A method, system, equipment and medium for unmanned aerial vehicle (UAV) inspection route planning

Abstract

This application provides a method, system, device, and medium for planning UAV inspection routes, relating to the field of UAV inspection route planning technology. The method includes: dividing a disaster area into multiple sub-regions based on its regional characteristics, with each sub-region corresponding to a first emergency coefficient; predicting the disaster evolution trend of each sub-region within a preset time period after the current moment based on the real-time status of each sub-region, and generating a second emergency coefficient for each sub-region based on the disaster evolution trend; generating a target emergency coefficient for each sub-region by combining the first and second emergency coefficients; generating an initial inspection route for each UAV by combining the location and remaining battery power of each UAV and the target emergency coefficient of each sub-region; and adjusting the initial inspection route and the target inspection route based on the correlation coefficient between each UAV. The technical effect of this application is: reducing the deviation between the inspection route and the actual emergency needs, and improving the execution effect of the inspection task.

Description

Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) inspection route planning technology, specifically to a UAV inspection route planning method, system, equipment, and medium. Background Technology

[0002] With the widespread application of drone technology in disaster emergency management, utilizing multiple drones to collaboratively conduct disaster area inspections has become an important technical means for post-disaster situational awareness and emergency decision-making. After a disaster occurs, how to efficiently plan inspection routes for multiple drones to quickly and comprehensively obtain disaster information is a key technical problem that urgently needs to be solved in the field of disaster emergency inspections.

[0003] Currently, traditional methods primarily plan inspection routes for multiple drones based on static factors such as shortest distance, least flight time, and lowest energy consumption. While these methods can achieve some coverage of disaster areas, they only consider the mathematical optimization objective of path planning and ignore the differences in urgency and dynamic changes in different areas under disaster emergency scenarios. This leads to discrepancies between the generated inspection routes and actual emergency needs, affecting the effectiveness of the inspection mission. Summary of the Invention

[0004] This application provides a method, system, equipment, and medium for planning unmanned aerial vehicle (UAV) inspection routes, which reduces the deviation between inspection routes and actual emergency needs and improves the execution effect of inspection tasks.

[0005] In a first aspect, this application provides a method for planning unmanned aerial vehicle (UAV) inspection routes, applied to a main control platform. The main control platform integrates multiple sub-platforms through an embedded framework. Each sub-platform corresponds to a different business module and has an independent data display interface. The method includes: In response to a user clicking on a disaster area on the 3D map of the main control platform, each sub-platform displays target parameters through a display interface based on its corresponding business module. These target parameters include: regional characteristics and real-time status of the disaster area; one or more of the following: the location, remaining battery power, and inspection routes of each drone; obtaining the regional characteristics of the disaster area; dividing the disaster area into multiple sub-regions based on the regional characteristics, with each sub-region corresponding to a first emergency coefficient; obtaining the real-time status of each sub-region; predicting the disaster evolution trend of each sub-region within a preset time period after the current moment based on the real-time status; generating a second emergency coefficient for each sub-region based on the disaster evolution trend; combining the first and second emergency coefficients to generate a target emergency coefficient for each sub-region; obtaining the location and remaining battery power of each drone; combining the location and remaining battery power of each drone with the target emergency coefficient of each sub-region to generate an initial inspection route for each drone; obtaining the correlation coefficient between each drone; adjusting the initial inspection route based on the correlation coefficient to generate a target inspection route for each drone; wherein the correlation coefficient is used to characterize the communication link strength between drones during the execution of the inspection task.

[0006] By adopting the above technical solution, multiple sub-platforms can synchronously display target parameters based on the disaster area coordinates clicked by the user on the 3D map of the main control platform through browser extensions. This solves the problem of data inconsistency among various business modules in traditional emergency command systems, improving the accuracy and real-time nature of decision-making information. Simultaneously, by combining the regional characteristics of the disaster area to generate a first urgency coefficient and predicting the disaster evolution trend based on real-time status to generate a second urgency coefficient, and then comprehensively forming the target urgency coefficient for each sub-region, the UAV inspection route planning can dynamically reflect the urgency and development trend of the disaster, overcoming the limitations of traditional methods that only optimize based on static factors such as distance and energy consumption. Furthermore, by obtaining the correlation coefficient between UAVs to adjust the initial inspection route and generate the target inspection route, the communication and coordination needs of the UAV swarm during the execution of inspection tasks are fully considered, ensuring the effectiveness of multi-UAV collaboration, reducing the deviation between the inspection route and the actual emergency needs, and improving the execution effect of the inspection task.

[0007] Secondly, this application provides a UAV inspection route planning system, the system comprising: a response module, a first acquisition module, a second acquisition module, a combination module, a third acquisition module, and an adjustment module; wherein, The response module is used to respond to the user's click on the disaster area on the 3D map of the main control platform. Each sub-platform displays target parameters through a display interface based on its corresponding business module. The target parameters include the regional characteristics and real-time status of the disaster area; one or more of the following: the location, remaining power, and inspection route of each drone. The first acquisition module is used to acquire the regional characteristics of the disaster area and divide the disaster area into multiple sub-regions based on the regional characteristics. Each sub-region corresponds to a first urgency coefficient. The second acquisition module is used to acquire the real-time status of each sub-region and predict the disaster evolution trend of each sub-region within a preset time period after the current moment based on the real-time status. The first emergency coefficient and the second emergency coefficient are combined to generate a target emergency coefficient for each sub-region. The third acquisition module is used to acquire the location and remaining power of each UAV, and generate an initial inspection route for each UAV by combining the location and remaining power of each UAV and the target emergency coefficient of each sub-region. The adjustment module is used to acquire the correlation coefficient between each UAV, and adjust the initial inspection route according to the correlation coefficient to generate a target inspection route for each UAV. The correlation coefficient is used to characterize the communication link connectivity strength between each UAV during the execution of the inspection task.

[0008] Thirdly, this application provides an electronic device that adopts the following technical solution: it includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory so that the electronic device executes a computer program such as any of the above-described UAV inspection route planning methods.

[0009] Fourthly, this application provides a computer-readable storage medium that stores a computer program capable of being loaded by a processor and executing any of the above-mentioned UAV inspection route planning methods.

[0010] In summary, this application includes at least one of the following beneficial technical effects: By leveraging browser extensions, multiple sub-platforms can synchronously display target parameters based on the disaster area coordinates clicked by the user on the main control platform's 3D map. This resolves the data inconsistency issue among various business modules in traditional emergency command systems, improving the accuracy and real-time nature of decision-making information. Simultaneously, by generating a first urgency coefficient based on the regional characteristics of the disaster area and a second urgency coefficient based on real-time disaster evolution trends, a comprehensive target urgency coefficient for each sub-region is formed. This allows UAV inspection route planning to dynamically reflect the urgency and development trend of the disaster, overcoming the limitations of traditional methods that optimize solely based on static factors such as distance and energy consumption. Furthermore, by obtaining the correlation coefficients between UAVs to adjust the initial inspection route and generate the target inspection route, the communication and coordination needs of the UAV swarm during inspection missions are fully considered, ensuring the effectiveness of multi-UAV collaboration, reducing deviations between inspection routes and actual emergency needs, and improving the execution effect of inspection missions. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating a drone inspection route planning method provided in an embodiment of this application; Figure 2 This is a schematic diagram of a drone disaster inspection and command system provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an unmanned aerial vehicle (UAV) inspection route planning system provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0012] Explanation of reference numerals in the attached figures: 1000, electronic device; 1001, processor; 1002, communication bus; 1003, user interface; 1004, network interface; 1005, memory. Detailed Implementation

[0013] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0014] In the description of the embodiments in this application, words such as "illustrative," "for example," or "for example" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "illustrative," "for example," or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Rather, the use of words such as "illustrative," "for example," or "for example" is intended to present the relevant concepts in a specific manner.

[0015] Figure 1 This is a flowchart illustrating a drone inspection route planning method provided in an embodiment of this application. Figure 1 As shown, the method includes S101-S106: S101, in response to the user's click on the disaster area on the 3D map of the main control platform, each sub-platform displays the target parameters through the display interface based on the corresponding business module. The target parameters include the regional characteristics and real-time status of the disaster area; the location, remaining power, and inspection route of each drone, or one or more of these.

[0016] In practical scenarios of drone inspection route planning, the main control platform serves as a unified operational entry point for disaster emergency command personnel, undertaking multi-dimensional business functions such as disaster situation awareness, drone dispatch and management, and inspection data aggregation. Due to significant differences in the data sources, update frequencies, and display formats involved in these functions, implementing all functions on a single page would face problems such as excessive coupling between modules, difficulty in independent iteration, and increased system maintenance costs. Therefore, the main control platform adopts an embedded framework integration architecture, that is, embedding multiple sub-platforms with independent business logic into the same main control page through HTML embedded frames (iframes). Each sub-platform corresponds to different business modules and has an independent data display interface.

[0017] In this embodiment, the sub-platforms include, but are not limited to, a disaster situation sub-platform, a drone status sub-platform, and an inspection route sub-platform. The disaster situation sub-platform displays the regional characteristics and real-time status of the disaster area; the drone status sub-platform displays the location and remaining battery power of each drone; and the inspection route sub-platform displays the inspection routes and task progress of each drone. Each sub-platform is deployed on different servers or different service ports; therefore, there is cross-domain isolation between the embedded frameworks corresponding to each sub-platform, preventing direct data interaction via conventional script calls.

[0018] Under the aforementioned architecture, when emergency command personnel need to plan drone patrol routes for a specific disaster area, they first need to click on the target disaster area on the main control platform's 3D map. This allows each sub-platform to retrieve and display target parameters related to the disaster area from its respective connected backend service based on the geographical location information of the same disaster area. However, because each sub-platform runs in an isolated embedded framework, the coordinate information generated by clicking on the main control platform's 3D map cannot be directly transmitted to each sub-platform. Furthermore, the location parameters carried by each sub-platform when initiating data requests to its respective backend service are still default or historical values. This results in inconsistencies between the target parameters displayed by each sub-platform and the disaster area actually clicked by the user, leading to a disconnect between coordinate synchronization and data requests.

[0019] To address the aforementioned issues, this embodiment utilizes a browser extension deployed in the browser of the main control platform to achieve a complete closed loop from the acquisition of 3D map click coordinates to the synchronous display of target parameters on each sub-platform. The specific process is as follows.

[0020] The browser extension refers to a feature enhancement program built based on the browser extension development specification. It runs at the browser level rather than the page level, thus possessing cross-domain access permissions and the ability to simultaneously access and manipulate page content in embedded frames of different origins on the main control platform and various sub-platforms. In this embodiment, the browser extension comprises three core components: a content script, an injection script, and a backend service module. The content script is a script program automatically injected by the browser extension into the embedded frames of the main control platform page and various sub-platforms, running in a document object model environment shared with the page, capable of accessing page elements and communicating with the backend service module. The injection script is a script program dynamically loaded by the content script into the main execution environment of the page, capable of directly accessing 3D map instance objects and global variables on the page. The backend service module is an independent service script running in the browser background, responsible for receiving messages sent by the content script and coordinating the distribution of instructions to various embedded frames.

[0021] Once the browser extension loads along with the main control platform page, the content script is automatically injected into all embedded frames of the main control platform page and each sub-platform. On the main control platform page, the content script further loads the injected script into the main execution environment to gain access to the 3D map instance object. The 3D map instance object refers to the core running instance generated after the 3D map component in the main control platform is initialized, encapsulating functional interfaces such as map rendering, camera control, coordinate picking, and event listening. After the injected script is loaded, it first waits for the 3D map instance object to complete initialization through a polling mechanism. Specifically, it continuously checks at preset time intervals whether the 3D map instance object is ready in the main execution environment. Only after detecting that the instance object and its scene rendering components have completed initialization does it execute the subsequent event listening logic, thus avoiding listening operation failures due to the 3D map instance object not yet being loaded.

[0022] Once the 3D map instance object is ready, a script is injected to create a screen space event handler and bind it to the 3D map's scene canvas. This event handler listens for left-click actions performed by the user on the 3D map. When the user clicks on a disaster area, the event handler captures the screen pixel coordinates corresponding to the click and calls the coordinate picking method provided by the 3D map instance object to convert the screen pixel coordinates into Cartesian coordinates in 3D space. Subsequently, a coordinate transformation utility method is called to convert the Cartesian coordinates into geographic coordinates including longitude and latitude, thus obtaining the coordinate information of the disaster area.

[0023] To ensure that coordinate information can be reliably obtained by the content script, the injection script employs a dual exposure mechanism to transmit coordinate information: on the one hand, the injection script creates a custom event and encapsulates the coordinate information in the event's data payload, triggering the custom event to the page document object; on the other hand, the injection script writes the coordinate information along with the current timestamp into a global variable in the page's main execution environment. The content script correspondingly employs a dual acquisition mechanism to receive coordinate information: on the one hand, the content script listens for the aforementioned custom event and directly extracts the coordinate information from the event's data payload when the event is triggered; on the other hand, the content script continuously checks the global variable at a preset polling interval. When it detects coordinate information in the global variable and the difference between its timestamp and the current time is within a preset valid time window, it extracts the coordinate information and clears the global variable. The valid time window is used to filter out expired historical coordinate data, preventing the content script from reading old coordinates not generated by the current click operation.

[0024] After obtaining the coordinates of the disaster area, the content script sends the coordinates to the backend service module via the messaging interface provided by the browser extension. The messaging mechanism refers to a dedicated communication channel established between the content script and the backend service module by the browser extension. This channel is independent of the page's own script execution environment and is not subject to cross-domain isolation restrictions. Upon receiving the coordinates, the backend service module first verifies the integrity and validity of the longitude and latitude parameters. If the verification passes, the backend service module injects the coordinate processing script into all embedded frames in the current tab that match the target domain name through the script injection interface provided by the browser extension.

[0025] After the coordinate processing script is injected into the target embedded frame, it performs two core operations: local storage of coordinate information and data request interception. Regarding local storage of coordinate information, the script writes the coordinate information of the disaster area into a designated global variable within the execution environment of the embedded frame's page. This global variable constitutes a shared storage area accessible to all sub-platforms, allowing business scripts on each sub-platform to retrieve the latest disaster area coordinate information when subsequently initiating data requests. Regarding data request interception, the script intercepts two types of core network request initiation methods within the embedded frame. Specifically, it rewrites the request initiation method of the XMLHttpRequest object and the global fetch function in the page execution environment, ensuring that all subsequent data requests from each sub-platform to its respective backend service via these two methods are processed through the interception logic.

[0026] Through the above process, in the data requests that each sub-platform sends to its respective backend service after the user clicks on the disaster area, the location-related parameters have been uniformly replaced with the coordinate information of the disaster area. Each backend service returns the corresponding business data based on the consistent disaster area coordinate information. After receiving the returned data, each sub-platform displays the target parameters corresponding to the disaster area through its own independent display interface.

[0027] Based on the above embodiments, as an optional implementation, in S101, in response to the user clicking on the disaster area on the 3D map of the main control platform, each sub-platform displays the target parameters through the display interface based on the corresponding business module, specifically including S11-S16: S11 uses a browser extension to monitor clicks on the 3D map component in the main control platform and obtains the coordinate information of the disaster area.

[0028] The browser extension is injected into the main control platform page via content scripts to listen for clicks on the 3D map component within the main control platform. When a user clicks on a disaster area on the 3D map, the browser extension extracts the coordinate information of the disaster area from the event object of the 3D map component. This coordinate information includes the latitude and longitude range and center coordinates of the disaster area clicked by the user.

[0029] S12 sends the coordinate information to each sub-platform through a message passing mechanism and stores the coordinate information in a shared storage area.

[0030] After acquiring the coordinate information, the browser extension distributes it to each sub-platform via a messaging mechanism. This messaging mechanism refers to the cross-page communication channel provided by the browser extension. The browser extension's background script receives the coordinate information from the main control platform's content script and then forwards it to the content scripts injected into the pages of each sub-platform. Simultaneously, the browser extension stores the coordinate information in a shared storage area. This shared storage area is a unified storage space provided by the browser extension that is readable and writable by all content scripts. It is used to maintain the consistency of coordinate information across sub-platforms, ensuring that any subsequent data request from any sub-platform will receive the latest disaster area coordinate information.

[0031] S13 intercepts data requests sent from various sub-platforms to the server via browser extensions.

[0032] During normal operation, each sub-platform sends data requests to its respective server to obtain business data based on page initialization logic or user operations. The browser extension intercepts the data requests sent by each sub-platform to the server through the network request interception interface, preventing the original request from being sent directly, so that the request parameters can be modified later.

[0033] S14, based on the preset parameter mapping configuration, identify the location-related parameters in the data request.

[0034] The browser extension identifies location-related parameters in intercepted data requests based on a pre-defined parameter mapping configuration. This parameter mapping configuration refers to the data request parameter parsing rules pre-configured for each sub-platform, defining the names and formats of parameters representing geographic locations in data requests from each sub-platform. For example, one sub-platform might use latitude and longitude fields, while another might use a region code field. The browser extension matches the corresponding parameter mapping configuration to the sub-platform to which the current data request belongs, thereby accurately locating the parameters carrying location information within the data request.

[0035] S15: Replace the location-related parameters identified in the data request with coordinate information to generate a modified data request.

[0036] The browser extension reads coordinate information from the shared storage area, uses this coordinate information to replace the original values ​​of location-related parameters, and generates a modified data request. If the format of the location parameters defined in the parameter mapping configuration is inconsistent with the original format of the coordinate information, the browser extension will also convert the coordinate information according to the preset conversion rules in the parameter mapping configuration before performing the replacement.

[0037] S16 sends the modified data request to the server, enabling each sub-platform to obtain and display the corresponding target parameters based on the coordinate information of the disaster area; among them, the browser extension has cross-domain access permissions, which can simultaneously access and manipulate the content of sub-platforms from different sources.

[0038] The browser extension sends the modified data request to the server. After receiving the modified data request, the server returns business data corresponding to the coordinates of the disaster area. Each sub-platform displays the corresponding target parameters through its own display interface based on the returned data.

[0039] Users only need to perform a single click on the 3D map of the main control platform, and each sub-platform can automatically switch to the data view of the corresponding disaster area and display the target parameters corresponding to their respective business modules. Users do not need to manually operate each sub-platform one by one, which ensures the consistency of the data displayed by each sub-platform in terms of geographical location and improves the efficiency of multi-platform collaborative operation in disaster emergency scenarios.

[0040] S102, Obtain the regional characteristics of the disaster area, and divide the disaster area into multiple sub-regions based on the regional characteristics. Each sub-region corresponds to a first emergency coefficient.

[0041] Specifically, this embodiment first performs equidistant gridding on the disaster area, dividing it into multiple basic grid units with a preset grid side length. The grid side length is determined based on the total area of ​​the disaster area and the sensor coverage of the UAV. Then, for each basic grid unit, regional feature values ​​within its corresponding spatial range are extracted to form a feature vector for that basic grid unit. A feature vector is a multidimensional numerical sequence composed of values ​​from each feature dimension arranged in a fixed order; in this embodiment, it is a five-dimensional vector. Before clustering, this embodiment normalizes the feature vectors of all basic grid units, linearly mapping the values ​​of each feature dimension to the range of zero to one.

[0042] After normalization, this embodiment employs a density-based spatial clustering method to cluster the normalized feature vectors of all basic grid cells. The density-based method is chosen over centroid-based clustering because disaster areas typically exhibit irregular, continuous spatial distribution patterns, such as water areas extending along rivers or commercial land areas distributed along road networks. Density-based clustering can identify high-density connected regions of arbitrary shapes, making it more suitable for these distribution characteristics. After clustering, basic grid cells with similar regional characteristics and spatially adjacent to each other are grouped into the same cluster, each cluster forming a sub-region, thus dividing the disaster area into multiple sub-regions. For noise points generated during clustering—isolated basic grid cells that do not belong to any cluster—this embodiment assigns them to the sub-region closest to their geographical location based on spatial proximity, ensuring complete coverage of the disaster area.

[0043] After completing the sub-region division, this embodiment determines the corresponding first urgency coefficient for each sub-region. The first urgency coefficient is a numerical index obtained by quantifying the regional characteristics of the sub-region. It is used to characterize the urgency of the sub-region relative to other sub-regions in disaster emergency inspection. The larger the value, the more priority the sub-region needs to be allocated inspection resources. The calculation process of the first urgency coefficient is as follows: First, the characteristic statistical values ​​of each sub-region in five characteristic dimensions are determined. Among them, the terrain elevation dimension calculates the standard deviation of the average terrain elevation value of each basic grid unit in the sub-region and normalizes it into a terrain complexity score. The more complex the terrain, the higher the risk of disaster damage and the greater the difficulty of ground rescue. The land use type dimension maps the dominant land use type of the sub-region to a land use risk score according to the preset land use type risk comparison table. The risk scores corresponding to residential land and commercial land are higher than those of agricultural land and forest land. The building distribution density dimension and the population distribution density dimension take the mean of the corresponding values ​​of each basic grid unit in the sub-region. The critical infrastructure dimension takes the sum of the number of critical infrastructures in the sub-region. The scores for the five dimensions were then normalized to a range of zero to one. A weighted summation method was used to calculate the first urgency coefficient, which involves multiplying the normalized scores for each of the five dimensions by their respective weight coefficients and then summing the results. The weight coefficients reflect the relative importance of each feature dimension in the urgency assessment. The population density dimension has the highest weight coefficient because it is directly related to the safety of the affected people. The critical infrastructure dimension has the second highest weight coefficient because its operational status affects the efficiency of post-disaster relief and reconstruction. The weight coefficients for the building density dimension, land use risk score dimension, and terrain complexity score dimension decrease in that order, respectively reflecting the urgency of the inspection from the perspectives of physical damage degree, functional area sensitivity, and ground rescue accessibility.

[0044] Through the above process, the disaster area is divided into multiple sub-regions, and each sub-region corresponds to a first emergency coefficient.

[0045] Based on the above embodiments, as an optional implementation, in S102, the regional characteristics include topographic elevation data, land use type data, building distribution density data, population distribution density data, and critical infrastructure distribution data. According to these regional characteristics, the disaster area is divided into multiple sub-regions, specifically including S21-S24: S21. Based on topographic elevation data, extract the natural geographical boundaries within the disaster area, and use these natural geographical boundaries as initial boundaries to divide the disaster area into multiple initial blocks. Among these, the natural geographical boundaries include one or more of the following: ridgelines, water system boundaries, and main transportation lines.

[0046] Natural geographical boundaries refer to natural regional dividing lines formed by topography or artificial structures, including one or more of ridgelines, water system boundaries, and main transportation lines. Ridgelines often exhibit different slope aspects and water flow directions on either side, resulting in different disaster characteristics in floods, landslides, and other calamities. Water system boundaries, such as rivers and lakes, naturally separate the disaster propagation paths between their banks. Main transportation lines, such as highways and railways, act as barriers to disaster spread due to differences in roadbed elevation. This embodiment extracts ridgelines and water system boundaries through slope aspect and water flow analysis of topographic elevation data, and extracts main transportation lines by combining this with transportation network data of the disaster area. These natural geographical boundaries are then used as initial boundaries to divide the disaster area into multiple initial blocks. Using natural geographical boundaries as boundaries ensures that each initial block has relatively consistent topographical conditions.

[0047] S22, combining land use type data and building distribution density data, calculate the homogeneity of land use type and building distribution density in each initial block respectively, and split the initial blocks with land use type homogeneity lower than the preset homogeneity threshold or building distribution density homogeneity lower than the preset homogeneity threshold into multiple sub-blocks along the abrupt change line of land use type or building distribution density.

[0048] Specifically, this embodiment combines land use type data and building distribution density data to calculate the homogeneity of land use type and building distribution density within each initial block. The land use type homogeneity refers to the area proportion of the largest land use type within the initial block; a higher value indicates a more uniform and homogeneous land use type within the block. The building distribution density homogeneity refers to the normalized reciprocal of the standard deviation of the building distribution density within the initial block; a higher value indicates a more uniform building distribution within the block. When the land use type homogeneity or building distribution density homogeneity of an initial block is lower than a preset homogeneity threshold, it indicates significant characteristic differences within the block. This embodiment detects the spatial location lines where the land use type or building distribution density within the block changes drastically, uses these lines as abrupt change lines, and divides the initial block into multiple sub-blocks along these change lines, making the internal characteristics of each sub-block more homogeneous.

[0049] S23, calculate the regional feature similarity between adjacent initial blocks and / or subdivided blocks, merge adjacent blocks with regional feature similarity greater than a preset similarity threshold, and generate multiple sub-regions.

[0050] After the division and splitting in steps S21 and S22, there may be some adjacent initial blocks or subdivided blocks in the disaster area that, although separated by natural geographical boundaries or abrupt change lines, have very similar actual regional characteristics. Keeping these similar blocks separate would lead to an excessive number of sub-regions, increasing the computational complexity of subsequent inspection route planning. Therefore, this embodiment calculates the regional feature similarity between adjacent initial blocks and subdivided blocks. The regional feature similarity refers to the comprehensive similarity between two adjacent blocks in terms of regional characteristics such as topographic elevation, land use type, and building density. When the regional feature similarity between two adjacent blocks is greater than a preset similarity threshold, this embodiment merges them into one block. The merging process is iteratively executed until no more adjacent block pairs satisfying the merging conditions exist, ultimately generating multiple sub-regions. By splitting and then merging, both the homogeneity of features within each sub-region and the excessive expansion of the number of sub-regions are ensured.

[0051] S24. Based on the population distribution density data and critical infrastructure distribution data of each sub-region, generate the first emergency coefficient for each sub-region; wherein, the first emergency coefficient is positively correlated with the population distribution density and the quantity and type of critical infrastructure in the sub-region, and critical infrastructure includes one or more of medical institutions, schools, transportation hubs and energy facilities.

[0052] The first urgency coefficient measures the level of urgency required for a sub-region to address a disaster, based on the exposure levels of people and critical facilities within that sub-region. Critical infrastructure refers to facilities whose damage during a disaster would significantly impact public safety and rescue efforts, including one or more of medical institutions, schools, transportation hubs, and energy facilities. The first urgency coefficient is positively correlated with population density within the sub-region; denser populations indicate a higher potential risk of casualties. It is also positively correlated with the quantity and type of critical infrastructure; areas with more and more critical infrastructure have a higher priority in disaster response. In this embodiment, the first urgency coefficient for each sub-region is obtained by weighted summing of the normalized population density and the weighted normalized quantity of critical infrastructure.

[0053] S103, obtain the real-time status of each sub-region, predict the disaster evolution trend of each sub-region within a preset time period after the current time based on the real-time status, and generate the second emergency coefficient of each sub-region based on the disaster evolution trend.

[0054] Specifically, this embodiment first acquires the real-time status of each sub-region. The real-time status refers to dynamic data describing the disaster development and environmental conditions of each sub-region at the current moment, including disaster intensity, disaster spread rate, and meteorological environmental conditions. Disaster intensity is a quantitative value of the degree of disaster impact in each sub-region at the current moment, using different metrics depending on the type of disaster; for example, floods are characterized by inundation depth, earthquakes by peak ground acceleration, and fires by the percentage of burned area. The disaster spread rate refers to the change in the disaster impact area of ​​each sub-region per unit time, reflecting the speed of disaster spread within that sub-region. Meteorological environmental conditions include meteorological parameters such as wind speed, wind direction, rainfall, and temperature in each sub-region at the current moment. These meteorological parameters are collected in real-time by meteorological monitoring stations deployed within the disaster area. The aforementioned real-time status data is acquired and aggregated in real-time from the corresponding monitoring sensor network and meteorological data service interface by the backend service connected to the disaster situation sub-platform.

[0055] After obtaining the real-time status of each sub-region, this embodiment predicts the disaster evolution trend of each sub-region within a preset time period after the current moment based on the real-time status. The preset time period refers to a fixed time length extending into the future from the current moment. Its value is determined based on the estimated time required for the UAV to complete a full round of inspection. In this embodiment, it is set to two hours, that is, predicting the disaster evolution trend of each sub-region within the next two hours. The disaster evolution trend refers to the predicted trend of the disaster intensity of each sub-region changing over time within the preset time period, including three indicators: the predicted peak value of the disaster intensity, the estimated time to reach the predicted peak value, and the predicted disaster intensity at the end of the preset time period.

[0056] The disaster evolution trend prediction process employs a trend prediction method based on time-series data. This embodiment first retrieves the disaster intensity time series of each sub-region within a preset backtracking period prior to the current time from the historical database of the backend service. In this embodiment, the preset backtracking period is set to six hours, meaning the historical disaster intensity value series of each sub-region sampled at fixed time intervals over the past six hours is obtained. Subsequently, this embodiment uses the historical disaster intensity value series of each sub-region and the corresponding meteorological environmental condition time series as input features, inputting them into a pre-trained disaster evolution prediction model. The disaster evolution prediction model is a time-series prediction model built based on a Long Short-Term Memory (LSTM) network. This model learns the correlation between historical disaster intensity change patterns and meteorological environmental conditions, outputting the predicted disaster intensity values ​​for each sub-region at each prediction time point within the preset period. The LSTM network is chosen as the basic architecture of the prediction model because the time series of disaster intensity typically contains both short-term fluctuations and long-term trends. The LSTM network, through its internal gating mechanism, can effectively capture long-range dependencies in the time series, making it suitable for modeling the aforementioned mixed-time-scale evolution patterns. After the disaster evolution prediction model outputs the disaster intensity prediction value at each prediction time point, this embodiment extracts the predicted peak value of the disaster intensity, the expected time to reach the predicted peak value, and the predicted disaster intensity at the end of the preset time period to form the disaster evolution trend of each sub-region.

[0057] After obtaining the disaster evolution trend of each sub-region, this embodiment generates a second urgency coefficient for each sub-region based on the disaster evolution trend. The second urgency coefficient is a numerical index obtained by quantifying the disaster evolution trend of a sub-region. It is used to characterize the urgency of the disaster dynamic development level of the sub-region within a preset time period. The larger the value, the more likely the disaster situation of the sub-region will deteriorate in the future, and the more priority should be given to arranging inspections.

[0058] Based on the above embodiments, as an optional implementation, in S103, the real-time status includes disaster intensity data, disaster spread rate data, meteorological environment data, and disaster severity data for each sub-region. Based on the real-time status, the disaster evolution trend of each sub-region within a preset time period after the current moment is predicted. Based on the disaster evolution trend, the second emergency coefficient for each sub-region is generated, specifically including S31-S34: S31. Based on the disaster intensity data and disaster spread rate data of each sub-region, construct a single-domain disaster evolution model for each sub-region, and generate the initial evolution trend of each sub-region based on the single-domain disaster evolution model.

[0059] A single-domain disaster evolution model is a time-series prediction model that considers only the disaster state of a single sub-region and does not involve the mutual influence between adjacent sub-regions. The disaster intensity data refers to a quantitative indicator of the severity of the disaster in each sub-region at the current moment, such as water level in a flood or peak ground acceleration in an earthquake. The disaster spread rate data refers to the rate at which the disaster's impact area changes over time within each sub-region, such as the expansion rate of flood-inundated area or the growth rate of fire-damaged area. This embodiment uses historical time series disaster intensity data as input and disaster spread rate data as a rate of change constraint. It predicts the disaster intensity change curve of each sub-region over a future period under conditions unaffected by external propagation through time-series extrapolation. This curve represents the initial evolution trend. The initial evolution trend reflects the natural evolution trajectory of the disaster in each sub-region under isolated conditions.

[0060] S32. Based on the regional characteristics of each sub-region and the geographical connectivity between adjacent sub-regions, and combined with the meteorological environmental data of each sub-region, a disaster propagation impact matrix is ​​constructed between the sub-regions. The elements in the disaster propagation impact matrix are used to characterize the impact weight of the disaster state change of a sub-region on the disaster evolution of adjacent sub-regions. The impact weight is positively correlated with the degree of geographical connectivity between adjacent sub-regions and the strength of the disaster propagation driving factors indicated in the meteorological environmental data.

[0061] The disaster propagation impact matrix is ​​a square matrix with each sub-region as its row and column index. The elements in the matrix represent the weight of the impact of a sub-region's disaster state change on the disaster evolution of adjacent sub-regions. In this embodiment, the basic values ​​of the impact weights are determined based on the regional characteristics of each sub-region obtained in step S102 and the geographical connectivity between adjacent sub-regions. Geographical connectivity refers to the existence of geographical channels between two adjacent sub-regions that allow the propagation of disaster factors, such as river system connectivity, continuous terrain slopes, or road networks. The higher the degree of geographical connectivity, the greater the likelihood of a disaster propagating from one sub-region to an adjacent sub-region. Based on this, this embodiment adjusts the impact weights by incorporating meteorological environmental data for each sub-region. The meteorological environmental data includes meteorological elements such as wind speed and direction, rainfall, and temperature and humidity. These meteorological elements are important factors driving the cross-regional propagation of disasters. For example, strong winds accelerate the spread of fires downwind to sub-regions, and continuous heavy rainfall exacerbates the spread of upstream floods to downstream sub-regions. The impact weights are positively correlated with the degree of geographical connectivity and the strength of the disaster propagation driving factors indicated in the meteorological environmental data. For non-adjacent sub-regions, the corresponding elements in the disaster propagation impact matrix are set to zero.

[0062] S33, based on the disaster propagation impact matrix, perform cross-regional coupling correction on the initial evolution trend of each sub-region to generate the disaster evolution trend of each sub-region within a preset time period after the current time; wherein, the cross-regional coupling correction includes: for the target sub-region, multiply the initial evolution trend of each sub-region adjacent to the target sub-region by the corresponding impact weight in the disaster propagation impact matrix to obtain the propagation evolution amount of each adjacent sub-region to the target sub-region, and superimpose each propagation evolution amount to the initial evolution trend of the target sub-region to obtain the disaster evolution trend of the target sub-region.

[0063] Cross-regional coupling correction refers to incorporating the disaster propagation effects between adjacent sub-regions into the evolution prediction, so that the prediction results are no longer limited to the isolated evolution of a single sub-region, but reflect the comprehensive evolution trajectory after the mutual propagation and superposition of disasters between regions. Specifically, for any target sub-region, this embodiment extracts the influence weights corresponding to each adjacent sub-region from the disaster propagation influence matrix. The initial evolution trends of each adjacent sub-region are multiplied by their corresponding influence weights to obtain the propagation evolution amount of each adjacent sub-region to the target sub-region. This propagation evolution amount characterizes the increment contributed by the disaster state changes of adjacent sub-regions to the disaster evolution of the target sub-region through the propagation effect. The propagation evolution amounts are then superimposed onto the initial evolution trend of the target sub-region itself to obtain the disaster evolution trend of the target sub-region after cross-regional coupling correction. After performing the above correction on all sub-regions, the disaster evolution trend of each sub-region simultaneously includes the natural evolution of its own disaster and the superimposed influence of disaster propagation from adjacent sub-regions.

[0064] S34. Based on the disaster evolution trend of each sub-region, extract the predicted disaster peak intensity and predicted disaster deterioration rate of each sub-region within a preset time period, and combine them with the disaster severity data of each sub-region to generate the second emergency coefficient of each sub-region; wherein, the second emergency coefficient is positively correlated with the predicted disaster peak intensity, the predicted disaster deterioration rate and the disaster severity.

[0065] This embodiment extracts the predicted peak disaster intensity and predicted disaster deterioration rate within a preset time period from the disaster evolution trend curve of each sub-region. The predicted peak disaster intensity refers to the maximum disaster intensity value reached by the disaster evolution trend curve within the preset time period, reflecting the most severe disaster state that the sub-region may face in the future. The predicted disaster deterioration rate refers to the average rate of change of the disaster intensity during the rising phase of the disaster evolution trend curve within the preset time period, reflecting the speed at which the disaster deteriorates in the sub-region. This embodiment generates a second emergency coefficient by combining the disaster severity data of each sub-region, where the disaster severity data refers to the actual degree of loss currently suffered by each sub-region. The second emergency coefficient is positively correlated with the predicted peak disaster intensity, the predicted disaster deterioration rate, and the disaster severity; that is, the higher the predicted peak, the faster the deterioration, and the more severe the current disaster in the sub-region, the higher the second emergency coefficient, and the stronger the urgency of inspection.

[0066] S104, combining the first and second urgency coefficients, generates the target urgency coefficient for each sub-region.

[0067] Specifically, this embodiment uses an adaptive weighted fusion method to generate the target urgency coefficient. The target urgency coefficient is a fused numerical index obtained by comprehensively considering the urgency of the static regional characteristics of the sub-region and the urgency of the dynamic disaster evolution. It is used as the sole decision-making basis for the inspection priority and resource allocation ratio of each sub-region in subsequent inspection route planning. The calculation of the target urgency coefficient is not a simple fixed-proportion weighted sum of the first urgency coefficient and the second urgency coefficient. This is because the relative importance of static and dynamic factors in disaster emergency scenarios changes with the actual development stage of the disaster. In the initial stage of the disaster, the disaster intensity of each sub-region is generally low and the evolution trend is not yet clear. At this time, the inherent risk level reflected by the regional characteristics has a greater guiding significance for inspection priority, and the first urgency coefficient should have a higher fusion weight. However, in the rapid development stage of the disaster, the disaster intensity of some sub-regions is changing drastically. At this time, the impact of the disaster evolution trend on the urgency of inspection is significantly enhanced, and the second urgency coefficient should have a higher fusion weight.

[0068] To achieve the aforementioned adaptive weighting, this embodiment introduces a dynamic weight adjustment factor. The dynamic weight adjustment factor is a dimensionless adjustment parameter calculated based on the overall distribution characteristics of the second urgency coefficient across all sub-regions. It is used to adaptively adjust the relative weights of the first and second urgency coefficients in the fusion calculation. The dynamic weight adjustment factor is calculated as the ratio of the standard deviation to the mean of the second urgency coefficients across all sub-regions, i.e., the coefficient of variation. A larger coefficient of variation indicates a more significant difference in the disaster evolution trend among sub-regions, meaning a stronger spatial imbalance in disaster development. In this case, the dynamic factor contributes more to distinguishing the urgency levels of each sub-region, and the second urgency coefficient should receive a higher fusion weight. In this embodiment, the dynamic weight adjustment factor is mapped to a range of zero to one. The mapped value is used as the fusion weight of the second urgency coefficient, and one minus this value is used as the fusion weight of the first urgency coefficient. Subsequently, for each sub-region, the first urgency coefficient multiplied by its corresponding fusion weight and the second urgency coefficient multiplied by its corresponding fusion weight are added together to obtain the target urgency coefficient for each sub-region.

[0069] Based on the above embodiments, as an optional implementation, in S104, generating the target urgency coefficient for each sub-region by combining the first urgency coefficient and the second urgency coefficient specifically includes S41-S45: S41, normalize the first and second urgency coefficients of each sub-region respectively, and generate the normalized first urgency coefficient and normalized second urgency coefficient of each sub-region accordingly.

[0070] The normalization process employs the minimum-maximum normalization method, mapping the first and second urgency coefficients of all sub-regions to the range of zero to one, respectively, eliminating the differences in the dimensions and numerical scales of the two types of coefficients, and making subsequent weighted calculations comparable.

[0071] S42, based on the disaster evolution trend of each sub-region, determine the current disaster evolution stage of each sub-region, and the expected arrival time when the disaster evolution trend reaches the predicted disaster peak intensity.

[0072] The disaster evolution stage refers to the evolutionary state corresponding to the current moment on the disaster evolution trend curve. This embodiment determines this stage by analyzing the first and second derivative characteristics of the disaster evolution trend curve. When the disaster intensity is low and growing slowly, it is determined to be in the initial stage; when the disaster intensity continues to rise and the growth rate increases, it is determined to be in the development stage; when the disaster intensity is close to or near the predicted peak intensity, it is determined to be in the peak stage; and when the disaster intensity begins to decline, it is determined to be in the decline stage. The estimated arrival time refers to the time required for the disaster evolution trend curve to reach the predicted peak intensity from the current moment. For sub-regions already in the peak or decline stage, the estimated arrival time is set to zero.

[0073] S43. Based on the disaster evolution stage, assign corresponding first dynamic weight and second dynamic weight to each sub-region, and generate a time urgency factor for each sub-region based on the expected arrival time. The disaster evolution stage includes the initial stage, development stage, peak stage, and decline stage. The first dynamic weight corresponding to the initial stage and decline stage is greater than the second dynamic weight, and the second dynamic weight corresponding to the development stage and peak stage is greater than the first dynamic weight. The time urgency factor is negatively correlated with the expected arrival time.

[0074] The first dynamic weight refers to the weight assigned to the normalized first urgency coefficient, and the second dynamic weight refers to the weight assigned to the normalized second urgency coefficient. In the initial and decline stages, the disaster situation is relatively stable or tending to ease. At this time, the inspection priority should be determined more by the population and facility exposure levels of the sub-region itself; therefore, the first dynamic weight is greater than the second dynamic weight. In the development and peak stages, the disaster situation changes drastically or has reached its most severe state. At this time, the inspection priority should be determined more by the dynamic urgency of the disaster; therefore, the second dynamic weight is greater than the first dynamic weight. The time urgency factor refers to a moderating factor that measures the urgency of the disaster reaching its peak based on the estimated arrival time. The time urgency factor is negatively correlated with the estimated arrival time; that is, the shorter the estimated arrival time, the sooner the disaster will reach its peak intensity, the more urgent the inspection time window, and the higher the value of the time urgency factor. This embodiment uses the normalized reciprocal of the estimated arrival time as the calculation method for the time urgency factor.

[0075] S44. Calculate the weighted urgency coefficient of each sub-region based on the product of the first dynamic weight and the normalized first urgency coefficient, and the product of the second dynamic weight, the normalized second urgency coefficient, and the time urgency factor.

[0076] Multiplying the first dynamic weight by the normalized first urgency coefficient yields the static weighting term. Multiplying the second dynamic weight, the normalized second urgency coefficient, and the time urgency factor yields the dynamic weighting term. The sum of these two terms is the weighted urgency coefficient. The time urgency factor only affects the dynamic weighting term, not the static weighting term. This is because the time urgency factor reflects the time constraint of disaster evolution, which directly affects the urgency of dynamic urgency, while static importance does not change with the time progress of disaster evolution.

[0077] S45, calculate the coupling enhancement coefficient of each sub-region, and superimpose the coupling enhancement coefficient onto the weighted urgency coefficient to generate the target urgency coefficient of each sub-region; wherein, the coupling enhancement coefficient is determined based on the product of the normalized first urgency coefficient and the normalized second urgency coefficient, and is used to characterize the urgency superposition enhancement effect generated when a sub-region has both high static importance and high dynamic urgency.

[0078] This embodiment calculates the coupling enhancement coefficient of each sub-region and superimposes it onto the weighted urgency coefficient to generate the target urgency coefficient. The coupling enhancement coefficient is a quantitative indicator of the urgency enhancement effect generated when a sub-region simultaneously possesses high static importance and high dynamic urgency. The coupling enhancement coefficient is determined based on the product of the normalized first urgency coefficient and the normalized second urgency coefficient. Essentially, the product is larger when both coefficients are at high levels, and significantly lower when either coefficient is at a low level. For example, if a densely populated sub-region with concentrated critical infrastructure faces a rapidly deteriorating disaster situation, its actual urgency level far exceeds the simple weighted sum of the two dimensions. This is because high exposure combined with high dynamic urgency means that a large number of personnel and critical facilities are facing a rapidly approaching serious threat, requiring additional priority. This embodiment multiplies the coupling enhancement coefficient by a preset enhancement adjustment coefficient and then superimposes it onto the weighted urgency coefficient to obtain the target urgency coefficient for each sub-region. The target urgency coefficient refers to the final inspection urgency assessment value for each sub-region after comprehensively considering static importance, dynamic urgency, disaster evolution stage characteristics, time urgency, and the dual-dimensional coupling enhancement effect.

[0079] S105: Obtain the location and remaining battery power of each drone. Combine the location and remaining battery power of each drone with the target urgency coefficient of each sub-area to generate the initial inspection route for each drone.

[0080] Specifically, this embodiment first obtains the location and remaining battery power of each drone through the drone management sub-platform. Location refers to the geographic coordinates of each drone at the current moment, reported in real-time by the drone's onboard positioning module to the drone management sub-platform. Remaining battery power refers to the percentage of remaining usable energy of each drone's battery relative to its total battery capacity, reported in real-time by the drone's power management module. This embodiment calculates the maximum flight distance of each drone based on its remaining battery power and preset energy consumption parameters per unit distance. The energy consumption parameter per unit distance refers to the percentage of electricity consumed by the drone per unit distance flown under standard flight conditions; this parameter is pre-calibrated according to the drone's model parameters. The maximum flight distance represents the upper limit of the flight distance that each drone can complete with its current remaining battery power, providing a hard constraint for subsequent task allocation.

[0081] After obtaining the location and remaining battery power of each drone, this embodiment adopts a two-stage strategy to generate the initial inspection route for each drone. The first stage is sub-regional task allocation, and the second stage is intra-regional path planning.

[0082] In the sub-region task allocation phase, this embodiment needs to determine which drone will be responsible for inspecting each sub-region. This embodiment first constructs a task allocation cost matrix, where rows correspond to drones and columns correspond to sub-regions. Each element in the matrix represents the comprehensive cost of assigning a sub-region to a specific drone for inspection. The comprehensive cost consists of two parts: flight distance cost and emergency response cost. The flight distance cost is a normalized value calculated by dividing the Euclidean distance from the drone's current position to the geometric center of the sub-region by the drone's maximum flyable distance. It reflects the relative distance cost of the drone reaching the sub-region; a larger value indicates a higher proportion of flight distance to available flight resources. The emergency response cost is a normalized value minus the target urgency coefficient of the sub-region, reflecting the offsetting effect of the sub-region's urgency level on the cost. Sub-regions with higher target urgency coefficients have lower emergency response costs and are therefore more likely to be prioritized in the allocation process. The overall cost is a weighted sum of flight distance cost and emergency response cost, with the flight distance cost having a lower weight than the emergency response cost. This ensures that the inspection needs of high-urgency sub-areas are prioritized during the allocation process, while also taking into account the economic efficiency of flight distance.

[0083] After the task allocation cost matrix is ​​constructed, this embodiment uses the Hungarian algorithm to solve for the optimal task allocation scheme. The Hungarian algorithm is a classic combinatorial optimization algorithm for solving the minimum weight matching problem in bipartite graphs. It can find the allocation scheme that minimizes the sum of the comprehensive costs of all allocations in polynomial time. When the number of sub-regions exceeds the number of drones, some drones will be allocated to multiple sub-regions. This embodiment achieves multi-round allocation by iteratively executing the Hungarian algorithm: in each round of allocation, the allocated sub-regions are removed from the cost matrix, and the remaining flight distance of each drone is updated according to the estimated flight distance consumed by the allocated tasks, and the flight distance cost is recalculated until all sub-regions are allocated. During the iterative allocation process, if the remaining flight distance of a drone is insufficient to reach any unallocated sub-region and complete the minimum inspection coverage before returning to the nearest helipad, then the drone will no longer participate in subsequent rounds of allocation to avoid the drone being unable to return safely due to battery depletion.

[0084] After the sub-region task allocation phase is completed, each UAV obtains its own set of sub-regions to inspect. During the intra-region path planning phase, this embodiment plans specific flight paths for each UAV within its assigned set of sub-regions. For UAVs assigned multiple sub-regions, this embodiment first determines the order in which the UAV inspects each sub-region: starting from the UAV's current position, sub-regions are visited sequentially according to their target urgency coefficient from highest to lowest. When multiple sub-regions have similar target urgency coefficients, the sub-region closer to the UAV's current planned position is prioritized to reduce the transfer flight distance between sub-regions. After the access order is determined, an intra-region coverage flight path is generated for each sub-region. A coverage flight path refers to the path by which the UAV flies within a single sub-region according to a specific geometric pattern to achieve full coverage of the ground area of ​​that sub-region. This embodiment uses a reciprocating strip coverage mode to generate coverage flight paths, where the UAV flies back and forth along the long axis of the sub-region in parallel, equally spaced strips. The spacing between adjacent strips is determined based on the ground coverage width of the sensors on the UAV, ensuring a preset proportion of overlap between the coverage areas of adjacent strips to eliminate coverage blind spots. For sub-regions with a high urgency level, this embodiment reduces the spacing between adjacent strips to increase coverage density, enabling the sensor to collect more detailed disaster data within that sub-region. For sub-regions with a low urgency level, the strip spacing is appropriately increased to save flight distance and power resources.

[0085] Finally, in this embodiment, the transfer paths of each UAV from its current location to the first sub-region, the coverage flight paths within each sub-region, and the transfer paths between sub-regions are sequentially spliced ​​together to form the initial inspection routes of each UAV.

[0086] Based on the above embodiments, as an optional implementation method, in S105, adjusting the initial inspection route according to the correlation coefficient to generate the target inspection route for each UAV specifically includes S51-S55: S51. Based on the initial inspection route of each UAV, determine the predicted flight position of each UAV at each inspection time. Based on each predicted flight position, the communication performance parameters of each UAV, and the terrain feature data of each sub-region, calculate the correlation coefficient between any two UAVs at each inspection time. The correlation coefficient is negatively correlated with the predicted spatial distance between the two UAVs, positively correlated with the communication performance parameters, and affected by the attenuation of the terrain occlusion degree between the two UAVs.

[0087] This embodiment determines the predicted flight position of each UAV at each inspection time based on its initial inspection route, and calculates the correlation coefficient between any two UAVs at each inspection time accordingly. The inspection time refers to a series of time sampling points obtained by discretizing the entire inspection task period according to a preset time interval. The predicted flight position refers to the three-dimensional spatial coordinates of each UAV at each inspection time while flying along its initial inspection route. The correlation coefficient is a comprehensive index measuring the communication connectivity quality between any two UAVs, and its calculation requires consideration of three factors. The first factor is the predicted spatial distance between the two UAVs. The greater the spatial distance, the stronger the signal propagation attenuation, and the lower the correlation coefficient; therefore, the correlation coefficient is negatively correlated with the predicted spatial distance. The second factor is the communication performance parameters of each UAV, including indicators such as the transmit power, antenna gain, and receive sensitivity of the communication module carried by the UAV. The stronger the communication performance parameters, the higher the effective communication capability of the UAV, and the higher the correlation coefficient; therefore, the correlation coefficient is positively correlated with the communication performance parameters. The third factor is the degree of terrain obstruction between the two drones. This embodiment determines whether there are terrain obstacles such as mountains or buildings obstructing the line-of-sight propagation path between the two drones based on the terrain feature data of each sub-region. Terrain obstruction will increase signal diffraction and scattering loss, thus attenuating the correlation coefficient. This embodiment integrates the above three factors to calculate the correlation coefficient between any two drones at each inspection time.

[0088] S52. Based on the correlation coefficient, construct a time-varying communication topology graph. In the time-varying communication topology graph, identify weak communication edges with correlation coefficients lower than a preset connectivity strength threshold and the corresponding weak communication time periods. The time-varying communication topology graph uses each UAV as a node and the correlation coefficient between any two UAVs as the edge weight.

[0089] This embodiment constructs a time-varying communication topology graph based on correlation coefficients and identifies weak links in the communication network. The time-varying communication topology graph refers to a network topology structure constructed with each UAV as a node and the correlation coefficient between any two UAVs as the edge weight. Since the flight positions of each UAV change with the inspection time, the correlation coefficient also changes accordingly; therefore, the edge weights of this topology graph change dynamically over time. This embodiment traverses the correlation coefficients of all edges in the time-varying communication topology graph at each inspection time, identifies edges with correlation coefficients lower than a preset connectivity strength threshold as weak edges, and determines the time interval during which the correlation coefficient of such an edge remains below the preset connectivity strength threshold as the corresponding weak communication period. The preset connectivity strength threshold refers to the minimum correlation coefficient value required to ensure stable and reliable communication between two UAVs. The weak edges and their corresponding weak communication periods together indicate the specific locations and times of insufficient communication connectivity in the initial inspection route.

[0090] S53, for each weak UAV pair corresponding to a weak communication edge, weak route segments are determined on the initial inspection route during the weak communication period. Multiple candidate adjustment route segments are generated for each weak route segment. The connectivity gain value and inspection coverage loss value of each candidate adjustment route segment are calculated respectively. The connectivity gain value is used to characterize the improvement of the correlation coefficient between the weak UAV pairs by the candidate adjustment route segment relative to the weak route segment during the weak communication period. The inspection coverage loss value is used to characterize the decrease in inspection priority coverage caused by the candidate adjustment route segment relative to the weak route segment due to deviation from the sub-region with a higher target urgency coefficient.

[0091] This embodiment generates and evaluates candidate adjustment route segments for each communication weakness edge. The weak UAV pair refers to the two UAVs corresponding to the two ends of the communication weakness edge. In this embodiment, during the communication weakness period, corresponding route segments are extracted from the initial inspection routes of each weak UAV pair as weak route segments. These weak route segments are the specific flight paths that cause the communication weakness. This embodiment generates multiple candidate adjustment route segments for each weak route segment. These candidate adjustment route segments are alternative route schemes formed by adjusting the flight paths of one or two UAVs in the weak UAV pair during the communication weakness period, causing them to spatially approach or bypass terrain-obstructed areas. This embodiment calculates the connectivity gain and inspection coverage loss values ​​for each candidate adjustment route segment. The connectivity gain value refers to the increase in the correlation coefficient between the weak UAV pair during the communication weakness period relative to the original value after replacing the weak route segment with a candidate adjustment route segment. A higher connectivity gain value indicates a better improvement in communication connectivity by the candidate route. The inspection coverage loss value refers to the decrease in inspection priority coverage caused by the UAV missing or delaying the coverage of sub-areas with high target urgency coefficients due to the deviation of the candidate adjustment route segment from the initial inspection route. In this embodiment, the loss is quantified by calculating the difference between the candidate adjustment route segment and the weak route segment in the weighted sum of target urgency coefficients of each sub-area within the coverage range.

[0092] S54. Based on the connectivity gain and inspection coverage loss of each candidate adjustment route segment, calculate the comprehensive adjustment benefit of each candidate adjustment route segment, and select the candidate adjustment route segment with the largest comprehensive adjustment benefit to replace the corresponding weak route segment; where the comprehensive adjustment benefit is positively correlated with the connectivity gain value and negatively correlated with the inspection coverage loss value.

[0093] This embodiment calculates the comprehensive adjustment benefit of each candidate route segment based on connectivity gain and inspection coverage loss, and selects the optimal solution. The comprehensive adjustment benefit refers to the net benefit after comprehensively measuring communication connectivity improvement and inspection coverage loss. The comprehensive adjustment benefit is positively correlated with connectivity gain and negatively correlated with inspection coverage loss. That is, when communication connectivity improvement is similar, the candidate route with the least impact on inspection coverage is prioritized; when inspection coverage loss is similar, the candidate route with the greatest improvement in communication connectivity is prioritized. This embodiment selects the candidate route segment with the largest comprehensive adjustment benefit to replace the corresponding weak route segment, achieving the optimal balance between communication connectivity assurance and inspection coverage efficiency.

[0094] S55: After replacing all weak route segments, recalculate the correlation coefficient between each UAV at each inspection time after replacement. When all correlation coefficients are not lower than the preset connectivity strength threshold, the replaced inspection route is determined as the target inspection route for each UAV.

[0095] This embodiment performs global connectivity verification after replacing all weak route segments. Since replacing each weak route segment may change the flight position of the UAV within a corresponding time period, thus affecting the correlation coefficient between that UAV and other non-weak UAVs, it is necessary to recalculate the correlation coefficients between all UAVs at each inspection time after the replacement. When all correlation coefficients are not lower than the preset connectivity strength threshold, it indicates that the replaced inspection route can ensure the communication connectivity between all UAVs meets the collaborative operation requirements throughout the entire inspection task period. In this embodiment, the replaced inspection route is determined as the target inspection route for each UAV. If there are still cases where the correlation coefficient is lower than the preset connectivity strength threshold, steps S52 to S54 need to be re-executed for the newly generated weak communication edges until all correlation coefficients meet the requirements.

[0096] S106, obtain the correlation coefficient between each UAV, adjust the initial inspection route according to the correlation coefficient, and generate the target inspection route for each UAV; wherein, the correlation coefficient is used to characterize the communication link connectivity strength between each UAV during the execution of the inspection task.

[0097] Specifically, this embodiment first obtains the correlation coefficient between each UAV. The correlation coefficient is a numerical index that quantifies the communication link strength between any two UAVs during the execution of an inspection task. Its value ranges from zero to one; a higher value indicates a stronger communication link, while a lower value indicates a weaker link or even a risk of interruption. The correlation coefficient is not a static constant but a dynamic time-varying variable that changes with the flight positions of the two UAVs along their respective inspection routes. This embodiment discretizes the initial inspection routes of each UAV according to a preset time sampling interval to obtain the expected flight position of each UAV at each sampling time, and then calculates the correlation coefficient between any two UAVs at each sampling time.

[0098] The correlation coefficient at each sampling time is determined by three components: spatial distance, terrain occlusion, and electromagnetic interference. The spatial distance component is calculated based on the Euclidean distance between the expected flight positions of the two UAVs at that sampling time. In this embodiment, a negative exponential decay function is used to map the Euclidean distance to a component value between zero and one; that is, the larger the distance between the two UAVs, the smaller the spatial distance component. When the distance exceeds the maximum communication radius of the UAV communication module, the spatial distance component approaches zero. The terrain occlusion component is calculated based on whether there is terrain occlusion along the line of sight between the expected flight positions of the two UAVs. In this embodiment, the terrain elevation data obtained in step S101 is used to check point by point along the line connecting the two UAVs to see if the terrain elevation exceeds the height value of the corresponding position on the line. If terrain occlusion exists, the value of the terrain occlusion component is reduced according to the proportion of the length of the occluded segment to the total length of the line. The electromagnetic interference component is calculated based on the electromagnetic environment assessment values ​​of the area where the two UAVs are expected to fly. In disaster scenarios, damaged power facilities and communication base stations may generate abnormal electromagnetic radiation. In this embodiment, electromagnetic environment monitoring data of the disaster area is obtained from the disaster situation sub-platform to assess the electromagnetic interference intensity around the flight positions of the two UAVs. The stronger the electromagnetic interference, the lower the electromagnetic interference component. The correlation coefficient is the weighted product of the above three components. The reason for using the product form instead of the weighted sum form is that the connectivity of the communication link has a "bottleneck effect", that is, severe degradation of any component is enough to cause the communication link to be interrupted. The product form can ensure that when any component approaches zero, the overall correlation coefficient approaches zero, thereby accurately reflecting the risk of communication link interruption.

[0099] After obtaining the correlation coefficients between each UAV at each sampling time, this embodiment identifies communication weakness periods in the initial inspection route. A communication weakness period refers to a continuous time interval where the correlation coefficient between a particular UAV and all other UAVs is lower than a preset correlation coefficient threshold. This indicates that the UAV is in a communication isolation state during this period and cannot maintain effective communication with any other UAV or with the ground command center via relay from other UAVs. The correlation coefficient threshold is a preset minimum acceptable communication link connectivity value; values ​​below this threshold are considered unreliable.

[0100] After identifying weak communication periods, this embodiment adjusts the initial inspection route to eliminate or mitigate these periods. The adjustment strategy comprises two levels: timing adjustment and path offset. Timing adjustment refers to coordinating the inspection timing of the UAV in a weak communication period with its neighboring UAV with the highest correlation coefficient, without altering the geometry of the UAV's coverage flight path. This ensures that the flight times of the two UAVs entering areas with poor communication conditions are appropriately staggered or brought closer together, thereby reducing the spatial distance between the two UAVs during this period and improving the correlation coefficient. Specifically, this embodiment introduces local speed adjustment for UAVs in weak communication periods along their coverage flight path. This involves appropriately reducing the flight speed before entering the weak communication area to delay the entry time, or appropriately increasing the flight speed to pass through the area earlier, so that the flight times overlap with those of neighboring UAVs in the nearby area, thereby maintaining at least one usable inter-UAV communication link during this period.

[0101] When timing adjustments fail to raise the correlation coefficient during weak communication periods above the threshold, this embodiment further employs a path offset strategy. Path offset refers to locally geometrically shifting the coverage flight path of a UAV during a weak communication period along a corresponding path segment. This shifts the path segment a certain distance in the direction of a higher correlation coefficient, reducing the spatial distance between the UAV and adjacent UAVs or avoiding terrain-blocked areas when flying through the path segment. The magnitude of the path offset is limited by two constraints: first, the offset path must not exceed the boundary of the sub-area covered by the UAV to ensure coverage integrity; second, the additional flight distance generated by the offset must not exceed a preset proportion of the UAV's current remaining flight range to avoid insufficient battery power due to path offset. Under these constraints, this embodiment finds the minimum offset magnitude that makes the correlation coefficient exactly reach the threshold by iteratively increasing the offset magnitude and recalculating the correlation coefficient, thereby minimizing the impact of path offset on the original coverage flight path.

[0102] For communication gaps that cannot be eliminated even after timing adjustments and path offsets, this embodiment marks them as communication blind spots and records them in the supplementary information of the target inspection route. This allows the ground command center to make targeted communication support arrangements during mission execution, such as temporarily dispatching communication relay drones or activating satellite communication backup links.

[0103] After generating the target inspection routes for each drone, the process also includes: S107, when each UAV fails to complete the inspection of the entire disaster area within the preset time according to the target inspection route, the spatiotemporal correlation of the disaster situation between adjacent sub-regions is obtained based on the regional characteristics and disaster evolution trend of each sub-region; according to the spatiotemporal correlation of the disaster situation, the information gain coefficient of the inspection data generated by each UAV after performing inspection of each sub-region according to the target inspection route is determined for the uninspected sub-regions; according to the information gain coefficient, the target inspection route is adjusted to generate the final inspection route of each UAV; wherein, the information gain coefficient is used to characterize the amount of additional information provided by the inspection data of a sub-region to the disaster situation of adjacent sub-regions due to the spatiotemporal correlation of the disaster situation with adjacent sub-regions.

[0104] In real-world disaster emergency scenarios, factors such as the large area of ​​the disaster zone, the limited number of drones, or insufficient battery power for some drones may prevent them from completing the inspection coverage of all sub-regions of the disaster area within a preset timeframe when performing inspection tasks according to the target inspection route. Therefore, this embodiment obtains the spatiotemporal correlation of the disaster situation between adjacent sub-regions, calculates the information gain coefficient, and adjusts the target inspection route accordingly to generate the final inspection route.

[0105] This embodiment first estimates the completion time of all inspection tasks based on the total path length of the target inspection route and the standard cruising speed of each UAV. When the estimated completion time exceeds a preset duration, a subsequent adjustment process is triggered. This embodiment calculates the set of sub-areas that each UAV can actually cover within the preset duration, marks them as inspectable sub-areas, and marks the remaining sub-areas as uninspected sub-areas.

[0106] Subsequently, this embodiment obtains the spatiotemporal correlation of disaster situations between adjacent sub-regions. The spatiotemporal correlation of disaster situations refers to a comprehensive quantitative index of the spatial similarity and temporal synergy in disaster situations between two geographically adjacent sub-regions. The value ranges from zero to one; a higher value indicates greater synchronicity and consistency in the disaster situation changes between the two sub-regions. The spatiotemporal correlation of disaster situations is determined by the weighted geometric mean of the spatial correlation component and the temporal correlation component. The spatial correlation component is calculated based on the vector cosine similarity of the regional features obtained in step S102 between the two adjacent sub-regions. Sub-regions with more similar regional features have a higher probability of exhibiting similar disaster situations in the same disaster event. The temporal correlation component is calculated based on the Pearson correlation coefficient of the disaster intensity time series obtained in step S103 between the two adjacent sub-regions. A higher correlation coefficient indicates a more consistent trend in disaster evolution. The reason for using the geometric mean instead of the arithmetic mean is that the spatiotemporal correlation of disaster situations requires both spatial similarity and temporal synergy to be valid for inference value. The geometric mean can effectively lower the overall value when any component is low.

[0107] After obtaining the spatiotemporal correlation of disaster information, this embodiment calculates the information gain coefficient. The information gain coefficient is a quantitative indicator of the additional information provided by the inspection data of an inspectable sub-region due to the spatiotemporal correlation of disaster information with adjacent uninspected sub-regions. For each inspectable sub-region, this embodiment retrieves its adjacent uninspected sub-regions, multiplies the spatiotemporal correlation between the two by the normalized coverage density of the inspectable sub-region, and uses this as the information gain coefficient. Higher coverage density means more detailed and reliable inspection data, and higher inference value for adjacent regions. When an uninspected sub-region is adjacent to multiple inspectable sub-regions, the maximum value of each information gain coefficient is taken as the comprehensive information gain coefficient obtained for the uninspected sub-region, avoiding overestimation of inferred information due to information redundancy.

[0108] After calculating the information gain coefficient, this embodiment adjusts the target inspection route in two aspects. The first aspect is the reordering of inspection priorities. This embodiment weights and combines the target urgency coefficient of each inspectable sub-region with the sum of its information gain coefficients for all adjacent uninspected sub-regions to obtain the comprehensive inspection value. The comprehensive inspection value refers to the sum of the direct information value and the indirect inferred information value that can be obtained by inspecting a sub-region. This embodiment rearranges the inspection order from high to low according to the comprehensive inspection value, ensuring that the sub-regions that have not been inspected by the preset time limit are the sub-regions with the lowest comprehensive inspection value. The second aspect is the redistribution of coverage density. For inspectable sub-regions with high information gain coefficients, their coverage density is appropriately reduced, and the saved flight time is allocated to sub-regions that were originally uninspected but have high comprehensive inspection value. This is because, under conditions of high disaster spatiotemporal correlation, appropriately reducing coverage density has limited impact on inference accuracy, while redirecting flight time to uncovered sub-regions can obtain a greater overall information increment. A minimum threshold is set for the reduction in coverage density to maintain the basic availability of inspection data.

[0109] Figure 2 This is a schematic diagram of a drone disaster inspection and command system provided in an embodiment of this application, as shown below. Figure 2As shown, the main control platform occupies the core position in the upper half of the interface. It presents the overall monitoring situation intuitively in the form of a 3D panoramic map. The map not only marks the forest fire area with pulse animation effects and simulated user click interactions, but also clearly shows the location of the drone base and the blue dotted line inspection trajectory automatically planned by the system. Below the main control platform, three core business sub-platforms are arranged side by side. Sub-platform 1 on the left focuses on displaying disaster information, directly extracting key disaster data such as "forest fire", "spreading in the southeast direction", and "Level 1 emergency". Sub-platform 2 in the middle is responsible for real-time monitoring of equipment status, clearly listing the current battery level and standby or return status of drones A and B. Sub-platform 3 on the right intuitively displays the final inspection route generated by the system optimization, presenting the task distribution results to the command personnel in the form of a clear node list of "starting point No. 1 helipad, passing through the airspace above the core fire area, and ending point No. 2 observation station". The entire interface, through the linkage of the main and sub-platforms, perfectly interprets the complete closed-loop command process from disaster discovery, equipment monitoring to route planning and execution.

[0110] Based on the above method, this application also discloses a drone inspection route planning system, such as... Figure 3 As shown, Figure 3 This is a schematic diagram of the structure of a UAV inspection route planning system provided in an embodiment of this application. The system includes: a response module, a first acquisition module, a second acquisition module, a combination module, a third acquisition module, and an adjustment module; wherein, The response module responds to user clicks on a disaster area on the 3D map of the main control platform. Each sub-platform displays target parameters through a display interface based on its corresponding business module. These target parameters include the regional characteristics and real-time status of the disaster area; the location, remaining battery power, and inspection routes of each UAV; a first acquisition module acquires the regional characteristics of the disaster area and divides it into multiple sub-regions based on these characteristics, with each sub-region corresponding to a first urgency coefficient; a second acquisition module acquires the real-time status of each sub-region and predicts the disaster evolution trend of each sub-region within a preset time period after the current moment based on the real-time status, generating a second urgency coefficient for each sub-region based on the disaster evolution trend; a combination module combines the first and second urgency coefficients to generate a target urgency coefficient for each sub-region; a third acquisition module acquires the location and remaining battery power of each UAV and generates an initial inspection route for each UAV based on the location, remaining battery power, and target urgency coefficient of each sub-region; an adjustment module acquires the correlation coefficient between each UAV and adjusts the initial inspection route based on the correlation coefficient to generate a target inspection route for each UAV; wherein, the correlation coefficient is used to characterize the communication link connectivity strength between each UAV during the execution of the inspection task.

[0111] It should be noted that the system provided in the above embodiments is only illustrated by the division of the above functional modules. In actual applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. In addition, the system and method embodiments provided in the above embodiments belong to the same concept, and the specific implementation process can be found in the method embodiments, which will not be repeated here.

[0112] Please see Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 4 As shown, the electronic device 1000 may include: at least one processor 1001, at least one network interface 1004, a user interface 1003, a memory 1005, and at least one communication bus 1002.

[0113] The communication bus 1002 is used to realize the connection and communication between these components.

[0114] The user interface 1003 may include a display screen and a camera. Optionally, the user interface 1003 may also include a standard wired interface and a wireless interface.

[0115] The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0116] The processor 1001 may include one or more processing cores. The processor 1001 connects to various parts of the server using various interfaces and lines, and performs various server functions and processes data by running or executing instructions, programs, code sets, or instruction sets stored in the memory 1005, and by calling data stored in the memory 1005. Optionally, the processor 1001 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 1001 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 1001 and may be implemented as a separate chip.

[0117] The memory 1005 may include random access memory (RAM) or read-only memory. Optionally, the memory 1005 may include a non-transitory computer-readable storage medium. The memory 1005 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 1005 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 1005 may also be at least one storage device located remotely from the aforementioned processor 1001. Figure 4 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for a UAV inspection route planning method.

[0118] exist Figure 4In the electronic device 1000 shown, the user interface 1003 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 1001 can be used to call an application program stored in the memory 1005 for a UAV inspection route planning method. When executed by one or more processors, the electronic device performs one or more of the methods described in the above embodiments.

[0119] An electronic device readable storage medium stores instructions that, when executed by one or more processors, cause the electronic device to perform one or more of the methods described in the above embodiments.

[0120] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

[0121] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0122] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some service interfaces; indirect couplings or communication connections between devices or units may be electrical or other forms.

[0123] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0124] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0125] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, portable hard drives, magnetic disks, or optical disks.

[0126] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Other embodiments of this disclosure will be readily apparent to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for planning inspection routes using unmanned aerial vehicles (UAVs), characterized in that, The method is applied to a main control platform, which integrates multiple sub-platforms through an embedded framework. Each sub-platform corresponds to a different business module and has an independent data display interface. The method includes: In response to a user clicking on a disaster area on the 3D map of the main control platform, each of the sub-platforms displays target parameters through a display interface based on its corresponding business module. The target parameters include the regional characteristics and real-time status of the disaster area; the location, remaining battery power, and inspection routes of each UAV; The regional characteristics of the disaster area are obtained, and the disaster area is divided into multiple sub-regions based on the regional characteristics, with each sub-region corresponding to a first emergency coefficient; The real-time status of each sub-region is obtained, and based on the real-time status, the disaster evolution trend of each sub-region within a preset time period after the current time is predicted. Based on the disaster evolution trend, a second emergency coefficient of each sub-region is generated. By combining the first urgency coefficient and the second urgency coefficient, a target urgency coefficient for each of the sub-regions is generated; The location and remaining battery power of each UAV are obtained. Based on the location and remaining battery power of each UAV and the target urgency coefficient of each sub-region, the initial inspection route of each UAV is generated. Obtain the correlation coefficient between each of the UAVs, adjust the initial inspection route according to the correlation coefficient, and generate the target inspection route for each of the UAVs; wherein, the correlation coefficient is used to characterize the communication link connectivity strength between each UAV during the execution of the inspection task.

2. The UAV inspection route planning method according to claim 1, characterized in that, After generating the target inspection routes for each of the aforementioned UAVs, the process further includes: If each of the aforementioned drones fails to complete the inspection of the entire disaster area within the preset time according to the target inspection route, the spatiotemporal correlation degree of the disaster between adjacent sub-regions is obtained based on the regional characteristics and disaster evolution trend of each sub-region. Based on the spatiotemporal correlation of the disaster, determine the information gain coefficient of the inspection data generated by each of the UAVs after performing inspections on each of the sub-regions according to the target inspection route for the uninspected sub-regions. Based on the information gain coefficient, the target inspection route is adjusted to generate the final inspection route for each of the UAVs; wherein, the information gain coefficient is used to characterize the amount of additional information provided by the inspection data of a sub-region to the disaster situation of the adjacent sub-regions due to the spatiotemporal correlation of the disaster situation with the adjacent sub-regions.

3. The UAV inspection route planning method according to claim 1, characterized in that, In response to a user clicking on a disaster area on the 3D map of the main control platform, each of the sub-platforms displays target parameters through a display interface based on its corresponding business module, including: By monitoring click operations on the 3D map component in the main control platform through a browser extension, the coordinate information of the disaster area can be obtained. The coordinate information is sent to each of the sub-platforms via a message passing mechanism, and the coordinate information is stored in a shared storage area; The browser extension intercepts data requests sent from each of the sub-platforms to the server. Based on the preset parameter mapping configuration, identify the location-related parameters in the data request; The coordinate information is used to replace the location-related parameters identified in the data request to generate a modified data request. The modified data request is sent to the server, enabling each sub-platform to obtain and display the corresponding target parameters based on the coordinate information of the disaster area.

4. The UAV inspection route planning method according to claim 1, characterized in that, The regional characteristics include topographic elevation data, land use type data, building density data, population density data, and critical infrastructure distribution data. Based on these regional characteristics, the disaster area is divided into multiple sub-regions, including: Based on the topographic elevation data, the natural geographical boundary line within the disaster area is extracted, and the disaster area is divided into multiple initial blocks using the natural geographical boundary line as the initial boundary. Combining the land use type data and the building distribution density data, the land use type homogeneity and building distribution density homogeneity within each initial block are calculated respectively. Initial blocks with land use type homogeneity lower than a preset homogeneity threshold or building distribution density homogeneity lower than a preset homogeneity threshold are split into multiple sub-blocks along the abrupt change line of the land use type or the building distribution density. Calculate the regional feature similarity between adjacent initial blocks and / or subdivided blocks, and merge adjacent blocks with regional feature similarity greater than a preset similarity threshold to generate multiple sub-regions; Based on population density data and critical infrastructure distribution data within each sub-region, a first emergency coefficient is generated for each sub-region.

5. The UAV inspection route planning method according to claim 1, characterized in that, The real-time status includes disaster intensity data, disaster spread rate data, meteorological environment data, and disaster severity data for each sub-region. Based on the real-time status, the process of predicting the disaster evolution trend of each sub-region within a preset time period after the current moment, and generating a second emergency coefficient for each sub-region based on the disaster evolution trend, includes: Based on the disaster intensity data and disaster spread rate data of each sub-region, a single-domain disaster evolution model is constructed for each sub-region, and the initial evolution trend of each sub-region is generated based on the single-domain disaster evolution model. Based on the regional characteristics of each sub-region and the geographical connectivity between adjacent sub-regions, and combined with the meteorological environmental data of each sub-region, a disaster propagation and impact matrix between each sub-region is constructed; Based on the disaster propagation impact matrix, the initial evolution trend of each sub-region is cross-regionally coupled and corrected to generate the disaster evolution trend of each sub-region within a preset time period after the current moment; Based on the disaster evolution trend of each sub-region, the predicted disaster peak intensity and predicted disaster deterioration rate of each sub-region within the preset time period are extracted, and combined with the disaster severity data of each sub-region, a second emergency coefficient for each sub-region is generated.

6. The UAV inspection route planning method according to claim 1, characterized in that, The step of combining the first urgency coefficient and the second urgency coefficient to generate the target urgency coefficient for each of the sub-regions includes: The first and second urgency coefficients of each sub-region are normalized respectively to generate the normalized first urgency coefficient and normalized second urgency coefficient of each sub-region. Based on the disaster evolution trend of each sub-region, determine the current disaster evolution stage of each sub-region and the estimated arrival time when the disaster evolution trend reaches the predicted disaster peak intensity; Based on the disaster evolution stage, a corresponding first dynamic weight and second dynamic weight are assigned to each of the sub-regions, and a time urgency factor for each of the sub-regions is generated based on the expected arrival time. The weighted urgency coefficient of each sub-region is calculated based on the product of the first dynamic weight and the normalized first urgency coefficient, and the product of the second dynamic weight, the normalized second urgency coefficient, and the time urgency factor. Calculate the coupling enhancement coefficient for each sub-region, and superimpose the coupling enhancement coefficient onto the weighted urgency coefficient to generate the target urgency coefficient for each sub-region.

7. The UAV inspection route planning method according to claim 1, characterized in that, The step of adjusting the initial inspection route based on the correlation coefficient to generate the target inspection route for each of the UAVs includes: Based on the initial inspection route of each UAV, the predicted flight position of each UAV at each inspection time is determined. Based on the predicted flight position, the communication performance parameters of each UAV, and the terrain feature data of each sub-region, the correlation coefficient between any two UAVs at each inspection time is calculated. Based on the correlation coefficient, a time-varying communication topology graph is constructed. In the time-varying communication topology graph, weak communication edges with correlation coefficients lower than a preset connectivity strength threshold and corresponding weak communication time periods are identified. For each weak UAV pair corresponding to the weak communication edge, a weak route segment is determined on the initial inspection route during the weak communication period. Multiple candidate adjustment route segments are generated for each weak route segment, and the connectivity gain value and inspection coverage loss value of each candidate adjustment route segment are calculated respectively. Based on the connectivity gain value and the inspection coverage loss value of each candidate adjustment route segment, calculate the comprehensive adjustment benefit of each candidate adjustment route segment, and select the candidate adjustment route segment with the largest comprehensive adjustment benefit to replace the corresponding weak route segment. After replacing all weak route segments, the correlation coefficients between each UAV at each inspection time after replacement are recalculated. When all correlation coefficients are not lower than the preset connectivity strength threshold, the replaced inspection route is determined as the target inspection route for each UAV.

8. A drone inspection route planning system, characterized in that, The system includes: a response module, a first acquisition module, a second acquisition module, a combination module, a third acquisition module, and an adjustment module; wherein, The response module is used to respond to the user's operation of clicking on the disaster area on the 3D map of the main control platform. Each sub-platform displays target parameters through the display interface based on the corresponding business module. The target parameters include the regional characteristics and real-time status of the disaster area; the location, remaining power, and inspection route of each UAV, or one or more of these. The first acquisition module is used to acquire the regional characteristics of the disaster area, and divide the disaster area into multiple sub-regions according to the regional characteristics, with each sub-region corresponding to a first emergency coefficient; The second acquisition module is used to acquire the real-time status of each sub-region, predict the disaster evolution trend of each sub-region within a preset time period after the current time based on the real-time status, and generate a second emergency coefficient for each sub-region based on the disaster evolution trend. The combining module is used to combine the first urgency coefficient and the second urgency coefficient to generate the target urgency coefficient for each of the sub-regions; The third acquisition module is used to acquire the location and remaining power of each UAV, and generate the initial inspection route of each UAV by combining the location and remaining power of each UAV and the target urgency coefficient of each sub-region. The adjustment module is used to obtain the correlation coefficient between each of the UAVs, adjust the initial inspection route according to the correlation coefficient, and generate the target inspection route for each of the UAVs; wherein, the correlation coefficient is used to characterize the communication link connectivity strength between each UAV during the execution of the inspection task.

9. An electronic device, characterized in that, The device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer program is stored that can be loaded by a processor and executed as described in any one of claims 1-7.

Images