Power line inspection method and unmanned aerial vehicle automatic airport vehicle

Abstract

This invention relates to the field of power line inspection technology, specifically disclosing a power line inspection method and an automated unmanned aerial vehicle (UAV) airport vehicle. By real-time monitoring of the spectral efficiency and channel state information feedback delay time of multiple UAVs, abnormal spectral efficiency characteristic values ​​and feedback delay characteristic values ​​are calculated to assess the stability of the communication link and the risk of channel prediction errors. These characteristics are then used to construct a multi-dimensional congestion warning vector, which is input into a pre-trained deep reinforcement learning optimization decision model. The model outputs optimal frequency band allocation, time slot scheduling, and power control strategies, achieving adaptive optimization of the concurrent communication performance of multiple UAVs. This invention combines Haar wavelet transform, random matrix theory, and artificial intelligence technology to improve the system's robustness, resource utilization efficiency, and quality of service assurance capabilities in complex dynamic environments, ensuring the stability of UAV communication and thus accurately completing inspection tasks.

Description

Technical Field

[0001] This invention relates to the field of power line inspection technology, specifically to a power line inspection method and an unmanned aerial vehicle (UAV) automated airport vehicle. Background Technology

[0002] Power line inspections typically utilize drone teams, mounted on automated guided vehicles (AGVs). These drones usually transmit information wirelessly, which is susceptible to interference from various factors, affecting the timeliness of their response to inspection commands. In high-density multi-drone scenarios, achieving efficient spectrum resource utilization and intelligent collaborative scheduling among multiple drones has become a key technical challenge in current wireless communication system design. Existing technologies often rely on single metrics such as Channel Quality Indicator (CQI) or average throughput for resource allocation decisions, lacking a comprehensive evaluation mechanism for communication link stability and channel prediction reliability. Furthermore, traditional methods struggle to dynamically perceive network conditions and optimize resource strategies in real-time when facing sudden interference, spectrum efficiency fluctuations, and uncertainties in feedback delays, thus impacting overall system stability and service quality.

[0003] Shortcomings of existing technology: This invention proposes an innovative solution to the communication instability problem caused by abnormal fluctuations in spectral efficiency in concurrent multi-UAV communication. Traditional technologies often focus only on static indicators such as instantaneous channel state information (CSI) or average throughput, neglecting the trend of spectral efficiency changes over time and its potential threat to system stability. Especially in scenarios with sudden interference, enhanced UAV mobility, or rapid channel fading, spectral efficiency may fluctuate drastically. If this is not identified and addressed in time, it will lead to a sharp drop in communication performance or even connection interruption. However, existing resource scheduling algorithms lack effective stability assessment mechanisms and cannot provide early warning and intervention for such non-stationary communication behaviors, limiting the robustness and adaptability of the system in complex dynamic environments, making it difficult to complete inspection tasks with high quality. Summary of the Invention

[0004] To address the aforementioned technical problems, this technical solution provides a power line inspection method and an unmanned aerial vehicle (UAV) automated airport vehicle, thus resolving the issues raised in the background section.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows: Power line inspection methods include: S1: Acquire information on multiple drones controlled by the drone automated airport vehicle, and monitor the spectral efficiency and channel status information feedback delay time of the wireless communication of the multiple drones in real time. S2: Perform trend analysis on spectrum efficiency, calculate spectrum efficiency anomaly characteristic values ​​based on the stability of spectrum efficiency, and use them to evaluate the stability of wireless communication. S3: Calculate the feedback delay characteristic value based on the channel state information feedback delay time, which is used to assess whether there is a risk of prediction error in the channel; S4: Construct a multi-dimensional congestion warning vector from the spectral efficiency anomaly feature value and the feedback delay feature value, input it into the pre-trained optimization decision model for analysis, and output a resource allocation strategy; S5: Dynamically adjusts frequency band allocation, time slot scheduling, and power control parameters based on resource allocation strategies to achieve adaptive optimization of the concurrent communication performance of multiple UAVs.

[0006] Preferably, the evaluation of the stability of wireless communication specifically includes: The system acquires the spectral efficiency of multi-UAV wireless communication in real time, performs trend analysis on the spectral efficiency, calculates abnormal spectral efficiency characteristic values ​​based on the stability of the spectral efficiency, and determines whether the abnormal spectral efficiency characteristic values ​​are greater than or equal to a preset threshold. If they are, the wireless communication is unstable; otherwise, the wireless communication is stable.

[0007] Preferably, the process for obtaining the spectral efficiency anomaly characteristic value is as follows: Real-time acquisition of spectrum efficiency sequences in multi-UAV wireless communication; The Haar wavelet transform was applied to the spectral efficiency sequence for multi-scale decomposition to obtain detail coefficients and approximation coefficients. Calculate the energy distribution of wavelet detail coefficients at each level; Calculate the mean and standard deviation of all layers of Haar wavelet transform, and combine the energy distribution of wavelet detail coefficients of each layer to calculate the spectral efficiency anomaly characteristic value.

[0008] Preferably, the assessment of whether the channel has prediction error risk specifically includes: The feedback delay time of channel state information in multi-UAV wireless communication is acquired in real time. Based on the feedback delay time of channel state information, the feedback delay characteristic value is calculated. It is determined whether the feedback delay characteristic value is greater than or equal to a preset threshold. If it is, the channel has a prediction error risk; otherwise, the channel does not have a prediction error risk.

[0009] Preferably, the process for obtaining the feedback delay feature value is as follows: Collect a sequence of channel status information reported by multiple unmanned aerial vehicles (UAVs) within a preset time period; Construct the delay time matrix and then construct the delay covariance matrix based on random matrix theory; Calculate the maximum eigenvalue λmax and minimum eigenvalue λmin of the covariance matrix RΔT, and then calculate the feedback delay eigenvalue based on the maximum eigenvalue λmax and minimum eigenvalue λmin.

[0010] Preferably, the step of constructing a multi-dimensional congestion warning vector from the spectral efficiency anomaly feature value and the feedback delay feature value, and inputting it into a pre-trained optimization decision model for analysis, specifically includes: The model acquires spectral efficiency anomaly features and feedback delay features, and constructs a multi-dimensional congestion warning vector from these features. This vector serves as input to the optimization decision model. The model learns resource allocation strategies under different network conditions through pre-training. Upon receiving the current warning information, it evaluates the expected benefits of different actions based on a defined utility function, which comprehensively considers signal-to-noise ratio gain, interference cost, and energy consumption cost. Finally, the model outputs a set of optimal combinations of frequency band allocation, time slot scheduling, and power control parameters.

[0011] Preferably, the specific construction process of the optimization decision model is as follows: The pre-trained optimization decision model adopts a deep reinforcement learning architecture. During the training phase, it learns to establish a nonlinear mapping relationship between congestion warning vectors and optimal resource allocation strategies by simulating a large number of different network loads, channel conditions, and UAV behavior scenarios. In actual operation, the model dynamically adjusts the frequency band allocation strategy, time slot scheduling order, and power control level according to the real-time input congestion warning vectors to achieve continuous optimization of the concurrent communication performance of multiple UAVs. The model parameters are continuously updated through an online incremental learning mechanism to adapt to the dynamic changes in the network environment.

[0012] Preferably, the dynamic adjustment of frequency band allocation, time slot scheduling, and power control parameters based on the resource allocation strategy achieves adaptive optimization of the concurrent communication performance of multiple UAVs. Based on the resource allocation strategy output by the optimization decision model, the target frequency band with the least interference and the best channel quality is dynamically selected for UAV access allocation; the time slot scheduling priority and allocation ratio are dynamically adjusted based on the UAV service type and traffic requirements; at the same time, the transmit power level is adjusted in real time in combination with link budget information to reduce mutual interference between multiple UAVs while ensuring communication quality, thereby achieving multi-objective adaptive optimization of system throughput, connection stability and energy efficiency.

[0013] Preferably, the dynamic selection of the target frequency band with the least interference and the best channel quality for UAV access allocation specifically includes: By analyzing channel occupancy, noise intensity, and historical throughput data of each frequency band in real time, combined with spectrum scanning results of the UAV's location, the communication quality indicators of each available frequency band are evaluated, and UAVs are prioritized for allocation to the optimal non-overlapping frequency band in the current environment to reduce spectrum conflicts and improve spectrum utilization. The dynamic adjustment of time slot scheduling priority and allocation ratio based on UAV service type and traffic requirements specifically includes: identifying the service level of UAV service flows, setting corresponding scheduling priority queues according to the sensitivity of different services to latency and bandwidth, and dynamically adjusting the proportion of high-priority services in time resources. The real-time adjustment of transmit power level based on link budget information specifically includes: dynamically calculating the minimum transmit power required to meet communication quality based on signal strength and bit error rate information fed back from the receiver, ensuring link stability while reducing interference to other UAVs, achieving dual optimization of energy saving and anti-interference.

[0014] The unmanned aerial vehicle (UAV) automatically conducts inspections according to the aforementioned power line inspection methods.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) This invention introduces a method for calculating spectral efficiency anomaly features based on Haar wavelet transform, thus constructing a technical mechanism for multi-scale and refined evaluation of communication link stability. This method decomposes the real-time acquired spectral efficiency time series into multiple layers of wavelets, extracts detail coefficients at each scale, and calculates their energy distribution characteristics, thereby revealing the fluctuation characteristics of spectral efficiency at different time scales. Based on this, combined with the statistical characteristics of wavelet transform coefficients (such as mean and standard deviation), spectral efficiency anomaly features are further constructed to quantify the stability level of the current communication state. This feature can effectively identify abnormal trends such as sudden drops or periodic fluctuations in spectral efficiency, and is more forward-looking and accurate than the traditional judgment method based on instantaneous channel quality indicators. When this feature exceeds a preset threshold, the system can quickly determine that the current wireless communication is in an unstable state, thereby triggering a resource optimization process, dynamically adjusting frequency band allocation, time slot scheduling, and power control strategies, significantly improving the system's anti-interference capability and operational reliability in complex dynamic environments such as high interference, dense UAVs, and rapid channel changes. This mechanism not only enhances the network's ability to detect communication anomalies, but also provides solid data support for subsequent intelligent decision-making, demonstrating the innovation and advancement of this invention in wireless resource management.

[0016] (2) This invention constructs a feedback delay eigenvalue calculation method based on random matrix theory to statistically model the feedback delay of channel state information (CSI) reported by multiple UAVs within a preset time period, forming a spatially correlated delay time matrix. A covariance matrix is ​​then constructed using conjugate transpose operations, and its maximum and minimum eigenvalues ​​are further extracted to calculate the feedback delay eigenvalue. This eigenvalue accurately reflects the distribution characteristics and fluctuation degree of feedback delay in a multi-UAV environment, thereby achieving accurate identification of channel state prediction error risks. Compared to traditional methods relying on a single average delay index, this invention can more comprehensively characterize the dynamic behavior of feedback delay, improving the system's ability to judge the reliability of channel prediction. Based on this, the feedback delay eigenvalue and the spectral efficiency anomaly eigenvalue are jointly constructed into a multi-dimensional congestion warning vector, which is input into a pre-trained deep reinforcement learning optimization decision model. Combined with the defined resource allocation utility function, the model can output the optimal frequency band allocation, time slot scheduling, and power control strategies. This mechanism enables intelligent scheduling and adaptive optimization of wireless resources in multi-drone concurrent communication scenarios. While ensuring communication quality, it effectively reduces interference, increases throughput, and optimizes energy consumption. It significantly enhances the resource management capabilities and operational stability of wireless communication modules in complex and dynamic environments, demonstrating high practicality and innovation. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating the power line inspection method of the present invention. Figure 2 This is a flowchart illustrating the process of obtaining the spectral efficiency anomaly characteristic value according to the present invention. Figure 3 This is a flowchart illustrating the process of obtaining the feedback delay feature value according to the present invention. Detailed Implementation

[0018] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0019] Reference Figure 1 As shown, the power line inspection method includes the following steps: S1: Acquire information on multiple drones controlled by the drone automated airport vehicle, and monitor the spectral efficiency and channel status information feedback delay time of the wireless communication of the multiple drones in real time. S2: Perform trend analysis on spectrum efficiency, calculate spectrum efficiency anomaly characteristic values ​​based on the stability of spectrum efficiency, and use them to evaluate the stability of wireless communication. S3: Calculate the feedback delay characteristic value based on the channel state information feedback delay time, which is used to assess whether there is a risk of prediction error in the channel; S4: Construct a multi-dimensional congestion warning vector from the spectral efficiency anomaly feature value and the feedback delay feature value, input it into the pre-trained optimization decision model for analysis, and output a resource allocation strategy; S5: Dynamically adjusts frequency band allocation, time slot scheduling, and power control parameters based on resource allocation strategies to achieve adaptive optimization of the concurrent communication performance of multiple UAVs.

[0020] In S1, the system acquires data from multiple drones controlled by the automated airport vehicle, and monitors in real time the spectral efficiency and channel state information feedback delay time of the multi-drone wireless communication. Specifically, this includes: In this invention, real-time monitoring of the spectrum efficiency of multi-UAV wireless communication specifically includes the following acquisition and calculation process: through the physical layer monitoring module deployed at the wireless access point, the instantaneous transmission rate, channel bandwidth and signal quality parameters of each connected UAV are periodically acquired; combined with the amount of data successfully transmitted per unit time and the currently allocated spectrum resources, the actual spectrum efficiency value of each UAV is calculated, and this value is stored and updated in a time series to reflect the spectrum utilization under the current network conditions.

[0021] Furthermore, the process of real-time monitoring of channel state information feedback delay time includes: establishing a CSI feedback mechanism between the wireless access point and each UAV device, recording the time difference between the access point initiating a CSI request and receiving complete channel state information, forming a single feedback delay sample; statistically processing the feedback delay samples within multiple sampling periods to obtain the average feedback delay value of each UAV device in the current time period, and using it as the basis for assessing the risk of channel prediction error.

[0022] In S2, trend analysis is performed on spectral efficiency. Based on the stability of spectral efficiency, anomaly characteristic values ​​are calculated to assess the stability of wireless communication. Specifically, this includes: The system acquires the spectral efficiency of multi-UAV wireless communication in real time, performs trend analysis on the spectral efficiency, calculates abnormal spectral efficiency characteristic values ​​based on the stability of the spectral efficiency, and determines whether the abnormal spectral efficiency characteristic values ​​are greater than or equal to a preset threshold. If they are, the wireless communication is unstable; otherwise, the wireless communication is stable.

[0023] Reference Figure 2 As shown, the process for obtaining the abnormal spectral efficiency characteristic value is as follows: Real-time acquisition of spectrum efficiency sequences in multi-UAV wireless communication; The Haar wavelet transform was applied to the spectral efficiency sequence for multi-scale decomposition to obtain detail coefficients and approximation coefficients. The energy distribution of wavelet detail coefficients at each level is calculated using the following expression: ; in, This indicates the number of decomposition levels in the Haar wavelet transform. Indicates the number of detail coefficients. Indicates the first The first layer of Haar wavelet transform A detailed coefficient, Indicates the first Energy of layered Haar wavelet transform; Calculate the mean and standard deviation of all layers of Haar wavelet transform, and combine the energy distribution of wavelet detail coefficients of each layer to calculate the spectral efficiency anomaly characteristic value; The calculation expression for the spectral efficiency anomaly characteristic value is as follows: ; In the formula, This indicates anomaly characteristics in spectral efficiency. This represents the total number of decomposition levels in the Haar wavelet transform. This represents the mean of all layers of Haar wavelet transform. This represents the standard deviation of the Haar wavelet transform.

[0024] It should be noted that this invention achieves a refined assessment of wireless communication stability by introducing a computational mechanism for anomaly spectral efficiency feature values. This mechanism performs multi-scale decomposition of the real-time acquired spectral efficiency sequence using Haar wavelet transform, extracts the energy distribution characteristics of the detail coefficients at each level, and constructs anomaly spectral efficiency feature values ​​by combining the mean and standard deviation of the wavelet coefficients, thereby quantifying the stability state of the current communication link. When this feature value exceeds a preset threshold, the system is determined to be in an unstable state, triggering a resource optimization and adjustment process. Compared to traditional methods that rely solely on instantaneous channel quality indicators for judgment, this invention can detect abnormal fluctuations in communication performance earlier and more accurately, improving the system's robustness and adaptability in complex dynamic environments, demonstrating significant technological innovation and practical value.

[0025] In S3, based on the channel state information feedback delay time, a feedback delay characteristic value is calculated to assess whether the channel has prediction error risk. Specifically, this includes: The feedback delay time of channel state information in multi-UAV wireless communication is acquired in real time. Based on the feedback delay time of channel state information, the feedback delay characteristic value is calculated. It is determined whether the feedback delay characteristic value is greater than or equal to a preset threshold. If it is, the channel has a prediction error risk; otherwise, the channel does not have a prediction error risk.

[0026] Reference Figure 3 As shown, the process of obtaining the feedback delay feature value is as follows: Collect a sequence of channel status information reported by multiple unmanned aerial vehicles (UAVs) within a preset time period; Construct the delay time matrix, and then construct the delay covariance matrix based on random matrix theory: ; in, This represents the conjugate transpose operation. The covariance matrix representing the time delay. Represents the delay time matrix, Indicates the number of drones; Calculate the covariance matrix Maximum eigenvalue and minimum eigenvalue According to the largest eigenvalue and minimum eigenvalue The feedback delay characteristic value is calculated using the following expression: ; In the formula, This represents the feedback delay characteristic value.

[0027] It should be noted that this method collects CSI feedback delay times reported by multiple UAVs within a preset time period, constructs a delay time matrix, and uses random matrix theory to construct a delay covariance matrix. The maximum and minimum eigenvalues ​​are then extracted to calculate the feedback delay eigenvalues. This method fully considers the spatial correlation and dynamic fluctuations of feedback delay in a multi-UAV environment, effectively identifying channel state prediction error risks caused by excessive or abnormally distributed feedback delays. By comparing the feedback delay eigenvalues ​​with preset thresholds, it can accurately determine whether the current channel is in a state of prediction inaccuracy risk, thus providing a crucial decision-making basis for subsequent resource scheduling strategy adjustments. Compared to traditional methods that rely solely on average delay or fixed delay models, this invention has higher sensitivity and discrimination accuracy, significantly improving the reliability of channel prediction and the intelligence level of resource scheduling in highly dynamic, multi-UAV concurrent scenarios, demonstrating outstanding innovation and engineering application value.

[0028] In S4, the spectral efficiency anomaly feature values ​​and feedback delay feature values ​​are constructed into a multi-dimensional congestion warning vector, which is then input into a pre-trained optimization decision model for analysis, outputting a resource allocation strategy, specifically including: The model acquires spectral efficiency anomaly features and feedback delay features, and constructs a multi-dimensional congestion warning vector from these features. This vector serves as input to the optimization decision model. The model learns resource allocation strategies under different network conditions through pre-training. Upon receiving the current warning information, it evaluates the expected benefits of different actions based on a defined utility function, which comprehensively considers signal-to-noise ratio gain, interference cost, and energy consumption cost. Finally, the model outputs a set of optimal combinations of frequency band allocation, time slot scheduling, and power control parameters.

[0029] The specific construction process of the optimization decision model is as follows: The pre-trained optimization decision model adopts a deep reinforcement learning architecture. During the training phase, it learns to establish a nonlinear mapping relationship between congestion warning vectors and optimal resource allocation strategies by simulating a large number of different network loads, channel conditions, and UAV behavior scenarios. In actual operation, the model dynamically adjusts the frequency band allocation strategy, time slot scheduling order, and power control level according to the real-time input congestion warning vectors to achieve continuous optimization of the concurrent communication performance of multiple UAVs. The model parameters are continuously updated through an online incremental learning mechanism to adapt to the dynamic changes in the network environment.

[0030] In S5, frequency band allocation, time slot scheduling, and power control parameters are dynamically adjusted based on resource allocation strategies to achieve adaptive optimization of the concurrent communication performance of multiple UAVs. Specifically, this includes: Based on the resource allocation strategy output by the optimization decision model, the target frequency band with the least interference and the best channel quality is dynamically selected for UAV access allocation; the time slot scheduling priority and allocation ratio are dynamically adjusted based on the UAV service type and traffic requirements; at the same time, the transmit power level is adjusted in real time in combination with link budget information to reduce mutual interference between multiple UAVs while ensuring communication quality, thereby achieving multi-objective adaptive optimization of system throughput, connection stability and energy efficiency.

[0031] The dynamic selection of target frequency bands with minimal interference and optimal channel quality for UAV access allocation specifically includes: By analyzing channel occupancy, noise intensity, and historical throughput data of each frequency band in real time, combined with spectrum scanning results of the UAV's location, the communication quality indicators of each available frequency band are evaluated, and UAVs are prioritized for allocation to the optimal non-overlapping frequency band in the current environment to reduce spectrum conflicts and improve spectrum utilization. The dynamic adjustment of time slot scheduling priority and allocation ratio based on UAV service type and traffic requirements specifically includes: identifying the service level of UAV service flows (such as video streaming, voice calls, or task downloads), setting corresponding scheduling priority queues according to the sensitivity of different services to latency and bandwidth, and dynamically adjusting the proportion of high-priority services in time resources. The real-time adjustment of transmit power level based on link budget information specifically includes: dynamically calculating the minimum transmit power required to meet communication quality based on signal strength and bit error rate information fed back from the receiver, ensuring link stability while reducing interference to other UAVs, achieving dual optimization of energy saving and anti-interference.

[0032] The working principle of this invention is as follows: This invention aims to improve network performance and the intelligence level of resource scheduling in high-density multi-UAV scenarios. By real-time monitoring of spectral efficiency and channel state information (CSI) feedback delay time, spectral efficiency anomaly features and feedback delay features are constructed to assess communication stability and channel prediction error risk. The spectral efficiency anomaly features are obtained through multi-scale energy analysis of the spectral efficiency sequence using Haar wavelet transform, while the feedback delay features are calculated based on the maximum and minimum eigenvalues ​​of the multi-UAV feedback delay covariance matrix using random matrix theory. These two features are then used to construct a multi-dimensional congestion warning vector, which is input into a pre-trained deep reinforcement learning optimization decision model to output the optimal resource allocation strategy. During the training phase, this model learns resource allocation strategies under various network conditions, and during the runtime phase, it dynamically adjusts frequency bands, time slots, and power parameters based on real-time warning information. Adaptive optimization is achieved based on this strategy, including dynamic frequency band selection, service-aware time slot scheduling, and link budget-driven power control, thereby improving system throughput, connection stability, and energy efficiency while ensuring communication quality. This invention achieves multi-objective cooperative optimization, exhibits good robustness and adaptability, and is suitable for complex and dynamic wireless communication environments.

[0033] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0034] The unmanned aerial vehicle (UAV) automatically conducts inspections according to the aforementioned power line inspection methods.

[0035] Furthermore, this solution also proposes a storage medium on which a computer-readable program is stored, which executes the aforementioned power line inspection method when invoked.

[0036] It is understandable that the storage medium can be a magnetic medium, such as a floppy disk, hard disk, or magnetic tape; an optical medium, such as a DVD; or a semiconductor medium, such as a solid-state drive (SSD).

[0037] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for inspecting power lines, characterized in that, include: S1: Acquire information on multiple drones controlled by the drone automated airport vehicle, and monitor the spectral efficiency and channel status information feedback delay time of the wireless communication of the multiple drones in real time. S2: Perform trend analysis on spectrum efficiency, calculate spectrum efficiency anomaly characteristic values ​​based on the stability of spectrum efficiency, and use them to evaluate the stability of wireless communication. S3: Calculate the feedback delay characteristic value based on the channel state information feedback delay time, which is used to assess whether there is a risk of prediction error in the channel; S4: Construct a multi-dimensional congestion warning vector from the spectral efficiency anomaly feature value and the feedback delay feature value, input it into the pre-trained optimization decision model for analysis, and output a resource allocation strategy; S5: Dynamically adjusts frequency band allocation, time slot scheduling, and power control parameters based on resource allocation strategies to achieve adaptive optimization of the concurrent communication performance of multiple UAVs.

2. The power line inspection method according to claim 1, characterized in that, The evaluation of the stability of wireless communication specifically includes: The system acquires the spectral efficiency of multi-UAV wireless communication in real time, performs trend analysis on the spectral efficiency, calculates abnormal spectral efficiency characteristic values ​​based on the stability of the spectral efficiency, and determines whether the abnormal spectral efficiency characteristic values ​​are greater than or equal to a preset threshold. If they are, the wireless communication is unstable; otherwise, the wireless communication is stable.

3. The power line inspection method according to claim 2, characterized in that, The process for obtaining the abnormal spectral efficiency feature value is as follows: Real-time acquisition of spectrum efficiency sequences in multi-UAV wireless communication; The Haar wavelet transform was applied to the spectral efficiency sequence for multi-scale decomposition to obtain detail coefficients and approximation coefficients. Calculate the energy distribution of wavelet detail coefficients at each level; Calculate the mean and standard deviation of all layers of Haar wavelet transform, and combine the energy distribution of wavelet detail coefficients of each layer to calculate the spectral efficiency anomaly characteristic value.

4. The power line inspection method according to claim 3, characterized in that, The assessment of whether the channel has prediction error risk specifically includes: The feedback delay time of channel state information in multi-UAV wireless communication is acquired in real time. Based on the feedback delay time of channel state information, the feedback delay characteristic value is calculated. It is determined whether the feedback delay characteristic value is greater than or equal to a preset threshold. If it is, the channel has a prediction error risk; otherwise, the channel does not have a prediction error risk.

5. The power line inspection method according to claim 4, characterized in that, The process for obtaining the feedback delay feature value is as follows: Collect a sequence of channel status information reported by multiple unmanned aerial vehicles (UAVs) within a preset time period; Construct the delay time matrix and then construct the delay covariance matrix based on random matrix theory; Calculate the maximum eigenvalue λmax and minimum eigenvalue λmin of the covariance matrix RΔT, and then calculate the feedback delay eigenvalue based on the maximum eigenvalue λmax and minimum eigenvalue λmin.

6. The power line inspection method according to claim 5, characterized in that, The process of constructing a multi-dimensional congestion warning vector from spectral efficiency anomaly features and feedback delay features, and inputting it into a pre-trained optimization decision model for analysis, specifically includes: The model acquires spectral efficiency anomaly features and feedback delay features, and constructs a multi-dimensional congestion warning vector from these features. This vector serves as input to the optimization decision model. The model learns resource allocation strategies under different network conditions through pre-training. Upon receiving the current warning information, it evaluates the expected benefits of different actions based on a defined utility function, which comprehensively considers signal-to-noise ratio gain, interference cost, and energy consumption cost. Finally, the model outputs a set of optimal combinations of frequency band allocation, time slot scheduling, and power control parameters.

7. The power line inspection method according to claim 6, characterized in that, The specific construction process of the optimization decision model is as follows: The pre-trained optimization decision model adopts a deep reinforcement learning architecture. During the training phase, it learns to establish a nonlinear mapping relationship between congestion warning vectors and optimal resource allocation strategies by simulating a large number of different network loads, channel conditions and drone behavior scenarios. In actual operation, the model dynamically adjusts the frequency band allocation strategy, time slot scheduling order and power control level based on the real-time input congestion warning vector to continuously optimize the performance of concurrent communication of multiple UAVs. It also continuously updates the model parameters through an online incremental learning mechanism to adapt to the dynamic changes in the network environment.

8. The power line inspection method according to claim 7, characterized in that, The system dynamically adjusts frequency band allocation, time slot scheduling, and power control parameters based on resource allocation strategies to achieve adaptive optimization of the concurrent communication performance of multiple UAVs. Based on the resource allocation strategy output by the optimization decision model, the target frequency band with the least interference and the best channel quality is dynamically selected for UAV access allocation; the time slot scheduling priority and allocation ratio are dynamically adjusted based on the UAV service type and traffic requirements; at the same time, the transmit power level is adjusted in real time in combination with link budget information to reduce mutual interference between multiple UAVs while ensuring communication quality, thereby achieving multi-objective adaptive optimization of system throughput, connection stability and energy efficiency.

9. The power line inspection method according to claim 8, characterized in that, The dynamic selection of target frequency bands with minimal interference and optimal channel quality for UAV access allocation specifically includes: By analyzing the channel occupancy rate, noise intensity, and historical throughput data of each frequency band in real time, and combining the spectrum scanning results of the location of the UAV equipment, the communication quality indicators of each available frequency band are evaluated, and the UAV is preferentially allocated to the optimal non-overlapping frequency band in the current environment in order to reduce spectrum conflict and improve spectrum utilization. The specific steps of dynamically adjusting the time slot scheduling priority and allocation ratio based on UAV service type and traffic demand include: identifying the service level of UAV service flows, setting corresponding scheduling priority queues according to the sensitivity of different services to latency and bandwidth, and dynamically adjusting the proportion of high-priority services in time resources; the specific steps of adjusting the transmit power level in real time based on link budget information include: dynamically calculating the minimum transmit power required to meet communication quality based on the signal strength and bit error rate information fed back by the receiver, reducing interference to other UAVs while ensuring link stability, and achieving dual optimization of energy saving and anti-interference.

10. An unmanned aerial vehicle (UAV) for conducting power line inspection operations according to any one of claims 1-9.

Images