Low-altitude airspace dynamic decision-making system based on multimodal large model

Abstract

The present invention provides a low-altitude airspace dynamic decision-making system based on a multimodal large-scale model, relating to the technical field of low-altitude airspace traffic management. The map construction module of the present invention generates a first thermal disturbance coverage map, a second shear wind high-risk area map, a third turbulence risk flow area map, and a fourth sudden obstacle prediction area map. Aircraft can determine in real time whether they have entered a dangerous area and assess the safety of their flight path based on the superimposed risk value. When the superimposed risk value #imgabs0# of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft is in the medium-risk segment, the system generates a first-level avoidance strategy, adjusting the aircraft's altitude, speed, and heading. When the superimposed risk value #imgabs1# is in the high-risk segment, the system generates a second-level avoidance strategy, significantly increasing the flight altitude and reducing the speed to avoid the high-risk area.

Description

Technical Field

[0001] The present invention relates to the technical field of low-altitude airspace traffic management, and in particular to a low-altitude airspace dynamic decision-making system based on a multimodal large model. Background Art

[0002] As the low-altitude airspace operating environment matures, various flight platforms, including drones and low-altitude manned aircraft, are widely used in near-ground airspace between 120 and 1,000 meters for inspections, logistics, emergency response, and environmental monitoring. Compared to high-altitude routes, the low-altitude flight environment is more complex, and aircraft face significant weather interference and dynamic environmental changes during mission execution, placing higher demands on safety and intelligent path planning.

[0003] In traditional low-altitude flight missions, flight path planning mainly relies on static charts and limited meteorological data, making it difficult to accurately capture and evaluate nonlinear meteorological factors such as instantaneous changes in low-altitude thermal disturbances, shear winds, and turbulence mutations. For example:

[0004] Surface thermal disturbance effect: Due to factors such as differences in the heat capacity of ground materials and the urban heat island effect, there are strong vertical thermal flow disturbances at low altitudes, which can easily cause aircraft altitude fluctuations and attitude instability, especially in summer or in complex terrain.

[0005] Shear winds and sudden turbulence: Aircraft often encounter rapidly changing wind speeds and directions in areas near the Earth, such as mountains, buildings, and overpasses. These conditions can have a significant impact on small aircraft, potentially causing flight path deviations or even flight control failure.

[0006] Dynamic targets and sudden obstacles: There are a large number of dynamically changing targets (such as other aircraft and flocks of birds) in low-altitude airspace, and traditional rule-based path planning is difficult to identify and avoid them in a timely manner.

[0007] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0008] The purpose of the present invention is to provide a low-altitude airspace dynamic decision-making system based on a multimodal large model to solve the problems raised in the above background technology.

[0009] To achieve the above object, the present invention provides the following technical solutions:

[0010] The low-altitude airspace dynamic decision-making system based on a multi-modal large model includes:

[0011] The airspace modeling module is used to establish a dynamic three-dimensional floating coordinate system for the target operation airspace for multiple aircraft flight missions in the low-altitude airspace of 120 to 1000 meters, with the i-th aircraft as the target aircraft. It also synchronously collects meteorological data, remote sensing images, visual images, and navigation data related to the airspace where the i-th aircraft is located to establish a multi-source data set.

[0012] The trajectory analysis module is used to collect the first original flight trajectory of the i-th aircraft in real time, establish its first original flight trajectory in the dynamic three-dimensional floating coordinate system, and identify the environmental data of the j-th area in the first original flight trajectory to construct the heat flux of the j-th area. , shear wind strength , turbulence intensity and dynamic target density indicators and evaluate;

[0013] A map construction module is used to identify whether each trajectory point of the i-th aircraft in the first original flight trajectory crosses the risk area, obtain the first original flight trajectory path identification result, and construct a first thermal disturbance coverage map, a second shear wind high risk area map, a third turbulence risk basin map, and a fourth sudden obstacle prediction area map;

[0014] The flight risk superposition module constructs the superposition risk value of the jth area based on the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map. , and evaluate, , normal traffic, when Indicates that the superimposed risk is a medium risk segment and generates a first-level avoidance strategy; when It indicates that the superimposed risk is a high-risk segment and generates a secondary avoidance strategy.

[0015] Furthermore, the airspace modeling module includes a three-dimensional floating coordinate system construction unit and a multi-source data synchronization acquisition unit;

[0016] The three-dimensional floating coordinate system construction unit is used for a flight mission of multiple aircraft in a low-altitude airspace of 120 to 1000 meters, and is used to take the i-th aircraft, where i∈[1,n], n is the total number of aircraft operating simultaneously;

[0017] Using the current position of the i-th aircraft as a reference point, a dynamic three-dimensional floating coordinate system is established in real time. The dynamic three-dimensional floating coordinate system is adaptively adjusted as the heading and attitude of the i-th aircraft change to ensure that the dynamic three-dimensional floating coordinate system is aligned with the aircraft's mission path;

[0018] The multi-source data synchronous acquisition unit is used to synchronously collect meteorological data, remote sensing images, visual images and navigation data related to the airspace where the i-th aircraft is located, establish a multi-source data set, and construct and train a three-dimensional dynamic environment model. It uses timestamp alignment and spatial interpolation methods to perform multimodal fusion of various types of data to ensure that different data sources have a consistent time and space reference system in the coordinate domain of the i-th aircraft.

[0019] Furthermore, the trajectory analysis module includes a first trajectory acquisition unit, a ground heat flux identification unit, and a local mutation shear wind field identification unit;

[0020] The first trajectory acquisition unit is used to acquire the position, attitude angle, velocity vector and first original flight trajectory of the i-th aircraft in real time;

[0021] The ground heat flux identification unit is used to deploy heat flux sensors on the ground to collect heat flux data, define the spatial mapping of heat flux disturbance and ground temperature gradient based on the relationship between heat flux and ground temperature, and obtain the heat flux of the jth area. :

[0022] Where, is the ground temperature of the jth region, is the ambient temperature between 120 and 1000 meters above the ground in the jth area, and h is the heat transfer coefficient, which is determined based on the actual scenario: h = 5-25 W / m²·K for natural convection; h = 25-250 W / m²·K for forced convection under strong wind; h = 10-50 W / m²·K for indoor air flow; and h = 50-500 W / m²·K near industrial chimney heat sources.

[0023] According to the heat flux of the jth region , judge the abnormal level of heat flux, including:

[0024] When the heat flux in the jth region <100W / m 2 ; It is judged as the normal heat flux area;

[0025] When 100W / m 2 ≤ ≤300W / m 2 ; It is judged that there is thermal interference anomaly and marked as the first-order thermal disturbance area;

[0026] when >300W / m 2 ; It is judged that there is thermal interference anomaly and marked as a second-order thermal disturbance area.

[0027] Furthermore, the local sudden shear wind field identification unit is used to deploy a three-dimensional ultrasonic wind speed sensor group and a laser wind measurement radar device on the ground and in the flight channel to respectively collect wind speed vector data from the ground to an altitude of 1000 meters in the target area, and obtain the shear wind intensity of the jth area based on the wind speed difference of each altitude layer. :

[0028] Where, and The jth region is at the second height and the first height Wind speed vector, second height Set to 100m, the first height Set to 50m;

[0029] According to the shear wind intensity of the jth region , determine the shear wind disturbance level, including:

[0030] When the shear wind intensity in the jth region <1.0m / s·100m -1 ; It is judged to be an area of normal shear wind intensity;

[0031] When 1.0m / s·100m -1 ≤ ≤2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a first-order shear wind disturbance area;

[0032] when >2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a second-order shear wind disturbance area.

[0033] Furthermore, the trajectory analysis module further includes a turbulence region boundary identification unit and a dynamic target obstacle distribution identification unit;

[0034] The turbulence zone boundary identification unit is used to deploy anemometer equipment in the target area, collect wind speed instantaneous change data, and obtain the turbulence intensity of the jth area. :

[0035] Where, is the standard deviation of wind speed in the jth region, reflecting the severity of fluctuations, represents the average wind speed in the jth region;

[0036] According to the turbulence intensity of the j-th region , determine the turbulence risk level, including:

[0037] When the turbulence intensity in the jth region <0.1; judged as normal turbulence area;

[0038] When 0.1≤ ≤0.2; it is judged that there is turbulence anomaly and marked as the first-order turbulence disturbance area;

[0039] when >0.2; it is judged that there is turbulence anomaly and marked as a second-order turbulence disturbance area.

[0040] Furthermore, the dynamic target obstacle distribution identification unit is used to collect the trajectory, speed, size and distribution density data of the moving target in the target area by deploying laser radar, visual sensor, millimeter wave radar, ADS-B or V2X equipment, and construct a real-time distribution map of dynamic obstacles to obtain the dynamic target density index of the jth area. :

[0041] Where, is the density of dynamic obstacles per unit area in the jth region, in units of m 2 , Indicates the number of dynamic obstacles currently detected in the jth area, represents the horizontal projection area of the jth region;

[0042] According to the dynamic target density index of the j-th area , determine the flight obstacle avoidance risk level, including:

[0043] When the dynamic target density index of the jth region <0.1 pieces / m 2 ; It is judged as a normal obstacle avoidance risk area;

[0044] When 0.1 / m 2 ≤ ≤0.5 pieces / m 2 ; Marked as the first-order obstacle avoidance prompt area;

[0045] when >0.5 pieces / m 2 ; Marked as the second-order high-density obstacle avoidance warning area.

[0046] Furthermore, the atlas construction module includes a first original flight trajectory recognition unit and an atlas generation unit;

[0047] The first original flight trajectory identification unit is configured to identify a flight path segment in the first original flight trajectory that passes through the jth region, and obtain a heat flux of the jth region, a shear wind intensity of the jth region, a turbulence intensity of the jth region, and a dynamic target density index of the jth region corresponding to the flight path segment of the jth region;

[0048] Determine whether each trajectory point in the flight trajectory crosses the following risk areas, obtain the first original flight trajectory path recognition result, and perform graded marking, including:

[0049] First-order or second-order thermal disturbance zone, first-order or second-order shear wind disturbance zone, first-order or second-order turbulence disturbance zone, first-order obstacle avoidance prompt zone or second-order high-density obstacle avoidance warning zone;

[0050] The first original flight trajectory path identification result is used to construct the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map.

[0051] Furthermore, the flight risk superposition module includes a flight risk superposition identification unit and a hierarchical response unit;

[0052] The flight risk superposition identification unit is used to calculate the superposition risk value of the jth area according to the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk flow area map and the fourth sudden obstacle prediction area map by the following formula: :

[0053] Where, Score the heat flux of the jth region, where the normal heat flux region is 0 points, the first-order thermal disturbance region is 1 point, and the second-order thermal disturbance region is 2 points; Score the shear score of the jth region, where the normal shear wind intensity region is 0 points, the first-order shear wind disturbance region is 1 point, and the second-order shear wind disturbance region is 2 points; is the turbulence intensity score of the jth region, where the normal turbulence region is 0 points, the first-order turbulence disturbance region is 1 point, and the second-order turbulence disturbance region is 2 points; Score the dynamic target density of the jth area, with the normal obstacle avoidance risk area being 0 points, the first-order obstacle avoidance prompt area being 1 point, and the second-order high-density obstacle avoidance warning area being 2 points;

[0054] when <2, indicating the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For low-risk segments, the jth region is marked in the dynamic three-dimensional floating coordinate system and is visualized as green;

[0055] when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For the medium-risk section, the jth area is marked in the dynamic three-dimensional floating coordinate system and is set to yellow for visualization;

[0056] when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For high-risk segments, the jth area is marked in the dynamic three-dimensional floating coordinate system and is visualized in red.

[0057] Furthermore, the hierarchical response unit is configured to maintain the current flight state when the i-th aircraft passes through the j-th area according to the first original flight trajectory and identifies the flight path segment as a low-risk segment;

[0058] When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a medium-risk segment, a first-level avoidance strategy is generated, including:

[0059] Increase the vertical flight altitude of the i-th aircraft by 20m, reduce the flight speed by 20-30% while maintaining a stable wall, including reducing it from 15m / s to 8-10m / s, and adjust the heading angle to ±5-10° to avoid the second-order high-density obstacle avoidance warning zone or the second-order turbulence disturbance zone;

[0060] When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a high-risk segment, a secondary avoidance strategy is generated, including:

[0061] Increase the vertical flight altitude of the i-th aircraft by 50-80m, reduce the flight speed by 40-50% while maintaining a stable wall, including reducing it from 15m / s to 6-7m / s, and adjust the heading angle to ±15-30° to avoid the second-order high-density obstacle avoidance warning area or the second-order turbulence disturbance area; determine whether there is a continuous band of risk voxels in the area where the current original path segment is located in the high-risk section; if so, call the path reconstruction algorithm to plan a new flight path segment with the lowest risk cost; splice the new path segment into the aircraft trajectory control system, issue the path points in real time, and simultaneously mark the current path segment as a high-risk passage record area.

[0062] Compared with existing technologies, the present invention offers the following advantages: The low-altitude airspace dynamic decision-making system, through real-time acquisition of multi-source data (including meteorological data, remote sensing imagery, visual imagery, and navigation data), establishes a dynamic three-dimensional floating coordinate system, and, combined with the aircraft's real-time trajectory, accurately identifies meteorological factors such as thermal disturbances, shear winds, and sudden turbulence changes, as well as dynamic targets and unexpected obstacles in the low-altitude airspace. This precise identification capability significantly improves the aircraft's adaptability to complex low-altitude environments, effectively avoiding dangerous factors that could cause the aircraft to deviate from its trajectory or cause flight control failure, thereby enhancing flight safety.

[0063] The present invention generates the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk flow area map and the fourth sudden obstacle prediction area map through the map construction module. The aircraft can judge in real time whether it has entered the dangerous area and evaluate the safety of the flight path based on the superimposed risk value. When the superimposed risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft is When in the medium risk zone, the system generates a first-level avoidance strategy to adjust the aircraft's altitude, speed, and heading; when the risk value is superimposed When the aircraft enters a high-risk zone, the system generates a secondary avoidance strategy, significantly increasing altitude and reducing speed to avoid the high-risk area. This intelligent path planning technology dynamically adapts to changes in risk in the low-altitude environment, ensuring optimal execution of the flight mission.

[0064] The present invention generates a superimposed risk value This technology enhances the aircraft's autonomous decision-making capabilities in complex environments through intelligent response mechanisms (such as primary and secondary avoidance strategies). Especially in dynamically changing low-altitude airspace, the aircraft can automatically adjust flight parameters to avoid unstable airflow, dynamic targets, and obstacles without human intervention. This technology enhances the autonomous obstacle avoidance capabilities of low-altitude aircraft, which is particularly important in densely populated and complex low-altitude flight environments such as urban and mountainous areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 The figure is a flow chart of the low-altitude airspace dynamic decision-making system based on the multimodal large model of the present invention. DETAILED DESCRIPTION

[0066] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to specific embodiments.

[0067] It should be noted that, unless otherwise defined, the technical or scientific terms used in the present invention should have the usual meanings understood by people with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0068] Example 1: Please refer to Figure 1 The present invention provides a technical solution: a low-altitude airspace dynamic decision-making system based on a multi-modal large model, comprising:

[0069] The airspace modeling module is used to establish a dynamic three-dimensional floating coordinate system for the target operation airspace for multiple aircraft flight missions in the low-altitude airspace of 120 to 1000 meters, with the i-th aircraft as the target aircraft. It also synchronously collects meteorological data, remote sensing images, visual images, and navigation data related to the airspace where the i-th aircraft is located to establish a multi-source data set.

[0070] The trajectory analysis module is used to collect the first original flight trajectory of the i-th aircraft in real time, establish its first original flight trajectory in the dynamic three-dimensional floating coordinate system, and identify the environmental data of the j-th area in the first original flight trajectory to construct the heat flux of the j-th area. , shear wind strength , turbulence intensity and dynamic target density indicators and evaluate;

[0071] A map construction module is used to identify whether each trajectory point of the i-th aircraft in the first original flight trajectory crosses the risk area, obtain the first original flight trajectory path identification result, and construct a first thermal disturbance coverage map, a second shear wind high risk area map, a third turbulence risk basin map, and a fourth sudden obstacle prediction area map;

[0072] The flight risk superposition module constructs the superposition risk value of the jth area based on the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map. , and evaluate, , normal traffic, when Indicates that the superimposed risk is a medium risk segment and generates a first-level avoidance strategy; when It indicates that the superimposed risk is a high-risk segment and generates a secondary avoidance strategy.

[0073] In this embodiment, the trajectory analysis module extracts meteorological factors such as heat flux, shear wind intensity, and turbulence intensity in real time, enabling accurate modeling of the aircraft's airspace environment. This allows for the effective identification and assessment of thermal disturbance zones, shear wind belts, and turbulence risk areas, enhancing risk perception for low-altitude flight. The system generates multiple risk maps (first, a thermal disturbance coverage map; second, a shear wind high-risk zone map; third, a turbulence risk flow domain map; and fourth, a sudden obstacle prediction zone map) through the map construction module. The flight risk overlay module then implements comprehensive calculation and dynamic assessment of multi-source risk, thus addressing the issue of a single risk model failing to cover complex scenarios. Based on the real-time assessment results, the system categorizes risk levels and automatically generates graded avoidance strategies, including level one (medium risk) and level two (high risk). This enables proactive avoidance capabilities in flight path planning, significantly reducing the instability rate and collision risk of aircraft during low-altitude missions. The system can sense and predict the dynamic target density and potential sudden obstacles in low-altitude airspace. It is particularly suitable for low-altitude application scenarios with high dynamics, complex terrain, or changeable weather, enhancing the stability and mission completion rate during flight mission execution.

[0074] Example 2: This example is explained in Example 1. Please refer to Figure 1 ,Specifically, the airspace modeling module includes a three-dimensional floating coordinate system ,construction unit and a multi-source data synchronous ,acquisition unit;

[0075] The three-dimensional floating coordinate system construction unit is used for a flight mission of multiple aircraft in a low-altitude airspace of 120 to 1000 meters, and is used to take the i-th aircraft, where i∈[1,n], n is the total number of aircraft operating simultaneously;

[0076] Using the current position of the i-th aircraft as a reference point, a dynamic three-dimensional floating coordinate system is established in real time. The dynamic three-dimensional floating coordinate system is adaptively adjusted as the heading and attitude of the i-th aircraft change to ensure that the dynamic three-dimensional floating coordinate system is aligned with the aircraft's mission path;

[0077] The multi-source data synchronous acquisition unit is used to synchronously collect meteorological data, remote sensing images, visual images and navigation data related to the airspace where the i-th aircraft is located, establish a multi-source data set, and construct and train a three-dimensional dynamic environment model. It uses timestamp alignment and spatial interpolation methods to perform multimodal fusion of various types of data to ensure that different data sources have a consistent time and space reference system in the coordinate domain of the i-th aircraft.

[0078] In this embodiment, the three-dimensional floating coordinate system construction unit uses the current position of the aircraft as the reference origin, and establishes a three-dimensional coordinate system that changes adaptively with the aircraft's heading and attitude in real time, so that the constructed environmental model is always aligned with the aircraft's mission direction, avoiding the problem of mismatch between the static geographic coordinate system and the aircraft's motion state, thereby significantly improving the real-time and adaptability of spatial modeling. The multi-source data synchronization acquisition unit realizes the unified spatiotemporal reference system mapping of multiple types of heterogeneous data such as remote sensing images, visual images, meteorological data and navigation data in the aircraft coordinate domain through timestamp alignment and spatial interpolation strategies, effectively solving the modeling deviation problem caused by time / space asynchrony of multimodal data in traditional systems. This module is particularly suitable for dealing with nonlinear disturbances caused by factors such as urban heat island effect, complex terrain or sudden airflow during low-altitude flight, and provides more stable and reliable environmental modeling support for aircraft to perform tasks such as inspections and emergency responses.

[0079] Example 3: This example is explained in Example 1. Please refer to Figure 1 ,Specifically, the trajectory analysis module includes a first trajectory acquisition unit, a ground heat flux ,identification unit and a local mutation shear wind field identification unit;

[0080] The first trajectory acquisition unit is used to acquire the position, attitude angle, velocity vector and first original flight trajectory of the i-th aircraft in real time;

[0081] The ground heat flux identification unit is used to deploy heat flux sensors on the ground to collect heat flux data, define the spatial mapping of heat flux disturbance and ground temperature gradient based on the relationship between heat flux and ground temperature, and obtain the heat flux of the jth area. :

[0082] Where, is the ground temperature of the jth region; is the ambient temperature between 120 and 1000 meters above the ground in the jth region, extracted from the meteorological data related to the airspace where the i-th aircraft is located;

[0083] h is the heat transfer coefficient, which is determined based on the actual scenario: h=5-25W / m²·K for natural convection; h=25-250W / m²·K for forced convection under strong wind; h=10-50W / m²·K for indoor air flow; and h=50-500W / m²·K near industrial chimney heat sources.

[0084] According to the heat flux of the jth region , judge the abnormal level of heat flux, including:

[0085] When the heat flux in the jth region <100W / m 2 ; It is judged as the normal heat flux area;

[0086] When 100W / m 2 ≤ ≤300W / m 2 ; It is judged that there is thermal interference anomaly and marked as the first-order thermal disturbance area;

[0087] when >300W / m 2 ; It is judged that there is thermal interference anomaly and marked as a second-order thermal disturbance area.

[0088] In this embodiment, the ground heat flux identification unit deploys heat flux sensors in the flight area and establishes a spatial mapping relationship between heat flux and ground temperature gradient. It introduces the heat transfer coefficient h for the first time to classify heat transfer mechanisms in different natural or man-made environments, effectively improving the ability to identify thermal disturbance factors such as urban heat island effects and complex terrain radiation heat sources.

[0089] The module categorizes heat flux anomaly levels into three levels: normal, first-order thermal disturbance, and second-order thermal disturbance. Using heat flux thresholds (<100W / m², 100-300W / m², and >300W / m²) as criteria, it provides quantitative metrics that can be directly integrated into path risk assessment and avoidance strategy generation. This enables a layered response mechanism for the flight system, enhancing flight safety and strategic intelligence. This module is widely applicable to typical low-altitude thermal anomaly environments, such as near-Earth urban thermal zones, areas surrounding industrial heat sources, and thermal stagnation zones in complex terrain. It can significantly improve the stability and reliability of small flight platforms performing missions in summer, high-temperature zones, and high-obstacle environments.

[0090] Example 4: This example is explained in Example 1. Please refer to Figure 1 Specifically, the local sudden shear wind field identification unit is used to deploy a three-dimensional ultrasonic wind speed sensor group and a laser wind measurement radar device on the ground and in the flight channel to collect wind speed vector data from the ground to an altitude of 1000 meters in the target area, and obtain the shear wind intensity of the jth area based on the wind speed difference at each altitude layer. :

[0091] Where, and The jth region is at the second height and the first height Wind speed vector, second height Set to 100m, the first height Set to 50m;

[0092] According to the shear wind intensity of the jth region , determine the shear wind disturbance level, including:

[0093] When the shear wind intensity in the jth region <1.0m / s·100m -1 ; It is judged to be an area of normal shear wind intensity;

[0094] When 1.0m / s·100m -1 ≤ ≤2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a first-order shear wind disturbance area;

[0095] when >2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a second-order shear wind disturbance area.

[0096] In this embodiment, by deploying three-dimensional ultrasonic wind speed sensors and laser wind radars on the ground and along flight paths, continuous wind speed vector data collection is achieved from the ground to an altitude of 1,000 meters. This effectively covers key airspaces such as densely populated areas, valley passages, and bridge and tunnel exits where shear winds are prevalent, thereby improving the vertical resolution and dynamic response capabilities of wind field identification. The local sudden shear wind field identification unit constructs highly differential shear wind intensity to meet the needs of identifying non-uniform wind fields: This unit introduces a wind speed difference calculation model, calculating shear wind intensity based on the wind speed vector difference between 50 and 100 meters in the jth region. This accurately captures local disturbances caused by sudden wind speed changes and is particularly suitable for identifying flight hazards where sudden vertical airflow is unevenly distributed.

[0097] By setting the shear wind intensity threshold (<1.0m / s·100m -1 Normal: 1.0–2.5 m / s·100 m -1 First order, >2.5m / s·100m -1 This recognition unit is particularly suitable for flight scenarios involving complex terrain or high-wind variability, such as mountain ridge wind outlets, wind corridors between urban high-rise buildings, and coastal wind pressure zones. It effectively improves the path-keeping and flight control stability capabilities of drones and small flight platforms during sudden wind disturbances, reducing safety risks such as attitude drift and trajectory yaw.

[0098] Example 5: This example is explained in Example 1. Please refer to Figure 1 ,Specifically, the trajectory analysis module also includes a turbulence zone boundary ,recognition unit and a dynamic target obstacle distribution recognition unit;

[0099] The turbulence zone boundary identification unit is used to deploy anemometer equipment in the target area, collect wind speed instantaneous change data, and obtain the turbulence intensity of the jth area. :

[0100] Where, is the standard deviation of wind speed in the jth region, reflecting the severity of fluctuations, represents the average wind speed in the jth region;

[0101] According to the turbulence intensity of the j-th region , determine the turbulence risk level, including:

[0102] When the turbulence intensity in the jth region <0.1; judged as normal turbulence area;

[0103] When 0.1≤ ≤0.2; it is judged that there is turbulence anomaly and marked as the first-order turbulence disturbance area;

[0104] when >0.2; it is judged that there is turbulence anomaly and marked as a second-order turbulence disturbance area.

[0105] In this embodiment, a dimensionless turbulence intensity index is constructed by collecting the standard deviation and mean wind speed of the jth region. This index accurately characterizes the intensity and severity of local airflow disturbances, providing key quantitative support for identifying low-altitude unsteady wind fields. Based on the numerical range of turbulence intensity, risk classification rules are set (<0.1 is a normal area, 0.1–0.2 is a first-order disturbance area, and >0.2 is a second-order disturbance area). This allows for rapid classification and spatial boundary delineation of different levels of turbulence risk areas, providing a decision-making basis for subsequent track corrections and attitude adjustments. Based on real-time wind speed data input and a sliding time window processing mechanism, this unit can dynamically update the distribution range of turbulent areas. It is suitable for short-term turbulence bursts caused by sudden weather changes, terrain changes, and the heat island effect, improving the system's overall response speed and predictive ability to unstable airflow areas. By identifying and accurately locating the boundaries of turbulent areas, the system can implement real-time avoidance of high-risk path segments, flight altitude adjustment, or delayed entry, effectively reducing the risks of vibration, yaw, attitude loss of control, etc. caused by airflow disturbances, and improving the overall track smoothness and flight control stability.

[0106] Example 6: This example is explained in Example 1. Please refer to Figure 1Specifically, the dynamic target obstacle distribution identification unit is used to collect the trajectory, speed, size and distribution density data of the moving target in the target area by deploying laser radar, visual sensor, millimeter wave radar, ADS-B or V2X equipment, and build a real-time distribution map of dynamic obstacles to obtain the dynamic target density index of the jth area. :

[0107] Where, is the density of dynamic obstacles per unit area in the jth region, in units of m 2 , Indicates the number of dynamic obstacles currently detected in the jth area, represents the horizontal projection area of the jth region;

[0108] Using target detection algorithms, dynamic targets can be identified and extracted from remote sensing and visual images of the airspace surrounding the i-th aircraft. Image differencing techniques or optical flow methods, both of which are relatively mature, can be used to extract the motion trajectory of dynamic targets and determine their speed, direction, size, and other attributes. For each detected dynamic target, its position in the remote sensing image (usually through coordinate calibration) is combined with the horizontal projection area of the target region to calculate the target density within a specific area. In image processing, the image is typically divided into several grids (for example, with a resolution of per square meter), and the dynamic target density within each grid (i.e., the number of dynamic targets per unit area) is calculated. Dynamic target information obtained from remote sensing images can be fused with data from other sensors (such as lidar and millimeter-wave radar) to improve target recognition accuracy. The dynamic target density metric (e.g., "targets / m²" in the figure) is calculated based on this fused data and can be used to construct a real-time dynamic obstacle distribution map, reflecting the distribution of dynamic targets within the target area.

[0109] According to the dynamic target density index of the j-th area , determine the flight obstacle avoidance risk level, including:

[0110] When the dynamic target density index of the jth region <0.1 pieces / m 2 ; It is judged as a normal obstacle avoidance risk area;

[0111] When 0.1 / m 2 ≤ ≤0.5 pieces / m 2 ; Marked as the first-order obstacle avoidance prompt area;

[0112] when >0.5 pieces / m 2 ; Marked as the second-order high-density obstacle avoidance warning area.

[0113] In this embodiment, the unit leverages the high-precision distance measurement capabilities of lidar, the morphological recognition characteristics of visual sensors, the all-weather penetration advantages of millimeter-wave radar, and the broadcast sensing capabilities of ADS-B and V2X to efficiently detect and supplement information about multiple types of dynamic obstacles (such as drones, bird flocks, and airborne objects) in complex airspace, effectively alleviating the limitations of single sensors, such as susceptibility to obstruction and high false alarm rates. By counting the number of dynamic obstacles per unit area, a dynamic target density index (unit: number / m²) is established to quantitatively assess the spatial density of obstacles in the flight path. Combined with a defined risk grading standard (<0.1 is considered normal, 0.1–0.5 is in the warning zone, and >0.5 is in the alert zone), this provides the aircraft with a clear and controllable basis for obstacle avoidance strategy decisions. Through dynamic target distribution maps and real-time risk level assessment, the system can implement automated route adjustments, obstacle avoidance path reconstruction, or waiting mechanisms during flight path planning or navigation. It is particularly suitable for low-altitude multi-aircraft collaborative operations or urban air traffic (UAM) scenarios, effectively preventing airspace congestion, collision risks, or attitude disturbances.

[0114] Example 7: This example is explained in Example 1. Please refer to Figure 1 ,Specifically, the atlas construction module includes a first original flight trajectory ,recognition unit and an atlas generation unit;

[0115] The first original flight trajectory identification unit is configured to identify a flight path segment in the first original flight trajectory that passes through the jth region, and obtain a heat flux of the jth region, a shear wind intensity of the jth region, a turbulence intensity of the jth region, and a dynamic target density index of the jth region corresponding to the flight path segment of the jth region;

[0116] Determine whether each trajectory point in the flight trajectory crosses the following risk areas, obtain the first original flight trajectory path recognition result, and perform graded marking, including:

[0117] First-order or second-order thermal disturbance zone, first-order or second-order shear wind disturbance zone, first-order or second-order turbulence disturbance zone, first-order obstacle avoidance prompt zone or second-order high-density obstacle avoidance warning zone;

[0118] The first original flight trajectory path identification result is used to construct the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map.

[0119] In this embodiment, this module leverages the first original flight trajectory identification unit to dynamically extract heat flux, shear wind intensity, turbulence intensity, and dynamic target density indicators from regions traversed within the flight path. This module then constructs spatial correlations between trajectory and disturbance points, enabling trajectory-level identification and classification of multiple risk factors, significantly improving the ability to analyze risk areas in a refined manner. It constructs maps of thermal disturbance coverage, high-shear wind risk areas, turbulence risk flow areas, and sudden obstacle prediction areas, respectively. This provides a visual representation of disturbance factors within the airspace operating environment, providing intuitive path avoidance guidance for the flight path planning system and enabling proactive detours and dynamic closure strategies for potentially high-risk airspace. By marking each flight path segment's passage through various risk areas, it categorizes each trajectory segment's disturbance level (e.g., first-order or second-order thermal disturbance zones). This provides a data foundation for subsequent path safety assessment, flight performance evaluation, and mission allocation and scheduling. This system is particularly suitable for intelligent trajectory planning and risk avoidance guidance in complex low-altitude scenarios.

[0120] Example 8: This example is explained in Example 1. Please refer to Figure 1 ,Specifically, the flight risk superposition module includes a flight risk superposition ,identification unit and a hierarchical response unit;

[0121] The flight risk superposition identification unit is used to calculate the superposition risk value of the jth area according to the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk flow area map and the fourth sudden obstacle prediction area map by the following formula: :

[0122] Where, Score the heat flux of the jth region, where the normal heat flux region is 0 points, the first-order thermal disturbance region is 1 point, and the second-order thermal disturbance region is 2 points; Score the shear score of the jth region, where the normal shear wind intensity region is 0 points, the first-order shear wind disturbance region is 1 point, and the second-order shear wind disturbance region is 2 points; is the turbulence intensity score of the jth region, where the normal turbulence region is 0 points, the first-order turbulence disturbance region is 1 point, and the second-order turbulence disturbance region is 2 points; Score the dynamic target density of the jth area, with the normal obstacle avoidance risk area being 0 points, the first-order obstacle avoidance prompt area being 1 point, and the second-order high-density obstacle avoidance warning area being 2 points;

[0123] when <2, indicating the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For low-risk segments, the jth region is marked in the dynamic three-dimensional floating coordinate system and is visualized as green;

[0124] when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For the medium-risk section, the jth area is marked in the dynamic three-dimensional floating coordinate system and is set to yellow for visualization;

[0125] when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For high-risk segments, the jth area is marked in the dynamic three-dimensional floating coordinate system and is visualized in red.

[0126] In this embodiment, four risk scores are defined for heat flux, shear wind, turbulence, and dynamic obstacles. Disturbance information from different physical dimensions and sensor sources is normalized and uniformly weighted to construct a flight overlay risk value for the jth region. This achieves a fusion of complex disturbance information, effectively enhancing the comprehensiveness and scientific nature of airspace risk identification. By mapping the overlay risk values into three dynamic color codes—green (low risk), yellow (medium risk), and red (high risk)—the risk level of each region is annotated in real time within a dynamic three-dimensional floating coordinate system, creating a visual flight risk map. This provides an intuitive risk-based decision-making framework for aircraft path avoidance, route reconstruction, and emergency dispatch. Identifying and labeling risk areas at the granularity of flight path segments enables differentiated risk assessments for different segments within the same path, avoiding a one-size-fits-all approach for the entire path. This improves the accuracy and flexibility of flight path planning and dynamic corrections. By setting up hierarchical response units, the system can automatically trigger different levels of risk avoidance response strategies (such as flight altitude adjustment, speed limit, route deviation recommendation or abort instruction) according to the risk level of the flight path segment, building an active, multi-level and scalable flight safety assurance mechanism, and improving the system's real-time autonomous response capabilities in complex low-altitude scenarios.

[0127] Example 9: This example is explained in Example 8. Please refer to Figure 1 Specifically, the hierarchical response unit is configured to maintain the current flight state of the i-th aircraft when it passes through the j-th area in the first original flight trajectory, if it identifies the flight path as a low-risk segment. This unit can automatically trigger appropriate avoidance measures based on the risk level of the flight path segment. In low-risk segments, the aircraft maintains its current flight state to avoid ineffective maneuvers. In medium- and high-risk segments, the aircraft adjusts flight parameters such as altitude, speed, and heading angle to ensure it safely avoids potential danger zones. This intelligent response effectively reduces safety hazards during flight.

[0128] When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a medium-risk segment, a first-level avoidance strategy is generated, including:

[0129] Increase the vertical flight altitude of the i-th aircraft by 20m, reduce the flight speed by 20-30% while maintaining a stable wall, including reducing it from 15m / s to 8-10m / s, and adjust the heading angle to ±5-10° to avoid the second-order high-density obstacle avoidance warning zone or the second-order turbulence disturbance zone;

[0130] When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a high-risk segment, a secondary avoidance strategy is generated, including:

[0131] Increase the vertical flight altitude of the i-th aircraft by 50-80m, reduce the flight speed by 40-50% while maintaining a stable wall, including reducing it from 15m / s to 6-7m / s, and adjust the heading angle to ±15-30° to avoid the second-order high-density obstacle avoidance warning area or the second-order turbulence disturbance area; determine whether there is a continuous band of risk voxels in the area where the current original path segment is located in the high-risk section; if so, call the path reconstruction algorithm to plan a new flight path segment with the lowest risk cost; splice the new path segment into the aircraft trajectory control system, issue the path points in real time, and simultaneously mark the current path segment as a high-risk passage record area.

[0132] In this embodiment, a hierarchical response unit generates precise flight strategies for different risk levels. For example, for medium-risk sections, the generated first-level avoidance strategy involves a moderate increase in altitude and a reduction in speed to minimize the aircraft's impact radius. For high-risk sections, the second-level avoidance strategy is more aggressive, requiring a significant increase in altitude and a significant reduction in speed to further ensure the aircraft stays away from areas with high obstacle density or turbulent disturbances. This meticulous, graded response provides comprehensive safety protection for the aircraft. When the aircraft enters a high-risk section and detects a continuous band of risky voxels, the system can invoke a path reconstruction algorithm in real time to plan a new path with the lowest risk cost. This dynamic path reconstruction capability enables the aircraft to flexibly adjust its route to avoid risky areas while ensuring efficient mission completion. The system splices the new avoidance path segments into the aircraft's trajectory control system in real time and simultaneously distributes path points, enabling real-time monitoring and adjustment of the aircraft's path. The system also marks high-risk areas passed through and automatically references them during subsequent flights, forming a complete safety monitoring record. This accumulated record will facilitate long-term data accumulation and optimization of flight missions, enhancing the intelligence level of flight missions. Through a hierarchical response mechanism, the system can intelligently adjust aircraft action strategies based on the actual conditions of different airspace risks, supporting large-scale, multi-aircraft collaborative operations. This will enhance the intelligence level of airspace management and adapt to the complex demands of future flight missions.

[0133] It should be noted that all calculation formulas in this application document utilize, including but not limited to, regression analysis within machine learning algorithms to deeply analyze the collected parameters and identify their natural trends and interrelationships. Professional software, such as Python's Scikit-learn library or the R language, is used to automatically generate mathematical models that match the data. Model performance is then objectively evaluated through methods such as cross-validation, combined with continuous feedback and optimization to ensure that the created formulas truly reflect the inherent laws of the data, thereby guaranteeing their validity and accuracy, and ensuring that the calculation process complies with the constraints of natural laws rather than being based on artificially set rules.

[0134] The technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0135] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0136] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

[0137] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A low-altitude airspace dynamic decision-making system based on a multimodal large model, characterized by: include: The airspace modeling module is used to establish a dynamic three-dimensional floating coordinate system for the target operation airspace for multiple aircraft flight missions in the low-altitude airspace of 120 to 1000 meters, with the i-th aircraft as the target aircraft. It also synchronously collects meteorological data, remote sensing images, visual images, and navigation data related to the airspace where the i-th aircraft is located to establish a multi-source data set. The trajectory analysis module is used to collect the first original flight trajectory of the i-th aircraft in real time, establish its first original flight trajectory in the dynamic three-dimensional floating coordinate system, and identify the environmental data of the j-th area in the first original flight trajectory to construct the heat flux of the j-th area. , shear wind strength , turbulence intensity and dynamic target density indicators and evaluate; A map construction module is used to identify whether each trajectory point of the i-th aircraft in the first original flight trajectory crosses the risk area, obtain the first original flight trajectory path identification result, and construct a first thermal disturbance coverage map, a second shear wind high risk area map, a third turbulence risk basin map, and a fourth sudden obstacle prediction area map; The flight risk superposition module constructs the superposition risk value of the jth area based on the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map. , and evaluate, , normal traffic, when Indicates that the superimposed risk is a medium risk segment and generates a first-level avoidance strategy; when It indicates that the superimposed risk is a high-risk segment and generates a secondary avoidance strategy.

2. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 1 is characterized in that: The airspace modeling module includes a three-dimensional floating coordinate system construction unit and a multi-source data synchronization acquisition unit; The three-dimensional floating coordinate system construction unit is used for a flight mission of multiple aircraft in a low-altitude airspace of 120 to 1000 meters, and is used to take the i-th aircraft, where i∈[1,n], n is the total number of aircraft operating simultaneously; With the current position of the i-th aircraft as the reference point, a dynamic three-dimensional floating coordinate system is established in real time. The dynamic three-dimensional floating coordinate system is adaptively adjusted as the heading and attitude of the i-th aircraft change to ensure that the dynamic three-dimensional floating coordinate system is aligned with the aircraft's mission path; The multi-source data synchronous acquisition unit is used to synchronously collect meteorological data, remote sensing images, visual images and navigation data related to the airspace where the i-th aircraft is located, establish a multi-source data set, and construct and train a three-dimensional dynamic environment model. It uses timestamp alignment and spatial interpolation methods to perform multimodal fusion of various types of data to ensure that different data sources have a consistent time and space reference system in the coordinate domain of the i-th aircraft.

3. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 2 is characterized by: The trajectory analysis module includes a first trajectory acquisition unit, a ground heat flux identification unit and a local mutation shear wind field identification unit; The first trajectory acquisition unit is used to acquire the position, attitude angle, velocity vector and first original flight trajectory of the i-th aircraft in real time; The ground heat flux identification unit is used to deploy heat flux sensors on the ground to collect heat flux data, define the spatial mapping of heat flux disturbance and ground temperature gradient based on the relationship between heat flux and ground temperature, and obtain the heat flux of the jth area. : Where, is the ground temperature of the jth region, is the ambient temperature between 120 and 1000 meters above the ground in the jth area, and h is the heat transfer coefficient, which is determined based on the actual scenario: h = 5-25 W / m²·K for natural convection; h = 25-250 W / m²·K for forced convection under strong wind; h = 10-50 W / m²·K for indoor air flow; and h = 50-500 W / m²·K near industrial chimney heat sources. According to the heat flux of the jth region , judge the abnormal level of heat flux, including: When the heat flux in the jth region <100W / m 2 ; It is judged as the normal heat flux area; When 100W / m 2 ≤ ≤300W / m 2 ; It is judged that there is thermal interference anomaly and marked as the first-order thermal disturbance area; when >300W / m 2 ; It is judged that there is thermal interference anomaly and marked as a second-order thermal disturbance area.

4. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 3 is characterized by: The local sudden shear wind field identification unit is used to deploy a three-dimensional ultrasonic wind speed sensor group and a laser wind measurement radar device on the ground and in the flight channel to collect wind speed vector data from the ground to an altitude of 1000 meters in the target area, and obtain the shear wind intensity of the jth area based on the wind speed difference at each altitude layer. : Where, and The jth region is at the second height and the first height Wind speed vector, second height Set to 100m, the first height Set to 50m; According to the shear wind intensity of the jth region , determine the shear wind disturbance level, including: When the shear wind intensity in the jth region <1.0m / s·100m -1 ; It is judged to be an area of normal shear wind intensity; When 1.0m / s·100m -1 ≤ ≤2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a first-order shear wind disturbance area; when >2.5m / s·100m -1 ; It is judged that there is shear wind anomaly and marked as a second-order shear wind disturbance area.

5. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 3 is characterized by: The trajectory analysis module also includes a turbulence zone boundary identification unit and a dynamic target obstacle distribution identification unit; The turbulence zone boundary identification unit is used to deploy anemometer equipment in the target area, collect wind speed instantaneous change data, and obtain the turbulence intensity of the jth area. : Where, is the standard deviation of wind speed in the jth region, reflecting the severity of fluctuations, represents the average wind speed in the jth region; According to the turbulence intensity of the j-th region , determine the turbulence risk level, including: When the turbulence intensity in the jth region <0.1; judged as normal turbulence area; When 0.1≤ ≤0.2; it is judged that there is turbulence anomaly and marked as the first-order turbulence disturbance area; when >0.2; it is judged that there is turbulence anomaly and marked as a second-order turbulence disturbance area.

6. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 5 is characterized by: The dynamic target obstacle distribution identification unit is used to collect the trajectory, speed, size and distribution density data of the moving target in the target area by deploying laser radar, visual sensor, millimeter wave radar, ADS-B or V2X equipment, and build a real-time distribution map of dynamic obstacles to obtain the dynamic target density index of the jth area. : Where, is the density of dynamic obstacles per unit area in the jth region, in units of m 2 , Indicates the number of dynamic obstacles currently detected in the jth area, represents the horizontal projection area of the jth region; According to the dynamic target density index of the j-th area , determine the flight obstacle avoidance risk level, including: When the dynamic target density index of the jth region <0.1 pieces / m 2 ; It is judged as a normal obstacle avoidance risk area; When 0.1 / m 2 ≤ ≤0.5 pieces / m 2 ; Marked as the first-order obstacle avoidance prompt area; when >0.5 pieces / m 2 ; Marked as the second-order high-density obstacle avoidance warning area.

7. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 4 is characterized by: The atlas construction module includes a first original flight trajectory recognition unit and an atlas generation unit; The first original flight trajectory identification unit is configured to identify a flight path segment in the first original flight trajectory that passes through the jth region, and obtain a heat flux of the jth region, a shear wind intensity of the jth region, a turbulence intensity of the jth region, and a dynamic target density index of the jth region corresponding to the flight path segment of the jth region; Determine whether each trajectory point in the flight trajectory crosses the following risk areas, obtain the first original flight trajectory path recognition result, and perform graded marking, including: First-order or second-order thermal disturbance zone, first-order or second-order shear wind disturbance zone, first-order or second-order turbulence disturbance zone, first-order obstacle avoidance prompt zone or second-order high-density obstacle avoidance warning zone; The first original flight trajectory path identification result is used to construct the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk basin map and the fourth sudden obstacle prediction area map.

8. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 7 is characterized by: The flight risk superposition module includes a flight risk superposition identification unit and a hierarchical response unit; The flight risk superposition identification unit is used to calculate the superposition risk value of the jth area according to the first thermal disturbance coverage map, the second shear wind high risk area map, the third turbulence risk flow area map and the fourth sudden obstacle prediction area map by the following formula: : Where, Score the heat flux of the jth region, where the normal heat flux region is 0 points, the first-order thermal disturbance region is 1 point, and the second-order thermal disturbance region is 2 points; Score the shear score of the jth region, where the normal shear wind intensity region is 0 points, the first-order shear wind disturbance region is 1 point, and the second-order shear wind disturbance region is 2 points; is the turbulence intensity score of the jth region, where the normal turbulence region is 0 points, the first-order turbulence disturbance region is 1 point, and the second-order turbulence disturbance region is 2 points; Score the dynamic target density of the jth area, with the normal obstacle avoidance risk area being 0 points, the first-order obstacle avoidance prompt area being 1 point, and the second-order high-density obstacle avoidance warning area being 2 points; when <2, indicating the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For low-risk segments, the jth region is marked in the dynamic three-dimensional floating coordinate system and is visualized as green; when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For the medium-risk section, the jth area is marked in the dynamic three-dimensional floating coordinate system and is set to yellow for visualization; when , represents the superposition risk value of the flight path segment passing through the jth area in the first original flight trajectory of the i-th aircraft For high-risk segments, the jth area is marked in the dynamic three-dimensional floating coordinate system and is visualized in red.

9. The low-altitude airspace dynamic decision-making system based on a multimodal large model according to claim 8 is characterized by: The hierarchical response unit is configured to maintain the current flight state when the i-th aircraft passes through the j-th area according to the first original flight trajectory and identifies the flight path segment as a low-risk segment; When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a medium-risk segment, a first-level avoidance strategy is generated, including: Increase the vertical flight altitude of the i-th aircraft by 20m, reduce the flight speed by 20-30% while maintaining a stable wall, including reducing it from 15m / s to 8-10m / s, and adjust the heading angle to ±5-10° to avoid the second-order high-density obstacle avoidance warning zone or the second-order turbulence disturbance zone; When the i-th aircraft passes through the j-th area in the first original flight trajectory, if the flight path segment is identified as a high-risk segment, a secondary avoidance strategy is generated, including: Increase the vertical flight altitude of the i-th aircraft by 50-80m, reduce the flight speed by 40-50% while maintaining a stable wall, including reducing it from 15m / s to 6-7m / s, and adjust the heading angle to ±15-30° to avoid the second-order high-density obstacle avoidance warning area or the second-order turbulence disturbance area; determine whether there is a continuous band of risk voxels in the area where the current original path segment is located in the high-risk section; if so, call the path reconstruction algorithm to plan a new flight path segment with the lowest risk cost; splice the new path segment into the aircraft trajectory control system, issue the path points in real time, and simultaneously mark the current path segment as a high-risk passage record area.

Images