Glacier debris flow monitoring method and system applied to unmanned aerial vehicle inspection

Abstract

The application provides a glacier debris flow monitoring method and system applied to unmanned aerial vehicle inspection, relates to the technical field of natural disaster monitoring, and first generates a dynamically adapted trajectory scheme based on a glacier area risk evolution prediction result and an unmanned aerial vehicle flight constraint condition, then drives an unmanned aerial vehicle multi-source sensing system to collect multi-dimensional data according to the dynamically adapted trajectory scheme, generates a time-sequenced glacier area risk perception data chain, then constructs a risk transmission node network by mining and analyzing the glacier area risk perception data chain, constructs a multi-dimensional monitoring response linkage mechanism based on the risk transmission node network, generates a risk monitoring linkage execution scheme, and finally reversely injects field implementation feedback data into the trajectory adaptation process to update scheme parameters. The application realizes the comprehensiveness, accuracy and timeliness of glacier debris flow monitoring, and effectively reduces disaster risks.

Description

Technical Field

[0001] This invention relates to the field of natural disaster monitoring technology, and more specifically, to a method and system for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections. Background Technology

[0002] Glacier regions, due to their unique geological and climatic conditions, are highly susceptible to natural disasters such as debris flows, posing a serious threat to the surrounding ecological environment, infrastructure, and the safety of people's lives and property. Traditional methods for monitoring glacier debris flows mainly rely on ground-based fixed monitoring equipment, such as rain gauges and displacement sensors. While these devices can acquire relevant data about glacier areas to some extent, they suffer from limitations in monitoring range and data acquisition dimensions. For example, ground-based fixed monitoring equipment struggles to comprehensively obtain crucial information such as the surface morphology and internal structure of glaciers, and the stability and reliability of the equipment face challenges under complex terrain and harsh weather conditions.

[0003] Meanwhile, existing monitoring methods lack the ability to dynamically analyze and predict the evolution of risks in glacier areas, and cannot identify risk transmission paths and key nodes in a timely and accurate manner, making it difficult to take effective preventive and response measures before disasters occur. In addition, traditional monitoring methods are relatively lagging in response mechanisms, making it difficult to achieve coordinated integration of multi-dimensional response measures such as adjusting drone inspection parameters, linking ground monitoring equipment, and pushing early warning information, thus failing to meet the real-time and accuracy requirements of glacier debris flow monitoring. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections, the method comprising: Based on the risk evolution prediction results of glacier areas and the flight constraints of UAVs, dynamic collaborative adaptation processing of inspection trajectories is performed to generate dynamic adaptation trajectory schemes that include trajectory node distribution, flight speed adjustment rules, and collaborative parameters of sensing devices. Based on the dynamic adaptation trajectory scheme, the UAV multi-source perception system is driven to perform hierarchical and precise perception operations, collect multi-dimensional data on the surface morphology, internal structure and surrounding environment of different risk characteristic areas in the glacier region, and generate a time-seriesd glacier region risk perception data chain. By triggering risk transmission node mining and correlation analysis through the risk perception data chain in glacier areas, core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk evolution process are identified, and a risk transmission node network including node correlation and transmission time sequence characteristics is constructed. A multi-dimensional monitoring and response linkage mechanism is constructed based on the risk transmission node network, integrating response measures such as UAV inspection parameter adjustment, ground monitoring equipment linkage, and early warning information push, and generating targeted risk monitoring linkage execution plans. Feedback data from the on-site implementation of the risk monitoring and linkage execution plan is injected into the dynamic collaborative adaptation process of the inspection trajectory, and the parameters of the dynamic adaptation trajectory plan are updated in combination with the real-time risk evolution.

[0005] Furthermore, embodiments of the present invention also provide a glacier debris flow monitoring system applied to unmanned aerial vehicle (UAV) inspections, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to execute the above-described method for monitoring glacial debris flows applied to UAV inspections by executing the machine-executable instructions.

[0006] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, a processor of a glacier debris flow monitoring system applied to UAV inspection reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the glacier debris flow monitoring system applied to UAV inspection to execute the above-described glacier debris flow monitoring method applied to UAV inspection.

[0007] Based on the above, by dynamically and collaboratively adapting the inspection trajectory based on the predicted risk evolution results of glacier areas and the flight constraints of UAVs, a dynamic adaptation trajectory scheme that closely matches the actual risk situation and the flight capabilities of UAVs can be generated. According to this dynamic adaptation trajectory scheme, the UAV multi-source perception system is driven to perform hierarchical and precise perception operations. It can comprehensively collect multi-dimensional data such as the surface morphology, internal structure, and surrounding environment of different risk characteristic areas in the glacier area, and generate a time-seriesd risk perception data chain for the glacier area. By triggering risk transmission node mining and correlation analysis through the risk perception data chain, the core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk evolution process can be accurately identified, and a risk transmission node network containing node correlation relationships and transmission time sequence characteristics can be constructed, thereby presenting the risk transmission path and key links. Based on this risk transmission node network, a multi-dimensional monitoring and response linkage mechanism is constructed, integrating various response measures such as UAV inspection parameter adjustment, ground monitoring equipment linkage, and early warning information push. It generates targeted risk monitoring linkage execution plans, realizing the coordinated and efficient operation of monitoring and response. The on-site implementation feedback data is injected back into the dynamic collaborative adaptation process of the inspection trajectory, and the parameters of the dynamic adaptation trajectory plan are updated in combination with the real-time risk evolution situation. This significantly improves the comprehensiveness, accuracy, and timeliness of glacier debris flow monitoring and effectively reduces disaster risks. Attached Figure Description

[0008] Figure 1This is a schematic diagram of the execution flow of the glacier debris flow monitoring method applied to UAV inspection provided in an embodiment of the present invention.

[0009] Figure 2 This is a schematic diagram of exemplary hardware and software components of a glacier debris flow monitoring system applied to drone inspection, provided in an embodiment of the present invention. Detailed Implementation

[0010] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a flowchart illustrating a method for monitoring glacial debris flows using drone inspections, provided by an embodiment of the present invention. The following is a detailed description of this method for monitoring glacial debris flows using drone inspections.

[0011] Step S110: Based on the risk evolution prediction results of the glacier area and the flight constraints of the UAV, perform dynamic collaborative adaptation processing of the inspection trajectory to generate a dynamic adaptation trajectory scheme that includes trajectory node distribution, flight speed adjustment rules, and collaborative parameters of sensing devices.

[0012] In this embodiment, a typical glacier area will be used as the monitoring target. This typical glacier area has historically experienced small-scale glacial debris flow disasters and is currently in the summer melting period, posing a certain possibility of risk evolution. When performing step S110, it is first necessary to comprehensively consider the risk situation of the glacier area and the flight conditions of the UAV in order to formulate a reasonable inspection trajectory plan.

[0013] Step S111: Integrate historical disaster evolution data, real-time topographic data of glacier areas, UAV flight performance parameters and real-time meteorological data to form a trajectory dynamic collaborative adaptation basic dataset. The trajectory dynamic collaborative adaptation basic dataset covers the topographic relief morphology, historical risk evolution path, maximum flight altitude of UAVs, endurance and real-time wind force and direction influencing factors of glacier areas.

[0014] In this embodiment, historical disaster evolution data of the typical glacier region over the past ten years can be collected, including the time, scale, impact range, and environmental conditions of each disaster. Real-time topographic data of the glacier region is acquired using specialized topographic surveying equipment. This data accurately reflects the topographic undulations of the glacier, such as the height of peaks, the depth of valleys, and the slope of the glacier surface. Simultaneously, flight performance parameters of the UAV used are acquired, covering its maximum flight altitude, maximum flight speed, endurance, and minimum turning radius. Furthermore, real-time meteorological data, including real-time wind speed, wind direction, temperature, and humidity, is collected from meteorological stations deployed around the glacier region. The above data is summarized and organized, removing duplicate and invalid data to form a basic dataset for dynamic trajectory collaborative adaptation. This basic dataset comprehensively reflects various conditions in the glacier region and the limiting factors affecting UAV flight.

[0015] Step S112: Extract core sensitive features of glacial debris flow risk evolution from the trajectory dynamic collaborative adaptation basic dataset. The core sensitive features include the density of surface fissures in glaciers, the stress transmission path inside glaciers, the infiltration range of water bodies around glaciers, the degree of loosening of strata particles, and the risk evolution rate.

[0016] Next, the trajectory dynamic collaborative adaptation dataset formed in step S111 is analyzed in depth, and feature extraction algorithms are used to extract core sensitive features. For the density of surface fractures in glaciers, the distribution of fractures on the glacier surface is identified by analyzing historical disaster evolution data and real-time topographic data, and the number of fractures per unit area is calculated to characterize the fracture development density. The stress transmission path inside the glacier is obtained through simulation and analysis of relevant data on the internal structure of the glacier, understanding the direction and mode of stress transmission within the glacier. The infiltration range of water bodies around the glacier is determined by analyzing data such as water level changes, flow velocity, and stratum permeability. The degree of stratum particle looseness is assessed through analysis of stratum samples and research on relevant geological data. The risk evolution rate is calculated and inferred based on the time interval between disasters in historical disaster data and the development process of each disaster. These core sensitive features directly reflect the evolution of glacier debris flow risk and are key basis for risk prediction.

[0017] Step S113: Construct a risk evolution prediction model for glacier regions based on the extracted core sensitive features. The risk evolution prediction model for glacier regions has a built-in risk level classification module and an evolution trend prediction module. The model is trained using historical disaster evolution data, so that the model can output risk evolution prediction results for different time dimensions based on real-time topographic data and environmental data.

[0018] Then, a risk evolution prediction model for glacier regions is constructed based on the core sensitive features extracted in step S112. This model includes a risk level classification module and an evolution trend prediction module. The risk level classification module classifies the risk of glacier regions into different levels, such as low risk, medium risk, and high risk, according to the numerical range of the core sensitive features. The evolution trend prediction module predicts the evolution trend of glacier region risk in different time dimensions (such as short-term, medium-term, and long-term) by learning from historical data. When constructing the model, a suitable machine learning algorithm, such as a neural network algorithm, is selected. Historical disaster evolution data is used as training data to train the model. During training, the model parameters are continuously adjusted to improve the model's prediction accuracy. After multiple training and optimizations, the glacier region risk evolution prediction model can accurately output risk evolution prediction results in different time dimensions based on the input real-time terrain and environmental data.

[0019] Step S114: Extract UAV flight constraints from the trajectory dynamic collaborative adaptation basic dataset, and transform the UAV flight constraints into hard rules for trajectory adaptation. The UAV flight constraints include maximum flight altitude limit, minimum turning radius, endurance time threshold, and flight prohibition areas under severe weather conditions.

[0020] Simultaneously, UAV flight constraints are extracted from the trajectory dynamic collaborative adaptation dataset. Based on the UAV's flight performance parameters, the maximum flight altitude limit is determined, i.e., the altitude the UAV cannot exceed during flight. The minimum turning radius is the minimum space required for the UAV to turn, and this parameter directly affects trajectory planning. The endurance time threshold is determined based on the UAV's battery capacity and energy consumption; the UAV's flight time cannot exceed this threshold, otherwise it will be unable to complete the inspection mission and return safely. Through analysis of real-time meteorological data, flight prohibition zones under severe weather conditions are identified. For example, under conditions such as strong winds, heavy rain, and dense fog, certain areas are unsuitable for UAV flight and are marked as flight prohibition zones. These flight constraints are transformed into hard rules for trajectory adaptation. These rules must be strictly followed during trajectory planning to ensure the UAV's flight safety and the successful completion of the inspection mission.

[0021] Step S115: Correlate the parameters of risk core area distribution, risk evolution direction, and evolution rate in the risk evolution prediction results with the UAV flight constraints to preliminarily plan the core coverage area and direction of the inspection trajectory, wherein the core coverage area covers the risk core area distribution and the direction satisfies the UAV flight constraints.

[0022] After obtaining the risk evolution prediction results and the UAV flight constraints, a correlation analysis is performed between the two. The distribution of the core risk areas in the risk evolution prediction results is the key focus of the inspection. Therefore, the core coverage area of ​​the initially planned inspection trajectory must include these core risk areas to ensure comprehensive monitoring of high-risk areas. Simultaneously, the direction and rate of risk evolution are considered to ensure the inspection trajectory follows the direction of risk evolution, enabling timely understanding of risk development. When planning the trajectory, UAV flight constraints, such as maximum flight altitude limits and minimum turning radius, must be strictly adhered to to ensure the trajectory remains within the UAV's flight capabilities. For example, if the core risk area is located at the top of a mountain, and the UAV's maximum flight altitude cannot reach that summit, the trajectory needs to be adjusted to a suitable altitude that covers the core risk area. Through the above correlation analysis, the core coverage area and trajectory of the inspection trajectory are preliminarily determined.

[0023] Step S116: Based on the preliminary planned inspection trajectory, the trajectory node distribution density is set according to the risk evolution rate of the core risk area. The trajectory node distribution density is adjusted in the positive direction of the evolution rate. The faster the evolution rate, the denser the trajectory node distribution in the area. At the same time, dynamic tracking nodes are set in the risk evolution direction.

[0024] Based on the preliminary inspection trajectory planned in step S115, the distribution density of trajectory nodes is further set. The distribution of nodes is determined according to the risk evolution rate of the core risk area. The faster the risk evolution rate, the more rapidly the risk changes in that area, requiring more frequent monitoring; therefore, the trajectory node distribution should be denser. For example, for areas with a fast risk evolution rate, a trajectory node is set at shorter intervals; while for areas with a slow risk evolution rate, the distance between nodes can be appropriately increased. Simultaneously, dynamic tracking nodes are set along the risk evolution direction. These nodes can dynamically adjust their positions as the risk evolves to ensure continuous monitoring of the latest risk developments. The setting of dynamic tracking nodes needs to be combined with the risk evolution prediction results, determining the initial position and adjustment rules of the nodes based on the predicted risk evolution direction and speed.

[0025] Step S117: Based on the distribution of trajectory nodes and the risk characteristics in the risk evolution prediction results, construct flight speed adjustment rules. The core risk area adopts a set low-speed flight mode, the risk diffusion and influence area adopts a set uniform speed flight mode, and the risk transition and correlation area adjusts the flight speed mode according to the set gradient.

[0026] Based on the trajectory node distribution determined in step S116 and the risk characteristic classification in the risk evolution prediction results, flight speed adjustment rules are constructed. The glacier area is divided into different risk characteristic regions, such as the core risk region, the risk diffusion impact region, and the risk transition correlation region. For the core risk region, due to the need for detailed monitoring and data collection, a low-speed flight mode is adopted to ensure that the UAV has sufficient time for comprehensive perception operations in this region. The risk diffusion impact region has a relatively low risk level, but still requires continuous monitoring; therefore, a uniform speed flight mode is adopted. The risk level of the risk transition correlation region is between that of the core region and the diffusion impact region, and its flight speed is adjusted according to a set gradient, that is, the flight speed gradually increases from the end closer to the core region to the end closer to the diffusion impact region. The above flight speed adjustment rules can improve the inspection efficiency of the UAV while ensuring the monitoring effect, ensuring that the inspection task is completed within the endurance time.

[0027] Step S118: Combining the perception needs of different risk characteristic areas with the performance of the multi-source perception devices carried by the UAV, construct the perception device collaborative parameters, start the multi-device synchronous acquisition mode in the core risk area, start the corresponding devices in the risk diffusion and impact area according to the perception priority, and determine the acquisition frequency and data transmission priority of each device.

[0028] Considering the perception needs of different risk areas and the performance of the multi-source perception devices carried by the UAVs, collaborative parameters for perception devices are constructed. The core risk areas require the most comprehensive and detailed data; therefore, a multi-device synchronous acquisition mode is initiated, such as simultaneously activating optical imaging devices, acoustic detection devices, electromagnetic penetration devices, and environmental perception devices to obtain multi-dimensional data on glacier surface morphology, internal structure, and surrounding environment. In areas affected by risk spread, corresponding devices are activated according to perception priority; for example, optical imaging devices and environmental perception devices are prioritized to obtain surface morphology and environmental data. Simultaneously, the acquisition frequency and data transmission priority of each device are determined based on its performance and the importance of the data. For critical data in the core areas, a higher acquisition frequency and data transmission priority are set to ensure timely and accurate transmission back to the ground control center. For auxiliary data in non-core areas, the acquisition frequency and transmission priority can be appropriately reduced to save UAV energy and communication bandwidth.

[0029] Step S119: Construct an initial dynamic adaptation trajectory scheme based on trajectory node distribution, flight speed adjustment rules, and sensing device collaborative parameters. Perform collaborative verification between the initial dynamic adaptation trajectory scheme and the real-time evolution status output by the glacier region risk evolution prediction model. The verification process determines whether the trajectory nodes in the initial dynamic adaptation trajectory scheme cover the risk area indicated by the real-time evolution status.

[0030] Based on the trajectory node distribution, flight speed adjustment rules, and sensing device coordination parameters determined in the previous steps, an initial dynamic adaptation trajectory scheme is constructed. This initial dynamic adaptation trajectory scheme specifies in detail the UAV's flight trajectory, flight speed at each node, and the operating parameters of the sensing devices. Then, the initial dynamic adaptation trajectory scheme is collaboratively validated with the real-time evolution situation output by the glacier region risk evolution prediction model. The real-time evolution situation reflects the latest development of glacier region risks. The rationality and effectiveness of the scheme are judged by comparing whether the trajectory nodes in the initial dynamic adaptation trajectory scheme can cover the risk area indicated by the real-time evolution situation. If it is found that the trajectory nodes do not completely cover the risk area, the initial scheme needs to be adjusted and optimized.

[0031] Step S1110: Integrate the trajectory node coordinates, flight speed adjustment parameters, sensing device operating parameters, and trajectory dynamic adjustment rules after the collaborative verification is passed, and generate a dynamically adapted trajectory scheme.

[0032] After collaborative verification, for the verified initial dynamic adaptation trajectory scheme, information such as trajectory node coordinates, flight speed adjustment parameters, sensing device operating parameters, and trajectory dynamic adjustment rules are integrated to generate the final dynamic adaptation trajectory scheme. The trajectory dynamic adjustment rules specify how to adjust the trajectory during the inspection process when encountering changes in real-time risk evolution or other special circumstances, such as adding or deleting trajectory nodes or adjusting flight speed. This dynamic adaptation trajectory scheme can achieve precise inspection of glacier areas based on the actual conditions of the glacier region and the flight constraints of the UAV.

[0033] Step S120: Based on the dynamic adaptation trajectory scheme, drive the UAV multi-source perception system to perform hierarchical precise perception operations, collect multi-dimensional data on the surface morphology, internal structure, and surrounding environment of different risk characteristic areas in the glacier region, and generate a time-seriesd glacier region risk perception data chain.

[0034] After generating a dynamically adaptable trajectory scheme, the UAV's multi-source perception system is driven to operate according to this scheme. The UAV flies along the trajectory specified in the dynamically adaptable trajectory scheme, and performs hierarchical precise perception operations based on different risk characteristic areas during flight, collecting multi-dimensional data, and finally generating a time-seriesd risk perception data chain for the glacier area.

[0035] Step S121: Analyze the trajectory node distribution, risk feature division, flight speed adjustment rules and sensing device coordination parameters in the dynamic adaptation trajectory scheme, and divide the glacier inspection area into three sensing levels according to risk features: core risk triggering sensing area, risk transmission intermediate sensing area and risk diffusion terminal sensing area.

[0036] First, a detailed analysis of the dynamic adaptation trajectory scheme is conducted, extracting the distribution information of trajectory nodes to understand the location and sequence of each node the UAV needs to pass through. Simultaneously, the scheme's risk characteristic classification, flight speed adjustment rules, and sensing device coordination parameters are clarified. Based on this information, the glacier inspection area is divided into three sensing levels according to different risk characteristics. The core risk triggering sensing zone is the area with the highest risk level and the greatest likelihood of disaster, typically including areas with dense surface fissures and unstable internal structures. The risk transmission intermediate sensing zone is the intermediate area connecting the core risk triggering sensing zone and the risk diffusion terminal sensing zone, where risks are transmitted and diffused. The risk diffusion terminal sensing zone is the area where the risk ultimately affects the glacier; its risk level is relatively low, but monitoring is still necessary. This classification enables the UAV to adopt different sensing strategies for areas with different risk levels during inspections, improving the accuracy and efficiency of sensing.

[0037] Step S122: Construct differentiated perception strategies for different perception levels. For the core risk trigger perception zone, a multi-source device synchronous acquisition strategy is adopted, activating optical imaging equipment, acoustic detection equipment, electromagnetic penetration equipment, and environmental perception equipment simultaneously. The acquisition frequency is set according to the parameter standards corresponding to the core risk trigger perception zone in the dynamic adaptation trajectory scheme. For the risk transmission intermediate perception zone, a multi-source device alternating acquisition strategy is adopted, alternately activating optical imaging equipment and acoustic detection equipment at preset time intervals, with auxiliary activation of environmental perception equipment. The acquisition frequency is set according to the parameter standards corresponding to the risk transmission intermediate perception zone in the dynamic adaptation trajectory scheme. For the risk diffusion terminal perception zone, a single-device acquisition strategy is adopted, activating only the optical imaging equipment for surface morphology acquisition. The environmental perception equipment acquires data according to the parameter standards corresponding to the risk diffusion terminal perception zone in the dynamic adaptation trajectory scheme. The device activation strategy and data acquisition frequency during the acquisition process are dynamically adjusted based on the remaining flight time of the UAV.

[0038] For the different perception levels defined in step S121, differentiated perception strategies are constructed. Due to its importance and high risk, the core risk-triggered perception zone employs a multi-source device synchronous acquisition strategy. Within this area, the UAV simultaneously activates optical imaging equipment, acoustic detection equipment, electromagnetic penetration equipment, and environmental perception equipment. These devices work collaboratively to collect relevant glacier data from multiple angles. The acquisition frequency is strictly set according to the parameter standards corresponding to the core risk-triggered perception zone in the dynamic adaptation trajectory scheme to ensure that any subtle changes within this area can be captured. The risk transmission intermediate perception zone employs a multi-source device alternating acquisition strategy, activating optical imaging equipment and acoustic detection equipment alternately at preset time intervals, such as periodically, while simultaneously activating environmental perception equipment. This ensures comprehensive data acquisition while saving UAV energy consumption. The acquisition frequency is also set according to the parameter standards corresponding to this area in the scheme. The risk diffusion terminal perception zone employs a single-device acquisition strategy, primarily activating optical imaging equipment to collect surface morphology data, while environmental perception equipment collects data according to the parameter standards corresponding to this area in the scheme. Throughout the acquisition process, the device activation strategy and data acquisition frequency are dynamically adjusted based on the UAV's remaining flight time. If there is sufficient remaining battery life, data can be collected according to the normal strategy and frequency; if there is limited remaining battery life, the data collection frequency should be reduced or some unnecessary equipment should be turned off to ensure that the drone can complete the inspection mission and return safely.

[0039] Step S123: Generate hierarchical control instructions for the multi-source sensing system based on the differentiated sensing strategy. The hierarchical control instructions determine the device startup sequence, acquisition frequency, data resolution and data transmission format of each sensing level, and associate the corresponding trajectory node coordinates and flight speed parameters.

[0040] Based on the differentiated perception strategy determined in step S122, hierarchical control instructions for the multi-source perception system are generated. These control instructions specify in detail the activation sequence, acquisition frequency, data resolution, and data transmission format of each perception device at different perception levels. For example, in the core risk-triggered perception zone, the control instructions require the environmental perception device to be activated first, followed by the optical imaging device, acoustic detection device, and electromagnetic penetration device in sequence, to ensure that the environmental conditions are understood before detailed data acquisition begins. The acquisition frequency, data resolution, and data transmission format are also set according to the parameter standards of that area. Simultaneously, the control instructions are associated with the corresponding trajectory node coordinates and flight speed parameters, enabling the UAV to activate the perception devices for data acquisition according to the corresponding control instructions and fly at the set flight speed when it reaches the designated trajectory node. This ensures the accuracy and timeliness of data acquisition, closely coordinating with the UAV's flight trajectory.

[0041] Step S124: Control the UAV to move sequentially according to the trajectory node distribution in the dynamic adaptation trajectory scheme. At each trajectory node, activate the multi-source sensing device according to the corresponding perception level control command, and synchronously record the acquisition time and the UAV's real-time position coordinates.

[0042] After generating hierarchical control commands, these are sent to the UAV's control system, which in turn controls the UAV to move sequentially according to the trajectory node distribution in the dynamically adapted trajectory scheme. Starting from the initial node, the UAV flies along the planned trajectory, passing through each trajectory node in turn. Upon reaching each trajectory node, the UAV's control system activates the corresponding multi-source sensing devices to collect data based on the perception hierarchy control command for that node. Simultaneously, the UAV's built-in clock and positioning system record the data acquisition time and the UAV's real-time position coordinates. This information is stored and transmitted along with the acquired data to accurately determine the time and location of data acquisition during subsequent data analysis and processing.

[0043] Step S125: Collect glacier surface morphology data at each sensing level using optical imaging equipment. In the core risk trigger sensing zone, focus on collecting the extension direction and width changes of surface fissures and the accumulation state of materials around the fissures. In the risk transmission intermediate sensing zone and the risk diffusion terminal sensing zone, focus on collecting the overall morphological changes of the surface.

[0044] During the drone's flight, the optical imaging equipment operates according to control commands, collecting glacier surface morphology data at various sensing levels. For the core risk-triggered sensing zone, the optical imaging equipment focuses on surface fissures, clearly recording the fissure extension direction, width variations, and surrounding material accumulation through high-resolution image acquisition. This data reflects the stability of the glacier surface and is crucial for assessing risk evolution trends. The risk transmission intermediate sensing zone and the risk diffusion terminal sensing zone focus on collecting data on overall surface morphological changes. Through continuous image capture, the overall contour and slope changes of the glacier surface are observed to understand the diffusion and impact of risks in these areas. The image data acquired by the optical imaging equipment is compressed and encoded for storage and transmission.

[0045] Step S126: Collect data on the internal structure of the glacier using acoustic detection equipment, transmit acoustic signals of a specific frequency band into the glacier, receive and record the waveform changes of the signals after reflection by different media inside the glacier, and increase the signal transmission power in the core risk trigger sensing area according to the parameter standards in the dynamic adaptation trajectory scheme.

[0046] The acoustic detection equipment activates when the UAV reaches a designated trajectory node, transmitting acoustic signals in a specific frequency band into the glacier. These signals can penetrate the glacier surface and enter its interior. When the signals encounter different media within the glacier, such as ice, rock debris, or water, they are reflected. The acoustic detection equipment receives and records these reflected waveform changes. Different media exhibit different reflection characteristics of the acoustic signals; by analyzing the waveform changes, structural information within the glacier can be inferred, such as the distribution range and relative positions of different media. In the core risk-triggered sensing zone, to further detect the structural details within the glacier, the transmission power of the acoustic signals is increased according to the parameter standards in the dynamic adaptation trajectory scheme, thereby improving signal penetration and resolution.

[0047] Step S1261: Analyze the acoustic detection parameters of the corresponding sensing level in the dynamic adaptation trajectory scheme; set the frequency band acoustic signal transmission parameters of the core risk trigger sensing area in the dynamic adaptation trajectory scheme; set the frequency band parameters of the risk transmission intermediate sensing area in the dynamic adaptation trajectory scheme; set the frequency band parameters of the risk diffusion terminal sensing area in the dynamic adaptation trajectory scheme.

[0048] Before activating the acoustic detection equipment, the acoustic detection parameters corresponding to each sensing level in the dynamic adaptation trajectory scheme are first analyzed. For the core risk-triggered sensing zone, the signal frequency, transmission power, pulse width, and other parameters of the acoustic detection equipment are set according to the corresponding frequency band acoustic signal transmission parameters in the scheme. These parameters are determined based on the characteristics of the glacier's internal structure and the detection requirements in that area, ensuring that the transmitted acoustic signal is most suitable for detecting the internal structure of that area. The risk transmission intermediate sensing zone and the risk diffusion terminal sensing zone are also set according to their respective corresponding frequency band parameters in the scheme to adapt to the detection needs of different areas.

[0049] Step S1262: Control the acoustic signal transmission module on the UAV to start according to the set parameters, and transmit continuous acoustic signals to the glacier area corresponding to the current inspection node. Based on the boundary distance of the monitoring area corresponding to the current inspection node, dynamically adjust the acoustic signal transmission power.

[0050] Based on the acoustic detection parameters obtained in step S1261, the acoustic signal transmission module on the UAV is activated according to the set parameters. The acoustic signal transmission module transmits continuous acoustic signals to the glacier area corresponding to the current inspection node. Simultaneously, the transmission power of the acoustic signal is dynamically adjusted according to the boundary distance of the monitoring area corresponding to the current inspection node. If the boundary distance is far, the transmission power is appropriately increased to ensure the signal reaches the boundary area and reflects back; if the distance is short, the transmission power is reduced to avoid excessively strong signals damaging the equipment or interfering with other signals.

[0051] Step S1263: Start the acoustic signal receiving module to synchronously receive acoustic signals reflected from different medium interfaces inside the glacier and secondary acoustic signals generated by the vibration of the internal structure of the glacier.

[0052] Simultaneously with the transmission of signals by the acoustic signal transmitting module, the acoustic signal receiving module is activated. This acoustic signal receiving module can synchronously receive acoustic signals reflected from different medium interfaces within the glacier, as well as secondary acoustic signals generated by vibrations of the glacier's internal structure. These signals carry rich information about the glacier's internal structure. The receiving module amplifies and filters these signals for subsequent analysis and recording.

[0053] Step S1264: Perform signal enhancement processing on the received acoustic signal, and use an adaptive filtering algorithm to reduce the interference signals generated by environmental noise and the vibration of the UAV itself, improve the signal-to-noise ratio of the effective signal, and capture the waveform characteristics generated by the reflection of the glacier's internal medium.

[0054] The received acoustic signals contain interference signals from environmental noise and the drone's own vibration, which can affect the analysis of glacier internal structure data. Therefore, signal enhancement processing is required for the received acoustic signals. An adaptive filtering algorithm is employed, which automatically adjusts the filter parameters according to the characteristics of the signal and noise, effectively reducing interference signals from environmental noise and the drone's own vibration. Through processing, the signal-to-noise ratio of the effective signal is improved, making the waveform characteristics generated by reflection from the glacier's internal medium clearer, facilitating subsequent extraction and analysis of these waveform characteristics.

[0055] Step S1265: Extract the waveform features of the enhanced acoustic signal. The waveform features include the peak shape, trough depth, waveform duration, waveform distortion degree, and signal attenuation rate. The waveform features reflect the differences in physical properties of different media inside the glacier.

[0056] Waveform features are extracted from the acoustic signal after enhancement processing in step S1264. The extracted waveform features include crest morphology, trough depth, waveform duration, waveform distortion, and signal attenuation rate. Crest morphology refers to the shape and size of the peaks in the waveform; different media reflect signals with different peak morphologies. Trough depth is the depth of the troughs in the waveform, which also reflects the characteristics of the medium. Waveform duration refers to the duration of a complete waveform, and signal attenuation rate is the rate at which the signal energy decays during propagation. These waveform features directly reflect the differences in physical properties of different media within a glacier, such as density and elastic modulus.

[0057] Step S1266: Analyze the distortion regions in the waveform features, identify waveform distortion caused by rupture, loosening or inhomogeneity of the glacier's internal structure, mark the signal reception time corresponding to each distortion region, and calculate the glacier's internal depth corresponding to the distortion region by combining the acoustic signal propagation speed.

[0058] The waveform features extracted in step S1265 are analyzed, with a focus on distorted regions. Waveform distortion is usually caused by internal structural fracturing, loosening, or inhomogeneity of the glacier. By identifying and analyzing distorted regions, structural problems within the glacier can be determined. The signal reception time corresponding to each distorted region is marked, and then the depth of the glacier corresponding to the distorted region is calculated based on the propagation speed of the acoustic signal within the glacier. Specifically, the difference between the signal transmission time and reception time is multiplied by the propagation speed, and then divided by 2 to obtain the distance from the distorted region to the glacier surface, i.e., the depth of the glacier.

[0059] Step S1267: Analyze the frequency distribution and continuous trend of waveform distortion of the acoustic signal in the core risk trigger sensing area to determine the dynamics of structural changes during the stress transmission process inside the glacier.

[0060] A more detailed analysis is needed of the acoustic signals from the core risk-triggered sensing area. This includes analyzing the frequency distribution of waveform distortion, specifically the frequency and degree of distortion across different frequency ranges, as well as the continuous trend of waveform distortion changes, such as whether the degree of distortion gradually increases or decreases, and whether the location of the distorted area has shifted. These analyses allow us to determine the dynamic structural changes during stress transmission within the glacier. For example, if the frequency distribution of waveform distortion gradually shifts towards higher frequencies and the degree of distortion gradually increases, it may indicate that stress is continuously accumulating within the glacier, and the structure is becoming increasingly unstable.

[0061] Step S1268: Combine the location coordinates of the current inspection node, associate the extracted waveform features, depth information corresponding to the distortion area, and signal attenuation data with the spatial location to form an acoustic detection data unit for the internal structure of the glacier corresponding to each inspection node.

[0062] The waveform features extracted in step S1265, the depth information corresponding to the distortion region calculated in step S1266, and the signal attenuation data are correlated with the location coordinates of the current inspection node. Each inspection node has its specific location coordinates. By mapping the above data to the location coordinates, an acoustic detection data unit for the glacier's internal structure is formed for each inspection node. This acoustic detection data unit for the glacier's internal structure contains detailed acoustic detection information about the glacier's internal structure at that node.

[0063] Step S1269: Perform signal superposition processing on the acoustic detection data units of the glacier internal structure collected multiple times at the same inspection node, integrate the acoustic detection data units of the glacier internal structure from all inspection nodes, and construct an acoustic detection data sequence according to the time sequence and spatial distribution of the inspection trajectory.

[0064] For the same inspection node, if multiple data acquisitions are conducted, the acoustic detection data units of the glacier's internal structure acquired from these multiple acquisitions are subjected to signal superposition processing. Signal superposition processing can improve the signal-to-noise ratio of the data, reduce the influence of random noise, and make the data more reliable. Then, the acoustic detection data units of the glacier's internal structure from all inspection nodes are integrated and arranged according to the temporal order and spatial distribution of the inspection trajectory to construct an acoustic detection data sequence. This acoustic detection data sequence can reflect the acoustic detection status of the glacier's internal structure at different times and locations.

[0065] Step S127: Collect data on the distribution of the internal medium of the glacier using an electromagnetic penetration device to obtain the distribution range and relative position of the ice layer, rock debris, and water body inside the glacier.

[0066] The electromagnetic penetration device activates according to control commands during the drone's flight. This device emits electromagnetic signals into the glacier's interior, which penetrate the glacier's surface and interact with the ice, rock debris, water, and other media within the glacier. By receiving and analyzing the reflection and refraction of these electromagnetic signals, the distribution range and relative positions of different media within the glacier can be determined. For example, ice and water exhibit different reflection characteristics of electromagnetic signals; analyzing these differences allows for the differentiation of ice and water distribution areas. Understanding the distribution of the internal media helps assess the likelihood of glacier slippage, rupture, and other risks.

[0067] Step S128: Collect temperature, humidity, wind, and air pressure data at each sensing level through environmental sensing devices. The temperature, humidity, wind, and air pressure data are used to analyze the driving role of environmental factors on the evolution of glacier risk.

[0068] Environmental sensing equipment operates continuously throughout the inspection process, collecting environmental data such as temperature, humidity, wind speed, and air pressure at various sensing levels. This data reflects the real-time environmental conditions of the glacier region. Changes in temperature and humidity affect the melting and freezing processes of glaciers, while wind strength and direction influence the transport and erosion of materials on the glacier surface. Changes in air pressure may be related to changes in weather systems. By analyzing this environmental data, we can understand the driving role of environmental factors in the evolution of glacier risks. For example, high temperature and humidity environments may accelerate glacier melting and increase the risk of debris flows.

[0069] Step S129: Integrate the data collected by all sensing devices in three dimensions according to the collection time, UAV location coordinates, and sensing level to construct a time-series risk perception data chain for glacier areas. Each data unit in the glacier area risk perception data chain contains collection scene information and associated data identifiers.

[0070] After data collection from all sensing devices is completed, the collected data is integrated and processed. The data is correlated and integrated according to three dimensions: collection time, UAV location coordinates, and sensing level. Data from the same time, location coordinates, and sensing level are grouped together to form a data unit. These data units are arranged chronologically to construct a time-seriesd glacier area risk perception data chain. Each data unit contains information about the collection scenario, such as the weather conditions and glacier surface features, as well as a related data identifier to identify the relationship between this data unit and other related data units. Through this three-dimensional correlation and integration, the data in the data chain has a clear spatiotemporal and hierarchical relationship, facilitating subsequent analysis and assessment of glacier area risks.

[0071] Step S130: Through risk perception data chain in glacier areas, risk transmission node mining and correlation analysis are triggered to identify core triggering nodes, transmission intermediary nodes and diffusion terminal nodes in the risk evolution process, and a risk transmission node network including node correlation and transmission time sequence characteristics is constructed.

[0072] After generating a time-series risk perception data chain for glacier regions, this data chain is used to trigger the mining and correlation analysis of risk transmission nodes. Through in-depth analysis of the data in the data chain, various nodes in the risk evolution process are identified, and a risk transmission node network is constructed.

[0073] Step S131: Extract the feature parameters of each sensing data unit from the risk perception data chain of the glacier area to form a risk feature parameter set. The feature parameters include surface morphological change parameters, internal structural feature parameters, environmental impact parameters, and time and space information of data collection.

[0074] First, the risk perception data chain for the glacier region is analyzed, and characteristic parameters are extracted from each perception data unit. Surface morphological change parameters include the number, length, width, and slope changes of fissures on the glacier surface; internal structural characteristic parameters include the distribution range, density, and elastic modulus of different media within the glacier; environmental impact parameters include temperature, humidity, wind force, and air pressure; and the temporal and spatial information of data acquisition is the acquisition time and UAV location coordinates corresponding to each data unit. These characteristic parameters are then aggregated to form a risk characteristic parameter set.

[0075] Step S132: Based on the spatial information in the risk feature parameter set, map all feature parameters to the geographic coordinate system of the glacier region to construct the spatial distribution features of the risk features. The spatial distribution features present the spatial location and distribution density of each feature parameter in the glacier region.

[0076] By utilizing the spatial information within the risk characteristic parameter set, all characteristic parameters are mapped to a geographic coordinate system of the glacier region. Using Geographic Information System (GIS) technology, each characteristic parameter is associated with its corresponding spatial location, constructing a spatial distribution feature of the risk characteristics. This spatial distribution feature reveals the spatial location and distribution density of each characteristic parameter within the glacier region. For example, by drawing contour maps or density maps, the distribution of surface fissures in the glacier in different areas and the differences in their distribution density can be visually observed.

[0077] Step S133: Analyze the time sequence of changes of each feature parameter in the spatial distribution features, extract the time points when the feature parameters show significant changes, mark them as risk evolution time nodes, and construct the risk evolution time sequence axis in chronological order.

[0078] A time-series analysis is performed on the spatial distribution characteristics of the risk features constructed in step S132 to study the changes of each feature parameter over time. By comparing the spatial distribution characteristics at different time points, the time points where significant changes in the feature parameters occur are extracted and marked as risk evolution time nodes. Then, these risk evolution time nodes are arranged in chronological order to construct a risk evolution time-series axis. The risk evolution time-series axis can show the changes in risk feature parameters at different times, reflecting the time process of risk evolution.

[0079] Step S134: On the risk evolution timeline, associate the spatial distribution characteristics of risk features corresponding to different time nodes, identify the regions where significant changes in feature parameters first occur, and the regions where significant changes in feature parameters first occur are the initial trigger regions of risk evolution. The set of feature parameters within the initial trigger regions constitutes the basic data of the core trigger nodes.

[0080] On the constructed risk evolution timeline, the spatial distribution characteristics of risk features at different time points are correlated. Through comparative analysis, the region where significant changes in characteristic parameters first occur during risk evolution is identified; this region is the initial triggering region of risk evolution. The initial triggering region is the source of risk evolution. The set of all characteristic parameters within the initial triggering region constitutes the basic data of the core triggering node, containing various characteristic information of the initial triggering region.

[0081] Step S135: Extract key driving feature parameters from the basic data of the core trigger node. The key driving feature parameters are feature parameters that are determined based on historical disaster data and have the highest correlation with subsequent changes in risk feature parameters. Define the core trigger node based on the key driving feature parameters and spatial location information.

[0082] From the core trigger node basic data obtained in step S134, key driving characteristic parameters are extracted. These key driving characteristic parameters are determined through analysis of historical disaster data. These parameters have the highest correlation with changes in subsequent risk characteristic parameters; that is, their changes directly lead to changes in other risk characteristic parameters. Through statistical analysis and correlation calculation of historical disaster data, those characteristic parameters that change first before the disaster occurs and have the greatest impact on changes in other parameters are identified and determined as key driving characteristic parameters. Then, combining these key driving characteristic parameters with their corresponding spatial location information, the core trigger node is defined.

[0083] For example, step S1351: Select the feature parameters that show the most significant changes in the initial stage of risk evolution from the basic data of the core trigger node, and use the feature parameters that show the most significant changes in the initial stage of risk evolution as candidate driving feature parameters of the core trigger node.

[0084] In the basic data of the core trigger nodes, feature parameters that show the most significant changes in the initial stage of risk evolution are selected according to the time sequence of risk evolution. These changes in feature parameters are early signals of the start of risk evolution and are used as candidate driving feature parameters for the core trigger nodes. For example, in the initial stage of risk evolution, the width of glacier surface fissures may be the first to show a significant increase, so the feature parameter of glacier surface fissure width would be selected as a candidate driving feature parameter.

[0085] Step S1352: Analyze the interaction between candidate driving feature parameters and identify the source feature parameters that can directly cause changes in other feature parameters. The source feature parameters are the key driving feature parameters.

[0086] Interaction relationship analysis is performed on the candidate driving characteristic parameters selected in step S1351. By constructing a causal relationship model among the characteristic parameters, it is analyzed which candidate driving characteristic parameters can directly induce changes in other characteristic parameters. The source characteristic parameters that can directly induce changes in other characteristic parameters are identified as key driving characteristic parameters. For example, if an increase in the width of surface fissures in a glacier leads to a redistribution of stress within the glacier, thereby causing changes in the internal structure of the glacier, then the width of surface fissures is one of the key driving characteristic parameters.

[0087] Step S1353: Determine the change threshold of the key driving characteristic parameter. The change threshold of the key driving characteristic parameter is a numerical threshold or percentage change rate threshold obtained based on the statistical analysis of historical disaster data. The change threshold of the key driving characteristic parameter is determined based on the change law of the characteristic parameter of the core triggering node in the historical disaster data.

[0088] Based on the changing patterns of characteristic parameters of core triggering nodes in historical disaster data, threshold values ​​for key driving characteristic parameters are determined. Through statistical analysis of a large amount of historical disaster data, the range and extent of change of key driving characteristic parameters before the disaster occurs are calculated, thereby determining numerical thresholds or percentage change rate thresholds. When the change of key driving characteristic parameters reaches or exceeds this threshold, it indicates that the risk evolution has entered a new stage, requiring close attention.

[0089] Step S1354: Record the time point when the key driving characteristic parameter reaches the change threshold. The time point when the key driving characteristic parameter reaches the change threshold is the start time of the core trigger node. The start time of the core trigger node serves as the starting benchmark of the risk evolution timeline.

[0090] In the risk perception data chain for glacier regions, changes in key driving characteristic parameters are closely monitored. When the change in a key driving characteristic parameter reaches the change threshold determined in step S1353, this time point is recorded; this time point is the activation time of the core trigger node. The activation time of the core trigger node is used as the starting benchmark of the risk evolution timeline, and other risk evolution time points are arranged with reference to this starting benchmark.

[0091] Step S1355: Determine the spatial region range corresponding to the key driving feature parameters. The spatial region range corresponding to the key driving feature parameters is the region where the key driving feature parameters show significant changes. The spatial region range corresponding to the key driving feature parameters is determined based on the spatial distribution characteristics of the feature parameters.

[0092] Based on the spatial distribution characteristics of the feature parameters, the spatial range corresponding to the key driving feature parameters is determined. This spatial range is the area where the key driving feature parameters show significant changes; the boundary of this region is delineated by analyzing the spatial distribution characteristics. For example, if the key driving feature parameter is the width of surface fissures in a glacier, then the spatial range corresponding to this key driving feature parameter is determined by analyzing areas where the fissure width increases significantly.

[0093] Step S1356: Integrate the types, change thresholds, start times, and corresponding spatial ranges of key driving characteristic parameters to form the core definition elements of the core trigger node.

[0094] By integrating information such as the type, change threshold, start time, and corresponding spatial range of key driving characteristic parameters, the core definition elements of the core trigger node are formed. These elements comprehensively describe the characteristics and attributes of the core trigger node and are the key content for defining the core trigger node.

[0095] Step S1357: Supplement the auxiliary definition elements of the core trigger node. The auxiliary definition elements include the changing trend, changing rate and correlation with the surrounding environmental parameters of the key driving characteristic parameters.

[0096] In addition to the core definition elements, auxiliary definition elements for the core trigger node also need to be added. The changing trend of the key driving characteristic parameters includes whether they gradually increase or gradually decrease, and the rate of change is the amount of change per unit time. The correlation with surrounding environmental parameters refers to the mutual influence between the changes in the key driving characteristic parameters and environmental parameters such as temperature, humidity, wind force, and air pressure. These auxiliary definition elements can further enrich the description of the core trigger node, making it more comprehensive and accurate.

[0097] Step S1358: Based on the core definition elements and auxiliary definition elements, construct the definition standard for core trigger nodes. The judgment conditions and description specifications of core trigger nodes are determined in the definition standard for core trigger nodes.

[0098] Based on the core and auxiliary defining elements, a definition standard for core trigger nodes is constructed. This standard clearly defines the criteria for determining a core trigger node, specifying which elements must be met for it to be considered a core trigger node. Simultaneously, a description specification is established, outlining how to describe and record core trigger nodes to ensure consistency and accuracy in their definition and description.

[0099] Step S1359: Based on the definition standard of core trigger nodes, formally define the core trigger node, clearly describe the connotation and extension of the core trigger node, and distinguish the core trigger node from other risk nodes.

[0100] Following the definition standards for core trigger nodes, a formal definition of core trigger nodes is provided. The connotation of core trigger nodes is described, namely their essential characteristics and attributes; their extension is clarified, namely the scope covered by core trigger nodes. Simultaneously, core trigger nodes are distinguished from other risk nodes, explaining their differences and connections to ensure no confusion arises in subsequent analysis and application.

[0101] Step S13510: Associate the definition information of the core trigger node with the corresponding feature parameter data, spatial location data and time series data to form a core trigger node data unit, and incorporate the core trigger node data unit into the risk transmission node network.

[0102] Finally, the definition information of the core triggering node is associated with its corresponding feature parameter data, spatial location data, and temporal series data to form a core triggering node data unit. This core triggering node data unit contains all relevant information about the core triggering node and is incorporated into the risk transmission node network.

[0103] Step S136: Track the propagation path of the key driving characteristic parameters corresponding to the core triggering node. In the subsequent time nodes of the risk evolution timeline, identify the regions where characteristic parameters change due to the influence of key driving characteristic parameters. The regions where characteristic parameters change due to the influence of key driving characteristic parameters are the intermediary regions of risk transmission. Analyze the change pattern of characteristic parameters within the intermediary regions, determine the transmission and transformation effect of the intermediary regions on key driving characteristic parameters, and define the transmission intermediary nodes based on the spatial location of the intermediary regions, the type of characteristic parameter change, and the time sequence relationship.

[0104] After defining the core triggering node, the propagation path of its key driving characteristic parameters is tracked. At subsequent time points on the risk evolution timeline, it is analyzed which regions' characteristic parameter changes are influenced by the key driving characteristic parameters of the core triggering node; these regions are identified as intermediate regions for risk transmission. The variation patterns of characteristic parameters within these intermediate regions are analyzed to understand how they transmit and transform key driving characteristic parameters. For example, the key driving characteristic parameters of the core triggering node may cause changes in the internal stress of the glacier in the intermediate region, thereby leading to changes in the surface morphology of that region. Based on the spatial location of the intermediate region, the type of characteristic parameter change, and its temporal relationship with the core triggering node, transmission intermediate nodes are defined.

[0105] Step S137: Continue to track the propagation path of the characteristic parameters corresponding to the transmission intermediary nodes, identify the region ultimately affected by the risk evolution, and define the diffusion terminal nodes based on the spatial location of the region ultimately affected by the risk evolution, the result of characteristic parameter changes, and the temporal relationship.

[0106] Next, the propagation path of characteristic parameters corresponding to the intermediate nodes of transmission is tracked to further identify the regions ultimately affected by risk evolution. Changes in characteristic parameters within these regions represent the terminal manifestation of risk evolution, such as the downstream areas affected by glacial debris flow disasters. Based on the spatial location of these regions, the results of characteristic parameter changes, and their temporal relationship with the intermediate nodes of transmission, diffusion terminal nodes are defined.

[0107] Step S138: Analyze the characteristic parameter transmission relationship between the core trigger node, the transmission intermediary node, and the diffusion terminal node, determine the transmission direction, transmission carrier, and transmission delay between the core trigger node, the transmission intermediary node, and the diffusion terminal node, and construct the node association matrix.

[0108] A thorough analysis is conducted on the characteristic parameter transmission relationships among the core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes. The transmission direction is determined, i.e., from which node the characteristic parameter is transmitted to which node; the transmission carrier refers to the specific medium or factor transmitting the characteristic parameter, such as stress or water flow; and the transmission delay is the time required for the characteristic parameter to be transmitted from one node to another. Based on these analytical results, a node correlation matrix is ​​constructed.

[0109] Step S1381: Extract the feature parameter sets and time series change data of the core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes from the basic data of the risk transmission node network, classify and organize them according to node type, and form a node feature time series dataset.

[0110] First, feature parameter sets and temporal variation data for each node are extracted from the basic data of the risk transmission node network. Core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes each have their unique feature parameter sets and temporal variation data. This data is then categorized and organized according to node type to form a node feature temporal dataset. For example, data from all core triggering nodes are grouped together, while data from transmission intermediary nodes and diffusion terminal nodes are categorized separately.

[0111] Step S1382: Compare the time-series change data of the characteristic parameters of the core trigger node and the adjacent transmission intermediary node, and identify the characteristic parameters that show consistent change trends between the two. The characteristic parameters with consistent change trends are the transmission carriers between the core trigger node and the adjacent transmission intermediary node.

[0112] By comparing the temporal changes of characteristic parameters of the core triggering node and adjacent intermediate transmission nodes, we can observe which characteristic parameters exhibit consistent trends. For example, if a characteristic parameter of the core triggering node begins to increase at a certain point in time, and a characteristic parameter of an adjacent intermediate transmission node also begins to increase at a subsequent point in time, with similar trends, the characteristic parameters exhibiting consistent trends are identified as the transmission carriers between the core triggering node and the adjacent intermediate transmission nodes.

[0113] Step S1383: Based on the temporal change data of the transmission carrier, determine the time when the transmission carrier in the core trigger node changes and the time when the corresponding transmission carrier in the transmission intermediary node changes, calculate the difference between the two times, and the difference between the two times is the transmission delay from the core trigger node to the transmission intermediary node.

[0114] Based on the temporal change data of the conduction carrier, the time points when the conduction carrier changes in the core trigger node and the corresponding time points when the conduction carrier changes in the intermediate nodes are identified. The difference between these two time points is calculated; this difference represents the conduction delay from the core trigger node to the intermediate node. The conduction delay reflects the time required for the feature parameters to be transmitted from the core trigger node to the intermediate node.

[0115] Step S1384: Based on the propagation direction of the conduction carrier, determine the conduction direction from the core trigger node to the conduction intermediary node, and mark the magnitude and type of change of the conduction carrier during the conduction process.

[0116] Based on the propagation path of the conduction carrier and the order of change of characteristic parameters, the direction of conduction from the core trigger node to the intermediate conduction node is determined. Simultaneously, the magnitude of change in the conduction carrier during the conduction process is recorded, i.e., the amount of change in the conduction carrier from the core trigger node to the intermediate conduction node, as well as the type of change, such as whether it is an increase or decrease, a linear change or a non-linear change, etc.

[0117] Step S1385: Analyze the characteristic parameter transmission relationships between the transmission intermediate node and subsequent transmission intermediate nodes, and between the transmission intermediate node and the diffusion terminal node in sequence, and determine the transmission carrier, transmission delay and transmission direction between each node pair.

[0118] Following a similar approach, the characteristic parameter transmission relationships between intermediate nodes and subsequent intermediate nodes, and between intermediate nodes and terminal nodes, are analyzed sequentially. For each pair of nodes, the transmission carrier, transmission delay, and transmission direction are determined. Through the above analysis, a comprehensive understanding of the relationships between various nodes in the entire risk transmission process can be achieved.

[0119] Step S1386: For nodes with multiple transmission paths, distinguish between primary and secondary transmission carriers. Primary transmission carriers are feature parameters that play a dominant role in the characteristic changes of subsequent nodes, while secondary transmission carriers are feature parameters that play an auxiliary role.

[0120] For nodes with multiple transmission paths, it is necessary to distinguish between primary and secondary transmission carriers. Primary transmission carriers are feature parameters that dominate the changes in subsequent node characteristics, and their changes have the greatest impact on subsequent nodes. Secondary transmission carriers are feature parameters that play a supporting role, and their impact on subsequent nodes is relatively small. The primary and secondary transmission carriers are determined by analyzing the contribution of node characteristic changes.

[0121] Step S1387: Record the type of conduction carrier, conduction delay duration, conduction direction, and change characteristics of conduction carrier between each node pair to form detailed data on node conduction relationships.

[0122] The information such as the type of transmission carrier, transmission delay, transmission direction, and changes in the transmission carrier between each node pair is recorded in detail to form detailed data on node transmission relationships. This data forms the basis for constructing the node association matrix.

[0123] Step S1388: Based on the detailed data of node transmission relationships, construct a node association matrix. The rows and columns of the node association matrix correspond to each node, and the elements of the node association matrix contain information on the transmission carrier, transmission delay, and transmission direction between corresponding node pairs.

[0124] A node association matrix is ​​constructed using detailed node transmission relationship data. The rows and columns of the matrix correspond to individual nodes, and the elements contain information such as the transmission medium, transmission delay, and transmission direction between corresponding node pairs. This matrix provides a clear visual representation of the transmission relationships between nodes.

[0125] Step S1389: Complete the data in the node association matrix, ensuring that all node pairs with a transmission relationship have corresponding elements in the node association matrix, and mark node pairs without a direct transmission relationship.

[0126] The constructed node association matrix is ​​checked to ensure that all node pairs with transmission relationships have corresponding matrix elements. If any transmission relationships are missing, they are supplemented to ensure completeness. For node pairs without direct transmission relationships, they are marked in the matrix to clarify that there is no direct transmission path between them.

[0127] Step S13810: Integrate the node association matrix with the spatial location data of each node, and supplement the spatial distance information between node pairs in the node association matrix so that the node association matrix contains both temporal transmission characteristics and spatial distribution characteristics.

[0128] Finally, the node association matrix is ​​integrated with the spatial location data of each node, and the spatial distance information between node pairs is added to the matrix. In this way, the node association matrix not only includes temporal transmission characteristics but also spatial distribution characteristics, which can more comprehensively reflect the relationships between nodes.

[0129] Step S139: Integrate the definition information, spatial location data, temporal characteristic data, and node association matrix of core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes to construct a risk transmission node network.

[0130] By integrating the definition information, spatial location data, temporal characteristic data, and node relationship matrix of core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes, a complete risk transmission node network is constructed. This risk transmission node network demonstrates the relationships and interactions between various nodes during the risk evolution process.

[0131] Step S140: Construct a multi-dimensional monitoring and response linkage mechanism based on the risk transmission node network, integrate response measures such as UAV inspection parameter adjustment, ground monitoring equipment linkage, and early warning information push, and generate a targeted risk monitoring linkage execution plan.

[0132] After constructing a risk transmission node network, a multi-dimensional monitoring and response linkage mechanism is built based on this network to generate a risk monitoring linkage execution plan.

[0133] Step S141: Analyze the spatial location data, temporal characteristic data, and node association matrix of the core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk transmission node network to identify the core impact area of ​​risk evolution, the main transmission path, and the temporal rhythm of risk diffusion.

[0134] First, the risk transmission node network is analyzed to extract spatial location data, temporal feature data, and node correlation matrix of core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes. Through analysis of this data, the core impact area of ​​risk evolution is identified, namely the area where the core triggering node is located; the main transmission path, i.e., the path through which feature parameters are transmitted from the core triggering node to the diffusion terminal node; and the temporal rhythm of risk diffusion, i.e., the time and speed required for the risk to spread from the core impact area to other areas.

[0135] Step S142: Based on the spatial location and risk characteristics of the core triggering node, determine the key monitoring area for UAV inspection, adjust the distribution density of inspection trajectory nodes in the key monitoring area, and increase the number of nodes in the key monitoring area.

[0136] Based on the spatial location and risk characteristics of the core triggering node, key monitoring areas for UAV inspections are determined. Within these key monitoring areas, the distribution density of inspection trajectory nodes is increased, i.e., the number of nodes is increased. This allows UAVs to collect data more frequently in these areas, improving the accuracy and timeliness of monitoring the core impact areas. For example, if the core triggering node is located in a specific area of ​​a glacier, more trajectory nodes are added around that area to monitor risk changes in that area in more detail.

[0137] Step S143: According to the conduction delay and conduction direction in the node association relationship matrix, predict the time when the risk spreads to each conduction intermediate node and diffusion terminal node, and plan in advance the inspection time for the UAV to reach each node. The planned inspection time for the UAV to reach each node is earlier than or equal to the predicted time when the risk spreads to the area where each node is located.

[0138] Utilize the conduction delay and conduction direction information in the node association relationship matrix to predict the time when the risk spreads to each conduction intermediate node and diffusion terminal node. Based on these predicted times, plan in advance the inspection time for the UAV to reach each node. Ensure that the planned inspection time for the UAV to reach each node is earlier than or equal to the predicted time when the risk spreads to the area where each node is located, so that before the risk spreads to the area, the UAV can complete data collection and monitoring tasks and timely grasp the risk situation.

[0139] Step S144: Combine the time rhythm of risk evolution and the requirements for adjusting UAV inspection parameters to generate a dynamic adjustment instruction for UAV inspection parameters. The dynamic adjustment instruction for UAV inspection parameters includes rules for adding and deleting trajectory nodes, flight speed adjustment parameters, and conditions for switching the working mode of sensing devices.

[0140] Combine the time rhythm of risk evolution and the requirements for adjusting UAV inspection parameters to generate a dynamic adjustment instruction for UAV inspection parameters. The rules for adding and deleting trajectory nodes specify the situations under which trajectory nodes need to be added or deleted, such as adding nodes when a new risk area is discovered and deleting nodes when the risk in certain areas decreases. The flight speed adjustment parameters are set according to the urgency of risk evolution and the time requirements of the inspection task, reducing the flight speed in key monitoring areas to increase data collection time; increasing the flight speed in non-key areas to improve inspection efficiency. The conditions for switching the working mode of sensing devices specify when the sensing devices switch their working mode according to different risk characteristic areas and data collection requirements, such as switching to the multi-device synchronous collection mode in the core risk area.

[0141] Step S145: Based on the spatial distribution of the risk conduction node network, collect the location data of the ground monitoring devices already deployed in the glacier area, associate the ground monitoring devices with the risk nodes, and define the scope of the ground monitoring devices that need to be linked.

[0142] According to the spatial distribution of the risk conduction node network, collect the location data of the ground monitoring devices already deployed in the glacier area. Match the locations of the ground monitoring devices with the locations of the risk nodes and associate the ground monitoring devices with the risk nodes. For example, associate the ground monitoring devices located near the core trigger node with the core trigger node. Through the above association, define the scope of the ground monitoring devices that need to be linked, that is, these ground monitoring devices will work in coordination with the UAV inspection system to jointly monitor the risk situation in the glacier area.

[0143] Step S1451: Extract the spatial coordinate data of all core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes from the risk transmission node network, classify and label them according to node type and risk characteristics, and form a set of risk node spatial coordinates.

[0144] First, spatial coordinate data of all nodes in the risk transmission node network are extracted, including core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes. This spatial coordinate data is then classified and labeled according to node type and risk characteristics; for example, core triggering nodes are labeled as high-risk nodes, transmission intermediary nodes as medium-risk nodes, and diffusion terminal nodes as low-risk nodes, thus forming a set of risk node spatial coordinates.

[0145] Step S1452: Collect basic information on the ground monitoring equipment deployed in the glacier area. The basic information of the ground monitoring equipment includes the equipment name, equipment type, spatial coordinates, monitoring parameter range, working status and communication address, forming a set of basic information of the ground monitoring equipment.

[0146] By interacting with the ground monitoring equipment management system, basic information on the ground monitoring equipment deployed within the glacier area is collected. This information includes equipment name, equipment type (such as water level gauges, geothermometers, crack gauges, etc.), spatial coordinates, monitoring parameter range (such as the measurement range of water level gauges, the temperature measurement range of geothermometers, etc.), operating status (such as normal operation, fault, offline, etc.), and communication address. This information is then compiled into a basic information set for the ground monitoring equipment.

[0147] Step S1453: Associate the set of spatial coordinates of risk nodes with the spatial coordinate data in the set of basic information of ground monitoring equipment, and calculate the spatial distance between each ground monitoring equipment and each risk node.

[0148] Using Geographic Information System (GIS) technology, the spatial coordinates of risk nodes are correlated with the spatial coordinate data in the basic information set of ground monitoring equipment. Through coordinate calculation, the spatial distance between each ground monitoring device and each risk node is determined. For example, the straight-line distance between a water level gauge and a core triggering node can be calculated.

[0149] Step S1454: For each risk node, select the ground monitoring device with the smallest spatial distance calculation result and establish a preliminary correlation between the risk node and the ground monitoring device.

[0150] For each risk node, the ground monitoring device with the smallest spatial distance to that risk node is selected from the basic information set of ground monitoring devices, and a preliminary correlation is established. This preliminary correlation is based on the principle of closest spatial distance, assuming that the ground monitoring device closest to the risk node can most directly reflect the risk situation of that node.

[0151] Step S1455: Analyze the monitoring parameter range of the preliminarily associated ground monitoring equipment, and determine whether the monitoring parameter range of the ground monitoring equipment can cover the characteristic parameter type of the corresponding risk node. If the monitoring parameter range can fully cover it, the preliminary association relationship is retained.

[0152] The monitoring parameter ranges of the initially associated ground monitoring equipment are analyzed to determine whether they can cover the characteristic parameter types of the corresponding risk nodes. For example, if the characteristic parameters of a risk node include the width and temperature of glacier surface fissures, and the initially associated ground monitoring equipment consists of fissure gauges and geothermometers, and their monitoring parameter ranges can cover these two characteristic parameter types, then the initial association is retained.

[0153] Step S1456: If the monitoring parameter range of the initially associated ground monitoring equipment cannot cover the characteristic parameter type of the corresponding risk node, the screening range is expanded, and ground monitoring equipment in the surrounding area is re-screened to select equipment whose monitoring parameters can be covered and whose spatial distance is less than the set distance threshold as associated equipment.

[0154] If the monitoring parameters of the initially associated ground monitoring equipment cannot cover the characteristic parameter types of the corresponding risk node, the screening scope is expanded. Within a certain area surrounding the risk node, ground monitoring equipment is re-screened. Equipment whose monitoring parameters cover the characteristic parameter types of the risk node and whose spatial distance is less than a set distance threshold is selected as the associated equipment. Setting a distance threshold ensures that the distance between the associated equipment and the risk node is within a reasonable range, thereby guaranteeing the validity of the monitoring data.

[0155] Step S1457: For areas where multiple risk nodes are clustered, integrate the risk node characteristic parameter requirements in the area and select ground monitoring equipment that can simultaneously cover the characteristic parameters of multiple risk nodes for association.

[0156] For areas with multiple clustered risk nodes, the characteristic parameter requirements of all risk nodes in the area should be integrated. Ground monitoring equipment capable of simultaneously covering the characteristic parameters of multiple risk nodes should be selected for association to improve the utilization efficiency of ground monitoring equipment and reduce redundant deployment. For example, in an area with multiple transmission mediator nodes, a ground monitoring device capable of simultaneously monitoring multiple parameters such as temperature, humidity, and stress should be selected and associated with these nodes.

[0157] Step S1458: According to the risk characteristics of the risk nodes, prioritize the associated ground monitoring devices. The devices associated with the core trigger nodes and the conduction intermediary nodes with significant risk characteristics are of the first priority level, and the devices associated with the nodes whose risk characteristic change range is lower than the set risk threshold are of the second priority level.

[0158] According to the risk characteristics of the risk nodes, prioritize the associated ground monitoring devices. The ground monitoring devices associated with the core trigger nodes and the conduction intermediary nodes with significant risk characteristics are listed as the first priority level. These devices need to ensure the priority of their working status and the timeliness of data transmission. The devices associated with the nodes whose risk characteristic change range is lower than the set risk threshold are of the second priority level, and their working priority is relatively low.

[0159] Step S1459: Based on the association relationship and the priority ranking, define the scope of the ground monitoring devices that need to be linked. The devices of the first priority level are set according to the full-process synchronous linkage mode, and the devices of the second priority level are set according to the on-demand linkage mode according to the risk evolution rhythm.

[0160] Based on the association relationship and the priority ranking, define the scope of the ground monitoring devices that need to be linked. The ground monitoring devices of the first priority level are set according to the full-process synchronous linkage mode, that is, during the entire UAV inspection process, these devices are always in a working state and maintain synchronous data interaction with the UAV inspection system. The devices of the second priority level are set according to the on-demand linkage mode according to the risk evolution rhythm, that is, according to the situation of risk evolution, these devices are started when needed for data collection and transmission to save energy and communication resources.

[0161] Step S14510: Record the names, spatial coordinates, monitoring parameters, priorities, and information of the associated risk nodes of the ground monitoring devices that need to be linked to form a ground monitoring device linkage list.

[0162] Record the names, spatial coordinates, monitoring parameters, priorities, and information of the associated risk nodes of the ground monitoring devices that need to be linked to form a ground monitoring device linkage list.

[0163] Step S146: For the associated ground monitoring devices, set the monitoring parameters and acquisition frequencies of the ground monitoring devices according to the type of characteristic parameters of the corresponding risk nodes, so that the ground monitoring data and the UAV perception data are complementary.

[0164] For the associated ground monitoring equipment, the monitoring parameters and acquisition frequency are set according to the characteristic parameter types of the corresponding risk nodes. For example, if the characteristic parameters of a risk node include water level and temperature, the monitoring parameters of the associated water level gauge and geothermometer are set to water level and temperature respectively, and an appropriate acquisition frequency is set according to the time rhythm of risk evolution. This ensures that the ground monitoring data and the UAV sensing data complement each other in terms of parameter type and acquisition time, so as to comprehensively reflect the risk situation in the glacier area.

[0165] Step S147: Generate ground monitoring equipment linkage instructions. The ground monitoring equipment linkage instructions include equipment start-up time, monitoring parameter settings, data transmission frequency and data association identifier, so that the ground monitoring equipment starts working synchronously according to the risk evolution rhythm.

[0166] Based on the settings of the ground monitoring equipment, a linkage command for the ground monitoring equipment is generated. This linkage command includes the equipment start-up time, i.e., the time when the equipment begins operation; monitoring parameter settings, i.e., the types and ranges of parameters that the equipment needs to monitor; data transmission frequency, i.e., the time interval at which the equipment transmits data to the ground control center; and data association identifiers, used to associate ground monitoring data with UAV perception data. Through these commands, the ground monitoring equipment can start operating synchronously according to the rhythm of risk evolution, coordinating with the UAV inspection system.

[0167] Step S148: Based on the risk characteristics and impact range of risk nodes, classify the early warning levels. Different early warning levels correspond to different early warning information push ranges and push content. Early warning information for the core risk impact area is pushed to all relevant units and personnel in the area, while early warning information for the risk diffusion impact area is pushed to the monitoring and management department.

[0168] Early warning levels are determined based on the risk characteristics and impact scope of risk nodes. These levels can be categorized into multiple grades, such as Level 1, Level 2, and Level 3. Different warning levels correspond to different information dissemination scopes and content. For areas with core risk impact, warning information needs to be disseminated to all relevant units and personnel within the area so they can take timely countermeasures. For areas affected by risk spread, warning information is disseminated to the monitoring and management department for further assessment and decision-making.

[0169] Step S149: Construct early warning information push rules, determine the push channels, push frequency and information update cycle for different early warning levels, and at the same time associate the temporal characteristics of the risk transmission node network to match the time rhythm of early warning information push and risk diffusion reaching the corresponding area.

[0170] Establish rules for pushing early warning information, specifying the channels for different warning levels, such as SMS, email, and app notifications; push frequency, i.e., the number of times and time intervals between warning messages; and information update cycle, i.e., the frequency of updating the warning information content. Simultaneously, link the rules for pushing early warning information with the temporal characteristics of the risk transmission node network to ensure that the timing of early warning information pushes matches the time rhythm of risk spread to the corresponding area. For example, before a risk is about to spread to a certain area, early warning information can be sent in advance through the appropriate push channels.

[0171] Step S1410: Integrate the UAV inspection parameter dynamic adjustment instructions, ground monitoring equipment linkage instructions, early warning level classification standards, and early warning information push rules to form a risk monitoring linkage execution plan that includes response measure activation conditions, execution process, and responsible entities.

[0172] This plan integrates dynamic adjustment commands for UAV inspection parameters, linkage commands for ground monitoring equipment, early warning level classification standards, and early warning information push rules to form a risk monitoring linkage execution scheme. This scheme includes the activation conditions for response measures (i.e., under what conditions the corresponding response measures are activated); the execution process (i.e., the steps and sequence of execution for each response measure); and the responsible entity (i.e., the unit or personnel responsible for implementing each response measure). Through this scheme, UAV inspections, ground monitoring equipment linkage, and early warning information push can be organically combined, improving the efficiency and accuracy of glacial debris flow risk monitoring.

[0173] Step S150: Inject the on-site implementation feedback data of the risk monitoring linkage execution plan into the dynamic collaborative adaptation process of the inspection trajectory, and update the parameters of the dynamic adaptation trajectory plan in combination with the real-time risk evolution situation.

[0174] After the risk monitoring and linkage implementation plan is implemented, on-site implementation feedback data is collected and injected back into the dynamic collaborative adaptation process of the inspection trajectory to update the parameters of the dynamic adaptation trajectory plan.

[0175] Step S151: Collect on-site implementation feedback data of the risk monitoring and linkage execution plan. The on-site implementation feedback data includes actual flight status data after the adjustment of UAV inspection parameters, working status data of sensing equipment, linkage response data of ground monitoring equipment, early warning information push feedback data, and actual situation data of on-site risk evolution.

[0176] The system collects on-site implementation feedback data of the risk monitoring and linkage execution plan through channels such as the drone's flight control system, the status monitoring module of the sensing equipment, the data transmission system of the ground monitoring equipment, and the early warning information push platform. Actual flight status data includes the drone's actual flight trajectory, flight speed, and flight altitude; sensing equipment operational status data includes equipment startup status, data collection quality, and equipment fault information; ground monitoring equipment linkage response data includes equipment startup time, data collection status, and data transmission quality; early warning information push feedback data includes the number of units and personnel receiving early warning information and their feedback; and the actual situation data of on-site risk evolution is the actual change in risk in the glacier area obtained through various monitoring methods.

[0177] Step S152: Classify and organize the on-site implementation feedback data, dividing it into UAV operation feedback data, perception data feedback data, ground equipment feedback data, early warning response feedback data, and risk situation feedback data according to data type. The data under each category is associated with the corresponding time and space information.

[0178] The collected field feedback data was categorized and organized according to data type, including UAV operation feedback data, perception data feedback data, ground equipment feedback data, early warning response feedback data, and risk situation feedback data. Within each category, the data underwent further processing and filtering to remove invalid and noisy data. Simultaneously, the data in each category was associated with corresponding time and spatial information for subsequent analysis and processing.

[0179] Step S153: Extract key parameters of the core area, evolution direction, and evolution rate of actual risk evolution from the risk situation feedback data, compare them with the previous risk evolution prediction results, identify the differences between the two, and analyze the reasons for the differences.

[0180] Key parameters such as the core area, direction, and rate of actual risk evolution are extracted from risk situation feedback data. These actual parameters are then compared with the risk evolution prediction results obtained from previous models for predicting glacier region risk evolution to identify discrepancies. For example, the core area of ​​actual risk evolution may differ from the predicted core area, or the actual evolution rate may be faster or slower than predicted. A deeper analysis reveals that these discrepancies may be due to inaccurate input data for the prediction model, unreasonable model parameter settings, or unforeseen changes in actual environmental conditions.

[0181] Step S154: Compare the UAV operation feedback data with the flight parameters in the dynamic adaptation trajectory scheme, analyze the adaptation relationship between trajectory node distribution, flight speed adjustment rules and actual flight state, calculate the coordinate deviation between the flight trajectory and the planned trajectory, and mark the area as the trajectory deviation area if the deviation is greater than the set tolerance threshold; compare the planned flight speed with the actual flight speed, and mark the speed adjustment as unreasonable if the speed difference is greater than the set speed tolerance threshold.

[0182] The actual flight trajectory and speed from the UAV's operational feedback data are compared with the flight parameters in the dynamically adapted trajectory scheme. The adaptation relationship between the trajectory node distribution and flight speed adjustment rules and the actual flight state is analyzed. The coordinate deviation between the actual flight trajectory and the planned trajectory is calculated, i.e., the distance between the actual position and the planned position at each trajectory node. If this deviation exceeds a set tolerance threshold, the area is marked as a trajectory deviation area. Simultaneously, the planned flight speed and the actual flight speed are compared, and the speed difference between the two is calculated. If the speed difference exceeds a set speed tolerance threshold, the speed adjustment is marked as unreasonable. Through these analyses, problems existing in the UAV's flight process are identified.

[0183] Step S155: Extract the sensing data quality parameters of different regions from the sensing data feedback data, analyze the matching relationship between the sensing device collaborative parameter settings and risk monitoring requirements, compare the signal-to-noise ratio, resolution or integrity indicators of the sensing data with the set quality threshold, and identify the substandard regions and the corresponding sensing device parameters.

[0184] Extract sensing data quality parameters from different regions from the feedback data, such as signal-to-noise ratio, resolution, and completeness. Analyze the matching relationship between the collaborative parameter settings of sensing devices and risk monitoring requirements to determine whether the current parameter settings can meet the risk monitoring requirements. Compare the quality parameters of the sensing data with the set quality thresholds. If the signal-to-noise ratio of the sensing data in a certain region is lower than the threshold, the resolution does not meet the requirements, or the data is incomplete, then that region is identified as a region with substandard sensing data, and the corresponding sensing device parameters, such as acquisition frequency and transmission power, are recorded.

[0185] Step S156: Based on the differences between the risk evolution prediction results and the actual risk situation, the deviation of the UAV operation feedback data, and the perception data quality issues, determine the core parameters that need to be adjusted during the dynamic collaborative adaptation processing of the inspection trajectory.

[0186] Taking into account the differences between the risk evolution prediction results and the actual risk situation, the deviation of the UAV operation feedback data, and the quality issues of the perception data, the core parameters that need to be adjusted in the dynamic collaborative adaptation process of the inspection trajectory are determined. These core parameters may include the input parameter weights of the risk evolution prediction model for glacier areas, the distribution density of trajectory nodes, the flight speed adjustment rules, and the collaborative parameters of the perception devices.

[0187] Step S157: Adjust the input parameter weights of the risk evolution prediction model for glacier regions to address the differences in risk evolution, optimize the prediction accuracy of the risk evolution prediction model for actual risk situations, and adjust the distribution density of trajectory nodes based on the core risk areas.

[0188] To address the discrepancy between the predicted risk evolution and the actual risk situation, the weights of the input parameters in the glacier region risk evolution prediction model are adjusted. Through analysis of historical and actual feedback data, it is determined which input parameters have a greater impact on the prediction results, and the weights of these parameters are increased; the weights of parameters with less influence are decreased. These adjustments optimize the model's accuracy in predicting the actual risk situation. Simultaneously, based on the location and extent of the core risk area, the distribution density of trajectory nodes is adjusted to ensure that the trajectory nodes more accurately cover the actual high-risk areas.

[0189] Step S158: To address the operational deviations of the UAV, the flight speed adjustment rules are corrected, and the smooth transition parameters between trajectory nodes are optimized to stabilize the UAV's flight status while avoiding terrain obstacles discovered during actual flight.

[0190] To address issues such as trajectory deviations and unreasonable speed adjustments identified in drone operation feedback data, the flight speed adjustment rules were revised. Based on actual flight data, flight speed parameters for different regions were adjusted to make the drone's flight speed more reasonable. Simultaneously, smooth transition parameters between trajectory nodes, such as turning radius and climb / descent rate, were optimized to make the drone more stable during flight and reduce turbulence and shaking. Furthermore, based on terrain obstacles discovered during actual flight, such as unexpected peaks and valleys, avoidance mechanisms were implemented in the trajectory planning, adjusting the position and sequence of trajectory nodes.

[0191] Step S159: To address the issue of perceived data quality, adjust the collaborative parameters of the sensing devices in different risk characteristic areas, optimize the acquisition frequency and data resolution settings, and improve the reliability and accuracy of the perceived data.

[0192] For areas where the quality of sensed data is substandard, adjust the collaborative parameters of the corresponding sensing devices in that area. For example, increase the acquisition frequency to improve the temporal resolution of the data, or increase the data resolution to obtain more detailed information. At the same time, optimize the operating mode of the sensing devices, such as adjusting the transmission power of acoustic detection devices and the signal frequency of electromagnetic penetration devices, to improve the reliability and accuracy of the sensed data.

[0193] Step S1510: Re-inject the adjusted parameters into the dynamic collaborative adaptation process of the inspection trajectory to generate an updated dynamic adaptation trajectory scheme, and apply the updated dynamic adaptation trajectory scheme to the next round of UAV inspection.

[0194] The adjusted parameters from steps S157, S158, and S159 are reinjected into the dynamic collaborative adaptation process of the inspection trajectory. Following steps S110 to S1110, an updated dynamic adaptation trajectory scheme is generated. This updated scheme is applied to the next round of UAV inspections. Through continuous feedback and adjustments, the dynamic adaptation trajectory scheme can better adapt to the actual evolution of risks in the glacier area, improving the effectiveness of risk monitoring.

[0195] The entire method for monitoring glacial debris flows using drones involves collecting data related to the glacier region, which may include sensitive information such as geographical location. To protect this sensitive data, data encryption technology is used to encrypt data during transmission and storage, ensuring that the data is not stolen or tampered with during transmission and is not illegally accessed during storage. Simultaneously, strict access control is implemented, allowing only authorized personnel to access this sensitive data to prevent data leakage. Furthermore, during data collection, the collection of irrelevant sensitive information is minimized, focusing only on data relevant to glacial debris flow monitoring to reduce the risk of privacy breaches at the source.

[0196] In one exemplary embodiment, a glacier debris flow monitoring system for drone inspection is provided. This system can be a terminal, server, etc., and its internal structure diagram can be as follows: Figure 2As shown, this glacial debris flow monitoring system for drone inspection includes a processor, memory, input / output interface, communication interface, display unit, and input device. The processor, memory, and input / output interface are connected via a system bus, and the communication interface, display unit, and input device are also connected to the system bus via the input / output interface. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides the environment for the operation of the operating system and computer programs in the non-volatile storage medium. The input / output interface is used for exchanging information between the processor and external devices. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, mobile cellular networks, near-field communication, or other technologies. When the computer program is executed by the processor, it implements a method for monitoring glacial debris flows using drone inspection. The display unit is used to generate a visually visible image and can be a display screen, projection device, or virtual reality imaging device. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, or it can be a button, trackball, or touchpad set on the shell of a glacier debris flow monitoring system used for drone inspections, or it can be an external keyboard, touchpad, or mouse, etc.

[0197] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspection, characterized in that, The method includes: Based on the risk evolution prediction results of glacier areas and the flight constraints of UAVs, dynamic collaborative adaptation processing of inspection trajectories is performed to generate dynamic adaptation trajectory schemes that include trajectory node distribution, flight speed adjustment rules, and collaborative parameters of sensing devices. Based on the dynamic adaptation trajectory scheme, the UAV multi-source perception system is driven to perform hierarchical and precise perception operations, collect multi-dimensional data on the surface morphology, internal structure and surrounding environment of different risk characteristic areas in the glacier region, and generate a time-seriesd glacier region risk perception data chain. By triggering risk transmission node mining and correlation analysis through the risk perception data chain in glacier areas, core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk evolution process are identified, and a risk transmission node network including node correlation and transmission time sequence characteristics is constructed. A multi-dimensional monitoring and response linkage mechanism is constructed based on the risk transmission node network, integrating response measures such as UAV inspection parameter adjustment, ground monitoring equipment linkage, and early warning information push, and generating targeted risk monitoring linkage execution plans. Feedback data from the on-site implementation of the risk monitoring and linkage execution plan is injected into the dynamic collaborative adaptation process of the inspection trajectory, and the parameters of the dynamic adaptation trajectory plan are updated in combination with the real-time risk evolution.

2. The method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections according to claim 1, characterized in that, The dynamic collaborative adaptation processing of the inspection trajectory based on the risk evolution prediction results of glacier areas and the flight constraints of UAVs generates a dynamic adaptation trajectory scheme that includes trajectory node distribution, flight speed adjustment rules, and sensing device collaborative parameters, including: By integrating historical disaster evolution data, real-time topographic data, UAV flight performance parameters, and real-time meteorological data in glacier areas, a basic dataset for dynamic collaborative adaptation of trajectories is formed. This dataset covers the topographic relief patterns, historical risk evolution paths, maximum flight altitude of UAVs, endurance, and real-time wind force and direction influencing factors in glacier areas. The core sensitive features of glacial debris flow risk evolution are extracted from the trajectory dynamic collaborative adaptation basic dataset. These core sensitive features include the density of surface fractures in glaciers, the stress transmission path inside glaciers, the infiltration range of water bodies around glaciers, the degree of loosening of strata particles, and the risk evolution rate. A risk evolution prediction model for glacier regions is constructed based on the extracted core sensitive features. The glacier region risk evolution prediction model has a built-in risk level classification module and an evolution trend prediction module. The glacier region risk evolution prediction model is trained by historical disaster evolution data, so that the glacier region risk evolution prediction model can output risk evolution prediction results in different time dimensions based on real-time topographic data and environmental data. Drone flight constraints are extracted from the trajectory dynamic collaborative adaptation basic dataset and transformed into hard rules for trajectory adaptation. The drone flight constraints include maximum flight altitude limit, minimum turning radius, endurance time threshold and flight prohibition area under severe weather conditions. The parameters of risk core area distribution, risk evolution direction, and evolution rate in the risk evolution prediction results are correlated with the UAV flight constraints to preliminarily plan the core coverage area and direction of the inspection trajectory. The core coverage area covers the risk core area distribution, and the direction satisfies the UAV flight constraints. Based on the preliminary planned inspection trajectory, the distribution density of trajectory nodes is set according to the risk evolution rate of the core risk area. The distribution density of trajectory nodes is adjusted in the positive direction according to the evolution rate. The faster the evolution rate, the denser the distribution of trajectory nodes. At the same time, dynamic tracking nodes are set in the direction of risk evolution. Based on the distribution of trajectory nodes and the risk characteristics in the risk evolution prediction results, flight speed adjustment rules are constructed. Low-speed flight mode is set for the core risk area, uniform-speed flight mode is set for the risk diffusion and impact area, and flight speed is adjusted according to the set gradient for the risk transition and correlation area. By combining the perception needs of different risk areas with the performance of multi-source perception devices carried by UAVs, collaborative parameters for perception devices are constructed. In the core risk area, a multi-device synchronous acquisition mode is activated, and in the risk spread and impact area, the corresponding devices are activated according to the perception priority. At the same time, the acquisition frequency and data transmission priority of each device are determined. An initial dynamic adaptation trajectory scheme is constructed based on trajectory node distribution, flight speed adjustment rules, and sensing device collaborative parameters. The initial dynamic adaptation trajectory scheme is then collaboratively verified with the real-time evolution situation output by the glacier region risk evolution prediction model. The verification process determines whether the trajectory nodes in the initial dynamic adaptation trajectory scheme cover the risk area indicated by the real-time evolution situation. The system integrates the coordinates of trajectory nodes, flight speed adjustment parameters, sensing device operating parameters, and trajectory dynamic adjustment rules after successful collaborative verification to generate a dynamically adapted trajectory scheme.

3. The method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections according to claim 1, characterized in that, The method, based on a dynamic trajectory adaptation scheme, drives the UAV multi-source perception system to perform hierarchical precise perception operations, collecting multi-dimensional data on the surface morphology, internal structure, and surrounding environment of different risk characteristic areas in the glacier region, generating a time-seriesd glacier region risk perception data chain, including: The distribution of trajectory nodes, risk characteristics, flight speed adjustment rules and sensing equipment coordination parameters in the dynamic adaptation trajectory scheme are analyzed. The glacier inspection area is divided into three sensing levels according to risk characteristics: core risk triggering sensing area, risk transmission intermediate sensing area and risk diffusion terminal sensing area. Differentiated perception strategies are constructed for different perception levels. The core risk trigger perception zone adopts a multi-source device synchronous acquisition strategy, which starts optical imaging equipment, acoustic detection equipment, electromagnetic penetration equipment and environmental perception equipment to work synchronously. The acquisition frequency is set according to the parameter standard corresponding to the core risk trigger perception zone in the dynamic adaptation trajectory scheme. The risk transmission intermediate sensing zone adopts a multi-source device alternating acquisition strategy, which alternately activates optical imaging equipment and acoustic detection equipment at preset time intervals, and assists in activating environmental sensing equipment. The acquisition frequency is set according to the parameter standard corresponding to the risk transmission intermediate sensing zone in the dynamic adaptation trajectory scheme. The risk diffusion terminal sensing area adopts a single device acquisition strategy, only activating the optical imaging device to acquire surface morphology data, while the environmental sensing device acquires data according to the parameter standards corresponding to the risk diffusion terminal sensing area in the dynamic adaptation trajectory scheme. The device activation strategy and data acquisition frequency during the acquisition process are dynamically adjusted based on the remaining flight time of the UAV. Based on the differentiated perception strategy, hierarchical control instructions for the multi-source perception system are generated. The hierarchical control instructions determine the device start-up sequence, acquisition frequency, data resolution and data transmission format of each perception level, and associate the corresponding trajectory node coordinates and flight speed parameters. The drone is controlled to move sequentially according to the trajectory nodes in the dynamic adaptation trajectory scheme. At each trajectory node, the multi-source sensing device is activated according to the corresponding perception level control command, and the acquisition time and the real-time position coordinates of the drone are recorded simultaneously. The glacier surface morphology data at each sensing level are collected using optical imaging equipment. The core risk-triggered sensing zone focuses on collecting the extension direction and width changes of surface fissures and the accumulation state of materials around the fissures. The risk transmission intermediate sensing zone and the risk diffusion terminal sensing zone focus on collecting the overall morphological changes of the surface. Acoustic detection equipment is used to collect data on the internal structure of glaciers, and specific frequency acoustic signals are transmitted into the glacier. The waveform changes of the signals after reflection by different media inside the glacier are received and recorded. The signal transmission power in the core risk trigger sensing area is increased according to the parameter standards in the dynamic adaptation trajectory scheme. Data on the distribution of the internal media of a glacier is collected using electromagnetic penetration equipment to obtain the distribution range and relative position of ice layers, rock debris, and water bodies within the glacier. Temperature, humidity, wind force, and air pressure data are collected at each sensing level through environmental sensing devices. These data are used to analyze the driving role of environmental factors in the evolution of glacier risk. All data collected by sensing devices are integrated in three dimensions according to the collection time, UAV location coordinates, and sensing level to construct a time-series risk perception data chain for glacier areas. Each data unit in the glacier area risk perception data chain contains information about the collection scene and associated data identifiers.

4. The method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections according to claim 3, characterized in that, The process of acquiring glacier internal structure data through acoustic detection equipment, transmitting acoustic signals of a specific frequency band into the glacier, and receiving and recording the waveform changes of the signals after reflection by different media within the glacier includes: The acoustic detection parameters of the corresponding perception level in the dynamic adaptation trajectory scheme are analyzed. The core risk trigger perception zone sets the frequency band acoustic signal transmission parameters of the core risk trigger perception zone in the dynamic adaptation trajectory scheme. The risk transmission intermediate perception zone sets the frequency band parameters of the risk transmission intermediate perception zone in the dynamic adaptation trajectory scheme. The risk diffusion terminal perception zone sets the frequency band parameters of the risk diffusion terminal perception zone in the dynamic adaptation trajectory scheme. The acoustic signal transmission module on the control drone is activated according to the set parameters to transmit continuous acoustic signals to the glacier area corresponding to the current inspection node. The acoustic signal transmission power is dynamically adjusted based on the boundary distance of the monitoring area corresponding to the current inspection node. The acoustic signal receiving module is activated to simultaneously receive acoustic signals reflected from different media interfaces inside the glacier and secondary acoustic signals generated by the vibration of the glacier's internal structure. The received acoustic signal is enhanced by using an adaptive filtering algorithm to reduce interference signals generated by environmental noise and the UAV’s own vibration, thereby improving the signal-to-noise ratio of the effective signal and capturing the waveform characteristics generated by reflection from the glacier’s internal medium. The waveform features of the enhanced acoustic signal are extracted. These waveform features include peak shape, trough depth, waveform duration, waveform distortion degree, and signal attenuation rate. The waveform features reflect the differences in physical properties of different media inside the glacier. The distortion regions in the waveform features are analyzed to identify waveform distortions caused by internal structural rupture, loosening, or inhomogeneity of the medium in the glacier. The signal reception time corresponding to each distortion region is marked, and the internal depth of the glacier corresponding to the distortion region is calculated by combining the acoustic signal propagation speed. For the acoustic signals in the core risk-triggered sensing area, analyze the frequency distribution and continuous trend of waveform distortion to determine the dynamics of structural changes during the stress transmission process inside the glacier. By combining the location coordinates of the current inspection node, the extracted waveform features, depth information corresponding to the distortion area, and signal attenuation data are correlated with the spatial location to form an acoustic detection data unit for the internal structure of the glacier corresponding to each inspection node. The acoustic detection data units of the glacier's internal structure collected multiple times at the same inspection node are processed by signal superposition. The acoustic detection data units of the glacier's internal structure collected from all inspection nodes are integrated and constructed according to the temporal order and spatial distribution of the inspection trajectory to form an acoustic detection data sequence.

5. The method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections according to claim 1, characterized in that, The process involves mining and analyzing risk transmission nodes triggered by a risk perception data chain in glacier regions. This identifies core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk evolution process, and constructs a risk transmission node network that includes node relationships and transmission time sequence characteristics. The feature parameters of each sensing data unit are extracted from the risk perception data chain of the glacier area to form a risk feature parameter set. The feature parameters include surface morphological change parameters, internal structural feature parameters, environmental impact parameters, and time and space information of data collection. Based on the spatial information in the risk feature parameter set, all feature parameters are mapped to the geographic coordinate system of the glacier region to construct the spatial distribution features of the risk features. The spatial distribution features present the spatial location and distribution density of each feature parameter in the glacier region. Analyze the temporal sequence of changes in each characteristic parameter in the spatial distribution characteristics, extract the time points when the characteristic parameters show significant changes, mark them as risk evolution time nodes, and construct a risk evolution timeline in chronological order; On the risk evolution timeline, the spatial distribution characteristics of risk features corresponding to different time nodes are associated, and the regions where significant changes in feature parameters first occur are identified. The regions where significant changes in feature parameters first occur are the initial triggering regions of risk evolution. The set of feature parameters within the initial triggering regions constitutes the basic data of the core triggering nodes. Key driving feature parameters are extracted from the basic data of the core triggering node. These key driving feature parameters are determined based on historical disaster data and are the feature parameters that have the highest correlation with subsequent changes in risk feature parameters. The core triggering node is defined based on the key driving feature parameters and spatial location information. Track the propagation path of key driving characteristic parameters corresponding to core triggering nodes, and identify the regions where characteristic parameters change due to the influence of key driving characteristic parameters in subsequent time nodes of the risk evolution timeline. The regions where characteristic parameters change due to the influence of key driving characteristic parameters are the mediating regions of risk transmission. Analyze the variation patterns of characteristic parameters within the intermediary region, determine the transmission and transformation effects of the intermediary region on key driving characteristic parameters, and define transmission intermediary nodes based on the spatial location of the intermediary region, the type of characteristic parameter variation, and the temporal relationship. Continue to track the propagation path of the characteristic parameters corresponding to the transmission intermediary nodes, identify the region ultimately affected by risk evolution, and the changes in characteristic parameters within the region ultimately affected by risk evolution are the terminal manifestations of risk evolution. Based on the spatial location of the region ultimately affected by risk evolution, the results of characteristic parameter changes, and the temporal relationship, define the diffusion terminal nodes. Analyze the characteristic parameter transmission relationships among core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes to determine the transmission direction, transmission carrier, and transmission delay among core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes, and construct a node association matrix; By integrating the definition information, spatial location data, temporal characteristic data, and node association matrix of core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes, a risk transmission node network is constructed.

6. The method for monitoring glacial debris flows using unmanned aerial vehicle (UAV) inspections according to claim 5, characterized in that, The analysis examines the characteristic parameter transmission relationships among the core triggering node, the conduction intermediary node, and the diffusion terminal node, determines the conduction direction, conduction carrier, and conduction delay among the core triggering node, the conduction intermediary node, and the diffusion terminal node, and constructs a node association matrix, including: The feature parameter sets and time series change data of core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes are extracted from the basic data of the risk transmission node network, and classified and organized according to node type to form a node feature time series dataset. By comparing the time-series changes of characteristic parameters of the core triggering node and the adjacent intermediate transmission nodes, we can identify characteristic parameters that show a consistent trend of change between the two. The characteristic parameters with consistent trends of change are the transmission carriers between the core triggering node and the adjacent intermediate transmission nodes. Based on the temporal change data of the transmission carrier, the time when the transmission carrier in the core trigger node changes and the time when the corresponding transmission carrier in the transmission intermediary node changes are determined. The difference between the two times is calculated as the transmission delay from the core trigger node to the transmission intermediary node. Based on the propagation direction of the conduction carrier, determine the direction of conduction from the core trigger node to the conduction intermediary node, and mark the magnitude and type of change of the conduction carrier during the conduction process; The characteristic parameter transmission relationships between the intermediate nodes and subsequent intermediate nodes, and between intermediate nodes and diffusion terminal nodes are analyzed sequentially to determine the transmission carrier, transmission delay, and transmission direction between each node pair. For nodes with multiple transmission paths, we distinguish between primary and secondary transmission carriers. Primary transmission carriers are feature parameters that play a dominant role in the characteristic changes of subsequent nodes, while secondary transmission carriers are feature parameters that play an auxiliary role. Record the type of conduction medium, conduction delay duration, conduction direction, and changes in conduction medium between each node pair to form detailed data on node conduction relationships; Based on the detailed data of node transmission relationships, a node association matrix is ​​constructed. The rows and columns of the node association matrix correspond to each node, and the elements of the node association matrix contain information on the transmission carrier, transmission delay, and transmission direction between corresponding node pairs. Complete the data in the node association matrix, ensuring that all node pairs with a direct relationship have a corresponding element in the node association matrix, while marking node pairs without a direct relationship. By integrating the node association matrix with the spatial location data of each node, and supplementing the node association matrix with spatial distance information between node pairs, the node association matrix can simultaneously contain temporal transmission characteristics and spatial distribution characteristics.

7. The method for monitoring glacial debris flows applied to UAV inspections according to claim 1, characterized in that, The multi-dimensional monitoring and response linkage mechanism built on the risk transmission node network integrates response measures such as UAV inspection parameter adjustment, ground monitoring equipment linkage, and early warning information push, generating targeted risk monitoring linkage execution plans, including: Analyze the spatial location data, temporal characteristic data, and node correlation matrix of the core triggering nodes, transmission intermediary nodes, and diffusion terminal nodes in the risk transmission node network to identify the core impact areas of risk evolution, the main transmission paths, and the temporal rhythm of risk diffusion. Based on the spatial location and risk characteristics of the core triggering nodes, the key monitoring areas for UAV inspections are determined, the distribution density of inspection trajectory nodes in the key monitoring areas is adjusted, and the number of nodes in the key monitoring areas is increased. Based on the transmission delay and transmission direction in the node association matrix, the time for risk to spread to each transmission intermediary node and diffusion terminal node is predicted, and the inspection time of the UAV to each node is planned in advance. The planned inspection time of the UAV to each node is earlier than or equal to the predicted time for the risk to spread to the area where each node is located. Based on the time rhythm of risk evolution and the need to adjust UAV inspection parameters, dynamic adjustment instructions for UAV inspection parameters are generated. The dynamic adjustment instructions for UAV inspection parameters include rules for adding and deleting trajectory nodes, flight speed adjustment parameters, and conditions for switching the working mode of sensing devices. Based on the spatial distribution of the risk transmission node network, the location data of the ground monitoring equipment already deployed in the glacier area are collected, the ground monitoring equipment is associated with the risk nodes, and the scope of ground monitoring equipment that needs to be linked is defined. For the associated ground monitoring equipment, the monitoring parameters and acquisition frequency of the ground monitoring equipment are set according to the characteristic parameter types of the corresponding risk nodes, so that the ground monitoring data and the UAV perception data complement each other; Generate ground monitoring equipment linkage instructions, which include equipment startup time, monitoring parameter settings, data transmission frequency, and data association identifiers, enabling the ground monitoring equipment to start working synchronously according to the rhythm of risk evolution; Based on the risk characteristics and influence scope of risk nodes, divide the early warning levels. Different early warning levels correspond to different early warning information push scopes and push contents. The early warning information in the core risk impact area is pushed to all relevant units and personnel within the area, and the early warning information in the risk diffusion impact area is pushed to the monitoring and management department; Construct early warning information push rules, determine the push channels, push frequencies, and information update cycles for different early warning levels, and at the same time associate the timing characteristics of the risk conduction node network to match the time rhythm of early warning information push and risk diffusion reaching the corresponding area; Integrate the dynamic adjustment instructions for UAV inspection parameters, the ground monitoring equipment linkage instructions, the early warning level division criteria, and the early warning information push rules to form a risk monitoring linkage execution plan that includes the startup conditions, execution processes, and responsible entities of response measures; 8. The method for monitoring glacial debris flows applied to UAV inspections according to claim 7, characterized in that, Based on the spatial distribution of the risk conduction node network, collect the location data of the ground monitoring equipment deployed in the glacier area, associate the ground monitoring equipment with the risk nodes, and define the scope of the ground monitoring equipment that needs to be linked, including: Extract the spatial coordinate data of all core trigger nodes, conduction intermediary nodes, and diffusion terminal nodes from the risk conduction node network, classify and mark them according to node types and risk characteristics to form a risk node spatial coordinate set; Collect the basic information of the ground monitoring equipment deployed in the glacier area. The basic information of the ground monitoring equipment includes equipment name, equipment type, spatial coordinates, monitoring parameter range, working status, and communication address to form a ground monitoring equipment basic information set; Associate the spatial coordinate data in the risk node spatial coordinate set with the spatial coordinate data in the ground monitoring equipment basic information set, and calculate the spatial distance between each ground monitoring equipment and each risk node; For each risk node, screen out the ground monitoring equipment with the smallest spatial distance calculation result and establish a preliminary association relationship between the risk node and the ground monitoring equipment; Analyze the monitoring parameter range of the preliminarily associated ground monitoring equipment, and judge whether the monitoring parameter range of the ground monitoring equipment can cover the characteristic parameter types of the corresponding risk node. If the monitoring parameter range can completely cover, retain the preliminary association relationship; If the monitoring parameter range of the preliminarily associated ground monitoring equipment cannot cover the characteristic parameter types of the corresponding risk node, expand the screening range, re-screen the ground monitoring equipment in the surrounding area, and select the equipment whose monitoring parameters can cover and the spatial distance is less than the set distance threshold as the associated equipment; For areas with multiple risk nodes clustering, integrate the characteristic parameter requirements of the risk nodes in the area and select the ground monitoring equipment that can cover the characteristic parameters of multiple risk nodes for association; Based on the risk characteristics of risk nodes, the associated ground monitoring equipment is prioritized. Equipment associated with core trigger nodes and transmission intermediary nodes with significant risk characteristics is classified as first-priority, while equipment associated with nodes whose risk characteristics change less than a set risk threshold is classified as second-priority. Based on the correlation and priority ranking, the scope of ground monitoring equipment that needs to be linked is defined. First-priority equipment is set up in the full-process synchronous linkage mode, and second-priority equipment is set up in the on-demand linkage mode according to the risk evolution rhythm. Record the names, spatial coordinates, monitoring parameters, priorities, and associated risk node information of the ground monitoring equipment that needs to be linked, and form a list of ground monitoring equipment to be linked.

9. The method for monitoring glacial debris flows applied to UAV inspections according to claim 1, characterized in that, The process of back-injecting on-site implementation feedback data of the risk monitoring and linkage execution plan into the dynamic collaborative adaptation processing of the inspection trajectory, and updating the parameters of the dynamic adaptation trajectory plan in conjunction with the real-time risk evolution situation, includes: Collect on-site implementation feedback data of the risk monitoring and linkage execution plan. The on-site implementation feedback data includes actual flight status data after the adjustment of UAV inspection parameters, working status data of sensing equipment, linkage response data of ground monitoring equipment, early warning information push feedback data, and actual situation data of on-site risk evolution. The on-site feedback data is classified and organized into data types, including UAV operation feedback data, perception data feedback data, ground equipment feedback data, early warning response feedback data, and risk situation feedback data. The data in each category is associated with corresponding time and space information. Extract key parameters such as the core area, direction, and rate of actual risk evolution from risk situation feedback data, compare them with previous risk evolution prediction results, identify the differences between the two, and analyze the reasons for the differences. The drone operation feedback data is compared with the flight parameters in the dynamic adaptation trajectory scheme. The adaptation relationship between the trajectory node distribution, flight speed adjustment rules and actual flight status is analyzed. The coordinate deviation between the flight trajectory and the planned trajectory is calculated. If the deviation is greater than the set tolerance threshold, the area is marked as the trajectory deviation area. The planned flight speed is compared with the actual flight speed. If the speed difference is greater than the set speed tolerance threshold, it is marked as unreasonable speed adjustment. Extract the quality parameters of the sensing data from different regions from the feedback data of the sensing data, analyze the matching relationship between the collaborative parameter settings of the sensing devices and the risk monitoring requirements, compare the signal-to-noise ratio, resolution or integrity indicators of the sensing data with the set quality thresholds, and identify the substandard regions and the corresponding sensing device parameters. Based on the differences between the risk evolution prediction results and the actual risk situation, the deviation of the drone operation feedback data, and the quality problems of the perception data, the core parameters that need to be adjusted in the dynamic collaborative adaptation process of the inspection trajectory are determined. To address the differences in risk evolution, the weights of the input parameters of the risk evolution prediction model for glacier regions are adjusted to optimize the prediction accuracy of the model for actual risk situations. At the same time, the distribution density of trajectory nodes is adjusted based on the core risk areas. To address the operational deviations of drones, the flight speed adjustment rules were revised, and the smooth transition parameters between trajectory nodes were optimized to stabilize the drone's flight status while avoiding terrain obstacles discovered during actual flight. To address the issue of perceived data quality, the collaborative parameters of sensing devices in different risk characteristic areas were adjusted, and the collection frequency and data resolution settings were optimized to improve the reliability and accuracy of the perceived data. The adjusted parameters are re-injected into the dynamic collaborative adaptation process of the inspection trajectory to generate an updated dynamic adaptation trajectory scheme, which is then applied to the next round of UAV inspections.

10. A glacier debris flow monitoring system applied to unmanned aerial vehicle (UAV) inspections, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the glacial debris flow monitoring method for unmanned aerial vehicle (UAV) inspection as described in any one of claims 1 to 9 by executing the machine-executable instructions.

Images