Approximate Inspection Method and System for Energy Storage Thermal Runaway Unmanned Aerial Vehicles

Abstract

This invention provides a method and system for the re-inspection of thermal runaway unmanned aerial vehicles (UAVs) in energy storage, belonging to the field of emergency detection for energy storage UAVs. The method includes: responding to a thermal runaway alarm, performing a re-inspection task, and acquiring on-site data; calculating a comprehensive spatial risk value based on the on-site data to construct a dynamic hazard domain; generating a graded approach strategy and generating a corresponding flight path based on the graded approach strategy; controlling the UAV to perform graded approach flight along the flight path according to the graded approach strategy, and updating the dynamic hazard domain, graded approach strategy, and flight path during the graded approach flight until reaching the re-inspection position to perform a re-inspection scan; after completing the re-inspection scan, generating a return path based on the latest dynamic hazard domain and controlling the UAV to return. This invention, by constructing a dynamic hazard domain and an update mechanism, transforms the UAV's operating mode from a fixed route to an adaptive safe approach, thereby effectively ensuring the safety of the UAV platform in the dynamic propagation of thermal runaway.

Description

Technical Field

[0001] This invention relates to the field of emergency detection technology for energy storage drones, specifically to a method and system for close-range re-inspection of energy storage thermal runaway drones. Background Technology

[0002] With the large-scale and rapid development of new energy industries represented by photovoltaics and wind power, intelligent inspection technology using drones has become a core means of operation and maintenance for wind and solar power plants to ensure their safe, stable, and efficient operation. Currently, the application of drones is mostly concentrated on automated inspections along preset routes. By comparing historical infrared thermal images, abnormal temperature rises in battery clusters or connection points are identified, providing a preliminary ability to detect potential thermal runaway hazards. However, based on a static, preset route and fixed operating mode, it lacks real-time perception and autonomous avoidance capabilities regarding the dynamic spread of thermal runaway risks at the site, resulting in the drone's inability to guarantee its own safety during close approach. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for approaching and re-inspecting energy storage thermal runaway drones, which guide drones to approach the accident site in a risk-assessing and graded manner when thermal runaway occurs in an energy storage system, and complete the re-inspection task.

[0004] To achieve the above objectives, this invention provides a method for the re-inspection of an energy storage thermal runaway unmanned aerial vehicle (UAV) upon approach, comprising: responding to a thermal runaway alarm and executing a re-inspection task; acquiring on-site data during the execution of the re-inspection task; calculating a comprehensive spatial risk value based on the on-site data to construct a dynamic hazard domain; generating a graded approach strategy based on the dynamic hazard domain and generating a corresponding flight path based on the graded approach strategy; controlling the UAV to perform graded approach flight along the flight path according to the graded approach strategy, and updating the dynamic hazard domain, the graded approach strategy, and the flight path based on the real-time acquired on-site data during the graded approach flight, until reaching the re-inspection location to perform a re-inspection scan; after completing the re-inspection scan, generating a return path based on the latest dynamic hazard domain and controlling the UAV to return.

[0005] Optionally, acquiring on-site data includes: controlling the UAV to fly to a preset initial safe reconnaissance position; at the initial safe reconnaissance position, activating the onboard infrared thermal imaging camera, lidar, and visible light camera; controlling the infrared thermal imaging camera, lidar, and visible light camera to perform segmented scanning of the target area to collect thermal imaging data, three-dimensional point cloud data, and visible light image data of the target area, and simultaneously recording the UAV's pose data.

[0006] Optionally, the segmented scanning includes: performing a first scan of the front of the target region; performing a second scan of the side of the target region; and performing a third scan of the top of the target region.

[0007] Optionally, the step of calculating the spatial comprehensive risk value based on the field data includes: normalizing the acquired temperature data, temperature rise rate data, smoke concentration data, structural deformation data, and wind direction influence data; and weighting and summing the normalized data according to preset weights to obtain the comprehensive risk value of the sampling point.

[0008] Optionally, the construction of the dynamic hazard domain includes: determining spatial locations that meet preset judgment conditions based on the spatial comprehensive risk value; dividing the main heat source region, the smoke-covered region, and the structural deformation region based on the thermal characteristics, smoke characteristics, and structural deformation characteristics corresponding to the spatial locations; and spatially fusing the main heat source region, the smoke-covered region, and the structural deformation region to obtain the dynamic hazard domain.

[0009] Optionally, generating a tiered approach strategy and generating a corresponding flight path based on the tiered approach strategy includes: determining multiple candidate approach locations in the spatial region outside the dynamic hazard zone; calculating the overall suitability of each candidate approach location, the overall suitability being determined based on the target visibility of the corresponding candidate approach location, the clearance distance between the candidate approach location and the boundary of the dynamic hazard zone, the target observation angle, the image transmission link quality, and the overall risk value at the corresponding location; selecting the candidate approach location with the highest overall suitability as the next approach location; and planning and generating a corresponding flight path based on the current location of the UAV and the next approach location.

[0010] Optionally, updating the dynamic hazard domain, the graded approach strategy, and the flight path based on real-time acquired field data includes: during the graded approach flight, determining whether the real-time acquired field data meets preset update trigger conditions; when the update trigger conditions are met for multiple consecutive calculation cycles, and the time since the last update exceeds a preset update time window, triggering a dynamic hazard domain update; recalculating the spatial comprehensive risk value based on the latest field data to update the dynamic hazard domain; redetermining the safety distance constraints, flight speed constraints, and observation angle constraints corresponding to the current approach phase based on the updated dynamic hazard domain; and replanning the flight path corresponding to the current phase based on the updated constraint parameters.

[0011] Optionally, the graded approach flight includes multiple consecutive approach phases; after each approach phase is completed, the updated dynamic hazard zone is subjected to a safety check; when it is confirmed that the updated dynamic hazard zone maintains a preset safe distance from the flight path corresponding to the next approach phase, the UAV is controlled to enter the next approach phase.

[0012] Optionally, the preset update triggering conditions include any one of the following: the change in the current comprehensive risk value within an adjacent calculation period exceeds a preset risk jump threshold; the real-time distance between the UAV and the boundary of the dynamic hazard domain is less than a preset dynamic safe distance threshold; the real-time image transmission link quality is lower than a preset minimum link quality threshold; the boundary of the dynamic hazard domain expands toward the direction of the UAV's current flight path.

[0013] On the other hand, the present invention provides an energy storage thermal runaway drone approach re-inspection system for implementing an energy storage thermal runaway drone approach re-inspection method. The system includes a control module, the control module including a memory, a processor and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the energy storage thermal runaway drone approach re-inspection method.

[0014] The above technical solution, by constructing a dynamic hazard domain and updating mechanism, transforms the UAV operation mode from a fixed route to an adaptive safe approach, thereby effectively ensuring the safety of the UAV platform at the site of dynamic thermal runaway propagation and successfully completing the data re-inspection and collection at key locations.

[0015] Other features and advantages of the present invention will be described in detail in the following detailed description section. Attached Figure Description

[0016] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the following detailed description to explain the invention, but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of the method for close-range re-inspection of energy storage thermal runaway drones.

[0017] Figure 2 This is a flowchart for constructing dynamic hazard domains. Detailed Implementation

[0018] The following is in conjunction with the appendix Figure 1 - Appendix Figure 2 The specific implementation methods of the embodiments of the present invention will be described in detail below. It should be understood that the specific implementation methods described herein are only for illustrating and explaining the embodiments of the present invention, and are not intended to limit the embodiments of the present invention.

[0019] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0020] In the process of realizing this invention, the inventors of this application discovered that the static fixed operation mode adopted by the prior art relies on a preset route, which poses a safety hazard of the drone entering a sudden danger zone during the approach process.

[0021] Example 1 Reference Figures 1-2 This is the first embodiment of the present invention, which provides a method for close-range re-inspection of an energy storage thermal runaway unmanned aerial vehicle, comprising: S100: In response to thermal runaway alarm, perform a retest task and acquire field data during the retest task.

[0022] In the embodiments of this application, acquiring on-site data includes: controlling the UAV to fly to a preset initial security reconnaissance position; at the initial security reconnaissance position, activating the onboard infrared thermal imaging camera, lidar, and visible light camera; controlling the infrared thermal imaging camera, lidar, and visible light camera to perform segmented scanning of the target area to collect thermal imaging data, three-dimensional point cloud data, and visible light image data of the target area, and simultaneously recording the UAV's pose data.

[0023] In embodiments of this application, segmented scanning includes: performing a first scan of the front of the target area; performing a second scan of the side of the target area; and performing a third scan of the top of the target area.

[0024] It should be noted that the first scan is used to acquire information on the frontal thermal anomalies and flame jets of the target area; the second scan is used to supplement information on the status of laterally obstructed areas and adjacent energy storage cabinets; and the third scan is used to acquire information on the direction of thermal plume diffusion, the accumulation of smoke at the top, and the deformation of the top structure. By using segmented scanning, the spatial features of the target area can be acquired as completely as possible while maintaining a safe distance, avoiding information loss caused by a single close-range acquisition.

[0025] In this step, thermal runaway alarms can originate from the battery management system (BMS) within the energy storage system, temperature sensors, smoke sensors, fire alarm linkage systems, or manually confirmed alarms. For example, when the internal temperature of the energy storage cabinet is detected to exceed a preset temperature threshold, and the temperature rise rate continues to exceed a preset rate threshold within a consecutive preset time window, a thermal runaway retest task is triggered.

[0026] Specifically, the initial safe reconnaissance position is preferentially set on the upwind or side-upwind side of the thermal runaway target, and a preset initial safe distance is maintained from the thermal runaway target to reduce the probability of the UAV being affected by high-temperature jets, smoke diffusion and local deflagration.

[0027] Furthermore, the initial safety distance can be preset based on the type of energy storage cabinet, cabinet size, thermal runaway level, on-site passage width, and wind speed conditions.

[0028] Specifically, the pose data includes the UAV's spatial position coordinates, attitude angles, heading angles, and corresponding timestamps; thermal imaging data, 3D point cloud data, and visible light image data are all time-synchronized with the pose data to establish a field data model under a unified spatial coordinate system.

[0029] Specifically, the lidar uses a rotating scanning method to acquire point cloud data of the target area, and constructs a spatial model of the target area based on the laser echo intensity, echo continuity, and point cloud density.

[0030] The aforementioned scheme, by controlling the UAV to fly to a pre-set initial safe reconnaissance position on the upwind side and performing segmented scanning of the target area, can systematically acquire thermal imaging, 3D point cloud, visible light images, and synchronous pose data. This design ensures that in the early stages of thermal runaway, the UAV can quickly acquire multi-dimensional, high-precision structured data from the scene at a safe distance, providing complete and reliable input for subsequent operations and overcoming the shortcomings of missing information from a single sensor or the high risk of manual close inspection.

[0031] S200: Based on on-site data, calculate the comprehensive spatial risk value to construct a dynamic hazard domain.

[0032] In the embodiments of this application, calculating the comprehensive risk value of a space includes: normalizing the acquired temperature data, temperature rise rate data, smoke concentration data, structural deformation data, and wind direction influence data; and weighting and summing the normalized data according to preset weights to obtain the comprehensive risk value of the sampling point.

[0033] In the embodiments of this application, constructing a dynamic hazard domain includes: determining spatial locations that meet preset judgment conditions based on the comprehensive spatial risk value; dividing the main heat source region, smoke-covered region, and structural deformation region based on the thermal characteristics, smoke characteristics, and structural deformation characteristics corresponding to the spatial locations; and spatially fusing the main heat source region, smoke-covered region, and structural deformation region to obtain the dynamic hazard domain.

[0034] Specifically, let any sampling point in space be... Its overall risk value is , The calculation formula is as follows:

[0035] in, Indicates sampling point The corresponding normalized temperature value, This represents the normalized value of the rate of temperature rise. This represents the normalized value of smoke concentration. This represents the normalized value of structural deformation. This represents the normalized value of the wind direction effect. This represents the corresponding weight coefficients, and each weight coefficient satisfies:

[0036] The weighting coefficients can be pre-set based on the relative importance of the risks of heat source ejection, smoke obstruction, and structural collapse in the energy storage thermal runaway scenario. For example, the weights of temperature and temperature rise rate can be higher than other factors. The weighting coefficient for temperature can be set to 0.3, and the weighting coefficient for temperature rise rate can be set to 0.25 to highlight the core hazardous characteristics during the thermal runaway propagation process. The weight for smoke concentration affects visible light and lidar observations and can be set to 0.15. The weight for wind direction is used to reflect the direction of the spread of the hazardous area and can be set to 0.1. The weight for structural deformation is used to reflect the tendency of cabinet instability and local damage, which is an important secondary risk and can be set to 0.2.

[0037] For example, a set of high-risk points is constructed based on spatial points whose comprehensive spatial risk value exceeds a preset risk threshold, and a main heat source region is formed based on spatial clustering. The main heat source region can be represented as:

[0038] in, This represents a collection of thermal hazard zones. This indicates a preset danger threshold.

[0039] Furthermore, the system identifies smoke-obscured areas based on point cloud data and visible light image data.

[0040] Specifically, when the laser echo intensity decreases, the point cloud dispersion increases, the stability of the point cloud in consecutive frames decreases, or the image clarity decreases within the target area, it is determined that there is smoke obstruction in the corresponding area, and the corresponding area is classified as a smoke hazard area.

[0041] For example, a smoke hazard zone can be represented as:

[0042] in, Indicates a smoke-prone area. Point The corresponding point cloud confidence score is obtained by quantifying the echo intensity attenuation, local dispersion, and position jitter amplitude of the point cloud at that location into numerical features, and then calculating them by weighting them according to preset weights. This indicates the threshold for detecting smoke.

[0043] Furthermore, based on the spatial deviation between the current point cloud model and the preset standard energy storage cabinet model, the structural deformation region is identified. The structural deformation region can be represented as:

[0044] in, Indicates the current midpoint of the point cloud. spatial coordinates, This represents the coordinates of the corresponding point in the standard model. Indicates the reference length.

[0045] When the degree of structural deformation exceeds the preset structural deformation threshold, the corresponding area is classified as a structural danger zone.

[0046] For example, a structurally hazardous area can be represented as:

[0047] in, Indicates a structurally hazardous area. This represents the structural deformation threshold.

[0048] After identifying the main heat source area, smoke hazard area, and structural hazard area, the system spatially merges the three types of areas to obtain a dynamic hazard domain.

[0049] Specifically, the dynamic hazard domain can be represented as:

[0050] in, This indicates a dynamic danger domain.

[0051] In a preferred embodiment of this application, a safety expansion process is performed on the boundary of the dynamic hazard zone to form a safety buffer zone. The safety buffer zone can be set with different expansion distances according to the type of hazard. For example, a larger expansion distance is used for heat source areas, a medium expansion distance is used for smoke areas, and the expansion distance for structurally hazardous areas is determined according to the possible collapse direction.

[0052] The above-mentioned scheme calculates a comprehensive spatial risk value by normalizing and weighting multi-source data such as temperature, temperature rise rate, smoke concentration, structural deformation, and wind direction. Based on this, it divides the area into three types of dangerous zones: main heat source, smoke obstruction, and structural deformation, and completes spatial fusion to construct a dynamic danger domain and add a safety buffer zone. This can accurately quantify the multi-dimensional risk distribution in energy storage thermal runaway scenarios, define the scope and spread trend of dangerous areas, and provide intuitive and quantifiable spatial safety boundaries for UAV close-in flights, thereby improving the comprehensiveness and accuracy of risk identification.

[0053] S300: Based on dynamic hazard domains, it generates graded approach strategies and corresponding flight paths based on these strategies.

[0054] It should be noted that the tiered approach strategy is used to determine the safety distance constraints, flight speed constraints, observation angle constraints, and phase switching conditions corresponding to different approach phases. The flight path is used to generate a space flight trajectory that meets the corresponding constraints under the current approach phase.

[0055] In the embodiments of this application, a corresponding spatial access cost map is generated based on the dynamic hazard domain, and areas with a comprehensive risk value exceeding a preset no-entry threshold are marked as impassable areas; the spatial access cost map is used to characterize the risk access cost corresponding to the UAV flying in different spatial areas.

[0056] Specifically, the three-dimensional space corresponding to the target area is divided into multiple voxel grid cells. Based on the comprehensive risk value of each voxel grid cell, the distance between it and the boundary of the dynamic hazard domain, and the degree of smoke obscuration, the passage cost value of the corresponding voxel grid cell is calculated to generate a spatial passage cost map.

[0057] In the embodiments of this application, a hierarchical approach strategy is generated, and a corresponding flight path is generated based on the hierarchical approach strategy, including: determining multiple candidate approach positions in the spatial region outside the dynamic hazard zone; calculating the comprehensive suitability of each candidate approach position, the comprehensive suitability being determined based on the target visibility of the corresponding candidate approach position, the clearance distance between the candidate approach position and the boundary of the dynamic hazard zone, the target observation angle, the image transmission link quality, and the comprehensive risk value at the corresponding position; selecting the candidate approach position with the highest comprehensive suitability as the next approach position; and planning and generating a corresponding flight path based on the current position of the UAV and the next approach position.

[0058] It should be noted that the flight path is generated based on the spatial access cost map, and high-cost areas and impassable areas are avoided first.

[0059] For example, the tiered approach strategy includes an initial reconnaissance phase, a mid-range verification phase, and a near-field re-inspection phase. Different phases correspond to different safety distance constraints, flight speed constraints, and observation angle constraints. For instance, the initial reconnaissance phase focuses on obtaining the overall hazard domain boundary, the mid-range verification phase focuses on confirming the main heat source, smoke diffusion, and structural deformation boundaries, and the near-field re-inspection phase primarily aims to complete detailed observations and defect confirmation.

[0060] Specifically, let the candidate proximity position be... Overall suitability It can be represented as:

[0061] in, Indicates the target visibility at the candidate location; This indicates the clearance margin between the candidate location and the boundary of the danger zone; Indicates the suitability of the target observation angle; Indicates the quality of the image transmission link; This represents the overall risk value corresponding to the candidate position; This represents the corresponding weighting coefficient.

[0062] Preferred target visibility The clearance margin can be determined based on the target's proportion in the image, the degree of occlusion, and the edge sharpness. It can be determined based on the shortest distance from the candidate location to the danger boundary; suitability of the observation angle. The location can be determined based on the relative orientation of the candidate location to the center of the main heat source; image transmission link quality. The location can be determined based on a combination of factors including the signal strength, bit error rate, and delay. The system selects from multiple candidate locations. The largest candidate position is chosen as the next approach position.

[0063] Furthermore, when planning the flight path based on the UAV's current position and next approach position, a path cost function is constructed, and the flight path with the minimum cost is selected as the target flight path. Path cost function It can be represented as:

[0064] in, Indicates the flight path; k represents the path sampling point number. This represents the k-th sampling point on the path. This represents the overall risk value at that point; This indicates the distance from the point to the boundary of the danger zone; This indicates the path curvature at that point; This indicates the image transmission link quality at that point; This represents the corresponding weight coefficient; N represents the number of path sampling points.

[0065] Furthermore, seeking to make The shortest path is used as the target flight path for the UAV. The flight path is preferably generated by combining segmented direct flight with intermediate correction nodes, so that the UAV can re-check the environmental conditions after each segment of flight.

[0066] The above scheme generates a graded approach strategy and corresponding flight path based on a dynamic hazard domain. It quantitatively evaluates and selects the optimal next approach position through a comprehensive suitability function, and plans the optimal flight trajectory using a path cost function. This achieves a progressive and optimized approach from the global safe zone to the local re-inspection position. It automatically balances multiple constraints such as risk, safe distance, flight smoothness, and communication quality along the path. It avoids blind or fixed-path approaches, automatically finding an approach path that offers both safety and observational effectiveness, thus improving the intelligence and safety of the re-inspection operation.

[0067] S400: Based on the graded approach strategy, the UAV is controlled to perform graded approach flight along the flight path. During the graded approach flight, the dynamic hazard domain, graded approach strategy and flight path are updated based on real-time acquired field data until the re-inspection position is reached to perform re-inspection scanning.

[0068] In the embodiments of this application, the UAV advances step by step along the flight path according to the graded approach strategy. After each approach phase is completed, the on-site data of the current phase is reviewed, and it is determined whether the dynamic hazard domain, graded approach strategy and flight path need to be updated, so as to prevent the UAV from continuing to advance along the failure path during the continuous evolution of the thermal runaway scenario.

[0069] In the embodiments of this application, updating the dynamic hazard domain, graded approach strategy, and flight path based on real-time acquired field data includes: during graded approach flight, determining whether the real-time acquired field data meets preset update trigger conditions; when multiple consecutive calculation cycles (which can be 3 or 4) (which can be 2 seconds or 3 seconds) meet the update trigger conditions, and the time since the last update exceeds a preset update time window, triggering dynamic hazard domain update; recalculating the spatial comprehensive risk value based on the latest field data to update the dynamic hazard domain; redetermining the safety distance constraint, flight speed constraint, and observation angle constraint corresponding to the current approach stage based on the updated dynamic hazard domain; and replanning the flight path corresponding to the current stage based on the updated constraint parameters.

[0070] In this step, the graded approach flight includes multiple consecutive approach phases; after each approach phase is completed, the updated dynamic hazard zone is checked for safety; when it is confirmed that the updated dynamic hazard zone and the corresponding flight path of the next approach phase maintain a preset safe distance, the UAV is controlled to enter the next approach phase.

[0071] The preset update trigger conditions include any of the following: the change in the current comprehensive risk value in adjacent calculation cycles exceeds the preset risk jump threshold; the real-time distance between the UAV and the boundary of the dynamic hazard domain is less than the preset dynamic safe distance threshold; the real-time image transmission link quality is lower than the preset minimum link quality threshold; the boundary of the dynamic hazard domain expands towards the direction of the UAV's current flight path.

[0072] Specifically, let the comprehensive risk value at the current moment be... The comprehensive risk value at the previous moment was The change in the overall risk value It can be represented as:

[0073] When satisfied At that time, it is determined that the current risk has changed dramatically, among which, This indicates the preset risk threshold.

[0074] When the real-time distance between the drone and the boundary of the dynamic hazard zone is less than the dynamic safe distance threshold, the current propulsion process is paused, and the flight path is replanned. The dynamic safe distance determination condition can be expressed as:

[0075] in, This indicates the real-time distance between the drone and the boundary of the dynamic hazard zone. This indicates the dynamic safety distance threshold.

[0076] Furthermore, when the image transmission link quality is lower than the minimum link quality threshold, the drone is preferentially controlled to retreat to the previous safe node to avoid the drone entering a low-controllability state.

[0077] Furthermore, when the boundary of the dynamic hazard domain expands toward the current flight path, the expansion direction of the hazard domain boundary is recalculated, and a local offset correction is performed on the current flight path; when the local offset correction still fails to meet the preset safe distance constraint, or when the updated dynamic hazard domain spatially overlaps with the current flight path, the current flight path is deemed invalid, and flight path replanning is triggered.

[0078] In the embodiments of this application, the re-inspection position is the target position used to perform the final re-inspection scan after the UAV completes the step-by-step approach. The re-inspection position preferably meets the following conditions: it can simultaneously observe the front, top and adjacent cabinets of the thermal runaway cabinet; it is not located in the area with the densest smoke; it is not located in the main direction of high temperature jet; the image clarity meets the requirements for detail recognition; and the UAV can hover stably.

[0079] It should be noted that the re-inspection position is determined by the position with the highest overall suitability among the candidate approach positions. In other words, the re-inspection position is not an additional, independently set concept, but rather the final target approach position selected during the graded approach process.

[0080] In this embodiment, after the UAV arrives at the re-inspection position, it performs a re-inspection scan. The re-inspection scan includes visible light re-inspection, infrared thermal imaging re-inspection, and lidar re-inspection. Visible light re-inspection is used to collect door panel deformation, spray marks, and exposed open flames; infrared thermal imaging re-inspection is used to collect the temperature of the main heat source, the temperature of adjacent battery packs, the temperature of the cabinet door, the temperature of the busbar connection point, and the thermal diffusion boundary; lidar re-inspection is used to collect the cabinet bulge, cabinet door offset, top cover warping, and thermal deformation of adjacent equipment.

[0081] The above solution dynamically updates the danger zone, strategy, and path based on real-time data and preset triggering conditions (such as risk jumps and insufficient safety distance), enabling the UAV to respond in real time to the evolution of fire, smoke, and structural conditions at the thermal runaway site. Once the risk is detected to be expanding or the path is blocked, it can be immediately paused or replanned. The static planning is upgraded to dynamic adaptive closed-loop control, ensuring that the UAV maintains the effectiveness of its operating path and its own safety in a constantly changing high-risk environment, reducing the risk of accidents caused by information lag.

[0082] S500: After completing the re-inspection scan, it generates a return path based on the latest dynamic hazard zone and controls the drone to return.

[0083] In the embodiments of this application, after the re-inspection scan is completed, the return path is regenerated based on the latest dynamic hazard domain, instead of returning directly along the original entry path, in order to avoid thermal runaway propagation causing the original path to enter a new hazard area.

[0084] The return path prioritizes avoiding areas with major heat sources, high smoke concentrations, and significant structural deformation. It also prioritizes flight paths with stable image transmission links, open spaces, and the ability to quickly escape the danger zone. The return path also adheres to preset safety distance constraints to ensure sufficient safety margin between the UAV and the danger zone boundary during the return process.

[0085] Preferably, the return cost function corresponding to the return path can be expressed as:

[0086] in, Indicates the cost of the return route; This indicates the remaining energy consumption cost of the drone; , This represents the corresponding weight coefficient; N represents the number of path sampling points.

[0087] Specifically, the system evaluates multiple candidate return paths based on the updated dynamic hazard domain and selects the path with the lowest return cost. If the dynamic hazard domain continues to expand during the return process, a new return path is generated, prioritizing ensuring that the drone escapes the range of high-temperature jets and areas obscured by smoke.

[0088] Furthermore, during the return journey of the UAV, the image transmission link quality, positioning reliability, remaining battery power, and changes in the dynamic hazard domain are continuously monitored. When a decrease in image transmission link quality, insufficient positioning reliability, or continued expansion of the dynamic hazard domain boundary towards the return journey is detected, the return journey is replanned to improve the safety and stability of the return process.

[0089] After the mission is completed, the control platform automatically summarizes all data from the approach re-inspection process and generates a re-inspection report. The re-inspection report includes at least the alarm time, UAV takeoff time, approach trajectory, thermal images at each stage, visible light images, point cloud data, hazard zone change process, final re-inspection conclusion, and return trajectory, for use in subsequent accident analysis, emergency response review, and energy storage equipment maintenance decisions.

[0090] It should be noted that the various weighting coefficients, various thresholds and update time windows in this embodiment can be preset or adaptively adjusted according to the scale of the energy storage station, the type of energy storage cabinet, the model of the drone, the on-site environmental conditions and historical experimental data. This application does not limit them.

[0091] The above-mentioned plan, after completing the re-inspection and scanning, does not simply return along the original route, but selects the optimal evacuation route based on the latest developments of the accident. It fully considers the possible expansion of the danger zone or new risks during the return phase, prioritizes avoiding areas with high heat and dense smoke, and takes into account energy consumption and communication quality, ensuring that the UAV can safely and efficiently evacuate the danger zone in the final stage of the mission.

[0092] The present invention also provides an approach re-inspection system for an energy storage thermal runaway drone, which is used to implement an approach re-inspection method for an energy storage thermal runaway drone. The system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the approach re-inspection method for an energy storage thermal runaway drone.

[0093] This invention provides a storage medium storing a program that, when executed by a processor, implements a method for close-range re-inspection of an energy-storage thermal runaway unmanned aerial vehicle.

[0094] This invention provides a processor for running a program, wherein the program executes a method for re-inspecting an energy storage thermal runaway unmanned aerial vehicle.

[0095] This invention provides a device including a processor, a memory, and a program stored in the memory and executable on the processor. When the processor executes the program, it implements a method for close-range re-inspection of an energy-storage thermal runaway unmanned aerial vehicle (UAV). The device described herein can be a server, PC, tablet, mobile phone, etc.

[0096] This application also provides a computer program product that, when executed on a data processing device, is suitable for performing a method for close-range re-inspection of an energy storage thermal runaway unmanned aerial vehicle.

[0097] Those skilled in the art will understand that embodiments of this application can provide methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0099] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0100] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0101] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0102] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0103] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0104] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0105] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for close-range re-inspection of an energy storage thermal runaway unmanned aerial vehicle, characterized in that, include: In response to a thermal runaway alarm, a re-inspection task is performed, and on-site data is acquired during the re-inspection task. Based on the aforementioned field data, a comprehensive spatial risk value is calculated to construct a dynamic hazard domain; Based on the dynamic hazard domain, a graded approach strategy is generated, and a corresponding flight path is generated based on the graded approach strategy; According to the graded approach strategy, the UAV is controlled to perform graded approach flight along the flight path, and during the graded approach flight, the dynamic hazard domain, the graded approach strategy and the flight path are updated based on the real-time acquired field data until the re-inspection position is reached to perform re-inspection scanning; After completing the re-inspection scan, a return path is generated based on the latest dynamic hazard domain, and the drone is controlled to return to its home location.

2. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 1, characterized in that, The acquisition of on-site data includes: Control the drone to fly to the preset initial safe reconnaissance position; At the initial security reconnaissance position, the airborne infrared thermal imaging camera, lidar, and visible light camera are activated; The infrared thermal imaging camera, lidar, and visible light camera are controlled to perform segmented scanning of the target area to collect thermal imaging data, three-dimensional point cloud data, and visible light image data of the target area, and the pose data of the UAV is recorded simultaneously.

3. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 2, characterized in that, The segmented scanning includes: A first scan is performed on the area directly in front of the target region; A second lateral scan is performed on the target region; and A third scan is performed above the target area.

4. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 1, characterized in that, The calculation of the spatial comprehensive risk value based on the on-site data includes: The acquired temperature data, temperature rise rate data, smoke concentration data, structural deformation data, and wind direction influence data were normalized. The normalized data are weighted and summed according to preset weights to obtain the comprehensive risk value of the sampling point.

5. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 1, characterized in that, The construction of the dynamic hazard domain includes: Based on the comprehensive spatial risk value, spatial locations that meet the preset judgment conditions are determined; Based on the thermal characteristics, smoke characteristics and structural deformation characteristics corresponding to the spatial locations, the main heat source area, the smoke-covered area and the structural deformation area are divided. The main heat source area, the smoke-covered area, and the structural deformation area are spatially fused to obtain a dynamic hazard domain.

6. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 1, characterized in that, The generation of a tiered approach strategy, and the generation of a corresponding flight path based on the tiered approach strategy, includes: Multiple candidate approach locations were identified in the spatial region outside the dynamic hazard zone; The overall suitability of each candidate approach location is calculated. The overall suitability is determined based on the target visibility of the corresponding candidate approach location, the clearance distance between the candidate approach location and the boundary of the dynamic hazard zone, the target observation angle, the image transmission link quality, and the overall risk value at the corresponding location. Select the candidate approach location with the highest overall suitability as the next approach location; Based on the drone's current location and the next approach location, a corresponding flight path is planned and generated.

7. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 1, characterized in that, The updating of the dynamic hazard domain, the graded approach strategy, and the flight path based on real-time acquired field data includes: During the graded approach flight, it is determined whether the real-time acquired field data meets the preset update trigger conditions; When the update triggering condition is met for multiple consecutive calculation cycles, and the time since the last update exceeds a preset update time window, dynamic danger domain update is triggered. The spatial comprehensive risk value is recalculated based on the latest field data to update the dynamic hazard domain; Based on the updated dynamic hazard domain, the safety distance constraints, flight speed constraints, and observation angle constraints corresponding to the current approach phase are redefined. The flight path for the current stage is replanned based on the updated constraint parameters.

8. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 7, characterized in that, The graded approach flight includes multiple consecutive approach phases; After each approach phase is completed, a security check is performed on the updated dynamic hazard domain; Once it is confirmed that a preset safe distance is maintained between the updated dynamic hazard zone and the flight path corresponding to the next approach phase, the UAV is controlled to enter the next approach phase.

9. The method for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles according to claim 7, characterized in that, The preset update trigger condition includes any one of the following: The change in the current overall risk value within adjacent calculation periods exceeds the preset risk jump threshold; The real-time distance between the drone and the boundary of the dynamic hazard zone is less than a preset dynamic safety distance threshold. The real-time image transmission link quality is lower than the preset minimum link quality threshold. The boundary of the dynamic danger zone expands toward the direction of the UAV's current flight path.

10. A system for close-range re-inspection of energy storage thermal runaway unmanned aerial vehicles, characterized in that, The system includes a control module, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the approach re-inspection method for energy storage thermal runaway unmanned aerial vehicles according to any one of claims 1-9.

Images