Unmanned aerial vehicle real-time path planning system and method based on dynamic weight distribution and multi-source data fusion

Abstract

The invention relates to the technical field of unmanned aerial vehicle flight path planning, in particular to an unmanned aerial vehicle real-time path planning system and method based on dynamic weight distribution and multi-source data fusion. The system comprises a multi-source data fusion module, an integrated laser radar, a millimeter wave radar, a visual sensor and a Beidou positioning unit. The dynamic weight distribution module dynamically adjusts the weight coefficient of each sensor according to the environmental complexity, the threat level and the state of the unmanned aerial vehicle by adopting a mixed decision-making mechanism combining fuzzy logic and reinforcement learning; an improved RRT * algorithm and a Markov decision process are built in the real-time path planning module, and a global optimal path and a local obstacle avoidance track are generated by adopting a layered planning architecture; the unmanned aerial vehicle cooperative control module comprises a dual-redundancy flight control system and a dynamic obstacle avoidance unit; and the communication relay module supports 5G and low-orbit satellite dual-mode communication, updates an environment cognitive model of each unmanned aerial vehicle through federated learning, and realizes multi-source fusion real-time path planning based on dynamic weight distribution and the unmanned aerial vehicles.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle (UAV) flight path planning, and in particular to a real-time UAV path planning system and method based on dynamic weight allocation and multi-source data fusion. Background Art

[0002] Current drone path planning systems face significant bottlenecks: single sensors are susceptible to failure in the presence of dynamic obstacles, inclement weather, or electromagnetic interference, resulting in incomplete environmental modeling. Traditional algorithms struggle to quickly generate optimal paths that balance safety and energy efficiency in the face of dynamic threats (such as moving obstacles). The lack of a dynamic task allocation mechanism makes resource conflicts and response delays more likely. Long-distance communications are susceptible to interference, potentially exposing sensitive data to theft.

[0003] At the sensor level, fixed-weight fusion strategies are unable to adapt to sudden environmental changes (such as dense fog that weakens visual data), resulting in a decrease in perception confidence. At the planning algorithm level, the standard RRT* algorithm converges slowly in dynamic scenarios, and the Monte Carlo search lacks real-time performance (>500ms), making it difficult to respond to sudden threats. At the collaborative control level, centralized decision-making is susceptible to communication delays, and distributed strategies lack Nash equilibrium optimization, resulting in suboptimal multi-machine task allocation.

[0004] Existing systems mostly rely on offline environmental models and are unable to update dynamic obstacle information (such as vehicle trajectory prediction) in real time, resulting in a mismatch between the planned path and the actual environment. Summary of the Invention

[0005] To solve the above technical problems, the present invention provides a real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion, comprising:

[0006] Multi-source data acquisition and fusion module, used to obtain multi-source data from drone sensors, terrain data of the drone flight area, and drone flight condition data, and pre-process the multi-source data from drone sensors to construct a multi-source environmental perception data matrix;

[0007] The dynamic weight allocation module is used to generate a weight coefficient control sequence based on the multi-source fusion perception data matrix, the terrain data of the UAV flight area, and the UAV flight condition data, and dynamically adjust the weight coefficients of the UAV sensors;

[0008] The real-time path planning module is used to generate the UAV's global path and local obstacle avoidance trajectory through a hierarchical path planning model based on the multi-source environmental perception data matrix and weight coefficient control sequence, and dynamically optimize and smooth the global path to obtain the UAV's track data;

[0009] The UAV collaborative control module is used to perform real-time flight control and flight command calculation for a cluster of multiple UAVs based on the UAV track data output by the real-time path planning module;

[0010] The communication relay module is used to establish a dual-mode communication link with low-orbit satellites through the 5G network, transmitting the UAV flight status information to the implementation path planning module and the main control and collaborative learning module;

[0011] The master control and collaborative learning modules are used to build a three-dimensional digital twin model based on the terrain data of the drone flight area and the drone flight status information, and perform online updates and global optimization of each module based on the federated learning algorithm.

[0012] Furthermore, the multi-source data acquisition and fusion module includes a lidar unit, a millimeter-wave radar unit, a visual sensor unit and a Beidou satellite positioning unit; the multi-source data of the drone sensor includes lidar point cloud data, millimeter-wave radar trajectory prediction data, image recognition data, and Beidou satellite positioning data;

[0013] The preprocessing includes calibrating heterogeneous data of multi-source data of drone sensors through time synchronization algorithm and spatial registration theory, establishing a hidden Markov model to perform confidence assessment on the health status of drone sensors, eliminating data conflicts among the multi-source data of drone sensors according to the Dempster-Shafer evidence theory, and constructing an ARIMA-LSTM hybrid model to predict and compensate for the communication delay of each data among the multi-source data of drone sensors.

[0014] Furthermore, the dynamic weight allocation module includes a terrain analysis unit, a threat identification unit, a confidence adjustment unit and a Nash equilibrium collaborative decision unit;

[0015] The terrain analysis unit is used to calculate the terrain complexity of the UAV flight area using a fractal geometry algorithm based on the terrain data of the UAV flight area;

[0016] The threat identification unit is used to obtain the predicted threat level of dynamic obstacles through the spatiotemporal graph convolution network according to the terrain complexity of the drone's flight area;

[0017] The confidence adjustment unit is used to predict the threat level of the drone's flight area based on the terrain complexity and dynamic obstacles, build an adaptive adjustment model based on the Lyapunov index, and optimize the drone sensor weight distribution in real time;

[0018] The Nash equilibrium collaborative decision unit establishes a game payoff matrix among multiple unmanned aerial vehicles according to the Nash equilibrium theory, according to the terrain complexity of the unmanned aerial vehicle flight area and the dynamic obstacle prediction threat level, in combination with a multi-source fusion perception data matrix, and performs collaborative decision on multiple unmanned aerial vehicle tasks;

[0019] The spatio-temporal graph convolution network extracts obstacle motion pattern features based on a hierarchical spatio-temporal convolution structure by constructing a spatio-temporal adjacency matrix, taking laser radar point cloud data as spatial graph node features, and taking millimeter wave radar trajectory prediction data as time graph edge weights, predicts the obstacle motion trajectory within 3 seconds in the future through a time sliding window mechanism, and the dynamic obstacle prediction threat level is divided into five quantitative indexes according to the obtained unmanned aerial vehicle flight trajectory conflict probability in advance: when the collision time is less than 2 seconds and the distance threshold is less than 5 meters, it is an emergency level; when the collision time is less than 5 seconds and the distance threshold is less than 10 meters, it is a high-risk level; when the collision time is less than 10 seconds and the distance threshold is less than 20 meters, it is a warning level, when the collision time is greater than 10 seconds and the distance threshold is greater than 20 meters, it is a safe level and an invalid level;

[0020] The adaptive adjustment model based on Lyapunov index, by constructing an energy function to evaluate the confidence of each unmanned aerial vehicle sensor in real time, normalizes the terrain complexity of the unmanned aerial vehicle flight area, and when the terrain complexity of the unmanned aerial vehicle flight area exceeds 0.7, the weight of the laser radar unit is increased to 0.6-0.8, and when the threat level reaches the high-risk level or the emergency level, the weight of the millimeter wave radar unit is dynamically increased to 0.3-0.5;

[0021] The game payoff matrix includes flight task completion degree, flight energy consumption efficiency and flight safety distance, the game payoff matrix is optimized by iteration strategy, the Pareto optimal allocation scheme is obtained, and the weight coefficient control sequence is generated;

[0022] The finally generated terrain complexity of the unmanned aerial vehicle flight area and the dynamic obstacle prediction threat level are fed back to the real-time path planning module and the master control and collaborative learning module in real time through the 5G network.

[0023] Further, the hierarchical path planning model includes a global topology layer, a local obstacle avoidance layer and a trajectory optimization layer.

[0024] The global topology layer constructs a navigation topology map based on the lidar point cloud data and calculates the initial waypoint sequence of the drone using the improved Dijkstra algorithm. The path sampling density of the initial waypoint sequence of the drone is dynamically adjusted according to the terrain complexity of the drone's flight area. When the terrain complexity of the drone's flight area is greater than 0.6, the path sampling density is reduced to 0.5 meters. When there are dynamic obstacles or highly complex terrain in the flight area, the global topology layer further uses the improved RRT* algorithm to integrate environmental characteristics and historical path information to dynamically adjust the sampling strategy and generate the initial waypoint sequence.

[0025] The local obstacle avoidance layer constructs an evaluation fusion function by dynamically predicting the threat level of obstacles and replanning the initial waypoint sequence. When a high-risk threat is detected, the Monte Carlo tree search algorithm is activated, and an emergency avoidance path is generated within 200ms based on a node expansion strategy constrained by the UAV's dynamics. The evaluation fusion function includes the threat field strength value, energy consumption rate, and heading deviation.

[0026] The trajectory optimization layer is based on the dynamic adjustment mechanism of the node vector of the adaptive B-spline curve. It optimizes the curve order and control point density in real time according to the UAV flight status information fed back by the communication relay module. The continuity of the curve order meets the C3 order differentiability requirement. A dynamic correlation model between the UAV lift-to-drag ratio and the UAV power system is established. The UAV power parameter configuration is optimized through the backpropagation algorithm to obtain the final UAV track data, which is transmitted to the UAV collaborative control module in real time.

[0027] Furthermore, the improved RRT* algorithm is used for path planning at the global topology layer. It integrates environmental characteristics and historical path information to dynamically adjust the sampling strategy, adopts a bidirectional expansion mechanism to accelerate the path search process, performs preliminary smoothing optimization on the generated initial trajectory through optimal control theory, and sets an online learning unit to update the heuristic function parameters according to reinforcement learning feedback. The optimization results are used as the output of the global topology layer for further processing by the local obstacle avoidance layer and trajectory optimization layer, and the flight instructions are generated by the UAV collaborative control module.

[0028] Furthermore, the UAV collaborative control module uses a quaternion interpolation algorithm to perform fault-tolerant control on the flight command solution, optimizes the multi-rotor thrust distribution scheme based on the pseudo-inverse method, and the dynamic obstacle avoidance unit completes the obstacle avoidance action within 200ms through the fusion perception of the lidar unit and the visual sensor. The formation maintenance unit uses the virtual structure method to maintain the cluster geometric configuration, and the control instructions are distributed to each UAV node through the communication relay module.

[0029] Furthermore, the main control and collaborative learning module constructs a three-dimensional digital twin model containing the terrain data of the drone flight area and the drone flight status information, updates the drone track data of the real-time path planning module through the transfer learning framework, and integrates the multi-source environmental perception data matrix of the multi-source data acquisition and fusion module to generate a threat situation map.

[0030] Furthermore, the threat situation map is constructed by fusing lidar point cloud data, millimeter-wave radar trajectory prediction data, and visual image recognition data from a multi-source data acquisition and fusion module, using a spatiotemporal graph convolutional network to extract dynamic obstacle motion features. Combined with the terrain data of the drone flight area in the three-dimensional digital twin model, a federated learning algorithm is used to construct a three-dimensional threat field strength distribution model.

[0031] According to the three-dimensional threat field strength distribution model, the drone collision probability is calculated through Monte Carlo risk simulation, and the energy consumption risk is predicted based on the evaluation fusion function of the real-time path planning module. Finally, a multi-dimensional threat situation map including dynamic obstacle hot spots, energy risk contours and communication blind spot markers is obtained. It is superimposed and displayed through the augmented reality navigation interface of the main control and collaborative learning modules, providing a basis for dynamic weight allocation for the real-time path planning module.

[0032] Furthermore, the communication relay module is built based on a deep Q network, improves anti-interference capability through an orthogonal spatiotemporal modulation algorithm, applies game theory to dynamically allocate communication spectrum resources, and constructs a quantum key distribution channel to ensure data transmission security, and uploads the flight status data of the UAV collaborative control module to the main control and collaborative learning modules in real time.

[0033] The present application also provides a real-time path planning method based on dynamic weight allocation and multi-source data fusion of a UAV. The method is implemented based on the above-mentioned real-time path planning system for a UAV based on dynamic weight allocation and multi-source data fusion, and includes the following steps:

[0034] S1. Through multi-source data acquisition, obtain multi-source data from drone sensors, terrain data of the drone's flight area, and drone flight condition data. Preprocess the multi-source data from drone sensors to construct a multi-source environmental perception data matrix.

[0035] S2. Based on the multi-source environmental perception data matrix, terrain complexity and dynamic obstacle threat levels are calculated using fractal geometry algorithms and Lyapunov rule models, respectively. A multi-UAV weight allocation model is constructed using Nash equilibrium theory to derive a multi-UAV task coordination control sequence.

[0036] S3. Build a real-time path planning model to generate UAV trajectory data based on the multi-source environmental perception data matrix and the multi-UAV collaborative control sequence;

[0037] S4. Based on the drone track data obtained in step S3, the drone's flight attitude is controlled using a quaternion interpolation algorithm. The drone's multi-rotor thrust is distributed based on the pseudo-inverse method. The laser radar unit and visual sensor are combined to complete obstacle avoidance within 200ms and output real-time flight parameters.

[0038] S5. Based on real-time flight parameters and a multi-source environmental perception data matrix, a three-dimensional digital twin model containing terrain, obstacle, and meteorological data is constructed. Multi-source information is integrated to generate a dynamic threat situation map, and a federated learning algorithm is used to update the local model of each drone in real time.

[0039] The beneficial effects of this invention include quantifying terrain complexity through fractal geometry, automatically increasing LiDAR weight, and ensuring that point cloud data dominates perception. Obstacle trajectories are predicted using a spatiotemporal graph convolutional network, and weight adjustments are triggered by threat level classification. An ARIMA-LSTM hybrid model compensates for 30ms communication delays, and the Dempster-Shafer theory resolves sensor conflicts.

[0040] An improved Dijkstra algorithm reduces the sampling interval to 0.5 meters in complex environments, generating an initial path in less than 50 milliseconds. High-threat threats trigger a Monte Carlo tree search, which, combined with the drone's dynamic constraints, generates an emergency obstacle avoidance path within 200 milliseconds. Adaptive B-spline curves achieve C3-order smooth trajectory, and a lift-to-drag ratio model optimizes powertrain parameters, reducing energy consumption by 18%.

[0041] The improved RRT* shortens the path search time by 60% through a bidirectional expansion strategy. The online learning unit updates the heuristic function through reinforcement learning feedback to improve adaptability to dynamic environments.

[0042] Construct a three-dimensional benefit matrix of task completion, energy consumption, and safety distance to achieve Pareto optimal task allocation, ensure communication availability through redundant links between 5G and low-orbit satellites, and block data theft through quantum key distribution.

[0043] After each drone's local model is updated, it is encrypted and uploaded to the master control and collaborative learning modules, where it is aggregated to generate a global environmental awareness model, eliminating data silos. By integrating laser point clouds, visual semantic segmentation, and meteorological data, a physical information neural network is used to generate a three-dimensional threat field intensity distribution map, dynamically annotating obstacle hotspots and communication blind spots. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 The figure is a schematic diagram of the workflow of a real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion according to an embodiment of the present invention.

[0045] Figure 2The figure is a flow chart of a method for real-time path planning of a UAV based on dynamic weight allocation and multi-source data fusion according to an embodiment of the present invention.

[0046] Figure 3 This is a schematic diagram of the terminal device structure of a real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion according to an embodiment of the present invention.

[0047] Figure 4 This is a schematic diagram of the computer-readable storage medium structure of a real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion in an embodiment of the present invention.

[0048] In the figure, 200 - terminal device, 210 - memory, 211 - RAM, 212 - cache memory, 213 - ROM, 214 - program / utility, 215 - program module, 220 - processor, 230 - bus, 240 - external device, 250 - I / O interface, 260 - network adapter, 300 - program product. DETAILED DESCRIPTION

[0049] In order to make the objectives, technical solutions, and advantages of the present invention more clearly understood, the present invention is further described in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present invention and are not intended to limit the present invention. That is, the embodiments described herein are only some embodiments of the present invention, not all embodiments. Generally, the components of the embodiments of the present invention described and illustrated in the drawings herein may be arranged and designed in various different configurations.

[0050] Furthermore, the terms "comprises," "comprising," or any other variations thereof are intended to cover a non-exclusive inclusion such that a process, method, article, or machine that comprises a list of elements includes not only those elements but also other elements not expressly listed or inherent to such process, method, article, or machine. In the absence of more limitations, an element defined by the phrase "comprising a..." does not preclude the presence of additional identical elements in the process, method, article, or machine that comprises the element.

[0051] Example 1:

[0052] like Figure 1 As shown, this embodiment 1 provides a real-time path planning system based on dynamic weight allocation and multi-source fusion of UAVs, including:

[0053] Multi-source data acquisition and fusion module, used to obtain multi-source data from drone sensors, terrain data of the drone flight area, and drone flight condition data, and pre-process the multi-source data from drone sensors to construct a multi-source environmental perception data matrix;

[0054] The dynamic weight allocation module is used to generate a weight coefficient control sequence based on the multi-source fusion perception data matrix, the terrain data of the UAV flight area, and the UAV flight condition data, and dynamically adjust the weight coefficients of the UAV sensors;

[0055] The real-time path planning module is used to generate the UAV's global path and local obstacle avoidance trajectory through a hierarchical path planning model based on the multi-source environmental perception data matrix and weight coefficient control sequence, and dynamically optimize and smooth the global path to obtain the UAV's track data;

[0056] The UAV collaborative control module is used to perform real-time flight control and flight command calculation for a cluster of multiple UAVs based on the UAV track data output by the real-time path planning module;

[0057] The communication relay module is used to establish a dual-mode communication link with low-orbit satellites through the 5G network, transmitting the UAV flight status information to the implementation path planning module and the main control and collaborative learning module;

[0058] The master control and collaborative learning modules are used to build a three-dimensional digital twin model based on the terrain data of the drone flight area and the drone flight status information, and perform online updates and global optimization of each module based on the federated learning algorithm.

[0059] Specifically, the multi-source data acquisition and fusion module integrates lidar, millimeter-wave radar, visual sensor and Beidou positioning unit;

[0060] The dynamic weight allocation module adopts a hybrid decision-making mechanism combining fuzzy logic and reinforcement learning to dynamically adjust the weight coefficient of each sensor according to the complexity of the environment, the threat level and the status of the drone;

[0061] The real-time path planning module has a built-in improved RRT* algorithm and Markov decision process, using a hierarchical planning architecture to generate the global optimal path and local obstacle avoidance trajectory;

[0062] The UAV collaborative control module includes a dual-redundant flight control system and a dynamic obstacle avoidance unit;

[0063] The communication relay module supports dual-mode communication between 5G and low-orbit satellites;

[0064] The master control and collaborative learning modules deploy a collaborative learning platform based on digital twins, and update the environmental cognition model of each drone through federated learning.

[0065] Specifically, taking the forest fire monitoring scenario as an example, the application of the dynamic weight allocation module in the forest fire monitoring scenario is as follows: when the drone passes through a thick smoke area (environmental complexity 0.85), the fuzzy logic automatically increases the weight of the lidar with strong smoke resistance to 0.78; at the same time, a mobile fire truck is detected in the direction of the fire spread (threat level 4), and the reinforcement learning module increases the weight of the millimeter wave radar by 0.4 based on historical obstacle avoidance data, and reduces the weight of the visual sensor to 0.1 in real time to avoid misjudgment caused by smoke interference.

[0066] Specifically, taking the urban logistics distribution scenario as an example, the application of the real-time path planning module in urban logistics distribution is as follows: when the drone flies from a warehouse to an office building, the improved RRT algorithm adopts bidirectional expansion to generate a forward path tree from the warehouse end (the sampling bias is the commercial district sky corridor), and a reverse path tree from the office building end (the sampling bias is the landing platform area). When the two trees are 8 meters apart, the connection mechanism is triggered, which shortens the search time by 52% compared with traditional RRT.

[0067] Specifically, the dynamic weight allocation module uses fractal geometry to calculate terrain complexity and assess environmental conditions. It uses a spatiotemporal graph convolutional network to identify dynamic obstacles and predict threat levels. It also constructs an adaptive adjustment model based on the Lyapunov exponent to optimize sensor weight allocation and applies Nash equilibrium theory to make collaborative decisions for multiple UAV missions.

[0068] The spatiotemporal graph convolutional network processes dynamic obstacle data by constructing a spatiotemporal adjacency matrix, using the lidar point cloud sequence and millimeter-wave radar trajectory prediction data as spatial graph node features and temporal graph edge weights, respectively. A layered spatiotemporal convolutional structure is used to extract obstacle motion pattern features, and a time sliding window mechanism is used to predict obstacle motion trajectories within the next 3 seconds. The threat level is divided into five quantitative indicators based on the trajectory conflict probability: when the collision time TTC is less than 2 seconds and the distance threshold is less than 5 meters, it is an emergency level; when TTC is less than 5 seconds and the distance threshold is less than 10 meters, it is a high-risk level; when TTC is less than 10 seconds and the distance threshold is less than 20 meters, it is a warning level; and the rest are safe and invalid levels.

[0069] The Lyapunov exponent adaptive adjustment model constructs an energy function to evaluate the confidence of each sensor in real time and normalize the environmental complexity. When the environmental complexity exceeds 0.7, the lidar weight is automatically increased to the range of 0.6-0.8. When the threat level reaches high or above, the millimeter-wave radar weight is dynamically increased by 0.3-0.5.

[0070] Among them, Nash equilibrium collaborative decision-making establishes a game benefit matrix between drones. The benefit function includes three dimensions: task completion, energy efficiency and safety distance. The Pareto optimal allocation plan is achieved through iterative strategy optimization. The final quantitative results of environmental complexity and threat level are fed back to the main control and collaborative learning modules in real time through the 5G link.

[0071] Specifically, taking the highway inspection scenario as an example, the workflow of the spatiotemporal graph convolutional network in the highway inspection scenario is as follows:

[0072] First, the 64-line point cloud scanned by the lidar is clustered into 150 nodes (each node represents a vehicle / roadblock), and the movement trajectory of the truck tracked by the millimeter-wave radar is used as the time edge weight. Ten consecutive frames of data form a space-time matrix. Then, after recognizing that the truck will change lanes in 2.3 seconds and the conflict probability reaches 87%, the system determines it as an emergency-level threat (TTC = 1.8 seconds, distance 4.2 meters) and immediately increases the weight of the millimeter-wave radar by 0.45.

[0073] Specifically, taking the energy control scenario as an example, the application of the Lyapunov adaptive model in the energy control scenario is: when the battery power drops to 28%, the model calculates the sensor energy consumption derivative dV / dt = -0.32 (exceeding the threshold of -0.2), and automatically reduces the high-power consumption lidar weight from 0.6 to 0.3, and increases the visual weight to 0.7, extending the battery life by 12 minutes.

[0074] Specifically, taking the disaster search and rescue scenario as an example, the decision-making process of Nash equilibrium collaboration in disaster search and rescue is as follows: set up three drones to respectively undertake life detection, material delivery, and terrain mapping tasks, and calculate them through the payoff matrix:

[0075] The safety distance weight of the life detector is 0.7 (close-range scanning is required);

[0076] The energy efficiency weight of the material delivery aircraft is 0.6 (it can fly with a full load);

[0077] The task completion weight of the surveying and mapping machine is 0.8 (meeting the full coverage requirement);

[0078] Based on the weight strategy of the three drones, the Pareto optimal solution was finally generated: the altitude of the life detection aircraft was reduced to 50 meters, the path of the material delivery aircraft was shortened by 300 meters, and the scanning interval of the surveying and mapping aircraft was expanded by 2 meters.

[0079] Specifically, the multi-source data fusion module calibrates heterogeneous data through microsecond-level time synchronization and sub-meter-level spatial registration. A hidden Markov model of sensor health status is established for confidence assessment. Dempster-Shafer evidence theory is used to resolve multi-source data conflicts. An ARIMA-LSTM hybrid model is constructed to predict and compensate for communication delays. The processed fused data is then fed into the real-time path planning module to generate a trajectory. For example, in a bridge inspection scenario, a lidar scan of the bridge pier generates a 200,000-point cloud, while a visual sensor captures high-definition images of the cracks. The SIFT algorithm is then used to extract the pier rivet feature points (matching 56 key points per frame). RANSAC is then used to remove mismatched points (such as interference caused by attached moss), resulting in a final registration error of less than 1.3 cm.

[0080] The specific process of communication delay compensation:

[0081] The ARIMA model predicts a baseline latency of 28ms (based on 100 sets of historical data);

[0082] LSTM network learning burst interference mode: when encountering electromagnetic interference, the predicted compensation value is +15ms;

[0083] The actual data packet transmission delay is 33ms, and the effective delay after compensation is reduced to 18ms.

[0084] The real-time path planning module adopts a layered architecture with a global topology layer to generate the initial path, a local obstacle avoidance layer to deal with dynamic obstacles, and a trajectory optimization layer to smooth the trajectory;

[0085] The global topology layer constructs a navigation topology map based on the 3D point cloud data output by the multi-source data fusion module. The improved Dijkstra algorithm is used to calculate the initial waypoint sequence. The path sampling density is dynamically adjusted according to the environmental complexity feedback from the dynamic weight distribution module. When the complexity is greater than 0.6, the sampling interval is reduced to 0.5 meters.

[0086] The local obstacle avoidance layer triggers path replanning based on the dynamic obstacle threat level predicted by the spatiotemporal graph convolutional network. When a high-risk threat is detected, the Monte Carlo tree search algorithm is activated, and a node expansion strategy based on the UAV's dynamic constraints is adopted to generate an emergency avoidance path within 200ms. The evaluation function integrates the weights of the three dimensions of threat field strength, energy consumption rate, and heading deviation. The weight coefficients are updated online by the main control and collaborative learning modules.

[0087] The trajectory optimization layer utilizes the node vector dynamic adjustment mechanism of the adaptive B-spline curve to optimize the curve order and control point density in real time according to the positioning accuracy feedback from the communication relay module. The curve continuity meets the C3-order differentiability requirement. At the same time, a dynamic correlation model between the lift-to-drag ratio and the power system is established. The power parameter configuration is optimized through the backpropagation algorithm. The optimized trajectory data is synchronized in real time to the UAV collaborative control module for execution.

[0088] Application of the global topology layer in power transmission inspection in mountainous areas:

[0089] Based on a terrain fractal dimension value of 0.75 (>0.6 threshold), the navigation map sampling interval was increased to 0.5 meters. An improved Dijkstra algorithm prioritized straight airspace between high-voltage power line towers, avoiding steep canyons. An initial path was generated containing 89 waypoints, with a computational time of 46 milliseconds. A local obstacle avoidance layer was implemented to address the sudden threat of bird swarms. The ST-GCN predicted that the swarm would intrude on the route within 1.2 seconds (high-risk level). A Monte Carlo tree search was initiated, generating three candidate paths based on the quadrotor's dynamic constraints:

[0090] Option A: Climb 15 meters (energy consumption +20%, safety factor 0.95); Option B: Turn right to avoid (course deviation 12°, safety factor 0.88); Option C: Slow down and hover (time consumption +8 seconds, safety factor 0.99); The evaluation function selects Option C (safety weight 0.6), and replanning is completed in 200 ms.

[0091] Example of track smoothing in the trajectory optimization layer:

[0092] In urban canyon areas with weak GPS signals (positioning error of 1.8 meters), a third-order B-spline curve (control point spacing of 2 meters) is used; after entering open areas (error <0.3 meters), a fifth-order curve (control point spacing of 0.5 meters) is switched. The power is optimized through a lift-to-drag ratio model: the rotor speed is automatically increased by 12% in the headwind phase, reducing energy consumption by 17%.

[0093] The communication relay module uses a deep Q network to achieve intelligent routing selection, improves anti-interference capabilities through orthogonal spatiotemporal modulation technology, applies game theory to dynamically allocate communication spectrum resources, and constructs a quantum key distribution channel to ensure data transmission security. It uploads the flight status data of the UAV collaborative control module to the main control and collaborative learning modules in real time.

[0094] The improved RRT* algorithm integrates environmental characteristics and historical path information to dynamically adjust the sampling strategy, adopts a bidirectional expansion mechanism to accelerate the path search process, smoothly optimizes the original trajectory through optimal control theory, and sets an online learning unit to update the heuristic function parameters according to reinforcement learning feedback. The optimization results are converted into flight instructions through the UAV collaborative control module.

[0095] The UAV collaborative control module uses a quaternion interpolation algorithm to achieve fault-tolerant control of flight command solution, optimizes the multi-rotor thrust distribution scheme based on the pseudo-inverse method, and the dynamic obstacle avoidance unit completes obstacle avoidance actions within 200ms through the fusion perception of lidar and visual sensors. The formation maintenance unit uses a virtual structure method to maintain the geometric configuration of the cluster, and control commands are distributed to each UAV node through the communication relay module.

[0096] Specifically, the dual-mode communication switching logic is as follows: taking the scenario of a drone operating on an offshore platform as an example, when the drone is operating on the offshore platform, the 5G signal strength drops to -85dBm; the DQN network evaluates the satellite link delay to be 120ms (<200ms threshold); the switching action is executed and quantum key distribution is started to generate a 256-bit encryption key.

[0097] The quadrotor's emergency obstacle avoidance maneuver was broken down into: the lidar detected the sudden appearance of the advertising balloon at a distance of 3.2 meters; the visual sensor confirmed the target as a lightweight object (non-metal) within 0.2 seconds.

[0098] The dynamic obstacle avoidance unit calculates the minimum avoidance torque: the output of the right front motor is increased to 85%; the output of the left rear motor is reduced to 45%; the vehicle avoids the obstacle by flying sideways at a 22° tilt angle; the entire process takes 180ms, and energy consumption is reduced by 35% (compared to a full-power climb).

[0099] The main control and collaborative learning modules construct a three-dimensional digital twin model that contains terrain features and dynamic obstacle information, updates the drone track data (decision-making strategy) of the real-time path planning module through the transfer learning framework, integrates the multi-source environmental perception data matrix of the multi-source data acquisition and fusion module to generate a threat situation map, and provides an augmented reality navigation interface and voice command interaction functions. The digital twin model and the drone collaborative control module establish a two-way data channel.

[0100] The threat situation map is generated by fusing lidar point clouds, millimeter-wave radar trajectory predictions, and visual semantic segmentation data from a multi-source data fusion module. A spatiotemporal graph convolutional network is used to extract dynamic obstacle motion characteristics. Combined with terrain elevation and meteorological data from the digital twin model, a physical information neural network is used to construct a three-dimensional threat field intensity distribution model.

[0101] Among them, the threat level assessment is based on the three-dimensional threat field strength distribution model, and the drone collision probability is calculated through Monte Carlo risk simulation. The energy consumption risk is predicted based on the evaluation fusion function of the real-time path planning module. The final multi-dimensional threat situation map contains dynamic obstacle hot zones, energy risk contours and communication blind spot markers, and is superimposed and displayed through the augmented reality navigation interface of the main control and collaborative learning modules, providing a dynamic weight allocation basis for the dynamic weight allocation module.

[0102] Generation of a three-dimensional threat situation map:

[0103] At the scene of a chemical plant leak: LiDAR builds a 3D model of the tank; millimeter-wave radar tracks the spread of toxic gases; and visual sensors identify cracks at the source of the leak.

[0104] Physical information neural network calculates threat field strength:

[0105] Leakage point core area: threat value 0.92 (red hot zone);

[0106] 50 meters downwind: threat value 0.75 (yellow warning);

[0107] Downwind safe zone: Threat value 0.31 (green).

[0108] The AR navigation interface overlays and displays the disaster avoidance route (bypassing the western safety passage).

[0109] Federated learning model update:

[0110] Drone No. 1 learned in a dense fog environment that the confidence of the millimeter-wave radar increased by 22% when the humidity was greater than 80%. This data was encrypted and uploaded to the main control and collaborative learning modules. After the global model was updated, it was distributed to all drones to improve the perception consistency of the cluster in inclement weather.

[0111] Example 2

[0112] like Figure 2 As shown, based on Example 1, Example 2 of the present invention proposes a real-time path planning method based on dynamic weight allocation and multi-source data fusion of UAV based on an active power filter current tracking control system based on an improved butterfly algorithm.

[0113] Specifically, the method includes the following steps:

[0114] S1. Through multi-source data acquisition, obtain multi-source data from drone sensors, terrain data of the drone's flight area, and drone flight condition data. Preprocess the multi-source data from drone sensors to construct a multi-source environmental perception data matrix.

[0115] S2. Based on the multi-source environmental perception data matrix, terrain complexity and dynamic obstacle threat levels are calculated using fractal geometry algorithms and Lyapunov rule models, respectively. A multi-UAV weight allocation model is constructed using Nash equilibrium theory to derive a multi-UAV task coordination control sequence.

[0116] S3. Build a real-time path planning model to generate UAV trajectory data based on the multi-source environmental perception data matrix and the multi-UAV collaborative control sequence;

[0117] S4. Based on the drone track data obtained in step S3, the drone's flight attitude is controlled using a quaternion interpolation algorithm. The drone's multi-rotor thrust is distributed based on the pseudo-inverse method. The laser radar unit and visual sensor are combined to complete obstacle avoidance within 200ms and output real-time flight parameters.

[0118] S5. Based on real-time flight parameters and a multi-source environmental perception data matrix, a three-dimensional digital twin model containing terrain, obstacle, and meteorological data is constructed. Multi-source information is integrated to generate a dynamic threat situation map, and a federated learning algorithm is used to update the local model of each drone in real time.

[0119] Example 3

[0120] like Figure 3 As shown, based on Example 1, this Example 3 proposes a terminal device of a real-time path planning system based on dynamic weight allocation and multi-source data fusion of UAVs, and the terminal device 200 includes at least one memory 210, at least one processor 220 and a bus 230 connecting different platform systems.

[0121] The memory 210 may include a readable medium in the form of a volatile memory, such as a RAM 211 and / or a cache memory 212 , and may further include a ROM 213 .

[0122] Among them, the memory 210 also stores a computer program, and the computer program can be executed by the processor 220, so that the processor 220 executes any of the above-mentioned embodiments of the present application. A real-time path planning system based on dynamic weight allocation and multi-source data fusion of drones, its specific implementation method is consistent with the implementation method and the technical effect achieved in the embodiment of the above method, and some contents are not repeated here. The memory 210 can also include a program / utility 214 having a set (at least one) of program modules 215. Such program modules include but are not limited to: an operating system, one or more application programs, other program modules, and program data. Each of these examples or some combination may include the implementation of a network environment.

[0123] Accordingly, the processor 220 may execute the aforementioned computer programs, as well as the program / utility 214 .

[0124] The bus 230 may represent one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus architectures.

[0125] The terminal device 200 can also communicate with one or more external devices 240, such as keyboards, pointing devices, Bluetooth devices, etc., and can also communicate with one or more devices that can interact with the terminal device 200, and / or communicate with any device that enables the terminal device 200 to communicate with one or more other computing devices (such as routers, modems, etc.). Such communication can be carried out through the I / O interface 250. In addition, the terminal device 200 can also communicate with one or more networks (such as local area networks (LANs), wide area networks (WANs) and / or public networks, such as the Internet) through the network adapter 260. The network adapter 260 can communicate with other modules of the terminal device 200 through the bus 230. It should be understood that although not shown in the figure, other hardware and / or software modules can be used in conjunction with the terminal device 200, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage platforms.

[0126] Example 4

[0127] like Figure 4 As shown, based on Example 1, this embodiment proposes a computer-readable storage medium for a real-time path planning system based on dynamic weight allocation and multi-source data fusion for drones. The computer-readable storage medium stores instructions that, when executed by a processor, implement any of the aforementioned real-time path planning systems based on dynamic weight allocation and multi-source data fusion for drones. The specific implementation methods and technical effects achieved are consistent with those described in the aforementioned system embodiments, and some details are not repeated here.

[0128] Figure 4The program product 300 provided in this embodiment for implementing the above-mentioned system is shown. It can adopt a portable compact disc read-only memory (CD-ROM) and include program code, and can be run on a terminal device, such as a personal computer. However, the program product 300 of the present invention is not limited to this. In this embodiment, the readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or in conjunction with an instruction execution system, device or device. The program product 300 can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0129] A computer-readable storage medium may include a data signal transmitted in baseband or as part of a carrier wave, carrying readable program code. This transmitted data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, which can transmit, transmit, or transfer a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof. The program code for performing the operations of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java, C++, etc., as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a standalone software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. Where a remote computing device is involved, the remote computing device may be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., through the Internet using an Internet service provider).

[0130] The present invention is explained from the perspectives of purpose of use, effectiveness, progress and novelty. The practical progress it has is in compliance with the functional enhancement and use requirements emphasized by the Patent Law. The above description and drawings of this application are only preferred embodiments of this application and are not intended to limit this application. Therefore, all structures, devices, features, etc. that are similar or identical to those of this application, that is, all equivalent replacements or modifications made in accordance with the scope of this patent application, should fall within the scope of protection of this patent application.

[0131] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion, characterized in that: include: Multi-source data acquisition and fusion module, used to obtain multi-source data from drone sensors, terrain data of the drone flight area, and drone flight condition data, and pre-process the multi-source data from drone sensors to construct a multi-source environmental perception data matrix; The dynamic weight allocation module is used to generate a weight coefficient control sequence based on the multi-source fusion perception data matrix, the terrain data of the UAV flight area, and the UAV flight condition data, and dynamically adjust the weight coefficients of the UAV sensors; The real-time path planning module is used to generate the UAV's global path and local obstacle avoidance trajectory through a hierarchical path planning model based on the multi-source environmental perception data matrix and weight coefficient control sequence, and dynamically optimize and smooth the global path to obtain the UAV's track data; The UAV collaborative control module is used to perform real-time flight control and flight command calculation for a cluster of multiple UAVs based on the UAV track data output by the real-time path planning module; The communication relay module is used to establish a dual-mode communication link with low-orbit satellites through the 5G network, transmitting the UAV flight status information to the implementation path planning module and the main control and collaborative learning module; The master control and collaborative learning modules are used to build a three-dimensional digital twin model based on the terrain data of the drone flight area and the drone flight status information, and perform online updates and global optimization of each module based on the federated learning algorithm.

2. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 1 is characterized in that: The multi-source data acquisition and fusion module includes a lidar unit, a millimeter-wave radar unit, a visual sensor unit and a Beidou satellite positioning unit; the multi-source data of the drone sensor includes lidar point cloud data, millimeter-wave radar trajectory prediction data, image recognition data, and Beidou satellite positioning data; The preprocessing includes calibrating heterogeneous data of multi-source data of drone sensors through time synchronization algorithm and spatial registration theory, establishing a hidden Markov model to perform confidence assessment on the health status of drone sensors, eliminating data conflicts among the multi-source data of drone sensors according to the Dempster-Shafer evidence theory, and constructing an ARIMA-LSTM hybrid model to predict and compensate for the communication delay of each data among the multi-source data of drone sensors.

3. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 2 is characterized in that: The dynamic weight allocation module includes a terrain analysis unit, a threat identification unit, a confidence adjustment unit and a Nash equilibrium collaborative decision unit; The terrain analysis unit is used to calculate the terrain complexity of the UAV flight area using a fractal geometry algorithm based on the terrain data of the UAV flight area; The threat identification unit is used to obtain the predicted threat level of dynamic obstacles through the spatiotemporal graph convolution network according to the terrain complexity of the drone's flight area; The confidence adjustment unit is used to predict the threat level of the drone's flight area based on the terrain complexity and dynamic obstacles, build an adaptive adjustment model based on the Lyapunov index, and optimize the drone sensor weight distribution in real time; The Nash equilibrium collaborative decision-making unit predicts the threat level of the terrain complexity and dynamic obstacles in the drone flight area, combines the multi-source fusion perception data matrix, and establishes a game payoff matrix among multiple drones based on Nash equilibrium theory to make collaborative decisions on multiple drone tasks; The spatiotemporal graph convolutional network constructs a spatiotemporal adjacency matrix, uses lidar point cloud data as spatial graph node features, and uses millimeter-wave radar trajectory prediction data as time graph edge weights. It extracts obstacle motion pattern features based on a hierarchical spatiotemporal convolutional structure, and predicts obstacle motion trajectories within the next 3 seconds through a time sliding window mechanism. The dynamic obstacle prediction threat level is divided into five levels of quantitative indicators based on the pre-obtained UAV flight trajectory conflict probability: when the collision time is less than 2 seconds and the distance threshold is less than 5 meters, it is an emergency level; when the collision time is less than 5 seconds and the distance threshold is less than 10 meters, it is a high-risk level; when the collision time is less than 10 seconds and the distance threshold is less than 20 meters, it is a warning level; when the collision time is greater than 10 seconds and the distance threshold is greater than 20 meters, it is a safe level and an invalid level. The adaptive adjustment model based on the Lyapunov index constructs an energy function to evaluate the confidence of each drone sensor in real time and normalizes the terrain complexity of the drone's flight area. When the terrain complexity of the drone's flight area exceeds 0.7, the weight of the lidar unit is increased to 0.6-0.

8. When the threat level reaches a high-risk or emergency level, the weight of the millimeter-wave radar unit is dynamically increased to 0.3-0.

5. The game payoff matrix includes flight mission completion, flight energy efficiency, and flight safety distance. The game payoff matrix is ​​optimized through an iterative strategy to obtain a Pareto optimal allocation solution and generate a weight coefficient control sequence. The resulting terrain complexity of the drone's flight area and the predicted threat level of dynamic obstacles are fed back to the real-time path planning module and the master control and collaborative learning module in real time through the 5G network.

4. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 3 is characterized in that: The hierarchical path planning model includes a global topology layer, a local obstacle avoidance layer, and a trajectory optimization layer; The global topology layer constructs a navigation topology map based on the lidar point cloud data and calculates the initial waypoint sequence of the drone using the improved Dijkstra algorithm. The path sampling density of the initial waypoint sequence of the drone is dynamically adjusted according to the terrain complexity of the drone's flight area. When the terrain complexity of the drone's flight area is greater than 0.6, the path sampling density is reduced to 0.5 meters. When there are dynamic obstacles or highly complex terrain in the flight area, the global topology layer dynamically adjusts the sampling strategy through the improved RRT* algorithm, integrating environmental features and historical path information to generate the initial waypoint sequence. The local obstacle avoidance layer constructs an evaluation fusion function by dynamically predicting the threat level of obstacles and replanning the initial waypoint sequence. When a high-risk threat is detected, the Monte Carlo tree search algorithm is activated, and an emergency avoidance path is generated within 200ms based on a node expansion strategy constrained by the UAV's dynamics. The evaluation fusion function includes the threat field strength value, energy consumption rate, and heading deviation. The trajectory optimization layer is based on the dynamic adjustment mechanism of the node vector of the adaptive B-spline curve. It optimizes the curve order and control point density in real time according to the UAV flight status information fed back by the communication relay module. The continuity of the curve order meets the C3 order differentiability requirement. A dynamic correlation model between the UAV lift-to-drag ratio and the UAV power system is established. The UAV power parameter configuration is optimized through the backpropagation algorithm to obtain the final UAV track data, which is transmitted to the UAV collaborative control module in real time.

5. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 4 is characterized in that: The improved RRT* algorithm is used for path planning at the global topology layer. It integrates environmental characteristics and historical path information to dynamically adjust the sampling strategy, adopts a bidirectional expansion mechanism to accelerate the path search process, and performs preliminary smooth optimization on the generated initial trajectory through optimal control theory. An online learning unit is set up to update the heuristic function parameters based on reinforcement learning feedback. The optimization results are used as the output of the global topology layer for further processing by the local obstacle avoidance layer and trajectory optimization layer, and flight instructions are generated by the UAV collaborative control module.

6. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 5 is characterized in that: The UAV collaborative control module uses a quaternion interpolation algorithm to perform fault-tolerant control on the flight command solution, optimizes the multi-rotor thrust distribution scheme based on the pseudo-inverse method, and the dynamic obstacle avoidance unit completes obstacle avoidance within 200ms through the fusion perception of the lidar unit and the visual sensor. The formation maintenance unit uses the virtual structure method to maintain the cluster geometric configuration, and the control commands are distributed to each UAV node via the communication relay module.

7. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 6 is characterized in that: The main control and collaborative learning module constructs a three-dimensional digital twin model containing the terrain data of the drone flight area and the drone flight status information, updates the drone track data of the real-time path planning module through the transfer learning framework, and integrates the multi-source environmental perception data matrix of the multi-source data acquisition and fusion module to generate a threat situation map.

8. The UAV real-time path planning system based on dynamic weight allocation and multi-source data fusion according to claim 7 is characterized in that: The threat situation map is constructed by fusing lidar point cloud data, millimeter-wave radar trajectory prediction data, and image recognition data from a multi-source data acquisition and fusion module. It uses a spatiotemporal graph convolutional network to extract dynamic obstacle motion features. Combined with the terrain data of the drone flight area in the three-dimensional digital twin model, a federated learning algorithm is used to construct a three-dimensional threat field intensity distribution model. According to the three-dimensional threat field strength distribution model, the drone collision probability is calculated through the Monte Carlo risk simulation algorithm, and the energy consumption risk is predicted based on the evaluation fusion function of the real-time path planning module. Finally, a multi-dimensional threat situation map including dynamic obstacle hot spots, energy risk contours and communication blind spot markers is obtained. It is superimposed and displayed through the augmented reality navigation interface of the main control and collaborative learning modules, providing a dynamic weight allocation basis for the dynamic weight allocation module.

9. A real-time path planning method based on dynamic weight allocation and multi-source data fusion for unmanned aerial vehicles, the method being implemented based on a real-time path planning system for unmanned aerial vehicles based on dynamic weight allocation and multi-source data fusion as claimed in any one of claims 1 to 8, characterized in that: The following steps are involved: S1. Through multi-source data acquisition, obtain multi-source data from drone sensors, terrain data of the drone's flight area, and drone flight condition data. Preprocess the multi-source data from drone sensors to construct a multi-source environmental perception data matrix. S2. Based on the multi-source environmental perception data matrix, terrain complexity and dynamic obstacle threat levels are calculated using fractal geometry algorithms and Lyapunov rule models, respectively. A multi-UAV weight allocation model is constructed using Nash equilibrium theory to derive a multi-UAV task coordination control sequence. S3. Build a real-time path planning model to generate UAV trajectory data based on the multi-source environmental perception data matrix and the multi-UAV collaborative control sequence; S4. Based on the drone track data obtained in step S3, the drone's flight attitude is controlled using a quaternion interpolation algorithm. The drone's multi-rotor thrust is distributed based on the pseudo-inverse method. The laser radar unit and visual sensor are combined to complete obstacle avoidance within 200ms and output real-time flight parameters. S5. Based on real-time flight parameters and a multi-source environmental perception data matrix, a three-dimensional digital twin model containing terrain, obstacle, and meteorological data is constructed. Multi-source information is integrated to generate a dynamic threat situation map, and a federated learning algorithm is used to update the local model of each drone in real time.

Images