A method for identifying the optimal landing point of a magnetically attracted seismic source from an unmanned aerial vehicle (UAV).

Abstract

This invention relates to a method for identifying the optimal landing point of a magnetically attached seismic source for unmanned aerial vehicles (UAVs), belonging to the field of geophysical exploration technology. Addressing the problems of excitation failure and equipment damage caused by the "blind landing" of existing UAV seismic sources, this method utilizes a multimodal sensor module consisting of an RGB camera, LiDAR, and multispectral camera mounted on the UAV to collect data such as surface slope, elevation variance, obstacle height, and geological material. A weighted scoring model is constructed based on four criteria: coupling, safety, proximity, and feasibility. Three task modes—high precision, efficiency priority, and safety in complex terrain—are adopted for dynamic weight adaptation to quantitatively screen candidate points with the best comprehensive score. An integrated spindle-shaped seismic source ensures stable landing attitude, and the landing point is autonomously verified by the machine after determination. This method solves the problem of lack of environmental perception, achieves intelligent landing point selection, avoids re-explosive firing operations, and significantly improves the accuracy, reliability, and efficiency of UAV seismic source operations.

Description

Technical Field

[0001] This invention relates to the field of efficient seismic source excitation technology in geophysical exploration, specifically a method for identifying the optimal landing point of a magnetically attracted seismic source using an unmanned aerial vehicle (UAV). Background Technology

[0002] The success of high-precision seismic exploration heavily depends on the spatial accuracy of the seismic source excitation point and the quality of its coupling with the Earth's surface. While traditional manual seismic source deployment methods rely on the experience of technicians for on-site selection and simple processing of the landing point, they suffer from inherent drawbacks such as extremely low operational efficiency, high cost, and inability to be implemented in dangerous or uninhabited areas. The emergence of unmanned aerial vehicle (UAV) seismic source technology has brought revolutionary hope for achieving large-area, high-efficiency seismic source deployment. It can easily reach areas inaccessible to humans, greatly expanding the operational boundaries of geophysical exploration.

[0003] However, existing UAV seismic source operation modes suffer from serious functional deficiencies in terms of "intelligence." The core problem can be summarized as "positioning without perception; deployment without decision-making." Current technology relies excessively on the precise positioning and hovering capabilities of UAVs. The standard operating procedure typically involves the UAV using GNSS or other navigation systems to reach a pre-set theoretical coordinate point based on an ideal planar grid, and then deploying the seismic source. However, in real-world field exploration environments, especially in complex terrains such as mountains, hills, and deserts, the surface conditions directly below the pre-set theoretical coordinate point often vary greatly. The actual geological conditions at this location may be extremely unsuitable for seismic source activation; for example, it may contain large volumes of bedrock, sharp-surfaced gravel piles, soft, unsupported quicksand, or dense shrubbery. Due to a lack of high-precision terrain perception and real-time geological assessment capabilities, existing systems are essentially "blind" to these risks, leading to widespread "blind landings." Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a method for constructing a terrain adaptive assessment system based on multimodal sensor fusion and machine vision technology, thereby forming a method for precise control of seismic sources by UAVs.

[0005] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0006] A method for identifying the optimal landing point of a magnetically attracted seismic source from an unmanned aerial vehicle (UAV) comprises the following key steps:

[0007] S1. The UAV, equipped with a multimodal sensor module, flies to the airspace above the preset survey grid nodes;

[0008] S2. Collect surface environmental data of the area surrounding the preset point through the multimodal sensor module. The environmental data includes macroscopic topographic parameters, microscopic flatness parameters, obstacle parameters, and stratum material parameters.

[0009] S3. Quantitatively evaluate the collected environmental data based on a multi-criteria weighted scoring model. The multi-criteria include coupling criteria, safety criteria, proximity criteria, and feasibility criteria. Each criterion is divided into three scoring levels: 90-100 points for "excellent", 60-89 points for "medium", and 0-59 points for "good". The "medium" level is the minimum qualified threshold for the operation.

[0010] S4. Configure dynamic weights for the multiple criteria;

[0011] S5. Calculate the overall score for each candidate landing point. The formula for calculating the overall score is:

[0012] ;

[0013] Among them, S c S s S p S f These represent the scores for coupling, security, proximity, and feasibility, respectively; Wc is the high-weight coefficient for the coupling criterion; Wo is the low-weight coefficient shared by the other three criteria, and satisfies Wc+3Wo = 1.

[0014] S6. Select the candidate point with the highest comprehensive score and no less than "medium" level in each individual criterion as the optimal landing point, and instruct the UAV to execute the seismic source descent.

[0015] In one embodiment of the present invention, in step S2, the macroscopic topographic parameters of the land surface include slope, the microscopic flatness parameters include elevation variance, the obstacle parameters include obstacle height, and the stratum material parameters are obtained through multispectral data and correspond to stratum type coefficients, with the stratum type coefficients satisfying rock=1.0, hard soil=0.8, sandy soil=0.4, and vegetation=0.1.

[0016] In one embodiment of the present invention, the evaluation criteria for the coupling performance in step S3 are as follows: flat and hard terrain is "excellent", slightly sloping and soft terrain is "medium", and broken and extremely soft terrain is "good"; the coupling performance score is obtained through the formula:

[0017] ;

[0018] The calculation is performed, where a, b, c, and γ are weighting coefficients, and α, H, L, and soil are the slope, elevation variance, obstacle height, and stratum type coefficients, respectively. max H maxL max These are the preset maximum allowable slope, maximum allowable elevation variance, and maximum allowable obstacle height, respectively.

[0019] As one embodiment of the present invention, the evaluation criteria for the safety criteria in step S3 are as follows: "Excellent" is the absence of any dangerous elements, "Medium" is the presence of controllable low-risk elements, and "Good" is the presence of insurmountable obstacles or high-risk terrain.

[0020] As one embodiment of the present invention, the evaluation criteria for the proximity criterion in step S3 are as follows: a straight-line distance of less than 1 meter from the preset grid node is "excellent", a distance of ≥1 meter and <3 meters is "medium", and a distance of ≥3 meters is "good"; and the lateral distance between the ideal optimal landing point and the actual landing point is <2 meters and the longitudinal distance is no more than 1.5 meters. When the distance between the two is <0.5 meters, the distance error is negligible, and the screening process follows the principle of prioritizing the survey line.

[0021] As one embodiment of the present invention, the evaluation criteria for the feasibility criteria in step S3 are as follows: "Excellent" is defined as the UAV having a smooth flight path and stable attitude, "Medium" is defined as the path having slight obstacles but still being able to operate safely, and "Good" is defined as the path being blocked or unable to hover stably.

[0022] As one embodiment of the present invention, the dynamic weight adjustment includes three task modes, specifically:

[0023] High-precision exploration mode: The weight configuration is a=0.2, b=0.35, γ=0.4, c=0.05, with an emphasis on increasing the weight of stratigraphic type γ and the weight of elevation variance b;

[0024] Fast scan / efficiency priority mode: weight configuration is a=0.4, b=0.15, γ=0.2, c=0.25, which significantly increases the slope weight a;

[0025] Complex terrain safety mode: The weights are configured as a=0.35, b=0.1, γ=0.1, c=0.45, which significantly increases the obstacle height weight c and the slope weight a;

[0026] Weight adjustments are achieved through manual presets or intelligent system control. Intelligent system control automatically assesses the complexity of the environment and dynamically adjusts it based on terrain data continuously acquired during flight.

[0027] As one embodiment of the present invention, after the optimal landing point is determined, the system executes a machine autonomous verification process to confirm the feasibility of the decision through secondary perception and data analysis of the target area.

[0028] The beneficial effects of adopting the above technical solution are as follows:

[0029] This invention integrates an RGB camera, LiDAR, and a multispectral camera to construct a multidimensional environmental perception system, enabling comprehensive detection of macroscopic terrain, microscopic obstacles, and geological materials, thus completely solving the problem of "blindness in environmental perception."

[0030] A novel multi-criteria weighted scoring model was constructed, which transforms the complex assessment of landing suitability into a standardized and quantifiable scoring system, making the decision-making process objective and transparent, and providing core algorithmic support for achieving intelligent autonomous landing that adapts to all scenarios.

[0031] The scoring model has a built-in dynamic weight adjustment mechanism that can adaptively adjust the evaluation criteria according to different task modes such as high precision, high efficiency, and high security, which significantly enhances the system's practical flexibility and reliability in variable exploration environments. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the drone vision device in an embodiment.

[0033] Among them: 1 and 2 are RGB high-precision cameras; 3 is a multispectral imaging vision camera; 4 and 5 are high-resolution lidar. Detailed Implementation

[0034] To make the objectives, technical solutions, and advantages of the present invention clearer, the invention will be described clearly and completely below in conjunction with specific embodiments.

[0035] A method for identifying the optimal landing point of a magnetically attracted seismic source from an unmanned aerial vehicle (UAV), comprising the following steps:

[0036] S1. The UAV, equipped with a multimodal sensor module, flies to the airspace above the pre-set survey grid nodes; see [link / reference]. Figure 1 The multimodal sensor module includes at least RGB high-precision cameras 1 and 2, high-resolution lidar 4 and 5, and multispectral imaging camera 3. The UAV hovers stably at a preset altitude, and the hovering meets the requirements of horizontal drift < ±0.5 meters, vertical fluctuation < ±0.2 meters, and lasts for at least 3 seconds.

[0037] S2. Collect surface environmental data of the area surrounding the preset point through the multimodal sensor module. The environmental data includes macroscopic topographic parameters, microscopic flatness parameters, obstacle parameters, and stratum material parameters.

[0038] S3. Quantitatively evaluate the collected environmental data based on a multi-criteria weighted scoring model. The multi-criteria include coupling criteria, safety criteria, proximity criteria, and feasibility criteria. Each criterion is divided into three scoring levels: 90-100 points for "excellent", 60-89 points for "medium", and 0-59 points for "good". The "medium" level is the minimum qualified threshold for the operation.

[0039] S4. Configure dynamic weights for the multiple criteria, assign the highest weight to the coupling criterion, and assign the same and relatively lower weights to the safety criterion, proximity criterion, and feasibility criterion. The weight configuration can be adaptively adjusted according to the task mode.

[0040] S5. Calculate the overall score for each candidate landing point. The formula for calculating the overall score is:

[0041] ;

[0042] Among them, S c S s S p S f These represent the scores for coupling, security, proximity, and feasibility, respectively; Wc is the high-weight coefficient for the coupling criterion; Wo is the low-weight coefficient shared by the other three criteria, and satisfies Wc+3Wo = 1.

[0043] S6. Select the candidate point with the highest overall score and no individual criterion lower than "Medium" as the optimal landing point, and instruct the UAV to execute the seismic source deployment. If there is no qualified point in the initial area, trigger the replanning mechanism to guide the UAV to perform secondary data collection and identification until a landing point that meets the standards is found or the maximum number of iterations is reached.

[0044] In step S2, the macroscopic topographic parameters of the land surface include slope, the microscopic flatness parameters include elevation variance, the obstacle parameters include obstacle height, and the stratum material parameters are obtained through multispectral data and correspond to stratum type coefficients. The stratum type coefficients satisfy rock=1.0, hard soil=0.8, sandy soil=0.4, and vegetation=0.1.

[0045] The evaluation criteria for coupling in step S3 are as follows: flat and hard terrain is "excellent", slightly sloping and soft terrain is "medium", and broken and extremely soft terrain is "good"; the coupling score is calculated using the formula:

[0046] ;

[0047] The calculation is performed, where a, b, c, and γ are weighting coefficients, and α, H, L, and soil are the slope, elevation variance, obstacle height, and stratum type coefficients, respectively. max H max L max These are the preset maximum allowable slope, maximum allowable elevation variance, and maximum allowable obstacle height, respectively.

[0048] For the weighting coefficients a, b, c, and γ in the coupling scoring formula, this module designs a dynamic adjustment mechanism to enable the UAV seismic source system of this invention to adapt to diverse exploration task requirements and environmental conditions, achieving a leap from a single strategy to scenario-adaptive operation. This mechanism redistributes the priority of core evaluation dimensions, enabling the system to make reasonable decisions under different task objectives. Specifically, a is the slope weight, representing the control system's tolerance for macroscopic surface tilt; b is the elevation variance weight, representing the control system's sensitivity to microscopic surface smoothness; c is the stratigraphic type weight, representing the control system's avoidance intensity of local sharp obstacles; and γ is the obstacle height weight, representing the control system's avoidance intensity of local sharp obstacles. By adjusting schemes based on different task modes, the weights are specifically configured to adapt to diverse application scenario requirements.

[0049] Slope (α) is a classic geomorphological parameter describing the degree of terrain tilt. Its selection is directly derived from statics and it is the primary geometric parameter for ensuring the stability of the seismic source. Exploring changes in slope can provide reference for preventing seismic source slippage and reducing the impact of bottom pressure distribution: the greater the slope, the higher the risk of slippage. Even if the seismic source is stationary at the moment of contact, it may still move laterally or longitudinally under the subsequent excitation impact force, leading to point drift and deviation in energy direction transfer. On a slope, the pressure distribution on the bottom surface of the seismic source is uneven, with high pressure on one side and low pressure or even suspension on the other. This uneven contact will seriously degrade the coupling effect, leading to asymmetry in seismic waves and affecting data interpretation. Therefore, slope is the most direct and general indicator of macroscopic terrain stability. Other terrain parameters, such as aspect, have a much smaller direct impact on seismic source coupling than slope itself.

[0050] Elevation variance (H), derived from statistics and surface metrology, quantifies the dispersion of elevation values ​​relative to their average within a local area. It is a core quality parameter characterizing surface micro-smoothness and determining coupling tightness. The study of elevation variance directly determines the effective contact area and minimizes stress concentration at point contact points. An ideally flat surface can form a large, uniform, and tight contact with the seismic source's base. However, on an uneven surface (large H value), the seismic source's base can only contact the highest points of the surface. Numerous gaps cause impact energy to be buffered, absorbed, and scattered, rather than effectively transmitted downwards. Secondly, high elevation variance implies the presence of protruding rocks or depressions. Point contact with protrusions generates significant local stress, posing a risk of damaging the seismic source's base and causing energy to be released in undesirable forms at the contact point. Therefore, compared to single parameters such as maximum elevation difference, elevation variance more comprehensively and robustly reflects the overall micro-undulations of a region, making it one of the best parameter tools for evaluating smoothness.

[0051] Obstacle height (L) refers to the vertical height of a protruding object on the ground surface relative to the surrounding reference ground plane within the pre-selected landing point and its surrounding area. It is a prerequisite parameter characterizing the local insurmountability of the ground surface and ensuring the safety and integrity of coupling. Controlling the obstacle height prevents bridging effects and avoids equipment overturning and damage. If a considerable obstacle exists at the landing point, such as a rock exceeding the adjustable range of the seismic source base, the seismic source will directly support itself on the obstacle after landing, leaving most of the base suspended in the air and completely losing coupling with the ground surface, forming a bridging phenomenon. This will lead to complete excitation failure. Furthermore, if the seismic source lands on an inclined obstacle, it is highly likely to cause the seismic source to overturn, leading to the failure of the mission and potentially causing structural damage to the equipment. While the physical dimensions of the obstacle, such as volume and projected area, are important, height is the most direct and critical indicator of whether it poses a threat to the seismic source base. A low, flat rock may not affect coupling, while a tall, thin, and sharp rock is extremely dangerous. Therefore, obstacle height is the primary screening criterion.

[0052] The safety criteria in step S3 are as follows: "Excellent" is defined as having no dangerous elements (normal height of obstacles ≤ 0.15 meters) and no high-risk terrain; "Medium" is defined as having controllable low-risk elements; and "Good" is defined as having insurmountable obstacles (normal height of obstacles > 0.15 meters) or high-risk terrain.

[0053] The criteria for judging proximity in step S3 are as follows: a straight-line distance of less than 1 meter from the preset grid node is "excellent", a distance of ≥1 meter and <3 meters is "medium", and a distance of ≥3 meters is "good". The lateral distance between the ideal optimal landing point and the actual landing point is <2 meters and the longitudinal distance is no more than 1.5 meters. When the distance between the two is <0.5 meters, the distance error is negligible. The screening process follows the principle of prioritizing survey lines.

[0054] The evaluation criteria for the feasibility criteria in step S3 are as follows: "Excellent" means the UAV has a smooth flight path and stable attitude; "Medium" means the path has slight obstacles but it can still operate safely; and "Good" means the path is blocked or it cannot hover stably.

[0055] In terms of weight allocation, the coupling criterion is given the highest weight and highest priority because different terrain parameters all affect landing safety. The other three criteria are set to the same and relatively lower weights. The fundamental goal of these criteria is to ensure the maximum close contact between the seismic source device and the ground surface, so as to ensure that seismic wave energy can propagate efficiently and vertically underground, avoiding energy scattering and loss in non-target media. Based on this goal, the macroscopic geometry, microscopic geometry, and physical contact conditions of the ground surface are particularly important for achieving optimal coupling. The three parameters of slope (α), elevation variance (H), and obstacle height (L) are the optimal choices for quantifying coupling from these three dimensions, respectively.

[0056] The dynamic weight adjustment includes three task modes, specifically:

[0057] High-precision exploration mode: The weights are configured as a=0.2, b=0.35, γ=0.4, and c=0.05, with a focus on increasing the weight of stratigraphic type γ and the weight of elevation variance b. This high-precision exploration mode is necessary for exploration tasks in critical areas with high seismic signal quality requirements, such as in-depth analysis of fine geological structures and precise delineation of oil and gas reservoir boundaries. The core of this mode is to maximize the source excitation quality and data signal-to-noise ratio, prioritizing coupling effectiveness. In this mode, stratigraphic characteristics have a significant impact on energy coupling efficiency and wave propagation quality; therefore, it is crucial to increase the weights of stratigraphic type γ and elevation variance b. Specifically, hard bedrock, due to its higher γ value, provides a more stable energy transfer path; simultaneously, extremely low micro-undulations (i.e., a high b value) help reduce wave scattering and attenuation during propagation, thereby ensuring optimal energy coupling efficiency and wave propagation quality. Furthermore, between slope and surface hardness, the latter has a more significant impact on energy coupling and wave propagation quality. Sites with a slightly steeper slope but a hard bedrock surface have a higher priority than flat but soft soil sites. Although the influence of obstacles still needs to be considered, its impact is relatively minor as long as it does not cause the seismic source to be completely suspended. Based on the above analysis, the slope weight 'a' and obstacle height weight 'c' can be appropriately reduced in the adjustment strategy to further strengthen the decision-making weight of ground hardness and surface smoothness.

[0058] Fast Scan / Efficiency Priority Mode: Weights are configured as a=0.4, b=0.15, γ=0.2, c=0.25, significantly increasing the slope weight 'a'. When conducting large-scale area surveys and requiring rapid completion, the system will switch to fast scan mode. This mode prioritizes maximizing efficiency by optimizing decision-making processes and reducing dwell time at single points, thereby improving overall exploration progress. In this mode, terrain slope has a crucial impact on operational efficiency and flight safety: flat terrain facilitates stable hovering of the UAV, reducing attitude adjustment time; it also reduces the risk of roll-off after seismic source release, avoiding repetitive operations. Therefore, the slope weight 'a' needs to be significantly increased to ensure the system prioritizes flat areas for operations, guaranteeing smooth workflow and efficiency.

[0059] Accordingly, prioritizing efficiency, the requirements for surface micro-smoothness and medium hardness can be appropriately relaxed: regarding micro-smoothness, only the requirements for safe take-off and landing of UAVs and basic operational stability need to be met; regarding medium hardness, although it affects coupling quality, strict restrictions can be reduced to improve operational speed. Therefore, in this mode, the elevation variance weight b and the stratum type weight γ should be appropriately reduced to enhance the flexibility of point selection and accelerate the scanning process. Through the above weight adjustments, the system can achieve more efficient regional data acquisition within an acceptable accuracy range.

[0060] Complex Terrain Safety Mode: Weights are configured as a=0.35, b=0.1, γ=0.1, c=0.45, significantly increasing the obstacle height weight c and the slope weight a. When dealing with complex exploration areas characterized by fragmented terrain, dense obstacles, or dense vegetation, the system will switch to Complex Terrain Safety Mode to ensure operational safety and equipment integrity. This mode prioritizes avoiding significant obstacles and steep slopes, taking precedence over the pursuit of source coupling quality. In this mode, the obstacle height weight c and the slope weight a should be significantly increased, allowing the system to prioritize excluding areas with tall obstacles such as boulders, dense forests, deep ravines, or steep slopes during site selection, ensuring basic safety for UAV flight and source release. Simultaneously, to adapt to complex terrain, the requirements for surface material and micro-level smoothness can be minimized. As long as the source can be stably placed without slipping or damage, a relatively large range of variations in surface material and micro-level undulations is permissible. Therefore, the elevation variance weight b and the stratigraphic type weight γ need to be significantly reduced to avoid failing to select safe sites due to excessive pursuit of ground conditions.

[0061] The above-mentioned weight adjustment strategy can be implemented through two methods: manual preset and intelligent system control. 1) Manually Preset Mode: During the task planning phase, the operator selects a mode from the established modes based on the overall objectives of the exploration task. The system will automatically load the corresponding set of weight parameters {a, b, c, γ}.

[0062] 2) Intelligent System Adjustment: During drone flight, the system automatically assesses the complexity of the environment using continuously acquired terrain data (such as average slope and obstacle density) and dynamically mixes or switches the weight configurations of different modes. For example, when the system detects that it has entered an area where the obstacle density exceeds a threshold, it will automatically increase the weight of c and switch to a safe mode.

[0063] Taking a base weight parameter setting of a=0.25, b=0.25, γ=0.3, c=0.2 as an example, the weight parameters will be adjusted accordingly in different task modes, as detailed below:

[0064] 1) High-precision mode: The weights are adjusted to a=0.2, b=0.35, γ=0.4, c=0.05, with emphasis on the coefficients of b and γ.

[0065] 2) Efficiency-first mode: The weights are adjusted to a=0.4, b=0.15, γ=0.2, c=0.25, with emphasis on the coefficient a.

[0066] 3) Safety mode: The weights are adjusted to a=0.35, b=0.1, γ=0.1, c=0.45, with emphasis on the coefficients a and c.

[0067] Ultimately, the system selects the candidate point with the highest overall score and no individual criterion below "medium." The selected point will be the optimal landing point for actual execution, and the system will then instruct the UAV to lower the seismic source. If no suitable point is found in the initial area, a replanning mechanism is triggered, guiding the UAV to perform secondary data collection and identification until a suitable landing point is found within the expanded search area or the maximum number of iterations is reached. After pre-confirming the optimal landing point, the system automatically executes a rapid machine-autonomous verification process, ensuring the feasibility of the decision through secondary perception and data analysis of the target area.

[0068] Experimental results show that the accuracy and consistency of the seismic wave waveform generated after releasing the seismic source from the optimal landing point identified by the method described in this application are significantly better than those from landing points with lower scores within the same area.

Claims

1. A method for identifying the optimal landing point of a magnetically attracted seismic source from an unmanned aerial vehicle (UAV), characterized in that, It includes the following steps: S1. The UAV, equipped with a multimodal sensor module, flies to the airspace above the preset survey grid nodes; S2. Collect surface environmental data of the area surrounding the preset point through the multimodal sensor module. The environmental data includes macroscopic topographic parameters, microscopic flatness parameters, obstacle parameters, and stratum material parameters. S3. Quantitatively evaluate the collected environmental data based on a multi-criteria weighted scoring model. The multi-criteria include coupling criteria, safety criteria, proximity criteria, and feasibility criteria. Each criterion is divided into three scoring levels: 90-100 points is "excellent", 60-89 points is "medium", and 0-59 points is "good". The "medium" level is the minimum qualified threshold for the operation. S4. Configure dynamic weights for the multiple criteria; S5. Calculate the overall score for each candidate landing point. The formula for calculating the overall score is: ； Among them, S c S s S p S f These represent the scores for coupling, security, proximity, and feasibility, respectively; Wc is the high-weight coefficient for the coupling criterion; Wo is the low-weight coefficient shared by the other three criteria, and satisfies Wc+3Wo = 1. S6. Select the candidate point with the highest overall score and no less than "medium" level in each individual criterion as the optimal landing point, and instruct the UAV to execute the seismic source descent.

2. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, In step S2, the macroscopic topographic parameters of the land surface include slope, the microscopic flatness parameters include elevation variance, the obstacle parameters include obstacle height, and the stratum material parameters are obtained through multispectral data and correspond to stratum type coefficients. The stratum type coefficients satisfy rock = 1.0, hard soil = 0.8, sandy soil = 0.4, and vegetation = 0.

1.

3. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The evaluation criteria for coupling in step S3 are as follows: flat and hard terrain is "excellent", slightly sloping and soft terrain is "medium", and broken and extremely soft terrain is "good"; the coupling score is calculated using the formula: ； The calculation is performed, where a, b, c, and γ are weighting coefficients, and α, H, L, and soil are the slope, elevation variance, obstacle height, and stratum type coefficients, respectively. max H max L max These are the preset maximum allowable slope, maximum allowable elevation variance, and maximum allowable obstacle height, respectively.

4. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The safety criteria in step S3 are as follows: "Excellent" is defined as having no dangerous elements, "Medium" is defined as having controllable low-risk elements, and "Good" is defined as having insurmountable obstacles or high-risk terrain.

5. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The criteria for judging proximity in step S3 are as follows: a straight-line distance of less than 1 meter from the preset grid node is "excellent", a distance of ≥1 meter and <3 meters is "medium", and a distance of ≥3 meters is "good". The lateral distance between the ideal optimal landing point and the actual landing point is <2 meters and the longitudinal distance is no more than 1.5 meters. When the distance between the two is <0.5 meters, the distance error is negligible. The screening process follows the principle of prioritizing survey lines.

6. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The evaluation criteria for the feasibility criteria in step S3 are as follows: "Excellent" is defined as a smooth flight path and stable attitude of the UAV; "Medium" is defined as a path with slight obstruction but still safe to operate; and "Good" is defined as a path with obstruction or inability to hover stably.

7. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 3, characterized in that, The dynamic weight adjustment includes three task modes, specifically: High-precision exploration mode: The weight configuration is a=0.2, b=0.35, γ=0.4, c=0.05, with an emphasis on increasing the weight of stratigraphic type γ and the weight of elevation variance b; Fast scan / efficiency priority mode: weight configuration is a=0.4, b=0.15, γ=0.2, c=0.25, which significantly increases the slope weight a; Complex terrain safety mode: The weights are configured as a=0.35, b=0.1, γ=0.1, c=0.45, which significantly increases the obstacle height weight c and the slope weight a; Weight adjustments are achieved through manual presets or intelligent system control. Intelligent system control automatically assesses the complexity of the environment and dynamically adjusts it based on terrain data continuously acquired during flight.

8. The method for identifying the optimal landing point of a magnetically attracted seismic source for an unmanned aerial vehicle (UAV) according to claim 1, characterized in that, Once the optimal landing point is determined, the system executes a machine autonomous verification process, confirming the feasibility of the decision through secondary perception and data analysis of the target area.

Images