Power grid emergency disposal system based on low-altitude intelligent three-dimensional coordination

Abstract

This invention discloses a power grid emergency response system based on low-altitude intelligent three-dimensional collaboration, belonging to the field of dynamic energy consumption scheduling technology. By real-time sensing of environmental energy consumption and edge computing load status, it calculates a "three-dimensional energy-computing asymmetric coupling index." When encountering high energy consumption limits, the system dynamically generates a computing power offloading gain signal and a cluster topology reconstruction density factor, offloading complex calculations to the cloud or helicopters and reducing the density of non-core clusters, significantly reducing end-side power consumption to extend continuous operation time. Simultaneously, the system constructs a dynamic spatiotemporal damping feedback mechanism, forcing the simulation refresh rate of the rear command center to maintain spatiotemporal synchronization with the limited computing power at the front end, completely eliminating scheduling misalignment and obstacle avoidance conflicts caused by hidden frequency reduction due to batteries. In extreme failure scenarios, the system automatically degrades to forced landing or relay mode, ensuring the minimum survival of emergency network nodes and the absolute safety of collaborative logic.

Description

Technical Field

[0001] This invention relates to the field of dynamic energy consumption scheduling technology, specifically to a power grid emergency response system based on low-altitude intelligent three-dimensional collaboration. Background Technology

[0002] Low-altitude power grid emergency response technology, based on the collaboration of helicopters, drones, and AI, originated from traditional manual inspections of power lines traversing mountains and rivers, and early extensive helicopter patrols. With breakthroughs in aircraft miniaturization, multi-source sensors, and edge computing visual algorithms, this technology has advanced to an intelligent stage of "manned + unmanned, edge-cloud collaboration." It perfectly complements the long endurance and heavy payload of helicopters, the flexible micro-management of drones, and the second-level defect identification of AI. Currently, this technology is widely used in the fields of daily three-dimensional inspection of ultra-high voltage power grids, rapid damage assessment in disaster areas after extreme natural disasters (such as snowstorms and typhoons), non-contact obstacle removal in complex geographical environments, and communication relay in network outage environments, among other applications in the power and emergency rescue sectors.

[0003] In power grid emergency scenarios (such as severe cold and icing, high altitude and strong winds), drones not only have to overcome adverse weather conditions to maintain flight attitude, but their onboard high-definition gimbals, lidar, and AI edge computing modules are also major power consumers. This multi-layered pressure of "high-intensity flight + full payload + high computing power concurrency" quickly pushes the physical limits of existing lithium batteries, causing a precipitous drop in usable battery capacity. A flight time originally rated at 40 minutes may be reduced to just a few minutes or even a dozen minutes in actual disaster relief, triggering a low battery alarm and forcing the equipment to return frequently, directly disrupting the continuity of emergency reconnaissance and repair.

[0004] Due to the rapid power depletion under these extreme conditions, the drone's underlying battery management system (BMS) will forcibly trigger a self-preservation mechanism to avoid crashing. This mechanism subtly "reduces the frequency" of the AI ​​edge computing chip and may even dynamically cut off some broadband communication to save power. This imperceptible system-level compromise can lead to dropped frames in the footage transmitted from the drone, missed defect detection, and fatal delays in coordinate transmission. Meanwhile, the helicopter in the air and the AI ​​command center at the rear are still conducting tactical simulations and scheduling based on a "high-frequency, zero-latency" model. This spatiotemporal information asymmetry and computing power gap caused by "forced power saving" can trigger a serious chain reaction: the command center may issue incorrect delayed coordinates to the helicopter, resulting in a missed attack, or it may cause obstacle avoidance logic conflicts among multiple drones in the air, ultimately causing the originally precisely coordinated "three-dimensional collaborative network" to fall into a logical deadlock or degenerate into independent operations. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a power grid emergency response system based on low-altitude intelligent three-dimensional collaboration, in order to solve the problems in existing technologies.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a low-altitude intelligent three-dimensional collaborative power grid emergency response system, comprising: The environment and platform energy consumption sensing module is used to acquire the first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of low-altitude aircraft. The edge computing power and task load perception module is used to obtain a second set of state parameters that characterize the concurrency of visual recognition and the latency of communication interaction on the aircraft's end side. The three-dimensional energy computing asymmetric coupling solution module is used to perform time-domain feature fusion and normalization processing on the first state parameter set and the second state parameter set using a preset dynamic energy computing game mapping model, and calculate the three-dimensional energy computing asymmetric coupling index. The dynamic adaptive strategy generation module is used to generate the corresponding edge-cloud computing power offloading gain signal and cluster topology reconstruction density factor based on the three-dimensional computing asymmetric coupling index. The three-dimensional collaborative adaptive control module is used to perform adaptive linkage control on the low-altitude emergency collaborative state by constructing multi-dimensional control commands based on the collaborative confidence level characterized by the three-dimensional energy computing asymmetric coupling index and the quality characteristics of the underlying power status. The adaptive linkage control includes correcting the computing frequency matrix of the single-machine edge AI chip, dynamically adjusting the data transmission link parameters between the UAV and the high-altitude helicopter, and adjusting the spatial obstacle avoidance feedback frequency of the aircraft group in real time, and performing adaptive switching between multiple preset collaborative scheduling modes.

[0007] Preferably, the environment and platform energy consumption perception module acquires a first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of the low-altitude aircraft, including acquiring real-time collected wind resistance output power data of the aircraft's power motor and extracting time-domain features, and calculating a normalized kinetic energy depletion index used to quantify the attitude maintenance cost under extreme weather conditions. Real-time collected thermal and voltage drop waveform data of the battery management system are acquired and nonlinear dynamic analysis is performed to calculate the underlying charge warning multi-scale entropy used to quantify the discharge penalty effect under low temperature or high load; wherein, the first set of state parameters is composed of the normalized kinetic energy depletion index and the underlying charge warning multi-scale entropy.

[0008] Preferably, the edge computing power and task load perception module acquires a second set of state parameters characterizing the concurrency of visual recognition and communication interaction latency of the aircraft end side, including estimating the computing power occupancy spectrum of the AI ​​edge inference stream generated by the aircraft during the disaster damage assessment process, and calculating the inference computing power centroid offset rate used to characterize the overload trend of computing resources. The topological distance between the real-time extracted power grid fault topology manifold and the preset standardized disaster loss evolution model is compared to calculate the emergency task mapping manifold degree, which is used to characterize the task complexity and communication bandwidth occupancy deviation; wherein, the second set of state parameters consists of the inference computing power centroid offset rate and the emergency task mapping manifold degree.

[0009] Preferably, the process of the three-dimensional computing asymmetric coupling solution module calculating the three-dimensional computing asymmetric coupling index using the dynamic computing game mapping model includes using the normalized kinetic energy depletion index as a weight adjustment factor to perform dynamic weighted correction on the inference computing power center of gravity offset rate, so as to identify and filter out non-characteristic computing power disturbances induced by aircraft mechanical vibration, and generate corrected task computing power feedback parameters.

[0010] Preferably, the three-dimensional energy computing asymmetric coupling solution module further includes: comparing the underlying power warning multi-scale entropy with the preset forced frequency reduction warning threshold in the external configuration data; when the underlying power warning multi-scale entropy indicates that the current power drop risk exceeds the threshold, performing attenuation processing on the contribution weight of the emergency task mapping manifold in the dynamic energy computing game mapping model, so as to suppress the sudden change in computing power degradation that may be caused under high energy consumption conditions in advance.

[0011] Preferably, the three-dimensional computing asymmetric coupling solution module further includes: inputting the corrected task computing power feedback parameters and the emergency task mapping manifold parameters after weight attenuation into a preset nonlinear mapping function for solution, and outputting a three-dimensional computing asymmetric coupling index that characterizes the spatiotemporal matching determinism between the remaining computing power on the current end side and the high-frequency dispatch instructions of the command center.

[0012] Preferably, in the dynamic adaptive strategy generation module, multi-dimensional control instructions for driving linkage control are generated based on the three-dimensional computing asymmetric coupling index; the multi-dimensional control instructions include an end-cloud computing power offloading gain signal for adjusting computing power allocation, and a cluster topology reconstruction density factor for adjusting the spatial cluster distribution. The process of generating the edge-cloud computing power offloading gain signal includes: obtaining a real-time three-dimensional computing asymmetric coupling index and subtracting a preset static computing power balance benchmark value to obtain a first difference; inputting the first difference into a hyperbolic tangent activation function to perform mapping to obtain an activation output value; multiplying the activation output value by a preset maximum bandwidth transmission upper limit constant to obtain the edge-cloud computing power offloading gain signal; the process of generating the cluster topology reconstruction density factor includes: subtracting the real-time three-dimensional computing asymmetric coupling index from a unit value of 1 to obtain a second difference; performing a square operation on the second difference to obtain an attenuation ratio; applying the attenuation ratio to the relative safe communication distance and airspace occupancy value of non-core operation UAVs in the low-altitude collaborative network to generate the cluster topology reconstruction density factor.

[0013] Preferably, in the stereo cooperative adaptive control module, when the stereo energy computing asymmetric coupling index indicates that the current cooperative confidence is in the first preset warning range, the aircraft is determined to be in a "high energy consumption computing power limited" state: at this time, the amplitude of the end-cloud computing power offloading gain signal is increased, the complex three-dimensional point cloud modeling task on the UAV end side is forcibly cut off, the original video stream is compressed and offloaded to the upper-level helicopter node or the rear command center for cloud identification, so as to suppress the identification missed due to the forced frequency reduction caused by the battery management system; By adjusting the density factor of the swarm topology reconstruction, the flight speed and swarm density of non-core operation drones are reduced, thereby filtering out the obstacle avoidance calculation load in secondary airspace and reducing the overall computing power consumption of the swarm.

[0014] Preferably, in the three-dimensional collaborative adaptive control module, a dynamic spatiotemporal damping feedback mechanism based on the three-dimensional energy computing asymmetric coupling index is established. According to the rate of change of the coupling index, the update frequency of the coordinates and disaster images transmitted back by the UAV to the command center is adjusted in real time. This realizes the nonlinear synchronous coupling between the simulation and refresh speed of the digital twin model of the command center and the real physical bottleneck of the front-end UAV, thereby constructing a stable collaborative environment to prevent disconnection after the compromise of computing power.

[0015] Preferably, in the three-dimensional cooperative adaptive control module, the signal integrity and available battery life extreme value of the first state parameter set are monitored in real time; When it is determined that the communication of the first set of state parameters is interrupted or the battery health is lower than the preset crash threshold, the complex input of the dynamic energy computing game mapping model is automatically cut off, and the real-time collaborative mode is downgraded to the original route forced landing mode or the fixed-point hovering communication relay mode based on inertial navigation, so as to ensure the minimum survival and basic availability of emergency network nodes in extreme scenarios of dual failure of computing power and power.

[0016] This invention provides a power grid emergency response system based on low-altitude intelligent three-dimensional coordination. It has the following beneficial effects: (1) The low-altitude intelligent three-dimensional collaborative power grid emergency response system obtains a first set of state parameters consisting of a normalized kinetic energy depletion index and a multi-scale entropy of the underlying power warning through the environment and platform energy consumption perception module, and obtains a second set of state parameters consisting of the inference computing power centroid offset rate and the emergency task mapping manifold through the edge computing power and task load perception module. The system is then fused and processed by the three-dimensional energy computing asymmetric coupling solution module and outputs the three-dimensional energy computing asymmetric coupling index. Furthermore, the dynamic adaptive strategy generation module generates the corresponding edge cloud computing power offloading gain signal and cluster topology reconstruction density factor based on the index, so that the three-dimensional collaborative adaptive control module can determine when the aircraft is in a "high energy consumption and limited computing power" state. By increasing the amplitude of the edge computing power offloading gain signal, the complex preset 3D point cloud modeling task on the UAV end side is forcibly cut off, and the original video stream is compressed and offloaded to the upper-level helicopter node or the rear command center. At the same time, the flight speed and cluster density of non-core operation UAVs are reduced by using the cluster topology reconstruction density factor to filter out the obstacle avoidance computing load in secondary airspace. The above system mechanism accurately eliminates non-critical computing power loss and significantly reduces the overall power consumption of the cluster edge computing chip. It successfully breaks through the "power consumption wall" limit under severe weather and full load, effectively avoids frequent equipment return due to low power alarms, fundamentally extends the effective working time, and ensures the long-term continuity of the power grid post-disaster emergency repair network.

[0017] (2) In this low-altitude intelligent three-dimensional collaborative power grid emergency response system, when the risk of power drop indicated by the multi-scale entropy of the underlying power warning exceeds the preset forced frequency reduction warning threshold, the three-dimensional energy computing asymmetric coupling solution module performs attenuation processing on the contribution weight of the emergency task mapping manifold in the dynamic energy computing game mapping model in advance to suppress the sudden change in computing power degradation under high energy consumption. On this basis, the three-dimensional collaborative adaptive control module establishes a dynamic spatiotemporal damping feedback mechanism based on the three-dimensional energy computing asymmetric coupling index, and adjusts the update frequency of the UAV transmitting coordinates and disaster images back to the command center in real time according to the rate of change of the coupling index. This nonlinear synchronous coupling strategy based on the real physical bottleneck of the front end The system ensures that the simulation and refresh speed of the digital twin model of the command center is synchronized with the limited computing power frequency of the terminal side, completely eliminating the high-frequency scheduling command delay mismatch and obstacle avoidance logic conflict between multiple UAVs caused by the hidden "forced power saving". In addition, the system monitors the signal integrity and available endurance extreme value of the first state parameter set in real time. Under the extreme value red line state, it automatically cuts off complex inputs and downgrades the real-time collaborative mode to the original route forced landing mode or the fixed-point hovering communication relay mode based on inertial navigation. Ultimately, it ensures the minimum node survival and basic availability of the power grid emergency system under extreme computing power and power failure scenarios, and maintains the absolute logical security of multi-air-ground joint collaborative scheduling. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the framework structure of the power grid emergency response system based on low-altitude intelligent three-dimensional collaboration in this invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Please see Figure 1 This invention provides a power grid emergency response system based on low-altitude intelligent three-dimensional collaboration, comprising: The environment and platform energy consumption sensing module is used to acquire the first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of low-altitude aircraft. The edge computing power and task load perception module is used to obtain a second set of state parameters that characterize the concurrency of visual recognition and the latency of communication interaction on the aircraft's end side. The three-dimensional energy computing asymmetric coupling solution module is used to perform time-domain feature fusion and normalization processing on the first state parameter set and the second state parameter set using a preset dynamic energy computing game mapping model, and calculate the three-dimensional energy computing asymmetric coupling index. The dynamic adaptive strategy generation module is used to generate the corresponding edge-cloud computing power offloading gain signal and cluster topology reconstruction density factor based on the three-dimensional computing asymmetric coupling index. The three-dimensional collaborative adaptive control module is used to perform adaptive linkage control on the low-altitude emergency collaborative state by constructing multi-dimensional control commands based on the collaborative confidence level characterized by the three-dimensional energy computing asymmetric coupling index and the quality characteristics of the underlying power status. The adaptive linkage control includes correcting the computing frequency matrix of the single-machine edge AI chip, dynamically adjusting the data transmission link parameters between the UAV and the high-altitude helicopter, and adjusting the spatial obstacle avoidance feedback frequency of the aircraft group in real time, and performing adaptive switching between multiple preset collaborative scheduling modes.

[0021] The environment and platform energy consumption perception module acquires a first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of the low-altitude aircraft, including acquiring real-time wind resistance output power data of the aircraft's power motor and extracting time-domain features, and calculating the normalized kinetic energy depletion index used to quantify the attitude maintenance cost under extreme weather conditions. Real-time collected thermal and voltage drop waveform data of the battery management system are acquired and nonlinear dynamic analysis is performed to calculate the underlying charge warning multi-scale entropy used to quantify the discharge penalty effect under low temperature or high load; wherein, the first set of state parameters is composed of the normalized kinetic energy depletion index and the underlying charge warning multi-scale entropy.

[0022] The edge computing power and task load perception module acquires a second set of state parameters characterizing the concurrency of visual recognition and communication interaction latency of the aircraft end-side, including estimating the computing power occupancy spectrum of the AI ​​edge inference stream generated by the aircraft during the disaster damage assessment process, and calculating the inference computing power centroid offset rate used to characterize the overload trend of computing resources. The topological distance between the real-time extracted power grid fault topology manifold and the preset standardized disaster loss evolution model is compared to calculate the emergency task mapping manifold degree, which is used to characterize the task complexity and communication bandwidth occupancy deviation; wherein, the second set of state parameters consists of the inference computing power centroid offset rate and the emergency task mapping manifold degree.

[0023] The process of calculating the asymmetric coupling index of the three-dimensional energy computing using the dynamic energy computing game mapping model includes using the normalized kinetic energy depletion index as a weight adjustment factor to perform dynamic weighted correction on the inference computing power center of gravity offset rate, so as to identify and filter out non-characteristic computing power disturbances induced by the mechanical vibration of the aircraft, and generate corrected task computing power feedback parameters.

[0024] The three-dimensional energy computing asymmetric coupling solution module further includes: comparing the underlying power warning multi-scale entropy with the preset forced frequency reduction warning threshold in the external configuration data; when the underlying power warning multi-scale entropy indicates that the current power drop risk exceeds the threshold, performing attenuation processing on the contribution weight of the emergency task mapping manifold in the dynamic energy computing game mapping model, so as to suppress the sudden change in computing power degradation that may be caused under high energy consumption.

[0025] The three-dimensional computing asymmetric coupling solution module further includes: inputting the corrected task computing power feedback parameters and the emergency task mapping manifold parameters after weight attenuation into a preset nonlinear mapping function for solution, and outputting a three-dimensional computing asymmetric coupling index that characterizes the spatiotemporal matching determinism between the remaining computing power on the current end side and the high-frequency dispatch instructions of the command center.

[0026] In the dynamic adaptive strategy generation module, multi-dimensional control instructions for driving linkage control are generated based on the three-dimensional computing asymmetric coupling index. The multi-dimensional control instructions include an end-cloud computing power offloading gain signal for adjusting computing power allocation, and a cluster topology reconstruction density factor for adjusting the spatial cluster distribution. The process of generating the edge-cloud computing power offloading gain signal includes: obtaining a real-time three-dimensional computing asymmetric coupling index and subtracting a preset static computing power balance benchmark value to obtain a first difference; inputting the first difference into a hyperbolic tangent activation function to perform mapping to obtain an activation output value; multiplying the activation output value by a preset maximum bandwidth transmission upper limit constant to obtain the edge-cloud computing power offloading gain signal; the process of generating the cluster topology reconstruction density factor includes: subtracting the real-time three-dimensional computing asymmetric coupling index from a unit value of 1 to obtain a second difference; performing a square operation on the second difference to obtain an attenuation ratio; applying the attenuation ratio to the relative safe communication distance and airspace occupancy value of non-core operation UAVs in the low-altitude collaborative network to generate the cluster topology reconstruction density factor.

[0027] In the stereo cooperative adaptive control module, when the stereo energy computing asymmetric coupling index indicates that the current cooperative confidence is in the first preset warning range, it is determined that the aircraft is in a "high energy consumption computing power limited" state: at this time, the amplitude of the end-cloud computing power offloading gain signal is increased, the complex three-dimensional point cloud modeling task on the UAV end side is forcibly cut off, the original video stream is compressed and offloaded to the upper-level helicopter node or the rear command center for cloud identification, so as to suppress the identification missed due to the forced frequency reduction of the battery management system; By adjusting the density factor of the swarm topology reconstruction, the flight speed and swarm density of non-core operation drones are reduced, thereby filtering out the obstacle avoidance calculation load in secondary airspace and reducing the overall computing power consumption of the swarm.

[0028] In the aforementioned three-dimensional collaborative adaptive control module, a dynamic spatiotemporal damping feedback mechanism based on the three-dimensional energy computing asymmetric coupling index is established. According to the rate of change of this coupling index, the update frequency of the coordinates and disaster images transmitted back to the command center by the UAV is adjusted in real time. This realizes the nonlinear synchronous coupling between the simulation and refresh speed of the digital twin model of the command center and the real physical bottleneck of the front-end UAV, thereby constructing a stable collaborative environment to prevent disconnection after the compromise of computing power.

[0029] In the three-dimensional collaborative adaptive control module, the signal integrity and available battery life extreme value of the first state parameter set are monitored in real time. When it is determined that the communication of the first set of state parameters is interrupted or the battery health is lower than the preset crash threshold, the complex input of the dynamic energy computing game mapping model is automatically cut off, and the real-time collaborative mode is downgraded to the original route forced landing mode or the fixed-point hovering communication relay mode based on inertial navigation, so as to ensure the minimum survival and basic availability of emergency network nodes in extreme scenarios of dual failure of computing power and power.

[0030] Furthermore, the Chinese names that can be quantified numerically will now be uniformly identified using letter symbols: "Normalized Kinetic Energy Depletion Index" will be identified as AE, "Inference Computing Power Center of Gravity Shift Rate" will be identified as AC, "Energy Variance Contribution Rate" will be identified as W1, "Computing Power Variance Contribution Rate" will be identified as W2, "Weighted Energy Computing Fusion Value" will be identified as AS, and "Three-Dimensional Energy Computing Asymmetric Coupling Index" will be identified as AI.

[0031] The specific steps and logical explanation for the deduction of the technical effect in this embodiment are as follows: The first algorithmic formula, expressed in language, originates from the principal component variance calculation model in multivariate statistics. Its derivation logic is that the sum of W1 and W2 equals the natural number one, which is used to determine the weight benchmark. Secondly, the second algorithmic formula, expressed in language, originates from the linear weighted evaluation model. Its derivation logic is that AE is substituted into the product and multiplied by W1, AC is substituted into the product and multiplied by W2, and the sum of the two products yields the result S. In this formula, since AE and AC are both processed dimensionless ratios, and parameters W1 and W2 are dimensionless weights, the physical dimensions on both sides of the algorithm formula are completely consistent, both being dimensionless. Finally, the third algorithm formula, expressed in language, is initially derived from the exponential decay response model in physics. Its derivation logic is that the result of raising the natural constant to the power of negative AS is subtracted from the natural number one, ultimately yielding the result AI.

[0032] Regarding parameter analysis and range interpretation, since aircraft motors and chips cannot be in a state of absolute shutdown or infinite energy consumption in actual physical scenarios, AE and AC are both strictly greater than zero and do not exist in a state equal to zero or the natural number one. This limits AS to the range of positive numbers greater than zero.

[0033] After standardization by the third algorithm formula, the output AI value range is strictly limited to the open interval (0, 1) greater than zero and less than the natural number one. There is absolutely no output value of "0" or "1", which fully meets the technical requirements of data stability. When the output value I of the algorithm formula approaches 0, AE and AC in the corresponding algorithm formula show a synchronous decreasing trend. At this time, the trend is: the drone is in a low-energy consumption and computing power redundancy state, the battery management system does not trigger frequency reduction, and the system does not need to intervene in computing power offloading. When the output value AI of the algorithm formula approaches 1, AE and AC in the corresponding algorithm formula show a continuous increasing trend. At this time, the trend of the invention content is: the drone faces the dual bottleneck of power depletion and computing power overload, and the system forcibly triggers the end-cloud computing power offloading mechanism to transfer the core computing tasks to the backend.

[0034] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A power grid emergency response system based on low-altitude intelligent three-dimensional collaborative operation, characterized in that: include: The environment and platform energy consumption sensing module is used to acquire the first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of low-altitude aircraft. The edge computing power and task load perception module is used to acquire a second set of state parameters that characterize the concurrency of visual recognition and the latency of communication interaction on the aircraft's end side. The three-dimensional energy computing asymmetric coupling solution module is used to perform time-domain feature fusion and normalization processing on the first state parameter set and the second state parameter set using a preset dynamic energy computing game mapping model, and calculate the three-dimensional energy computing asymmetric coupling index. The dynamic adaptive strategy generation module is used to generate the corresponding edge-cloud computing power offloading gain signal and cluster topology reconstruction density factor based on the three-dimensional computing asymmetric coupling index. The three-dimensional collaborative adaptive control module is used to perform adaptive linkage control on the low-altitude emergency collaborative state by constructing multi-dimensional control commands based on the collaborative confidence level characterized by the three-dimensional energy computing asymmetric coupling index and the quality characteristics of the underlying power status. The adaptive linkage control includes correcting the computing frequency matrix of the single-machine edge AI chip, dynamically adjusting the data transmission link parameters between the UAV and the high-altitude helicopter, and adjusting the spatial obstacle avoidance feedback frequency of the aircraft group in real time, and performing adaptive switching between multiple preset collaborative scheduling modes.

2. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 1, characterized in that: The environment and platform energy consumption perception module acquires a first set of state parameters characterizing the physical compressive load and transient energy decay characteristics of the low-altitude aircraft, including acquiring real-time wind resistance output power data of the aircraft's power motor and extracting time-domain features, and calculating the normalized kinetic energy depletion index used to quantify the attitude maintenance cost under extreme weather conditions. Real-time collected thermal and voltage drop waveform data of the battery management system are acquired and nonlinear dynamic analysis is performed to calculate the underlying charge warning multi-scale entropy used to quantify the discharge penalty effect under low temperature or high load; wherein, the first set of state parameters is composed of the normalized kinetic energy depletion index and the underlying charge warning multi-scale entropy.

3. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 2, characterized in that: The edge computing power and task load perception module acquires a second set of state parameters characterizing the concurrency of visual recognition and communication interaction latency of the aircraft end-side, including estimating the computing power occupancy spectrum of the AI ​​edge inference stream generated by the aircraft during the disaster damage assessment process, and calculating the inference computing power centroid offset rate used to characterize the overload trend of computing resources. The topological distance between the real-time extracted power grid fault topology manifold and the preset standardized disaster loss evolution model is compared to calculate the emergency task mapping manifold degree, which is used to characterize the task complexity and communication bandwidth occupancy deviation; wherein, the second set of state parameters consists of the inference computing power centroid offset rate and the emergency task mapping manifold degree.

4. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 3, characterized in that: The process of calculating the asymmetric coupling index of the three-dimensional energy computing using the dynamic energy computing game mapping model includes using the normalized kinetic energy depletion index as a weight adjustment factor to perform dynamic weighted correction on the inference computing power center of gravity offset rate, so as to identify and filter out non-characteristic computing power disturbances induced by the mechanical vibration of the aircraft, and generate corrected task computing power feedback parameters.

5. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 4, characterized in that: The three-dimensional energy computing asymmetric coupling solution module further includes: comparing the underlying power warning multi-scale entropy with the preset forced frequency reduction warning threshold in the external configuration data; when the underlying power warning multi-scale entropy indicates that the current power drop risk exceeds the threshold, performing attenuation processing on the contribution weight of the emergency task mapping manifold in the dynamic energy computing game mapping model, so as to suppress the sudden change in computing power degradation that may be caused under high energy consumption.

6. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 5, characterized in that: The three-dimensional computing asymmetric coupling solution module further includes: inputting the corrected task computing power feedback parameters and the emergency task mapping manifold parameters after weight attenuation into a preset nonlinear mapping function for solution, and outputting a three-dimensional computing asymmetric coupling index that characterizes the spatiotemporal matching determinism between the remaining computing power on the current end side and the high-frequency dispatch instructions of the command center.

7. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 6, characterized in that: In the dynamic adaptive strategy generation module, multi-dimensional control instructions for driving linkage control are generated based on the three-dimensional computing asymmetric coupling index. The multi-dimensional control instructions include an end-cloud computing power offloading gain signal for adjusting computing power allocation, and a cluster topology reconstruction density factor for adjusting the spatial cluster distribution. The process of generating the edge-cloud computing power offloading gain signal includes: obtaining a real-time three-dimensional computing asymmetric coupling index and subtracting a preset static computing power balance benchmark value to obtain a first difference; inputting the first difference into a hyperbolic tangent activation function to perform mapping to obtain an activation output value; multiplying the activation output value by a preset maximum bandwidth transmission upper limit constant to obtain the edge-cloud computing power offloading gain signal; the process of generating the cluster topology reconstruction density factor includes: subtracting the real-time three-dimensional computing asymmetric coupling index from a unit value of 1 to obtain a second difference; performing a square operation on the second difference to obtain an attenuation ratio; applying the attenuation ratio to the relative safe communication distance and airspace occupancy value of non-core operation UAVs in the low-altitude collaborative network to generate the cluster topology reconstruction density factor.

8. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 7, characterized in that: In the stereo cooperative adaptive control module, when the stereo energy computing asymmetric coupling index indicates that the current cooperative confidence is in the first preset warning range, it is determined that the aircraft is in a "high energy consumption computing power limited" state: at this time, the amplitude of the end-cloud computing power offloading gain signal is increased, the complex three-dimensional point cloud modeling task on the UAV end side is forcibly cut off, the original video stream is compressed and offloaded to the upper-level helicopter node or the rear command center for cloud identification, so as to suppress the identification missed due to the forced frequency reduction of the battery management system; By adjusting the density factor of the swarm topology reconstruction, the flight speed and swarm density of non-core operation drones are reduced, thereby filtering out the obstacle avoidance calculation load in secondary airspace and reducing the overall computing power consumption of the swarm.

9. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 8, characterized in that: In the aforementioned three-dimensional collaborative adaptive control module, a dynamic spatiotemporal damping feedback mechanism based on the three-dimensional energy computing asymmetric coupling index is established. According to the rate of change of this coupling index, the update frequency of the coordinates and disaster images transmitted back to the command center by the UAV is adjusted in real time. This realizes the nonlinear synchronous coupling between the simulation and refresh speed of the digital twin model of the command center and the real physical bottleneck of the front-end UAV, thereby constructing a stable collaborative environment to prevent disconnection after the compromise of computing power.

10. The power grid emergency response system based on low-altitude intelligent three-dimensional collaboration according to claim 9, characterized in that: In the three-dimensional collaborative adaptive control module, the signal integrity and available battery life extreme value of the first state parameter set are monitored in real time. When it is determined that the communication of the first set of state parameters is interrupted or the battery health is lower than the preset crash threshold, the complex input of the dynamic energy computing game mapping model is automatically cut off, and the real-time collaborative mode is downgraded to the original route forced landing mode or the fixed-point hovering communication relay mode based on inertial navigation, so as to ensure the minimum survival and basic availability of emergency network nodes in extreme scenarios of dual failure of computing power and power.

Images