Police unmanned aerial vehicle management and control system for emergency response

Abstract

The application relates to the field of unmanned aerial vehicle control, and discloses a police unmanned aerial vehicle management and control system for handling emergencies, which comprises a sensing module, an allocation module, a planning module, a control module, a tracking module and a feedback module. The sensing module is used for collecting environmental data in real time and generating a three-dimensional environmental model; the allocation module optimizes a task allocation scheme of unmanned aerial vehicles; the planning module dynamically plans an optimal flight path of the unmanned aerial vehicles according to the task allocation result and the three-dimensional environmental model; the control module dynamically adjusts the attitude and flight height of the unmanned aerial vehicles; the tracking module tracks a target position in real time and predicts a motion trajectory of the target; and the feedback module returns the execution situation of the unmanned aerial vehicles, the target tracking result and a real-time scene picture to a command center. Through the cooperative work of the modules, the application realizes quick response and efficient handling of emergencies, is particularly suitable for police tasks in complex and dynamic environments, and has the advantages of high response efficiency, strong environmental adaptability and high resource utilization rate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle control, in particular to a police unmanned aerial vehicle management and control system for handling emergencies. BACKGROUND

[0002] In recent years, with the rapid development of unmanned aerial vehicle technology, its application in the field of public security has gradually become a research hotspot. Police unmanned aerial vehicles, with their flexibility and real-time response capabilities, have been widely used in tasks such as on-site monitoring of emergencies, target tracking, and regional alert. However, in actual application, due to the high dynamicity, complexity, and variability of emergencies, existing police unmanned aerial vehicle management and control systems still have many technical bottlenecks when facing complex scenarios.

[0003] Existing technologies usually rely on a single unmanned aerial vehicle or a simple multi-unmanned aerial vehicle coordination model, and have insufficient optimization capabilities for task allocation, path planning, and target tracking. In multi-target, multi-unmanned aerial vehicle task scenarios, due to the lack of effective dynamic allocation mechanisms, existing systems often have difficulty adjusting unmanned aerial vehicle resource allocation according to task priorities and real-time environments, resulting in low response efficiency. In addition, existing path planning methods mostly use preset paths or static planning, which are difficult to adjust in real time in dynamic environments, leading to unmanned aerial vehicles easily deviating from the target or increasing flight risks in scenarios with dense obstacles or changing wind speeds.

[0004] In terms of target tracking, existing systems mostly use simple positioning and tracking algorithms, which are difficult to accurately predict the movement of mobile targets, especially in complex target trajectory or multi-target interference scenarios, which can easily result in tracking errors or even target loss. In addition, in terms of data feedback and command linkage, existing systems usually only support simple one-way data transmission, and cannot provide efficient real-time command support, limiting the flexible application of unmanned aerial vehicles in emergencies.

[0005] In summary, existing technologies have technical defects such as low response efficiency, poor environmental adaptability, insufficient target tracking capability, and weak command linkage in the management and control of police unmanned aerial vehicles in emergency situations, and there is an urgent need for a systematic solution that can comprehensively solve the above problems. SUMMARY

[0006] In view of the deficiencies of the prior art, the present application provides a police unmanned aerial vehicle management and control system for handling emergencies, which solves the problems of low response efficiency, unreasonable task allocation, insufficient path planning capability, and inaccurate target tracking and data feedback of existing police unmanned aerial vehicles in complex dynamic environments.

[0007] To achieve the above purpose, the present application realizes the following technical scheme: a police unmanned aerial vehicle management and control system for handling emergencies, comprising:

[0008] a perception module for collecting environmental data in real time, the environmental data including obstacle distribution, wind speed, target dynamic position, etc., and generating a three-dimensional environmental model;

[0009] a distribution module for optimizing the UAV task distribution scheme based on the environmental data of the perception module and the priority of the police task;

[0010] a planning module for planning the optimal flight path of the UAV according to the task distribution result of the distribution module and the three-dimensional environmental model, the path planning avoiding obstacles and minimizing flight time;

[0011] a control module for dynamically adjusting the attitude and flight height of the UAV according to the flight path of the planning module;

[0012] a tracking module for tracking the target position in real time and predicting the trajectory of the dynamic target;

[0013] a feedback module for feeding back the task execution of the UAV, the target tracking result and the environmental data to the command center.

[0014] Preferably, the perception module includes a laser radar, an ultrasonic sensor and a wind speed sensor, the laser radar being used to collect obstacle distribution point cloud data, the ultrasonic sensor being used to perceive close-range obstacles, and the wind speed sensor being used to generate a wind speed distribution model.

[0015] Preferably, the distribution module realizes task distribution optimization based on an improved Hungarian algorithm, the optimization goal of the task distribution being to minimize the comprehensive cost of the UAV executing the task, the comprehensive cost being composed of the distance from the UAV to the target, the priority of the task and the influence weight of the wind speed and obstacles.

[0016] Preferably, the planning module realizes dynamic planning of the flight path based on a particle swarm optimization algorithm, the optimization goal of the flight path being:

[0017] minimizing the flight time of the UAV;

[0018] avoiding obstacles on the flight path;

[0019] reducing the interference of the wind speed on the flight of the UAV.

[0020] Preferably, the particle swarm optimization algorithm realizes path planning through the following steps:

[0021] initializing path nodes based on the three-dimensional environmental model;

[0022] iteratively updating the speed and position of the path nodes according to the current position of the UAV, the target position and the obstacle distribution;

[0023] The optimal path is selected by comparing the historical optimal position of the path node with the global optimal position.

[0024] Preferably, the control module dynamically adjusts the attitude of the UAV based on adaptive control theory, including the pitch angle, the yaw angle and the flight height, wherein the attitude state vector of the UAV is calculated in real time, a Lyapunov function is constructed, and the control input of the UAV is adjusted to ensure the stability of the flight attitude.

[0025] Preferably, the tracking module predicts the target trajectory by using an extended Kalman filtering algorithm, which includes the following steps:

[0026] According to the motion state of the target and the flight state of the UAV, the position of the target at the next moment is predicted;

[0027] Based on the observation data of the target collected by the sensor, the predicted position of the target is corrected;

[0028] The state of the target and the covariance matrix thereof are updated.

[0029] Preferably, the feedback module includes a real-time video return unit and a remote shouting unit, the video return unit is used for transmitting the live picture to the command center, and the remote shouting unit is used for conveying instructions or warning information to the target area through the UAV.

[0030] Preferably, the system re-triggers the distribution module and the planning module according to the real-time sensed environmental change data, so as to update the task distribution result and the flight path.

[0031] The application also provides a police UAV management and control method for handling emergency events, which includes the following steps:

[0032] Collecting environmental data to generate a three-dimensional environmental model;

[0033] According to the environmental data and the priority of the police task, the UAV task distribution scheme is optimized;

[0034] Based on the task distribution result and the three-dimensional environmental model, the optimal flight path of the UAV is dynamically planned;

[0035] According to the planned path, the attitude and the flight height of the UAV are dynamically adjusted to ensure that the UAV flies according to the planned path;

[0036] The position of the target is tracked in real time, and the motion trajectory of the target is predicted;

[0037] The execution of the UAV, the target tracking result and the real-time picture are returned to the command center through the feedback module.

[0038] The application provides a police UAV management and control system for handling emergency events.

[0039] Advantages:

[0040] 1、The present application can quickly allocate and handle the police task in the emergency through the real-time data acquisition of the perception module and the dynamic optimization of the distribution module. The planning module generates the optimal flight path, and the control module ensures the stable flight of the UAV. The overall process is compact and efficient, greatly shortens the task response time, and improves the response ability of the system in the emergency.

[0041] 2、The system can real-time perceive and dynamically update the complex conditions such as obstacles and wind speed in the environment through the three-dimensional environment model generated by the perception module. The planning module and the control module adjust the path and attitude according to the environmental data, ensure that the UAV can adapt to different scenes and dynamic changes, and enhance the robustness and flexibility of the system in complex environments.

[0042] 3、The tracking module of the present application uses extended Kalman filter algorithm to realize real-time tracking and trajectory prediction of dynamic targets, ensuring high precision and continuity of target position data. Whether the target is stationary or moving, accurate locking and continuous monitoring of the target can be realized, providing reliable support for task execution.

[0043] 4、The command center can obtain the flight state, target tracking result and live picture information of the UAV at any time through the real-time data back function of the feedback module, providing comprehensive decision basis for the commander. In addition, the remote shouting function can directly convey instructions or warning information to the target area, improving the effectiveness of command and intervention.

[0044] 5、The distribution module of the present application reasonably allocates UAV resources through dynamic optimization algorithm, which can maximize the utilization of system resources in complex scenes with multiple tasks and multiple UAVs. The path planning module further optimizes the flight path, reduces unnecessary flight distance and energy consumption, thereby significantly improving the utilization efficiency of resources and the task completion rate.

[0045] 6、The present application supports dynamic updating of task allocation and flight path according to real-time environmental changes. Through the continuous monitoring of the perception module and the closed-loop response of the feedback module, the system can quickly adjust the task scheme when the target position changes or the environment mutates, maintaining the continuity and accuracy of task execution. BRIEF DESCRIPTION OF DRAWINGS

[0046] Figure 1 The system architecture diagram of the present application;

[0047] Figure 2 The method flowchart of the present application. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the specification of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0049] Please refer to the drawings attached Figure 1 The present application provides a police unmanned aerial vehicle management and control system for emergency response, which realizes rapid response, accurate execution and real-time feedback of unmanned aerial vehicles in emergencies. The system can adapt to dynamic changes in complex environments, optimize unmanned aerial vehicle resource allocation and improve disposal efficiency. The system includes a perception module, an allocation module, a planning module, a control module, a tracking module and a feedback module. The modules of the system will be described in detail below.

[0050] Perception module

[0051] The perception module is one of the core components of the police unmanned aerial vehicle management and control system of the present application, mainly responsible for real-time collection of environmental data, providing data support and decision basis for the allocation, planning, control and tracking modules of the system. Through the cooperative work of multiple sensors, the module obtains data such as obstacle distribution, wind speed dynamics and target position of the emergency scene and its surroundings, and finally generates a three-dimensional environmental model.

[0052] As a technical implementation, this module aims to solve the problem of insufficient perception of dynamic environmental data by unmanned aerial vehicles in complex environments, and provides basic support for path planning and task allocation by constructing a high-precision three-dimensional environmental model in real time.

[0053] It should be noted that the perception module not only supports obstacle recognition, but also can dynamically monitor the wind speed changes in the environment and collect position data of moving targets. These functions are realized through the integration of different sensors and algorithms, and the specific technical implementation is as follows.

[0054] In this embodiment, the perception module is mainly composed of a laser radar (LIDAR), an ultrasonic sensor and a wind speed sensor, which cooperatively collect environmental information. For example, the data collection process is as follows:

[0055] The laser radar emits laser pulses and receives return signals to calculate the spatial position of obstacles and form point cloud data. As an option, the perception range of the laser radar can be adjusted according to the specific flight height of the unmanned aerial vehicle to ensure adaptability to different terrain and building environments.

[0056] Specifically, the point cloud data generated by the laser radar can be described as:

[0057]

[0058] wherein:

[0059] (x i ,y i ,z i ) represents the three-dimensional coordinates of each point;

[0060] d i represents the distance between the point measured by the lidar and the UAV;

[0061] R represents the sensing range of the lidar.

[0062] It can be understood that when the environment complexity is high, for example, there are high-density buildings or multi-layer obstacles, the resolution of the point cloud data has a greater impact on the accuracy of subsequent path planning. Therefore, the application preferably uses a high-precision lidar device to improve the reliability of the data.

[0063] In some embodiments, the ultrasonic sensor is mainly used to sense the obstacles around the UAV at a short distance, and to supplement the insufficient environment data within a short distance by the lidar. As an option, the ultrasonic sensor can be installed at the bottom and sides of the UAV to cover the sensing blind area in different directions.

[0064] Specifically, the distance detection formula of the ultrasonic sensor is:

[0065]

[0066] wherein:

[0067] v is the propagation speed of ultrasonic waves in air;

[0068] t is the time from emission to reception of the ultrasonic wave.

[0069] It should be noted that the accuracy of the ultrasonic sensor is greatly affected by environmental noise, so the embodiment reduces noise interference through a signal filtering algorithm, thereby improving the stability of distance detection.

[0070] In this embodiment, the wind speed sensor is used to measure the wind speed and its direction in the environment. Specifically, the vector representation of the wind speed is:

[0071] V wind =(v x ,v y ,v z )

[0072] wherein:

[0073] v x ,v y ,v z are the components of the wind speed on the three-dimensional coordinate axes, respectively.

[0074] As one implementation method, wind speed data is dynamically integrated into the three-dimensional environment model, providing real-time reference for the UAV's flight control and path planning. It is understood that wind speed has a direct impact on the flight stability of the UAV; therefore, this invention preferably uses a highly sensitive wind speed sensor.

[0075] It should be noted that the perception module in this invention generates a unified three-dimensional environment model by fusing the data collected by the aforementioned sensors. Specifically, the model integration process includes:

[0076] Data preprocessing: Denoising and completion of point cloud data, ultrasonic distance data, and wind speed data;

[0077] Unified data coordinates: Mapping data from different sensors to the UAV's global coordinate system;

[0078] Model generation: The environment is divided into zones based on the 3D grid method, and each grid cell records the state information of obstacles, wind speed and dynamic targets.

[0079] As one possible approach, the raster resolution of the 3D environment model can be adjusted according to the specific needs of the scene. For example, in highly complex scenes, the raster resolution can be increased to describe environmental features with finer granularity.

[0080] For example, the mathematical representation of the three-dimensional raster method is as follows:

[0081]

[0082] in:

[0083] G i,j,k Indicates the state of a three-dimensional grid cell;

[0084] G i,j,k =1 indicates that there is an obstacle or abnormal wind speed within the unit, G i,j,k =0 indicates that the cell is empty.

[0085] It should be noted that the 3D environment model not only supports static descriptions of obstacles, but can also be dynamically updated. For example, when a target moves within the model, its position is updated in real time based on feedback from the tracking module.

[0086] As an extension, the perception module can be further integrated with visual sensors to achieve image-based recognition of the environment. For example, a camera can capture characteristic landmarks in the environment to help determine the type and nature of obstacles. This extended functionality can provide richer reference information for subsequent allocation and planning.

[0087] The perception module of the present application provides precise environmental data support for the task execution of the police unmanned aerial vehicle through the cooperative work of multiple sensors and efficient data fusion processing, and can maintain efficient and stable operation in dynamic and complex emergencies. These features lay a solid foundation for the implementation of the subsequent allocation module and planning module.

[0088] Allocation module

[0089] The allocation module is one of the core components of the police unmanned aerial vehicle management and control system of the present application, and is mainly used for optimizing the allocation of unmanned aerial vehicle task resources when receiving multiple police task. The module generates the optimal task allocation scheme by comprehensively analyzing the environmental data provided by the perception module, the priority of the police task and the real-time position of the unmanned aerial vehicle. The implementation of the allocation module can ensure the accuracy of task allocation, thereby maximizing the use of unmanned aerial vehicle resources and improving the efficiency of emergency disposal.

[0090] It should be noted that the allocation module uses an optimization algorithm to realize dynamic adjustment of the task allocation scheme. In the present application, the allocation module is particularly suitable for handling the complex scenario of multiple unmanned aerial vehicles responding to multiple emergency police tasks at the same time, and its optimization objectives include minimizing the task response time and minimizing the task execution cost.

[0091] In this embodiment, the allocation module realizes task optimization allocation based on an improved Hungarian algorithm. The Hungarian algorithm is a classic method for efficiently solving task allocation problems, which can find the allocation scheme with the minimum total cost between tasks and resources. The present application improves the algorithm to make it dynamically adapt to environmental changes and the needs of multiple task priorities.

[0092] Specifically, the allocation module first constructs a cost matrix C of task allocation, which is defined as follows:

[0093] C ij = α · d ij + β · E ij + γ · P j

[0094] Wherein:

[0095] C ij represents the comprehensive cost of unmanned aerial vehicle i executing task j;

[0096] d ij represents the Euclidean distance between unmanned aerial vehicle i and task point j, defined as:

[0097]

[0098] Where (x i , y i , z i ) and (x jy j z j are three-dimensional coordinates of the UAV and the task point, respectively;

[0099] E ij represents an environmental impact factor, including wind speed, obstacle density, etc., which is calculated based on data provided by the perception module;

[0100] P j represents the priority of the task, and the higher the priority, the greater the weight of the task;

[0101] α, β, γ are weight coefficients for balancing the influence of the above factors.

[0102] It can be understood that the above cost matrix is dynamically updated, and the allocation module can adjust the task allocation scheme in real time as the environmental data provided by the perception module changes.

[0103] In one possible implementation, the allocation module first pre-processes the cost matrix to ensure that each task can obtain at least one response from a UAV. For example, for an invalid task allocation path (e.g., a path blocked by an obstacle and thus cannot be completed), the corresponding cost value can be set to infinity to avoid the path being selected.

[0104] It should be noted that the improved Hungarian algorithm realizes task allocation through the following steps:

[0105] Initial allocation: allocate the lowest cost path between each UAV and the task point to each UAV;

[0106] Row and column adjustment: adjust the values in the matrix to reduce allocation conflicts between tasks;

[0107] Optimization iteration: recalculate the allocation scheme according to the remaining tasks and unallocated UAVs until all tasks are completed.

[0108] In some embodiments, to improve the response speed of allocation, the allocation module can also use parallel computing technology to accelerate the Hungarian algorithm. Specifically, the grouping of task points can be processed in parallel, and after each group is allocated, global optimization integration is performed.

[0109] As an option, to deal with the situation where the priority of the task changes dynamically in an emergency, the allocation module also supports dynamic adjustment of the priority factor. For example, when the urgency of a certain task suddenly rises, the system will automatically increase its priority weight P j , so as to allocate resources preferentially.

[0110] It is particularly noted that the allocation module is not only suitable for multi- UAV multi-task allocation scenarios, but also can flexibly cope with special cases of a single UAV responding to multiple police situations. In this case, the system will generate an execution queue in order according to the priority of the task, ensuring that high-priority tasks are responded to first.

[0111] In a possible extended implementation, the allocation module can also combine machine learning techniques to further optimize the effect of task allocation. For example, by training a prediction model with historical police data, the environmental complexity of the task point can be estimated, and the environmental impact factor E ij can be more accurately assigned. This extended implementation is particularly suitable for areas with strong regularity of police distribution, such as specific business districts or scenic spots.

[0112] It can be understood that the design of the allocation module aims to form an efficient synergy with other modules. For example, through linkage with the planning module, the allocation module can further optimize the allocation scheme according to the flight path cost feedback from the planning module; through interaction with the tracking module, the allocation module can adjust the task allocation result according to the dynamic position of the target.

[0113] In summary, the allocation module of the present application realizes the efficiency and flexibility of UAV task allocation in emergency situations through improved Hungarian algorithm, dynamic weight adjustment and deep fusion of environmental data.

[0114] Planning module

[0115] The planning module is an important part of the police UAV management and control system of the present application, and its core function is to generate the optimal flight path for the UAV according to the task results provided by the allocation module and the three-dimensional environmental model generated by the perception module. The path generation of the planning module needs to meet multiple optimization objectives, including avoiding obstacles, reducing wind speed impact, shortening flight time, etc., to ensure that the UAV can efficiently and stably execute the task.

[0116] It should be noted that the design of the planning module is particularly suitable for complex dynamic environments and can dynamically adjust the path to respond to real-time changes. Through the path generation method based on particle swarm optimization (PSO) algorithm, the planning module can effectively balance the safety and execution efficiency of the path.

[0117] In this embodiment, the planning module realizes three-dimensional path planning for the UAV based on the particle swarm optimization algorithm. Particle swarm optimization is a commonly used global optimization method that can find the optimal solution to a problem through iterative updates of particles. In this embodiment, the path nodes are represented as the positions of particles, and the optimal path of the UAV is generated through the search process of the particle swarm algorithm.

[0118] Specifically, the optimization objective function of path planning can be represented as:

[0119] f(X) = a*T(X) + b*O(X) + g*D(X)

[0120] where:

[0121] T(X): total flight time of the path, defined as:

[0122]

[0123] where ||X i+1 -X i || represents the distance between adjacent path nodes, v drone represents the flight speed of the UAV.

[0124] O(X): obstacle risk on the path, calculated as:

[0125]

[0126] where W wind (X i ) is the wind speed influence at path point X i , represents the density of obstacles at the path point, and η is an adjustment coefficient.

[0127] D(X): distance between the end point of the path and the target point, defined as:

[0128] D(X) = ||X N -X target ||

[0129] where X N is the last node of the path, and X target is the coordinate of the target point.

[0130] As an option, the weight coefficients a, b, g of the optimization objective function can be adjusted according to the task requirements to prioritize certain constraint conditions. For example, in an environment with high wind speed, the value of b can be increased to pay more attention to the avoidance of wind speed influence.

[0131] In one possible implementation, the search process of the particle swarm optimization algorithm includes the following steps:

[0132] Initialize the particle swarm: initialize the position of each particle as a random path from the current coordinates of the UAV to the target point, and initialize the speed as a random vector.

[0133] Particle update rule: the position and speed of each particle are updated according to the following formula:

[0134] V k+1 = w*V k + c1*r1*(P k -Xk )+c2·r2·(G k -X k )

[0135] X k+1 =X k +V k+1

[0136] wherein:

[0137] V k and X k are the velocity and position of particle k, respectively;

[0138] P k denotes the historical optimal position of particle k;

[0139] G k denotes the global optimal position;

[0140] w, c1, c2 are inertia coefficients and learning factors, used to balance global search and local search;

[0141] r1, r2 are random numbers, used to introduce randomness in search.

[0142] Convergence criterion: the algorithm iterates until one of the following convergence conditions is met:

[0143] the path cost values of all particles tend to be stable;

[0144] the maximum number of iterations is reached.

[0145] It should be noted that the particle swarm optimization algorithm will evaluate the feasibility of the particle path in each iteration. For example, for path nodes that cross obstacles, the obstacle distribution data provided by the perception module will be used to modify the path to ensure its safety.

[0146] In some embodiments, to improve the real-time performance of path planning, the planning module can combine parallel computing technology to accelerate the particle swarm optimization algorithm. Specifically, the particle swarm can be grouped and evaluated in parallel, and the global optimal solution of each group of particles is used to update the global optimization result after integration. This approach is particularly suitable for the rapid response requirements of path planning in high dynamic environments.

[0147] In an extended scheme, the planning module can also introduce a multi-objective optimization mechanism to balance the resource allocation of path planning among multiple tasks. For example, when multiple drones need to complete different tasks simultaneously, the system can generate more optimal paths for high-priority tasks while appropriately compromising the optimization of low-priority task paths.

[0148] It can be understood that the implementation of the planning module needs to be closely coordinated with the perception module and the allocation module. The environmental data provided by the perception module is the basis for path planning, and the task allocation result of the allocation module directly determines the target position of path planning. The optimal path generated by the planning module is passed to the control module to ensure that the UAV flies accurately according to the planned path.

[0149] The planning module of the present application realizes the efficiency and dynamic adaptability of path generation through the three-dimensional path planning method based on the particle swarm optimization algorithm.

[0150] Control module

[0151] The control module is a key functional unit in the police UAV management and control system of the present application, and its main function is to dynamically adjust the flight attitude and height of the UAV according to the path generated by the planning module, to ensure that the UAV flies stably along the optimal path and effectively completes the task. This module realizes real-time response through adaptive control technology, and can maintain the stability and path following performance of the UAV in complex and dynamic environments.

[0152] It should be noted that the control module plays a role of connecting the upstream and downstream in the whole system, its input includes the environmental data provided by the perception module and the path information output by the planning module, and its output directly affects the execution effect of the tracking module and the feedback module. Through real-time adjustment of attitude and flight parameters, the control module ensures the flight safety of the UAV and the accuracy of task execution.

[0153] In this embodiment, the control module realizes real-time adjustment of flight attitude based on adaptive control theory. The flight attitude of the UAV includes pitch angle, yaw angle and roll angle, and the dynamic adjustment of these parameters is adjusted by control input. The implementation steps and principles of the control module are as follows:

[0154] Firstly, the flight state of the UAV is described by the following state equation:

[0155]

[0156] Where:

[0157] x(t) represents the state vector of the UAV, including position, velocity and attitude angle;

[0158] u(t) is the control input vector, including thrust, direction adjustment, etc.

[0159] d(t) is the external disturbance, such as wind speed influence;

[0160] A and B are system parameter matrices, which describe the dynamic model of the UAV and the influence of control input on the state, respectively.

[0161] Specifically, the control module performs stability analysis on the flight attitude by constructing a Lyapunov function V(x) to ensure the smoothness of the flight process and the rapidity of the control response. The Lyapunov function is defined as:

[0162]

[0163] wherein:

[0164] P is a symmetric positive definite matrix;

[0165] x T represents the transpose of the state vector.

[0166] As an option, to ensure the stability of the system, the control module designs the control input u(t) such that i.e.:

[0167]

[0168] By solving the above condition, the calculation formula of the control input u(t) can be expressed as:

[0169] u(t) = -Kx(t)

[0170] wherein K is the feedback gain matrix, the value of which is obtained by solving the Lyapunov equation.

[0171] In a possible implementation, to further improve the robustness of the system, the control module dynamically adjusts the feedback gain K to adapt to different environmental disturbances. For example, when the perception module detects an increase in wind speed, the control module can appropriately increase the gain value to enhance the suppression ability of external disturbances.

[0172] In some embodiments, the control module can also combine a path following control algorithm to achieve accurate tracking of the planned path. The goal of path following control is to minimize the deviation between the current position of the UAV and the reference point of the path, and the specific control law is:

[0173] u p (t) = K p ·e(t)

[0174] wherein:

[0175] u p (t) is the path following control input;

[0176] K p is the path following gain;

[0177] e(t) is the error vector between the current position and the reference point.

[0178] It is worth noting that the path following control and the attitude adjustment control are executed in parallel, and the two work together to ensure that the UAV can maintain stable attitude and fly efficiently along the planned path.

[0179] As an option, the execution process of the control module can introduce a predictive control mechanism to anticipate potential changes in the environment. For example, by predicting the wind speed data provided by the perception module, the control module can adjust the flight parameters in advance before the wind speed changes, thereby further improving the flight stability.

[0180] In an extended scheme, the control module can also incorporate a self-learning algorithm, such as a neural network-based adaptive control method. By learning the disturbance characteristics in the environment and the dynamic response of the UAV online, the control module can gradually optimize the control input to adapt to more complex environmental conditions. This extended scheme is particularly suitable for high-dynamic scenarios with frequent emergencies, such as mass events in urban dense areas.

[0181] It can be understood that the performance of the control module has a direct impact on the task execution effect of the entire system. Through close cooperation with the planning module and the perception module, the control module ensures that the UAV can quickly respond to environmental changes and complete the execution of the planned path with high precision.

[0182] Tracking module

[0183] The tracking module is an important part of the police UAV management and control system of the present application, and its main function is to track dynamic targets in real time and predict the motion trajectory of the target, providing accurate target position and motion information for the control module and the feedback module. The design of the tracking module is particularly suitable for high-dynamic target scenarios in complex environments, such as the flow of crowds or the movement of suspicious vehicles in emergencies. Through advanced prediction and correction techniques, the tracking module can maintain high-efficiency tracking ability in the case of variable target behavior.

[0184] It is worth noting that the tracking module is highly cooperative with the perception module and the control module. The perception module provides real-time observation data of the target, and the control module adjusts the flight path and attitude of the UAV based on the predicted trajectory output by the tracking module, ensuring the continuous tracking of the UAV to the target.

[0185] In this embodiment, the tracking module realizes real-time tracking and trajectory prediction of dynamic targets based on the Extended Kalman Filter (EKF) algorithm. EKF is a commonly used state estimation method for nonlinear systems, which can efficiently predict and correct the position of the target in a noisy environment.

[0186] Specifically, the state model of target motion can be described as:

[0187] x k= Fx k-1 + Gu k-1 + w k

[0188] wherein:

[0189] x k represents the state vector of the target at time k, including position, velocity and acceleration;

[0190] F is a state transition matrix, describing the change rule of the target state over time;

[0191] G is a control input matrix, describing the influence of the control quantity on the state;

[0192] u k-1 is the control input;

[0193] w k is the process noise, subject to a zero-mean Gaussian distribution.

[0194] The observation model of the target can be described as:

[0195] z k = Hx k + v k

[0196] wherein:

[0197] z k is the observation value, including the real-time position of the target;

[0198] H is an observation matrix, describing the relationship between the state vector and the observation value;

[0199] v k is the measurement noise, subject to a zero-mean Gaussian distribution.

[0200] It should be noted that the EKF algorithm is divided into two stages of prediction and correction, and the specific implementation is as follows:

[0201] In the prediction stage:

[0202]

[0203] wherein:

[0204] is the predicted state;

[0205] is the predicted covariance matrix;

[0206] Q is the process noise covariance matrix.

[0207] In the correction stage:

[0208]

[0209] wherein:

[0210] K k is the Kalman gain, used to weigh the error between the prediction and the observation;

[0211] is the updated state estimate;

[0212] R is the observation noise covariance matrix.

[0213] In one possible implementation, the tracking module can optimize the state transition matrix F in combination with the behavior model of the dynamic target. For example, for a uniformly moving target, the state transition matrix can be represented as:

[0214]

[0215] where Δt is the time step.

[0216] For non-uniformly moving targets, an acceleration term can be introduced, and the state vector can be expanded to accommodate a more complex motion model.

[0217] In some embodiments, the tracking module can also optimize the observation matrix H in combination with a machine learning model. For example, by analyzing the motion patterns of the target through deep learning methods, the parameters of H can be dynamically adjusted to improve the accuracy of the observation model.

[0218] As an alternative, to cope with complex scenarios of multi-target tracking, the tracking module can extend its functionality through a Multiple Target Tracking (MTT) algorithm. The specific implementation can use the Joint Probabilistic Data Association (JPDA) method to associate the observation values of multiple targets with state predictions, ensuring the tracking accuracy among multiple targets.

[0219] It should be noted that the output of the tracking module includes the predicted position of the target and the confidence of the current state, which can be used by the control module to adjust the flight attitude of the UAV, and by the feedback module to generate more accurate command decision information.

[0220] In an extended scheme, the tracking module can also combine visual perception technology to capture target images in real time through the camera on the UAV, and identify and locate the target through target feature point matching algorithms. This extended scheme is particularly suitable for identity confirmation tasks of dynamic targets, such as locking specific targets in a crowd.

[0221] It can be understood that the performance of the tracking module directly affects the response efficiency of the system to the target and the accuracy of task completion. Through high coordination with the perception module and the control module, the tracking module can provide stable and efficient target tracking services for the entire system.

[0222] Feedback module

[0223] The feedback module is an important part of the police unmanned aerial vehicle management and control system of the present application, and its main function is to real-time return the task execution situation, target tracking result and environmental data of the unmanned aerial vehicle to the command center, providing data support for command decision and task dynamic adjustment. Through the feedback module, the system can realize comprehensive monitoring of the unmanned aerial vehicle task execution state, and simultaneously conduct real-time information transmission to the target area.

[0224] It should be noted that the feedback module not only undertakes the responsibility of data return, but also directly intervenes in the target area through audio and video transmission and remote shouting function, especially suitable for rapid information communication in emergency disposal. This module works in coordination with the control module and the tracking module, and its output directly affects the decision efficiency of the command center.

[0225] In this embodiment, the feedback module mainly includes a real-time data transmission unit, a video return unit and a remote shouting unit, and the specific implementation of each unit is as follows:

[0226] In the real-time data transmission unit, the feedback module uses wireless communication technology to return the state data and environmental perception data of the unmanned aerial vehicle to the command center. Exemplarily, the content of data return includes the current coordinates, flight speed, attitude angle of the unmanned aerial vehicle and the real-time position of the target.

[0227] As an option, in order to ensure the stability of data transmission, this embodiment adopts a dynamic bandwidth allocation method, which prioritizes the transmission of high-priority data according to the importance and real-time requirements of data. For example, the target position and video stream data are given higher transmission priority, while non-real-time data (such as task completion records) can be delayed in transmission.

[0228] It should be noted that this dynamic bandwidth allocation mechanism can prioritize the transmission of critical data in the case of limited network resources, thereby improving the overall response efficiency of the system.

[0229] In the video return unit, this embodiment collects the live pictures through a high-definition camera and uses an efficient compression algorithm to return the video stream to the command center in real time. As a possible implementation, the resolution and frame rate of video return can be dynamically adjusted according to network conditions to ensure picture clarity while reducing data transmission delay.

[0230] In some embodiments, to enhance the transmission reliability of the video data, the video backhaul unit further integrates a forward error correction (FEC) technology, which can automatically correct errors occurring in the transmission process to a certain extent by embedding redundant information in the data stream.

[0231] In the remote shouting unit, the embodiment can deliver the instructions of the command center or make a warning shout in real time by audio intervention on the target area through the loudspeaker carried by the unmanned aerial vehicle. Specifically, the remote shouting unit includes a speech synthesis module and an audio amplification module, wherein the speech synthesis module converts the text instructions into audio signals.

[0232] As an option, the speech synthesis module adopts a deep learning-based speech synthesis technology, which can realize flexible output of multiple languages and tones. For example, the tone and speed of the synthesized speech can be adjusted according to different scene requirements, such as selecting high-speed and high-volume speech output during emergency evacuation.

[0233] It should be noted that, to avoid audio interference in the target area, the remote shouting unit can also dynamically adjust the output power of the loudspeaker according to the environmental noise level.

[0234] In a possible implementation, the feedback module works jointly with the tracking module, which can dynamically adjust the directional output of the loudspeaker according to the real-time position of the target, ensuring that the shouting content can be accurately delivered to the target area.

[0235] In some embodiments, to enhance the functional expandability of the feedback module, the system can integrate speech recognition and text analysis technologies. For example, the voice instructions of the command center can be converted into text through the speech recognition module and directly transmitted to the shouting unit of the unmanned aerial vehicle, realizing efficient delivery of voice commands.

[0236] It can be understood that the core value of the feedback module lies in its comprehensive presentation of the task state of the unmanned aerial vehicle and the direct intervention capability of the target area.

[0237] In a preferred embodiment of the present application, to adapt to the dynamic changes of the environment in the process of handling emergencies, the system can re-trigger the allocation module and the planning module according to the real-time sensed environmental data, to dynamically update the task allocation result and the flight path. Through this mechanism, the system can effectively respond to the task adjustment requirements caused by changes in target position, environmental interference (such as sudden obstacles or changes in wind speed), etc., thereby ensuring the efficient operation of the unmanned aerial vehicle.

[0238] Specifically, the perception module continuously collects environmental data through various sensors (such as laser radars, wind speed sensors, etc.), and triggers an event update signal when a significant change in environmental data is detected.

[0239] For example:

[0240] The target position deviates;

[0241] A new obstacle suddenly appears, affecting the feasibility of the original path;

[0242] The wind speed direction or intensity changes, affecting the flight stability of the UAV.

[0243] These trigger conditions can be implemented through a threshold judgment mechanism. For example, the deviation threshold of the target position can be defined as:

[0244]

[0245] When Δd>d threshold , the system determines that the target position change reaches the adjustment threshold, triggering dynamic update.

[0246] After the trigger signal reaches the allocation module, the system will recalculate the task allocation scheme. The allocation module is based on the improved Hungarian algorithm, combined with real-time environmental data to update the task cost matrix C ij .

[0247] The dynamic adjustment of the allocation module ensures that the UAV can respond to the most significant changes or the most urgent tasks first. For example, when the target priority changes, the allocation module will appropriately increase the weight of γ, and allocate resources to high-priority tasks first.

[0248] After the allocation result is updated, the planning module will generate the optimal flight path of the UAV according to the new task allocation result and environmental data. The planning module uses the particle swarm optimization algorithm to quickly iterate under dynamic constraints, ensuring the real-time and safety of path planning.

[0249] Specifically, when the perception module detects an update in obstacle position, the planning module will dynamically adjust the feasibility of the path node. For example, when a path node X i is determined to be impassable, the particle swarm algorithm will automatically exclude this node and recalculate the global optimal path.

[0250] As an option, the path update of the planning module can use a distributed computing method to improve the response speed of the algorithm in complex scenarios. For multi-UAV scenarios, each UAV can independently plan its path while maintaining communication with other UAVs to avoid path conflicts.

[0251] It is to be noted that the dynamic adjustment mechanism is the result of the cooperation of multiple modules. The perception module is responsible for providing trigger signals for environmental changes, and the allocation module and the planning module complete the dynamic update of task allocation and path planning respectively. Through this modular cooperative mode, the system can quickly adapt to dynamic changes in complex scenarios, ensuring the continuity and efficiency of the task.

[0252] The dynamic adjustment mechanism of the present embodiment is particularly suitable for the following scenarios:

[0253] Sudden target position changes, such as large-scale migration of crowds at the scene of an event;

[0254] Drastic changes in environmental conditions, such as sudden changes in wind speed caused by severe weather;

[0255] New obstacles appear on the scene, such as vehicle blockage or building collapse.

[0256] This dynamic adjustment mechanism significantly improves the adaptability of the system in complex environments, enabling the UAV to efficiently and accurately perform tasks in dynamic scenarios while ensuring the optimality of resource allocation and the real-time nature of path planning.

[0257] In summary, the present invention achieves rapid response, accurate disposal, and real-time feedback to emergency events through the cooperative work of various modules. The perception module collects environmental data in real time and generates a three-dimensional environmental model, the allocation module implements UAV task allocation based on dynamic optimization algorithms, the planning module generates an obstacle-avoiding and time-optimal flight path, the control module dynamically adjusts the UAV attitude to ensure flight stability, the tracking module achieves accurate tracking of dynamic targets through extended Kalman filtering, and the feedback module returns the UAV task execution and on-site audio and video data to the command center. The present invention is particularly suitable for complex and dynamic environments and can efficiently and flexibly complete police tasks.

[0258] Please refer to the attached Figure 2 Correspondingly, the present invention provides a police UAV management and control method for handling emergency events, which completes efficient response, task execution, and command support for emergency events through multi-module cooperation in a systematic process. The specific embodiments of the method of the present invention are described below in conjunction with the system workflow.

[0259] S1: Collect environmental data and generate a three-dimensional environmental model

[0260] The present method first collects real-time environmental data at the scene of an emergency event through a perception module, including obstacle distribution, wind speed dynamics, target position, etc. The perception module integrates the collected data through devices such as laser radar, ultrasonic sensor, and wind speed sensor to generate a three-dimensional environmental model.

[0261] During the collection process, the perception module can dynamically adapt to environmental changes, updating obstacles and target locations in real time. The generated three-dimensional environmental model provides basic data support for subsequent modules.

[0262] S2: Optimize UAV task allocation scheme

[0263] According to the environmental data generated by the perception module and the priority of the emergency task, the allocation module uses a dynamic optimization algorithm to calculate the task allocation scheme. The allocation module comprehensively analyzes the resource allocation between multiple UAVs and multiple task points, and prioritizes high-priority tasks.

[0264] Through the optimization calculation of the allocation module, the UAV can quickly adjust the allocation scheme according to the real-time task demand, laying the foundation for efficient disposal of emergencies.

[0265] S3: Dynamically plan the optimal flight path of the UAV

[0266] Based on the task allocation results output by the allocation module and the three-dimensional environmental model provided by the perception module, the planning module generates an optimal flight path for each UAV that avoids obstacles and reduces wind speed interference.

[0267] The planning module ensures that the UAV reaches the target location in the shortest time and safest path through dynamic path planning methods. During the planning process, the system can update the path in real time according to the changes in environmental data, adapting to complex and dynamic environmental scenarios.

[0268] S4: Dynamically adjust the attitude and flight height of the UAV

[0269] According to the planned path, the control module dynamically adjusts the flight attitude and height of the UAV to ensure accurate execution of the path flight task. The control module monitors the flight state of the UAV in real time and adjusts the pitch angle, yaw angle, and thrust parameters to maintain flight stability.

[0270] The control module can also adjust the attitude during flight according to changes in environmental conditions to avoid flight deviations caused by wind speed interference or obstacles.

[0271] S5: Real-time tracking of target location and prediction of target motion trajectory

[0272] The tracking module uses a multi-sensor fusion method to track the target location in real time and predicts its future trajectory based on the motion law of the target. The prediction data of the tracking module provides key support for UAV path adjustment.

[0273] When the target location changes significantly, the tracking module can quickly correct the target positioning information to ensure that the system always maintains high-precision tracking capability for the target.

[0274] S6: The execution and feedback data are returned to the command center

[0275] The feedback module returns the task execution of the unmanned aerial vehicle, target tracking results and real-time pictures on the scene to the command center, providing comprehensive decision support for the commander. The returned data includes the flight state of the unmanned aerial vehicle, target positioning information and video pictures.

[0276] In addition, the feedback module also supports the shouting intervention function for the target area, and transmits instructions or warning information to the target area through the loudspeaker carried by the unmanned aerial vehicle, further improving the command efficiency.

[0277] Through the above steps, the method of the present application can realize the whole process management and control of the emergency. The steps are connected with each other, forming a closed loop real-time response mechanism, so as to ensure the efficient operation and task completion of the system in a complex dynamic environment.

[0278] Although the embodiments of the present application have been shown and described, it can be understood by those skilled in the art that various changes, modifications, replacements and variations can be made to the embodiments without departing from the principles and spirits of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A police drone control system for responding to emergencies, characterized in that, include: The perception module is used to collect environmental data in real time, including obstacle distribution, wind speed, target dynamic position, and generate a three-dimensional environment model. The allocation module is used to optimize the drone task allocation scheme based on the environmental data and alarm task priorities of the perception module; The planning module is used to plan the optimal flight path of the UAV based on the task allocation result of the allocation module and the three-dimensional environment model. The path planning avoids obstacles and minimizes flight time. The control module is used to dynamically adjust the attitude and flight altitude of the UAV according to the flight path of the planning module; The tracking module is used to track the target's location in real time and predict the trajectory of dynamic targets; The feedback module is used to transmit the drone's mission execution status, target tracking results, and environmental data back to the command center; The perception module includes a lidar, an ultrasonic sensor, and a wind speed sensor. The lidar is used to collect point cloud data of obstacle distribution, the ultrasonic sensor is used to detect nearby obstacles, and the wind speed sensor is used to generate a wind speed distribution model. The allocation module optimizes task allocation based on an improved Hungarian algorithm. The optimization objective of the task allocation is to minimize the overall cost of the UAV performing the task. The overall cost is composed of the distance from the UAV to the target, the task priority, and the influence weights of wind speed and obstacles. The planning module uses a particle swarm optimization algorithm to dynamically plan the flight path, and the optimization objective of the flight path is: Minimize drone flight time; Avoid obstacles in the flight path; Reduce the interference of wind speed on drone flight.

2. The police drone control system for handling emergencies according to claim 1, characterized in that, The particle swarm optimization algorithm implements path planning through the following steps: Initialize path nodes based on a 3D environment model; Based on the drone's current location, the target location, and the distribution of obstacles, the speed and position of the path nodes are iteratively updated. The optimal path is selected by comparing the historical best position of the path node with the global best position.

3. The police drone control system for handling emergencies according to claim 1, characterized in that, The control module dynamically adjusts the UAV's attitude, including pitch angle, yaw angle, and flight altitude, based on adaptive control theory. Specifically, it constructs a Lyapunov function by calculating the UAV's attitude state vector in real time and adjusts the UAV's control input to ensure the stability of its flight attitude.

4. The police drone control system for handling emergencies according to claim 1, characterized in that, The tracking module uses an extended Kalman filter algorithm to predict the target trajectory. The extended Kalman filter algorithm includes the following steps: Based on the target's motion state and the UAV's flight state, predict the target's position at the next moment; Correct the predicted position of the target based on target observation data collected by sensors; Update the target state and its covariance matrix.

5. A police drone control system for handling emergencies according to claim 1, characterized in that, The feedback module includes a real-time video transmission unit and a remote broadcasting unit. The video transmission unit is used to transmit live footage to the command center, and the remote broadcasting unit is used to transmit instructions or warnings to the target area via drone.

6. A police drone control system for handling emergencies according to claim 1, characterized in that, The system re-triggers the allocation and planning modules based on real-time environmental change data to update the task allocation results and flight paths.

7. A method for managing police drones in response to emergencies, based on the system described in any one of claims 1-6, characterized in that, Includes the following steps: Collect environmental data and generate a 3D environment model; Optimize the drone mission allocation scheme based on environmental data and police task priorities; Based on the task allocation results and the three-dimensional environment model, the optimal flight path of the UAV is dynamically planned; Based on the planned path, the attitude and flight altitude of the drone are dynamically adjusted to ensure that the drone flies along the planned path; Real-time tracking of target location and prediction of target trajectory; The drone's performance, target tracking results, and real-time video feed are transmitted back to the command center via a feedback module.

Images