power must meet the landing requirements at the nearest take-off and landing point). An improved particle swarm optimization algorithm combined with the Lagrange dual method is used to solve the model. The improved particle swarm optimization algorithm reduces the inertia weight from 0.8 to 0.3, and the learning factor, cognitive factor, and social factor are 1.8 and 2.2 respectively, adapting to the path planning needs of complex urban environments. The Lagrange duality method addresses height stratification and building occlusion constraints to determine the optimal flight path, mission execution sequence, and resource scheduling strategy for drones. For example, mobile patrol drones can use gridded patrol routes to avoid areas obscured by tall buildings, while emergency response drones can plan the shortest path and avoid no-fly zones.

[0049] Dynamic scheduling and conflict resolution: A dynamic scheduling mechanism is triggered in real-time by monitoring changes in the situation, including when a target enters a building-obstructed area, the drone's communication link is interrupted, a new emergency task is inserted, or the drone's battery level drops below 25%. Based on rolling time-domain optimization, a 3-minute time window is set, and the scheduling model is resolved within each window. Conflict resolution is achieved through a combination of altitude-level hierarchical management and time-slice allocation, allowing drones at different altitudes in the same area to travel at different times to avoid path overlap. When a mobile patrol drone encounters signal obstruction, it switches to a backup communication link and adjusts its flight altitude to an area with good signal. When a new unauthorized drone is detected, routine patrol tasks in the surrounding area are suspended, and the nearest mobile patrol drone and emergency response drone are coordinated to handle the situation.

[0050] Prevention and Control Effectiveness Evaluation and Strategy Iteration: An evaluation index system for prevention and control effectiveness is established, including task completion timeliness, target identification accuracy, coverage density in key areas, resource utilization rate, system energy consumption level, and resident interference. Task completion timeliness is the ratio of actual response time to the preset standard time; target identification accuracy is the ratio of the number of correctly identified target types to the total number of detected targets; and resident interference is calculated by collecting feedback from surrounding residents and noise monitoring data. Through real-time data collection and analysis, the values ​​of each index are calculated. Based on the evaluation results, a random forest algorithm is used to iteratively optimize the scheduling strategy, adjusting patrol route density, target identification threshold, and drone flight altitude. For example, the frequency of patrols in key areas is increased during peak pedestrian traffic hours, while drone flight altitude and speed are reduced in residential areas to minimize noise interference.

[0051] Enhanced Emergency Response: A rapid response mechanism has been established for top-priority tasks, with five pre-planned urban emergency flight routes avoiding sensitive areas such as schools and hospitals, and one backup drone on standby. Upon detection of unauthorized drone flights or other emergencies, the emergency response drone will be activated within one minute, communicating in real-time with the ground command center via an encrypted communication link. A priority-based preemption mechanism will be employed, suspending lower-priority patrol tasks and dispatching nearby fixed-point monitoring drones to provide real-time video support. Emergency dispatch process data, including incident handling procedures, drone coordination methods, and resident feedback, will be recorded to provide data support for subsequent optimization of urban security strategies.

[0052] Table 2 is a comparison table of the effectiveness of urban low-altitude security dispatch: Table 2

[0053] Table 2 data highlights the core advantages of this invention in urban low-altitude security scenarios. Traditional scheduling methods, relying on fixed-point monitoring and manual dispatching, struggle to adapt to urban high-rise obstruction and dynamic pedestrian flow, resulting in low target identification accuracy, delayed response, insufficient coverage in key areas, and significant resource waste and resident interference. This application accurately identifies targets in complex urban environments through multi-source information fusion, allocates resources based on priority and urban environmental adaptability, plans optimal paths and avoids obstacles using an improved optimization algorithm, employs a dynamic adjustment mechanism to quickly respond to emergencies, and establishes an emergency response mechanism to meet the needs of urban emergency handling. High target identification accuracy and coverage density ensure security effectiveness, high resource utilization reduces operating costs, and low resident interference improves urban adaptability, fully meeting the diverse needs of low-altitude security in core urban areas.

[0054] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for optimizing the collaborative prevention and control scheduling of low-altitude multi-UAVs, characterized in that, include: Steps for collecting and fusing multi-source information in the prevention and control scenario: Collect geospatial data, target information, environmental parameters, and UAV status data in the low-altitude area; denoise the data using the Kalman filter algorithm; integrate the data using a weighted fusion algorithm to generate a unified prevention and control scenario situation dataset. Prevention and control task modeling and priority division steps: Based on the situation dataset, a prevention and control task model is constructed. According to the target threat level, task urgency, impact range, and resource demand difficulty, a task priority evaluation system is constructed, and three priority levels are divided: Level 1 urgent threat, Level 2 important prevention and control, and Level 3 routine patrol. The steps for modeling and initial allocation of resource constraints among drones are as follows: Establish a drone resource constraint model, including the flight time, type and number of carried equipment, communication coverage radius, data processing capability, and maximum flight speed of each drone; record the current position, remaining battery power, and assigned tasks of each drone; calculate the matching score based on the matching degree between task requirements and drone resources; use a greedy algorithm to perform initial resource allocation; and assign a suitable drone cluster to each task. The steps for constructing and solving the collaborative scheduling optimization model are as follows: With the goals of maximizing the completion rate of prevention and control tasks, minimizing task execution delay, and minimizing the total energy consumption of the system, a collaborative scheduling optimization model is constructed. Task dependency constraints, UAV collaborative operation rules, regional conflict avoidance constraints, and battery life constraints are introduced. The model is solved by combining the particle swarm optimization algorithm with the Lagrange dual method. The particle swarm optimization algorithm improves the convergence speed by dynamically adjusting the inertia weight and learning factor. Dynamic scheduling and conflict resolution steps: Real-time monitoring of changes in the situation of the prevention and control scenario, drone status updates, and task execution progress. When there are changes in the target movement trajectory, drone malfunctions, new task insertions, or sudden changes in environmental parameters, the dynamic scheduling mechanism is triggered. Based on the rolling time domain optimization idea, time windows are divided and the scheduling model is re-solved in each window. Conflict detection algorithms are used to identify drone path conflicts and resource competition conflicts. Conflicts are resolved through path replanning, dynamic adjustment of task priorities, activation of backup drones, and task splitting. Prevention and control effectiveness evaluation and strategy iteration: Establish a prevention and control effectiveness evaluation index system, calculate the values ​​of various indicators through real-time data collection and analysis, and use the random forest algorithm to iteratively optimize the scheduling strategy based on the evaluation results, adjusting task priority weights, resource allocation rules, collaborative scheduling parameters, and algorithm solution coefficients.

2. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, Also includes: Task priority calculation steps: Quantify task priority through indicators. The formula is: Pt=ω1×Tl+ω2×Ed+ω3×Sr+ω4×Rd, where Pt is the priority coefficient of the t-th task, ω1 is the target threat level weight, ω2 is the task urgency weight, ω3 is the impact range weight, ω4 is the resource requirement weight, and ω1+ω2+ω3+ω4=1, Tl is the target threat level, Ed is the task urgency, Sr is the impact range, and Rd is the resource requirement difficulty.

3. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, Also includes: The steps for evaluating the efficiency of collaborative scheduling are as follows: Quantify the collaborative execution effect among UAVs, using the following formula: ,in The efficiency coefficient for coordinated scheduling. The actual number of tasks completed. Average task execution efficiency refers to the number of tasks completed per unit of time. For resource utilization, This represents the theoretical maximum number of tasks that can be completed. This represents the optimal task execution efficiency threshold. For resource utilization rate.

4. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the information collection and fusion steps of the prevention and control scenario, geospatial data is acquired through satellite remote sensing and lidar scanning technology, target information is collected through infrared sensors, visible light cameras and radar equipment carried by UAVs, environmental parameters are collected collaboratively by ground sensors deployed in the prevention and control area and UAV-borne sensors, UAV status data is uploaded in real time through IoT communication protocols, the process noise covariance and observation noise covariance of the Kalman filter algorithm are dynamically adjusted according to the scenario characteristics, and the weights of the weighted fusion algorithm are dynamically allocated based on data credibility.

5. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the steps of modeling and prioritizing prevention and control tasks, the task execution requirements clearly define the type of drone, sensor configuration, and execution time range required for the task. The time window is set according to the prevention and control needs. The target threat level is determined by expert evaluation based on the target's behavioral characteristics and the degree of harm. The task urgency is calculated based on the time difference between the task deadline and the current time. The impact range is determined by the area delineated by the geographic information system. The difficulty of resource requirements is based on the number of drones, equipment types, and computing resources required to complete the task.

6. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the resource constraint modeling and initial allocation steps among drones, the maximum flight time of a drone is calculated using battery capacity and energy consumption models. The types of equipment carried include infrared cameras, high-definition cameras, radar, and communication relay equipment. The communication coverage is calculated based on the drone's flight altitude and the power of the communication equipment. The data processing capability is determined by processor performance and algorithm efficiency. The matching degree between task requirements and drone resources is calculated from four dimensions: equipment adaptability, flight time matching, communication coverage matching, and processing capability matching. The greedy algorithm prioritizes allocating the drone cluster with the best resources to high-priority tasks.

7. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the construction and solution steps of the collaborative scheduling optimization model, the task dependency constraint sets the task logic with a sequential execution order, the UAV collaboration rule specifies the communication collaboration method and action coordination requirements of the UAV cluster for the same task, the regional conflict avoidance constraint ensures that the UAV flight paths do not overlap and the flight altitudes do not conflict, the inertia weight of the particle swarm optimization algorithm adopts a linear decreasing strategy, the learning factor adopts an adaptive adjustment strategy, and the Lagrange dual method is used to handle the inequality constraints in the model, transforming the multi-objective optimization problem into a single-objective optimization problem to be solved.

8. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the dynamic scheduling and conflict resolution steps, the system is dynamically adjusted according to the complexity of the scenario. The conflict detection algorithm adopts a combination of spatial conflict detection based on grid partitioning and temporal conflict detection based on time slices. The path replanning adopts an algorithm combined with the artificial potential field method. The backup drone activation strategy sets the standby status, activation conditions, and replacement process of the backup drone.

9. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, In the steps of prevention and control effect evaluation and strategy iteration, the task completion time is the ratio of the actual task completion time to the preset completion time, the target capture rate is the ratio of the number of successfully captured targets to the total number of targets, the area coverage integrity is the ratio of the actual coverage area to the total area of ​​the prevention and control area, the resource utilization rate includes power utilization rate, equipment utilization rate, and drone utilization rate, the system energy consumption level is the sum of the flight energy consumption, equipment energy consumption, and communication energy consumption of all drones, and the training data of the random forest algorithm includes historical scheduling data, scene parameters, and evaluation results.

10. The low-altitude multi-UAV collaborative prevention and control scheduling optimization method according to claim 1, characterized in that, Also includes: Emergency response enhancement steps: For first-priority tasks, establish a rapid response mechanism, pre-plan emergency flight paths, reserve emergency drone resources, and simplify the scheduling decision-making process. When a first-priority task is triggered, the emergency drone cluster will start execution within 30 seconds. The real-time and security of data transmission will be ensured through a dedicated communication link. A priority preemption mechanism will be adopted to ensure that emergency tasks obtain the best resource allocation, while recording emergency scheduling process data.

Images