Unmanned aerial vehicle post-disaster operation method and device, electronic equipment and storage medium
By constructing a post-disaster emergency situation map through multi-source data fusion and adopting an incremental update mechanism, the problems of information fragmentation and high computing load in UAV post-disaster operations were solved, enabling real-time rescue path planning and efficient rescue.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA TELECOM CLOUD TECH CO LTD
- Filing Date
- 2025-12-03
- Publication Date
- 2026-04-24
Smart Images
Figure CN121918581A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of emergency communication and disaster relief technology, specifically relating to a method for post-disaster operations using unmanned aerial vehicles (UAVs), a device for post-disaster operations using UAVs, electronic equipment, and a computer-readable storage medium. Background Technology
[0002] Currently, the application of drones in post-disaster emergency operations still faces systemic technical bottlenecks. Existing solutions use drone-borne sensors, such as RGB-D cameras and LiDAR, to detect local environments and construct static maps, such as landslide areas and damaged buildings, and use ESDF and other map technologies for obstacle avoidance. This approach simplifies the environment to a static spatial distribution of obstacles, making path planning unable to adapt to real-time changes in rescue needs. Disaster assessment typically relies solely on local sensor data from drones, with disaster area data scattered across satellites, drones, ground sensors, and social media platforms. Format heterogeneity (UTM / ENU coordinate system differences, temporal deviations) leads to information fragmentation, making it difficult to efficiently allocate rescue resources. Complex disaster models (such as communication signal attenuation prediction and traffic congestion rate analysis) operate inefficiently on edge devices, and existing technologies employ a full map update strategy. Even with only local environmental changes (such as new landslides), a global reconstruction of the raster map is still required, resulting in excessive computational load that cannot meet the real-time response requirements of complex post-disaster scenarios. While existing technologies attempt to introduce deep learning algorithms, their deployment is limited by the computing power and energy consumption of edge computing devices, enabling only basic target recognition and tracking functions. Summary of the Invention
[0003] The purpose of this invention is to provide a method for post-disaster operations using unmanned aerial vehicles (UAVs), a device for post-disaster operations using UAVs, electronic equipment, and a corresponding computer-readable storage medium. This addresses the problem that existing technical solutions mainly rely on single sensor data or manual intervention, making it difficult to meet the real-time response requirements of complex post-disaster scenarios.
[0004] To solve the above-mentioned technical problems, the present invention is implemented as follows: In a first aspect, embodiments of the present invention provide a method for post-disaster operations using unmanned aerial vehicles (UAVs), applied in a cloud-based environment, the method comprising: Receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of the following: unmanned aerial vehicle, satellite system, ground sensor, and network; Based on the multi-source heterogeneous data, the target objects in the disaster area and the attribute information of the target objects are determined; Based on the target objects in the disaster area and the attribute information of the target objects, a post-disaster emergency situation map is constructed. Based on the data changes in the post-disaster emergency situation map, an incremental update data package is generated; The incremental update data packet is sent to the UAV so that the UAV can plan a post-disaster operation route based on the updated post-disaster emergency situation map.
[0005] Optionally, constructing a post-disaster emergency situation map based on target objects in the disaster area and the attribute information of the target objects includes: The disaster area is divided into a grid network on a digital map, and the grid network includes grid nodes; Each grid node in the grid network stores the attribute information of the corresponding target object; Based on the attribute information of the target object corresponding to the grid node, the influence relationship between each pair of grid nodes is determined; The influence relationship between the grid network and the pairs of grid nodes is determined as the post-disaster emergency situation map.
[0006] Optionally, generating incremental update data packets based on data changes in the post-disaster emergency situation map includes: Based on the data changes in the post-disaster emergency situation map, at least one affected target grid in the grid network is identified; The post-disaster emergency situation map is updated based on the attribute information of the target object corresponding to the at least one target grid, and an incremental update data packet is generated.
[0007] Optionally, the attribute information of the target object includes geographical feature attributes and dynamic disaster situation attributes; The geographic feature attributes include one or more of the following: building damage level, terrain complexity, and distribution of communication base stations; The dynamic disaster attributes include one or more of the following: population density, traffic congestion rate, and the trend of secondary disaster spread.
[0008] Secondly, embodiments of the present invention provide a drone post-disaster operation system, the system comprising a cloud and at least one drone; The cloud platform is used to receive multi-source heterogeneous data, which comes from at least one or more of at least one UAV, satellite system, ground sensor, and network. Based on the multi-source heterogeneous data, target objects in the disaster area and their attribute information are determined. Based on the target objects in the disaster area and their attribute information, a post-disaster emergency situation map is constructed. Based on data changes in the post-disaster emergency situation map, an incremental update data package is generated. The incremental update data package is then sent to the UAV, enabling the UAV to plan post-disaster operation paths based on the updated post-disaster emergency situation map. The drone is used to collect raw data from the disaster area; process the raw data to generate standardized metadata; upload the metadata to the cloud for the cloud to build or update the post-disaster emergency situation map; receive incremental update data packets from the cloud, and plan post-disaster operation paths based on the updated post-disaster emergency situation map.
[0009] Optionally, the cloud platform is further configured to divide the disaster area into a grid network on a digital map, the grid network including grid nodes; store attribute information of a corresponding target object in each grid node of the grid network; determine the influence relationship between pairs of grid nodes based on the attribute information of the target objects corresponding to the grid nodes; and determine the grid network and the influence relationship between pairs of grid nodes as the post-disaster emergency situation map.
[0010] Optionally, the drone is also used to identify and locate at least one target object in the raw data; and to continuously track the at least one target object to obtain change data of the target object's attribute information.
[0011] Thirdly, embodiments of the present invention provide a device for unmanned aerial vehicle (UAV) disaster recovery operations, applied in the cloud, the device comprising: A data receiving module is used to receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of at least one UAV, satellite system, ground sensor and network. The attribute information determination module is used to determine the target objects in the disaster area and the attribute information of the target objects based on the multi-source heterogeneous data. The situation map construction module is used to construct a post-disaster emergency situation map based on the target objects in the disaster area and the attribute information of the target objects; The data packet generation module is used to generate incremental update data packets based on the data changes in the post-disaster emergency situation map; The data packet delivery module is used to deliver the incremental update data packet to the UAV so that the UAV can plan a post-disaster operation path based on the updated post-disaster emergency situation map.
[0012] Optionally, the situation map construction module includes: The grid division submodule is used to divide the disaster area into a grid network on a digital map, wherein the grid network includes grid nodes; The information storage submodule is used to store the attribute information of the corresponding target object at each grid node of the grid network; The influence relationship determination submodule is used to determine the influence relationship between any two grid nodes based on the attribute information of the target object corresponding to the grid node; The situation map determination submodule is used to determine the influence relationship between the grid network and the pairs of grid nodes as the post-disaster emergency situation map.
[0013] Optionally, the data packet generation module includes: The target grid determination submodule is used to determine at least one affected target grid in the grid network based on data changes in the post-disaster emergency situation map. The data packet generation submodule is used to update the post-disaster emergency situation map based on the attribute information of the target object corresponding to the at least one target grid, and generate incremental update data packets.
[0014] Optionally, the attribute information of the target object includes geographical feature attributes and dynamic disaster situation attributes; The geographic feature attributes include one or more of the following: building damage level, terrain complexity, and distribution of communication base stations; The dynamic disaster attributes include one or more of the following: population density, traffic congestion rate, and the trend of secondary disaster spread.
[0015] Fourthly, embodiments of the present invention provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.
[0016] Fifthly, embodiments of the present invention provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.
[0017] The embodiments of the present invention have the following advantages: This invention provides a method for post-disaster operations using unmanned aerial vehicles (UAVs) in a cloud-based environment. The method includes: receiving multi-source heterogeneous data; the multi-source heterogeneous data originating from at least one or more of a UAV, satellite system, ground sensor, and network; determining target objects and their attribute information in the disaster area based on the multi-source heterogeneous data; constructing a post-disaster emergency situation map based on the target objects and their attribute information; generating incremental update data packets based on data changes in the post-disaster emergency situation map; and sending the incremental update data packets to the UAVs to enable them to plan post-disaster operation paths based on the updated post-disaster emergency situation map. The post-disaster emergency situation map constructed by fusing multi-source data significantly improves the completeness of information in the disaster area. The emergency situation map predicts disaster spread paths through quantified parameter information, supports proactive rescue decision-making, reduces UAV communication costs, and improves rescue efficiency. Attached Figure Description
[0018] Figure 1This is a flowchart of the steps of a method for post-disaster operations using unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention; Figure 2 This is a post-disaster emergency situation diagram of a drone post-disaster operation method provided in an embodiment of the present invention; Figure 3 This is a system architecture diagram for constructing a post-disaster emergency situation map of a method for post-disaster operations using unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention. Figure 4 This is a flowchart of the component interactions of a method for post-disaster operations using unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention. Figure 5 This is a flowchart of the single-drone situation map construction process for a drone post-disaster operation method provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the structure of a UAV post-disaster operation system provided in an embodiment of the present invention; Figure 7 This is a structural block diagram of a drone disaster recovery device provided in an embodiment of the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0020] The terms "first," "second," etc., used in the specification and claims of this invention are used to distinguish similar objects and are not used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.
[0021] The following description, in conjunction with the accompanying drawings, details a method for post-disaster operations using unmanned aerial vehicles (UAVs), a device for post-disaster operations using unmanned aerial vehicles (UAVs), an electronic device, and a computer-readable storage medium provided by the embodiments of the present invention, through specific examples and application scenarios.
[0022] Reference Figure 1 The diagram illustrates a flowchart of a method for post-disaster operations using unmanned aerial vehicles (UAVs) according to an embodiment of the present invention. The method may specifically include the following steps: Step 101: Receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of the following: unmanned aerial vehicle, satellite system, ground sensor, and network; Traditional maps only reflect the static distribution of obstacles and do not model the trend of disaster spread (such as the impact range of aftershocks and the direction of flood flow) and dynamic rescue needs (such as areas with communication disruptions and gathering points of trapped people). Current technologies rely on local sensor data, and drones cannot grasp in real time the areas of communication disruption, the distribution of trapped people, and the spread trend of secondary disasters in disaster areas. For example, drones may mistakenly enter high-risk areas due to failure to predict the risk of secondary disasters, delaying the reconstruction of communication networks and the delivery of supplies.
[0023] In this embodiment of the invention, the post-disaster emergency situation map constructed by multi-source data fusion significantly improves the completeness of information in the disaster area. Unmanned aerial vehicles (UAVs) can acquire real-time information on communication disruption areas, the location of trapped personnel, and the trend of disaster spread. The multi-source heterogeneous data comes from at least one or more of at least one UAV, satellite system, ground sensor, and network. These data differ significantly in physical source, data structure, spatiotemporal characteristics, and update frequency, collectively forming a multi-dimensional, multi-scale perception system for the post-disaster environment.
[0024] The heterogeneous characteristics of multi-source heterogeneous data are mainly reflected in the following aspects: spatial reference heterogeneity, where data exists in different coordinate systems. For example, satellite imagery uses the UTM global projection coordinate system, while UAV real-time perception data is usually based on the ENU local coordinate system with itself as the origin; temporal reference heterogeneity, where timestamps from different data sources may be out of sync, and the update frequency of the data varies greatly. Satellite data may be updated on a minute-by-minute basis, while UAV IMU data is updated on a millisecond-by-millisecond basis; data format heterogeneity, where data takes on various forms, including structured data, semi-structured data, and unstructured data, such as images, point clouds, text, and numerical streams; and semantic hierarchy heterogeneity, where the information carried by the data is at different levels, from low-level pixels and point clouds to mid-level target bounding boxes and location coordinates, and then to high-level semantic tags such as "base station damaged" and "personnel trapped".
[0025] Specifically, the multi-source heterogeneous data includes, but is not limited to, the following types: The UAV platform's perception data includes visual and environmental data, collected by sensors such as RGB-D cameras and LiDAR onboard the UAV. The RGB-D cameras simultaneously capture high-resolution color images and corresponding depth information, forming 3D point cloud data for accurate perception of the geometric structure and spatial location of targets (such as trapped personnel or damaged buildings). The LiDAR emits laser beams and receives their echoes to generate high-precision 3D environmental point cloud maps, exhibiting excellent penetration and ranging accuracy, especially in post-disaster environments with low visibility such as smoke and dust. Positioning and attitude data are provided by the UAV's integrated GPS receiver and inertial measurement unit (IMU). The GPS module provides the UAV's absolute geographical location information, while the IMU unit uses accelerometers and gyroscopes to measure the UAV's three-axis linear acceleration and three-axis angular velocity in real time. Through fusion calculation, the UAV's high-frequency, high-precision attitude, heading, velocity, and position changes are obtained, providing crucial input for subsequent data coordinate alignment and motion compensation.
[0026] Satellite remote sensing data. The satellite system provides wide-area aerial imagery of disaster areas, covering a large area and macroscopically reflecting the affected area, topographic changes, and the spread of large-scale secondary disasters (such as floods and landslides). This imagery data is typically encoded based on the universal transverse Mercator projection coordinate system. In addition, the satellite system can also provide meteorological early warning data, such as precipitation cloud images and temperature distribution maps, to predict the development trend of disasters.
[0027] Ground-based sensor network data. Various IoT sensors deployed on the ground in the disaster area constitute a sensor network used to collect localized, refined environmental parameters. These sensors may include: environmental sensors, such as temperature and humidity sensors, used to monitor fire risk or specific microclimates; structural monitoring sensors, such as displacement and tilt sensors, deployed on critical buildings or bridges to monitor their structural stability and damage trends; and temporary communication node status data, where deployed temporary base stations or relay devices report their own operational status, load capacity, and coverage information.
[0028] Network and edge computing metadata. This part of the data mainly comes from public networks, such as social media and distress signals. It is obtained by crawling the web or accessing public emergency platforms to obtain distress information, text reports or images and videos from social media, mobile phone signals, etc. This data contains rich semantic information, such as the location of trapped people and descriptions of emergency needs.
[0029] This embodiment collects raw data about the disaster area in real time or near real time from multiple distributed and heterogeneous data sources, providing a rich and comprehensive information foundation for subsequent data fusion and situation map construction. This step receives data from all the aforementioned sources through a data interface compatible with multiple communication protocols and sends it into the subsequent data fusion and processing flow. This overcomes the limitations of perspective and information fragmentation caused by relying on a single data source in existing technologies, laying a solid data foundation for constructing a globally consistent and complete post-disaster emergency situation map.
[0030] Step 102: Based on the multi-source heterogeneous data, determine the target objects in the disaster area and the attribute information of the target objects; the attribute information of the target objects includes geographical feature attributes and dynamic disaster attributes; the geographical feature attributes include one or more of building damage degree, terrain complexity and communication base station distribution; the dynamic disaster attributes include one or more of population density, traffic congestion rate and secondary disaster spread trend.
[0031] Data from satellites, drones, and ground sensors is difficult to integrate due to differences in format and spatiotemporal references, resulting in fragmented disaster information and a lack of a global view for rescue command. Disaster area data is scattered across satellites, drones, ground sensors, and social media platforms, and format heterogeneity (UTM / ENU coordinate system differences, temporal discrepancies) leads to information fragmentation. For example, wide-area terrain data from satellite imagery cannot be coordinated with the real-time communication base station status detected by drones, resulting in a lack of a global disaster view for the command center and difficulty in efficiently allocating rescue resources.
[0032] In this invention, based on multi-source heterogeneous data, data processing is required to obtain the target objects and their attribute information in the disaster area, thereby subsequently constructing a post-disaster situation map. This involves integrating high-resolution RGB-D images collected by UAV sensors, high-precision GPS and IMU positioning information, combined with wide-area disaster images provided by satellite imagery, ground sensor data (temperature, humidity, displacement monitoring), and social media distress signals. The data format is standardized to eliminate coordinate system and timestamp differences, constructing a globally consistent disaster environment representation. The aim is to perform in-depth processing and intelligent analysis on the aforementioned received multi-source heterogeneous data to identify, locate, and quantify various entities (i.e., target objects) in the disaster area environment that are crucial for emergency rescue decisions, and to accurately calculate their dynamically changing attribute information.
[0033] In one embodiment of the present invention, the target objects are the basic elements constituting a post-disaster emergency response, mainly including but not limited to: trapped personnel, i.e. individuals or groups in urgent need of rescue; critical facilities, such as communication base stations, hospitals, bridges, power facilities, etc., whose status directly affects rescue operations; obstacles and hazards, such as damaged buildings, landslides, fire areas, flooded areas, etc.; rescue resources, such as rescue vehicles already deployed, material delivery points, etc.; environmental phenomena, such as road blockage areas, the forefront of secondary disaster spread, etc.
[0034] Precise 3D localization and state estimation of target objects are achieved by fusing multiple data sources, elevating targets in 2D images to a precise 3D geographic space for location and state analysis: depth information fusion and 3D coordinate calculation. For identified targets, the system calls upon their corresponding RGB-D camera depth images or LiDAR point cloud data. By defining a mask region within the target bounding box and applying extreme mean filtering techniques to eliminate abnormal depth values caused by reflection and noise, the precise distance of the target relative to the camera is calculated. Combining the camera intrinsic parameter matrix, real-time attitude data provided by the UAV IMU, and absolute position provided by GPS, a coordinate transformation chain is used to transform the target from the camera coordinate system to the aircraft coordinate system, then to the ENU local coordinate system, and finally mapped to a global geodetic coordinate system such as UTM or WGS-84 to obtain its precise latitude, longitude, and altitude.
[0035] For motion state estimation, for dynamic targets (such as a moving trapped person), the system uses its three-dimensional coordinate sequence at different timestamps to perform state estimation through a Kalman filter or an extended Kalman filter. Assuming the target follows a constant velocity or constant acceleration model, the filter can optimally estimate the target's instantaneous velocity, acceleration, and other motion attributes, and effectively smooth out observation jitter caused by sensor noise.
[0036] Next, multi-source information fusion and attribute standardization are performed. Attribute fusion involves the cloud-based data fusion system associating and verifying information from different sources that points to the same target or geographical area. For example: binding drone edge metadata (target ID, category) with 3D positioning results (precise location); using satellite imagery to perform wide-area verification and macroscopic assessment of the damage range of "damaged buildings" identified by drones; associating abnormal data from ground displacement sensors with "collapsed areas" identified by satellites and drones to quantify their instability; and spatially semantically associating social media distress messages (such as "injured on the third floor of XX building") with "trapped people" identified by drones near the building to enhance their urgency attribute.
[0037] Attribute vector generation. Finally, a standardized attribute vector is generated for each target object or each grid region of focus. The dimensions of this vector can be dynamically expanded. Typical attributes include: basic attributes such as target ID, target category (e.g., "trapped person"), and spatial location (latitude, longitude, and altitude); status attributes such as movement speed, direction of movement, posture (e.g., "stationary" or "lying down"), damage level (quantified as a 0-1 value for buildings), and operational status (e.g., "normal" or "interrupted" for base stations); and semantic and urgency attributes such as urgency level determined based on multi-source information fusion (e.g., "high," "medium," or "low"), credibility score, and a summary of the associated distress message text.
[0038] Through this step, the cloud successfully transformed raw, chaotic, multi-source, heterogeneous data into a structured, semantic set of target attribute information that can be directly understood and processed by machines. This provides an accurate and high-quality data foundation for the subsequent construction of a post-disaster emergency situation map that can dynamically reflect the "state" and "potential."
[0039] Step 103: Construct a post-disaster emergency situation map based on the target objects in the disaster area and the attribute information of the target objects; Post-disaster emergency situation map: refers to a multi-dimensional dynamic view of the post-disaster environment that comprehensively represents the state of the disaster-stricken area, including the degree of building damage, areas of communication disruption, distribution of trapped people, the spread trend of secondary disasters (such as aftershocks and floods), and the demand for rescue resources.
[0040] Complex disaster models (such as communication signal attenuation prediction and traffic congestion rate analysis) are inefficient on edge devices. Furthermore, existing technologies employ a full map update strategy, requiring global reconstruction of the raster map even for localized environmental changes (such as new landslides), leading to excessive computational load, delayed disaster modeling response, and difficulty in meeting minute-level emergency response requirements. The inefficiency of complex disaster models on edge devices and the wasteful full-update mechanism further exacerbate the problem. A solution is to divide the disaster area into a raster network, with each raster storing geographical features such as building damage levels, terrain complexity, and communication base station distribution, as well as disaster development trends such as resident movement trends, traffic congestion rates, and disaster spread trends. This allows for the quantitative assessment of disaster severity and rescue priorities using a quantifiable model.
[0041] The steps in this embodiment involve organizing, associating, and enhancing the discretely distributed target objects and their attribute information identified in the preceding steps using a structured data model to generate a post-disaster emergency situation map that can simultaneously depict the current state ("state") of the disaster area and predict its future evolution trend ("potential"). This situation map is not a simple visual map, but a calculable, inferable, and dynamically updated digital environmental model.
[0042] In one embodiment, step 103 may include the following sub-steps: Sub-step S11: Divide the disaster area into a grid network on a digital map, wherein the grid network includes grid nodes; First, the cloud logically divides the entire disaster area's geographic space into a uniform grid network. Each grid serves as the basic information carrier and calculation unit for the situation map. The granularity of the grid can be configured according to task requirements and computing resources; for example, it can be set to 10 meters × 10 meters.
[0043] Sub-step S12: Store the attribute information of the corresponding target object in each grid node of the grid network; Subsequently, the system fuses and maps the attribute information of the target object to its corresponding grid, forming a basic situational awareness layer. The information stored in each grid node constitutes a multi-dimensional feature vector, mainly including: Geographical features: Building damage level, a quantified value (e.g., 0 to 1), determined based on a combination of satellite imagery and UAV visual analysis, where 0 indicates intact and 1 indicates completely destroyed; Terrain accessibility complexity, calculated based on digital elevation models, LiDAR point clouds, and visually identified obstacles (e.g., ruins, cracks), used to assess the ease of passage for UAVs or rescue personnel; Distribution and status of temporary communication base stations, indicating whether a communication base station exists within the grid and recording its operational status (e.g., "normal", "interrupted", "moderate load") and coverage radius.
[0044] Dynamic disaster attributes: personnel density, calculated based on the number and distribution of "trapped people" identified within the grid; traffic congestion rate, assessed based on visual analysis of the road area and ground sensor data; secondary disaster spread trend, for example, indicating whether the grid is on a flood spread path or a high-risk area for aftershocks causing landslides. Sub-step S13: Based on the attribute information of the target object corresponding to the grid node, determine the influence relationship between each pair of grid nodes; That is, the construction of the situation relationship matrix and the quantification of "potential" go beyond the static stacking of attributes and dynamically express the mutual influence between grids and the spread of disaster.
[0045] Reference Figure 2 This invention illustrates a post-disaster emergency situation diagram of a drone-based post-disaster operation method provided by an embodiment of the invention; in some embodiments provided by the invention, a situation relationship matrix is used in the cloud. Describing the influence between grids, this matrix is a high-order tensor, and its specific applications and construction methods are as follows: For situation map nodes, For grid The dynamic target in the matrix, each element Represents from grid i To grid j The connection relationships and influence weights. This is a "bidirectional edge," meaning... and eji Different weights can be used to represent asymmetric effects. Weights The value is dynamically calculated based on the attribute information of the target object and its movement trend.
[0046] For example, the flow of people; if target tracking detects people moving from a grid... i Collective move to grid j ,but The weight of "personnel demand transmission" will increase. Disaster spread potential: If the flood is predicted to originate from the grid based on the hydrological model... i Flow to grid j ,but The weight representing "flood threat" will be set to a high value. Communication coverage potential: if the grid... i The temporary base station signal can cover the grid. j ,but The weight representing "communication connectivity" is positive. Resource demand potential: if the grid... i (Hospital) urgently needs grid j The medicines at the (supply point) are... The weight of "material demand" in the middle is increased. The specific types of impact can be determined based on the actual type of disaster and relief needs; this invention does not limit this.
[0047] Sub-step S14: The influence relationship between the grid network and the pairs of grid nodes is determined as the post-disaster emergency situation map.
[0048] The final disaster emergency situation map is a composite data object containing grid network data and a situation relationship matrix. Through this matrix, the originally isolated grids are connected into a dynamic, directed weighted network. This network can formally express the "pressure," "flow," and "trend" within the environment, enabling the situation map to describe "what might happen next," that is, to achieve the prediction from "state" to "potential."
[0049] Step 104: Generate an incremental update data packet based on the data changes in the post-disaster emergency situation map; In one embodiment, step 104 may include the following sub-steps: Sub-step S21: Based on the data changes in the post-disaster emergency situation map, determine at least one affected target grid in the grid network; For the current mission area of the drone, detect data changes within the neighboring grid, recalculate the disaster model and situation relationship matrix only for the changed areas, and manage historical situation map versions using version control technology to reduce computational load and communication overhead.
[0050] Sub-step S22: Update the post-disaster emergency situation map based on the attribute information of the target object corresponding to the at least one target grid, and generate an incremental update data packet.
[0051] To ensure the real-time nature of the situation map while significantly optimizing computing and communication resources, an incremental update mechanism is adopted. In some embodiments provided by this invention, change detection is performed first. The system continuously runs the CDC (Change Data Capture, a data change detection technology that captures and processes only newly added or modified data by monitoring incremental changes in a database or data stream) process, monitoring the input target attribute information stream. When a change in the attribute of a target object or a group of target objects is detected (such as location movement or status change), or when new satellite imagery indicates that building damage in a certain area has intensified, an update is triggered. The time frame of the data update depends on the type of disaster and the coverage area of the operation, which is not limited by this invention.
[0052] Next, local recalculation: The system only recalculates the grids whose attributes have changed and their directly related neighboring grids in the situational relationship matrix. For example, if a person moves 50 meters, the system only needs to update the personnel density attributes of the original grid and the new grid, and recalculate the correlation weights of these two grids with the surrounding grids in the situational relationship matrix, without having to reconstruct the situational map of the entire disaster area.
[0053] Finally, versioned storage is implemented. Each global or major incremental update generates a new version of the situation map, which is managed using version control systems such as Git-LFS (an extension of the Git version control system for efficient storage and management of large files). This not only ensures the traceability of historical data and supports post-disaster recovery and analysis, but also enables the system to quickly synchronize to the latest state when communication is interrupted and then restored.
[0054] Step 105: The incremental update data packet is sent to the UAV so that the UAV can plan a post-disaster operation path based on the updated post-disaster emergency situation map.
[0055] In this embodiment of the invention, the cloud encapsulates the incremental update data packet and sends it to the drone via a communication link. The specific steps are as follows: First, path planning. After receiving this situation map, the path planning algorithm can not only avoid static obstacles with high terrain complexity ("state"), but also proactively avoid high-risk paths with high secondary disaster spread trend weights ("potential") by querying the situation relationship matrix, and prioritize flying to grid areas with high population density and high urgency weights, thereby achieving intelligent path planning that combines safety and efficiency.
[0056] In some embodiments provided by the present invention, the planning of post-disaster operation routes by UAVs based on the updated post-disaster emergency situation map includes resource scheduling. That is, the command center can also analyze the situational relationship of each region based on the global situation map, carry out collaborative task allocation of multiple UAVs, and dynamically schedule resources to go to the most urgently needed areas.
[0057] This invention provides a method for post-disaster operations using unmanned aerial vehicles (UAVs) in a cloud-based environment. The method includes: receiving multi-source heterogeneous data; the multi-source heterogeneous data originating from at least one or more of a UAV, satellite system, ground sensor, and network; determining target objects and their attribute information in the disaster area based on the multi-source heterogeneous data; constructing a post-disaster emergency situation map based on the target objects and their attribute information; generating incremental update data packets based on data changes in the post-disaster emergency situation map; and sending the incremental update data packets to the UAVs so that the UAVs can plan post-disaster operation paths based on the updated post-disaster emergency situation map. The post-disaster emergency situation map constructed by fusing multi-source data significantly improves the completeness of information in the disaster area. The emergency situation map predicts disaster spread paths through quantified parameter information, supports proactive rescue decision-making, reduces UAV communication costs, and improves rescue efficiency.
[0058] Reference Figure 3 This diagram illustrates a system architecture for constructing a post-disaster emergency situation map using a UAV post-disaster operation method provided by an embodiment of the present invention. It mainly includes the following core modules: Data Acquisition and Preprocessing Layer: Drones, equipped with onboard sensors such as RGB-D cameras, LiDAR, and GPS / IMU modules, acquire high-resolution images, 3D point clouds, and self-positioning information of the local environment in real time; Satellite system provides wide-area disaster area imagery and meteorological early warning data, with an update cycle of minutes; Ground sensor network, deploying temperature and humidity sensors, displacement monitoring equipment, and temporary communication nodes, reports environmental conditions in real time, such as landslide areas and traffic congestion rates; Edge computing nodes: deployed on drones, running lightweight target detection and tracking algorithms to generate metadata, including the location and status of trapped personnel and the status of base station damage.
[0059] Cloud-based collaborative processing layer: Data fusion system, based on Apache Arrow, enables distributed memory sharing of satellite, UAV, and ground sensor data, and uses quaternion interpolation to align spatiotemporal references and eliminate coordinate system differences; Post-disaster emergency situation map construction module, which rasterizes the disaster area environment, with each node storing geographical features, building damage degree (0-1 quantization), terrain complexity, and distribution of temporary communication base stations; Dynamic disaster attributes, such as population density, traffic congestion rate, and secondary disaster spread trend; Incremental update service, based on Change Data Capture (CDC) technology to detect local data changes, only updating the affected raster, and combined with version control (Git-LFS) to manage historical situation maps, supporting post-disaster review and task retrospective.
[0060] Mission planning and feedback layer: The path planning terminal receives incremental situational maps from the cloud and combines them with planning algorithms to generate safe and efficient flight paths.
[0061] Reference Figure 4 The diagram illustrates the component interaction flowchart of a drone-based post-disaster operation method provided by an embodiment of the present invention, specifically including: The drone uploads pre-processed metadata to the cloud via a dual 5G / satellite link, including the location of trapped personnel, base station locations, and their status; satellite and ground sensor networks push wide-area disaster data to the cloud fusion system; the cloud constructs a global post-disaster emergency situation map and pushes it to the terminal through lightweight incremental packages containing only changed areas; the path planning terminal generates the task path.
[0062] Reference Figure 5 This invention provides a flowchart illustrating the construction of a single UAV situation map in a UAV post-disaster operation method. First, multi-target recognition and tracking are performed. After the UAV enters the mission area, its onboard RGB-D camera begins real-time data acquisition. Edge computing nodes load a pre-trained YOLOv5 model to perform multi-target detection on the high-resolution images captured by the camera, identifying dynamic targets, trapped personnel, and damaged communication base stations, and outputting target bounding boxes and category labels. Subsequently, the StrongSORT algorithm continuously tracks the targets based on their location and appearance features.
[0063] Next, target depth estimation. Based on the depth image captured by the RGB-D camera and combined with the target's bounding box, the UAV edge computing system uses mask extraction and extreme mean filtering techniques to define a mask region at the center of the target bounding box. Combining the camera intrinsic parameter matrix and IMU attitude data, extreme depth values are filtered out to calculate the precise distance between the target and the UAV, and its relative position with respect to the camera coordinate system is calculated and mapped to the geodetic coordinate system.
[0064] Then, target state estimation is performed. Combining the target's 3D coordinates with time-series data, the target's velocity and acceleration are iteratively optimized using a Kalman filter. Assuming the target's motion follows a constant acceleration model, the filter predicts the current state and updates the estimate using new observation data, generating a meta-data package which is then uploaded to the cloud.
[0065] Target attribute standardization. Cloud-based fusion of multi-source data, including satellite imagery, UAV perception data, and ground sensor information, is used to standardize the target, including attributes such as target category, location, status, and urgency, generating a structured attribute vector.
[0066] Next, the post-disaster emergency situation map is constructed and incrementally updated. Based on standardized target attribute data, this method constructs a post-disaster emergency situation map, dividing the disaster area into a grid network. Each grid node stores geographical features and dynamic disaster attributes. Geographical features include building damage level, terrain accessibility complexity, and distribution of temporary communication base stations; dynamic disaster attributes include personnel status, traffic congestion rate, and the spread trend of secondary disasters.
[0067] This method uses a situational relationship matrix. Describe the effects between grids: For situation map nodes, For grid Dynamic objectives in For nodes To the node The directed edges are stored in the situational relationship matrix. Therefore, the bidirectional edges in the situational relationship matrix reflect the mutual influence of information between nodes by describing the movement trend of dynamic targets within the nodes, giving the graph the ability to express the "potential" of the environment. At the same time, by integrating various characteristics of the post-disaster emergency environment into the nodes, the ability of the situational map to express the "state" of the battlefield is enhanced. CDC technology is used to detect local data changes, and only affected grids are recalculated, combined with version control to achieve incremental updates.
[0068] Finally, the terminal flies to the mission area based on the situation map. The path planning terminal receives the post-disaster emergency situation map from the cloud and uses the path planning algorithm to plan to the mission target area, giving priority to high-priority areas, such as areas where trapped people are gathered and base stations with interrupted signals.
[0069] In one embodiment of the present invention, the solution can be further extended technically, for example, applied to multi-drone collaborative communication reconstruction and material delivery. In large-scale emergency scenarios after disasters such as earthquakes and floods, a single drone is insufficient to cover the efficient rescue needs of a wide disaster area. It is necessary to use multiple drones to work collaboratively, share disaster situation maps in real time, dynamically allocate communication restoration and material delivery tasks, and adapt to dynamic changes in the disaster area environment, such as aftershocks and new landslides.
[0070] First, distributed situational awareness images are aggregated. Each drone exchanges local situational image segments via a mesh network, which are then aggregated in the cloud to generate a global view. A consistent hashing algorithm is used to dynamically allocate grid computing tasks, avoiding single-point bottlenecks in the cloud. Next, elastic tasks are dynamically scheduled, with the cloud dynamically prioritizing tasks based on the severity of the disaster. In the primary task area (areas with concentrated trapped personnel and hospitals), three drones are allocated for coordinated coverage, with priority given to deploying temporary communication base stations. In the secondary task area (transportation hubs and material distribution points), two drones are allocated to deliver medical supplies.
[0071] Other extended embodiments based on the present invention are not listed here, but all fall within the scope of protection of the present invention, and the present invention does not limit them.
[0072] This invention employs an edge-cloud collaborative architecture design, deploying lightweight target detection and tracking algorithms on the UAV to generate target attribute metadata in real time. The cloud integrates multi-source heterogeneous data to construct a globally consistent post-disaster environmental representation. This method significantly enhances the UAV's real-time perception of the global post-disaster situation, achieving deep fusion and efficient collaboration of multi-source data. Simultaneously, it significantly reduces edge computing resource consumption, while the cloud's elastic expansion supports large-scale UAV swarm tasks, maintaining basic decision-making functions even in communication-unstable scenarios. The disaster area environment is rasterized, with each node integrating geographical features and dynamic disaster attributes. A situational relationship matrix is used to quantify the disaster spread trend, generating a multi-dimensional situational map reflecting both "state" and "potential." This method helps improve the UAV's real-time perception of the global post-disaster situation, facilitating path planning terminals to proactively avoid high-risk areas and fly to high-priority areas based on the situational map, significantly enhancing mission safety and decision-making efficiency. Local data change detection technology is used, recalculating only grids affected by dynamic targets or environmental changes. Combined with version control management of historical situational maps, lightweight generation and push of incremental update packages are achieved.
[0073] It should be noted that the drone post-disaster operation method provided in this embodiment of the invention can be executed by a drone post-disaster operation device, or a control module within that device for executing the method of loading drone post-disaster operations. This embodiment of the invention uses the drone post-disaster operation device executing the method of loading drone post-disaster operations as an example to illustrate the drone post-disaster operation method provided in this embodiment of the invention.
[0074] Reference Figure 6 The diagram shows a structural schematic of a drone disaster recovery system provided by an embodiment of the present invention, the system including a cloud and at least one drone; The cloud platform is used to receive multi-source heterogeneous data, which comes from at least one or more of at least one UAV, satellite system, ground sensor, and network. Based on the multi-source heterogeneous data, target objects in the disaster area and their attribute information are determined. Based on the target objects in the disaster area and their attribute information, a post-disaster emergency situation map is constructed. Based on data changes in the post-disaster emergency situation map, an incremental update data package is generated. The incremental update data package is then sent to the UAV, enabling the UAV to plan post-disaster operation paths based on the updated post-disaster emergency situation map. The drone is used to collect raw data from the disaster area; process the raw data to generate standardized metadata; upload the metadata to the cloud for the cloud to build or update the post-disaster emergency situation map; receive incremental update data packets from the cloud, and plan post-disaster operation routes based on the updated post-disaster emergency situation map.
[0075] Optionally, the cloud platform is further configured to divide the disaster area into a grid network on a digital map, the grid network including grid nodes; store attribute information of a corresponding target object in each grid node of the grid network; determine the influence relationship between pairs of grid nodes based on the attribute information of the target objects corresponding to the grid nodes; and determine the grid network and the influence relationship between pairs of grid nodes as the post-disaster emergency situation map.
[0076] Optionally, the drone is also used to identify and locate at least one target object in the raw data; and to continuously track the at least one target object to obtain change data of the target object's attribute information.
[0077] To reduce cloud load and enable rapid response, this step executes lightweight algorithms on edge computing nodes such as drones to process local sensor data in real time.
[0078] In some embodiments provided by this invention, the processing of the collected data by the UAV includes: Visual object detection involves loading and running a pre-trained deep learning model (such as YOLOv5, a single-stage object detection algorithm based on deep neural networks for real-time object recognition and localization, offering high detection speed and accuracy, and suitable for deployment on edge computing platforms) to perform frame-by-frame analysis of high-resolution video streams captured by a drone's RGB camera. This model can output bounding boxes and category labels (such as "person," "vehicle," and "building") for all target objects in the image with high confidence. To further associate temporal information and maintain the consistency of target identity, multi-target tracking algorithms (such as StrongSORT) are used. This algorithm associates the target bounding boxes detected by YOLOv5 with historical trajectories, predicts motion trajectories through adaptive Kalman filtering, and uses depth appearance features for re-identification, thereby assigning a unique and unchanging ID to each target, effectively solving the problem of identity jump caused by target occlusion and cross-walking.
[0079] The edge nodes output edge metadata. Instead of uploading the massive raw video stream, the edge nodes encapsulate the processing results into lightweight metadata, which includes attribute information such as location information, speed, and acceleration information.
[0080] An embodiment of the invention provides a UAV post-disaster operation system, including a cloud platform and at least one UAV. The cloud platform receives multi-source heterogeneous data. The multi-source heterogeneous data comes from one or more of at least one UAV, satellite system, ground sensor, and network. Based on the multi-source heterogeneous data, target objects and their attribute information in the disaster area are determined. Based on the target objects and their attribute information in the disaster area, a post-disaster emergency situation map is constructed. Based on data changes in the post-disaster emergency situation map, incremental update data packets are generated. The incremental update data packets are sent to the UAVs. The UAVs are used to collect raw data from the disaster area. The raw data is processed to generate standardized meta-data packets. The meta-data packets are uploaded to the cloud platform. The system receives incremental update data packets sent from the cloud platform and plans post-disaster operation paths based on the updated post-disaster emergency situation map. The post-disaster emergency situation map constructed by multi-source data fusion significantly improves the completeness of disaster area information. The emergency situation map predicts disaster spread paths through quantitative parameter information, supports forward-looking rescue decisions, reduces UAV communication costs, and improves rescue efficiency.
[0081] Reference Figure 7 The diagram illustrates a structural block diagram of a drone-based disaster recovery device according to an embodiment of the present invention. Applied to the cloud, it may specifically include the following modules: The data receiving module 701 is used to receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of at least one UAV, satellite system, ground sensor and network. The attribute information determination module 702 is used to determine the target object in the disaster area and the attribute information of the target object based on the multi-source heterogeneous data; The situation map construction module 703 is used to construct a post-disaster emergency situation map based on the target objects in the disaster area and the attribute information of the target objects; The data packet generation module 704 is used to generate incremental update data packets based on the data changes in the post-disaster emergency situation map; The data packet delivery module 705 is used to deliver the incremental update data packet to the UAV so that the UAV can plan a post-disaster operation path based on the updated post-disaster emergency situation map.
[0082] Optionally, the situation map construction module includes: The grid division submodule is used to divide the disaster area into a grid network on a digital map, wherein the grid network includes grid nodes; The information storage submodule is used to store the attribute information of the corresponding target object at each grid node of the grid network; The influence relationship determination submodule is used to determine the influence relationship between any two grid nodes based on the attribute information of the target object corresponding to the grid node; The situation map determination submodule is used to determine the influence relationship between the grid network and the pairs of grid nodes as the post-disaster emergency situation map.
[0083] Optionally, the data packet generation module includes: The target grid determination submodule is used to determine at least one affected target grid in the grid network based on data changes in the post-disaster emergency situation map. The data packet generation submodule is used to update the post-disaster emergency situation map based on the attribute information of the target object corresponding to the at least one target grid, and generate incremental update data packets.
[0084] Optionally, the attribute information of the target object includes geographical feature attributes and dynamic disaster attributes; the geographical feature attributes include one or more of building damage degree, terrain complexity and communication base station distribution; the dynamic disaster attributes include one or more of population density, traffic congestion rate and secondary disaster spread trend.
[0085] The drone disaster relief device in this embodiment of the invention can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, in-vehicle electronic devices, wearable devices, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), televisions (TVs), ATMs, or self-service machines, etc. This embodiment of the invention does not impose specific limitations.
[0086] The unmanned aerial vehicle (UAV) disaster recovery device in this embodiment of the invention can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this embodiment of the invention does not impose specific limitations.
[0087] The unmanned aerial vehicle (UAV) post-disaster operation device provided in this embodiment of the invention can achieve Figures 1 to 5 The various processes implemented by the UAV post-disaster operation device in the method embodiment will not be described again here to avoid repetition.
[0088] An embodiment of the invention provides a drone-based post-disaster operation device, applied in the cloud, comprising: a data receiving module for receiving multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of a drone, satellite system, ground sensor, and network; an attribute information determination module for determining target objects and their attribute information in the disaster area based on the multi-source heterogeneous data; a situation map construction module for constructing a post-disaster emergency situation map based on the target objects and their attribute information in the disaster area; a data packet generation module for generating incremental update data packets based on data changes in the post-disaster emergency situation map; and a data packet distribution module for distributing the incremental update data packets to the drone, enabling the drone to plan post-disaster operation paths based on the updated post-disaster emergency situation map. The post-disaster emergency situation map constructed by multi-source data fusion significantly improves the completeness of disaster area information. The emergency situation map predicts disaster spread paths through quantified parameter information, supports proactive rescue decisions, reduces drone communication costs, and improves rescue efficiency.
[0089] This invention also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described UAV post-disaster operation method embodiments and achieve the same technical effects. To avoid repetition, they will not be described again here.
[0090] The processor is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.
[0091] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of the present invention is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.
[0092] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of the present invention.
[0093] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of the present invention.
Claims
1. A method for post-disaster operations using unmanned aerial vehicles (UAVs), characterized in that, Applied to the cloud, the method includes: Receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of the following: unmanned aerial vehicle, satellite system, ground sensor, and network; Based on the multi-source heterogeneous data, the target objects in the disaster area and the attribute information of the target objects are determined; Based on the target objects in the disaster area and the attribute information of the target objects, a post-disaster emergency situation map is constructed. Based on the data changes in the post-disaster emergency situation map, an incremental update data package is generated; The incremental update data packet is sent to the UAV so that the UAV can plan a post-disaster operation route based on the updated post-disaster emergency situation map.
2. The method for post-disaster operations using unmanned aerial vehicles according to claim 1, characterized in that, The step of constructing a post-disaster emergency situation map based on target objects in the disaster area and their attribute information includes: The disaster area is divided into a grid network on a digital map, and the grid network includes grid nodes; Each grid node in the grid network stores the attribute information of the corresponding target object; Based on the attribute information of the target object corresponding to the grid node, the influence relationship between each pair of grid nodes is determined; The influence relationship between the grid network and the pairs of grid nodes is determined as the post-disaster emergency situation map.
3. The method for post-disaster drone operations according to claim 2, characterized in that, The step of generating incremental update data packets based on data changes in the post-disaster emergency situation map includes: Based on the data changes in the post-disaster emergency situation map, at least one affected target grid in the grid network is identified; The post-disaster emergency situation map is updated based on the attribute information of the target object corresponding to the at least one target grid, and an incremental update data packet is generated.
4. The method for post-disaster unmanned aerial vehicle (UAV) operations according to claim 1, characterized in that, The attribute information of the target object includes geographical feature attributes and dynamic disaster situation attributes; The geographic feature attributes include one or more of the following: building damage level, terrain complexity, and distribution of communication base stations; The dynamic disaster attributes include one or more of the following: population density, traffic congestion rate, and the trend of secondary disaster spread.
5. A drone-based disaster recovery system, characterized in that, The system includes a cloud platform and at least one drone; The cloud is used to receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of at least one drone, satellite system, ground sensor and network; Based on the multi-source heterogeneous data, target objects in the disaster area and their attribute information are determined; based on the target objects in the disaster area and their attribute information, a post-disaster emergency situation map is constructed. Based on the data changes in the post-disaster emergency situation map, an incremental update data packet is generated; the incremental update data packet is then sent to the UAV so that the UAV can plan a post-disaster operation path based on the updated post-disaster emergency situation map. The drone is used to collect raw data from the disaster area; process the raw data to generate standardized metadata; upload the metadata to the cloud for the cloud to build or update the post-disaster emergency situation map; receive incremental update data packets from the cloud, and plan post-disaster operation routes based on the updated post-disaster emergency situation map.
6. The UAV post-disaster operation system according to claim 5, characterized in that, The cloud platform is also used to divide the disaster area into a grid network on a digital map, the grid network including grid nodes; to store the attribute information of the corresponding target object in each grid node of the grid network; to determine the influence relationship between pairs of grid nodes based on the attribute information of the target object corresponding to the grid node; and to determine the grid network and the influence relationship between pairs of grid nodes as the post-disaster emergency situation map.
7. The UAV post-disaster operation system according to claim 5, characterized in that, The drone is also used to identify and locate at least one target object in the raw data; and to continuously track the at least one target object to obtain change data of the target object's attribute information.
8. A device for unmanned aerial vehicle (UAV) post-disaster operations, characterized in that, The device, applied in the cloud, includes: A data receiving module is used to receive multi-source heterogeneous data; the multi-source heterogeneous data comes from at least one or more of at least one UAV, satellite system, ground sensor and network. The attribute information determination module is used to determine the target objects in the disaster area and the attribute information of the target objects based on the multi-source heterogeneous data. The situation map construction module is used to construct a post-disaster emergency situation map based on the target objects in the disaster area and the attribute information of the target objects; The data packet generation module is used to generate incremental update data packets based on the data changes in the post-disaster emergency situation map; The data packet delivery module is used to deliver the incremental update data packet to the UAV so that the UAV can plan a post-disaster operation path based on the updated post-disaster emergency situation map.
9. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the UAV disaster recovery method as described in claims 1-4.
10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the UAV disaster recovery operation method as described in claims 1-4.