A method and system for monitoring the flight dynamics of unmanned aerial vehicles (UAVs) based on 3D maps.

Abstract

This invention discloses a method and system for dynamic monitoring of unmanned aerial vehicle (UAV) flights based on a 3D map, relating to the field of UAV monitoring technology. The method includes: acquiring a 3D map of the target area and low-altitude airspace control requirements, constructing a low-altitude 3D partition model describing permitted UAV activities; collecting historical flight data, calculating the comprehensive heat value of partition units through a multi-factor weighted evaluation model, and generating a 3D heat map; defining deployable areas, using a multi-objective optimization algorithm to solve for the optimal deployment scheme with the optimization objectives of maximizing the weighted total coverage and minimizing the number of monitoring points; and deploying monitoring equipment according to the scheme to achieve 3D dynamic monitoring of the target area. This invention effectively solves the problems of numerous blind spots, unreasonable resource allocation, and low monitoring efficiency inherent in traditional methods, achieving refined and intelligent management of airspace resources.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) monitoring technology, specifically to a method and system for monitoring the flight dynamics of UAVs based on three-dimensional maps. Background Technology

[0002] The rapid development of drone technology has led to the increasingly widespread application of drones in both civilian and commercial fields, and the resulting airspace safety management issues have become increasingly prominent. Traditional drone surveillance methods mainly rely on two-dimensional planar monitoring or single sensor detection, which makes it difficult to effectively monitor low-altitude, slow-speed, and small drones in three-dimensional space.

[0003] Existing monitoring systems based on fixed-point deployment have significant drawbacks: First, they suffer from large blind spots, making them ill-suited for dynamic monitoring needs in complex three-dimensional airspace environments. Second, resource allocation lacks scientific rigor, failing to optimize for actual airspace usage characteristics, resulting in low equipment utilization. Furthermore, existing solutions underutilize historical flight data, lack precise modeling of three-dimensional spatial thermal distribution, and fail to establish a dynamic evaluation system linked to airspace structure. This leads to a lack of data support and scientific basis for monitoring point deployment, resulting in inefficient monitoring systems, high operation and maintenance costs, and an inability to meet the demands of refined urban airspace management. Summary of the Invention

[0004] This invention addresses the technical problems in existing technologies, such as incomplete three-dimensional spatial monitoring coverage, lack of data support for monitoring point deployment, and low efficiency in monitoring resource allocation, by providing a method and system for dynamic monitoring of UAV flight based on a three-dimensional map.

[0005] The technical solution of the present invention to solve the above-mentioned technical problems is as follows:

[0006] In a first aspect, the present invention provides a method for monitoring the flight dynamics of a UAV based on a three-dimensional map, comprising:

[0007] Acquire 3D map information of the target area and construct a low-altitude 3D partition model in combination with airspace control rules. The low-altitude 3D partition model is used to describe the permitted activity space of the UAV.

[0008] Historical flight data of UAVs within the target area are collected, and a multi-factor weighted evaluation model is used to calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model. A three-dimensional heat map is then generated based on the comprehensive heat value.

[0009] Define the deployable area in the low-altitude three-dimensional zoning model, with the optimization objectives of maximizing the weighted total coverage of the three-dimensional heat map and minimizing the total number of monitoring points, and use the monitoring point deployment scheme as the optimization variable to solve the multi-objective optimization problem;

[0010] Based on the results of the multi-objective optimization, the layout scheme of the monitoring points is obtained, and the three-dimensional monitoring of the target area is carried out accordingly.

[0011] Secondly, the present invention provides a UAV flight dynamic monitoring system based on a three-dimensional map, comprising:

[0012] The airspace grid management module is used to acquire three-dimensional map information of the target area and construct a low-altitude three-dimensional partition model in combination with airspace control rules. The low-altitude three-dimensional partition model is used to describe the space in which UAVs are allowed to operate.

[0013] The flight heat analysis module is used to collect historical flight data of UAVs in the target area, calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model using a multi-factor weighted evaluation model, and generate a three-dimensional heat map based on the comprehensive heat value.

[0014] The optimized deployment decision module is used to define the deployable area in the low-altitude three-dimensional zoning model. The optimization objectives are to maximize the weighted total coverage of the three-dimensional heat map and minimize the total number of monitoring points. The multi-objective optimization solution is performed with the monitoring point deployment scheme as the optimization variable.

[0015] The three-dimensional dynamic monitoring module is used to obtain the layout scheme of the monitoring points based on the multi-objective optimization solution results, and to perform three-dimensional monitoring of the target area accordingly.

[0016] The beneficial effects of this invention are:

[0017] Compared to existing technologies, this invention firstly constructs a precise low-altitude three-dimensional partition model, realizing a digital description of the permitted activity space for UAVs and providing an accurate spatial benchmark for subsequent monitoring; secondly, it uses a multi-factor weighted evaluation model to generate a three-dimensional heat map, which can scientifically reflect the spatiotemporal characteristics of airspace use and provide data support for monitoring resource allocation; thirdly, it achieves intelligent deployment of monitoring points through a multi-objective optimization algorithm, minimizing equipment investment costs while ensuring coverage; and finally, it implements three-dimensional dynamic monitoring based on the optimization results, significantly improving the comprehensiveness and effectiveness of UAV monitoring in complex airspace environments and solving problems such as numerous monitoring blind spots, unreasonable resource allocation, and low monitoring efficiency in existing technologies. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for monitoring the flight dynamics of a UAV based on a 3D map, provided by the present invention;

[0019] Figure 2 This is a schematic diagram of the structure of a UAV flight dynamic monitoring system based on a three-dimensional map, provided by the present invention.

[0020] In the attached diagram, the components represented by each number are as follows:

[0021] Airspace grid management module 11, flight heat analysis module 12, optimized deployment decision module 13, and three-dimensional dynamic monitoring module 14. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0024] In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0025] Example 1, as Figure 1 As shown, this embodiment of the invention provides a method for monitoring the flight dynamics of a UAV based on a three-dimensional map, including:

[0026] S10: Obtain three-dimensional map information of the target area and construct a low-altitude three-dimensional partition model in combination with airspace control rules, wherein the low-altitude three-dimensional partition model is used to describe the permitted activity space of the UAV;

[0027] Specifically, three-dimensional map information of the target area is acquired, and a low-altitude three-dimensional partition model is constructed in conjunction with airspace control rules. This low-altitude three-dimensional partition model describes the permitted activity space for UAVs, including:

[0028] The boundary of the first activity space is determined based on the airspace control rules and the three-dimensional map information;

[0029] Based on the historical air traffic records and UAV registration information of the target area, the intrinsic parameter information of the UAV is extracted accordingly.

[0030] Traverse the intrinsic parameter information of the UAV, extract the minimum airspace distance constraint, and spatially shrink the first activity space boundary according to the minimum airspace distance constraint to obtain the second activity space boundary;

[0031] The low-altitude three-dimensional partition model is constructed by combining the three-dimensional map information with the boundary of the second activity space.

[0032] First, based on the 3D geographic information data and airspace control rules of the target area, the initial activity space boundary is determined. The target area refers to the specific airspace range requiring dynamic UAV surveillance, such as densely populated urban areas, airport airspace clearance zones, and airspace for major events—areas with clear regulatory needs. The 3D geographic information data includes geographical features such as terrain elevation and building distribution. The airspace control rules include airspace delineation rules such as suitable airspace and controlled airspace. Using Geographic Information System (GIS) spatial analysis technology, the 3D geographic information data is overlaid with the airspace control rules to generate a preliminary airspace activity boundary, i.e., the first activity space boundary. This first activity space boundary defines the maximum spatial range that UAVs may operate in under ideal conditions.

[0033] Secondly, based on the historical air traffic records and UAV registration information of the target area, the intrinsic parameters of the UAVs are extracted. The historical air traffic records refer to a structured data set of historical UAV flight activities recorded within this airspace over a preset historical time period, including flight paths, altitudes, timestamps, and mission types. The preset historical time period is set according to the specific analysis objective, such as the past year. The UAV registration information refers to the database of UAV identities and performance records filed with the management department, including the registered UAV's model number, maximum flight speed, endurance, wingspan, maximum ceiling, and performance parameters provided by the manufacturer.

[0034] Furthermore, the intrinsic parameter information of the UAV is traversed to extract the minimum airspace distance constraint. The minimum airspace distance constraint refers to the minimum three-dimensional spatial distance that the UAV must maintain from obstacles (including terrain, buildings, and other aircraft) during flight to ensure flight safety. Specifically, based on the performance characteristics and safety requirements of different UAV models, the minimum safe flight distance that they need to maintain is determined. Then, using the most conservative principle, the minimum airspace distance with the most stringent requirements among all UAV models is selected as the global constraint condition. Based on this constraint, the first activity space boundary is spatially narrowed to obtain the second activity space boundary that simultaneously satisfies airspace rules and safety distance requirements.

[0035] Finally, combining the 3D map information with the boundary of the second activity space, a low-altitude 3D partitioning model is constructed. Specifically, the space within the boundary of the second activity space is divided into uniform 3D partitioning units, each representing a basic spatial unit. The grid resolution is determined based on the monitoring accuracy requirements, typically set to 10-50 meters. Each partitioning unit contains attributes such as spatial location information, airspace type identifier, and safety level labeling, collectively forming a low-altitude 3D partitioning model used to describe the permitted activity space of the UAV.

[0036] The final low-altitude three-dimensional zoning model takes into account both the basic constraints of geographical environment and air traffic control system, and incorporates the safety requirements of actual UAV operation. It can accurately and comprehensively describe the permitted activity space of UAVs, providing a precise spatial basis for subsequent heat analysis and site optimization.

[0037] S20: Collect historical flight data of UAVs within the target area, calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model using a multi-factor weighted evaluation model, and generate a three-dimensional heat map based on the comprehensive heat value;

[0038] First, collect historical flight data of drones within the target area, including:

[0039] Based on a preset standard monitoring time window, determine the optimal sampling window length M, where M≥2;

[0040] Obtain the UAV flight logs for the first M standard monitoring time windows within the target area as raw flight data;

[0041] Define the data space boundary according to the second activity space boundary, and clean the original flight data accordingly to obtain the historical flight data.

[0042] Specifically, based on a preset standard monitoring time window, the optimal sampling window length M is determined, including:

[0043] Combining the standard monitoring time window, with 1 to N standard monitoring time windows as the acquisition window length, N differential flight log sets are iteratively obtained, where N > M;

[0044] Calculate the corresponding data entropy by traversing N sets of the differential flight logs, and plot the entropy-window length curve with the acquisition window length and the data entropy as the coordinate axes;

[0045] Elbow identification is performed based on the entropy-window length curve, and the corresponding sampling window length is extracted and determined as the optimal sampling window length M.

[0046] First, based on a preset standard monitoring time window, the optimal sampling window length M is determined. The standard monitoring time window is a reference time unit set for airspace monitoring data analysis, specifically determined according to airspace monitoring requirements, typically a fixed duration such as 6 or 12 hours. Collecting historical flight data of UAVs within the target area based on the preset standard monitoring time window ensures the timeliness and representativeness of the data. Since UAV activity usually exhibits obvious periodic patterns, such as daytime peaks and nighttime troughs, using a standardized time window for data collection effectively avoids data bias that may be caused by sampling at random time intervals, ensuring that the collected data accurately reflects the typical operating conditions of airspace use. Furthermore, the standard time window provides a unified time alignment benchmark for multi-source heterogeneous data, enabling the integration and analysis of monitoring data from different sources on the same time scale. Data collection based on the standard time window also allows for a sliding window processing mechanism, avoiding the repeated processing of all historical data, significantly improving computational efficiency and reducing storage costs while ensuring the accuracy of data analysis.

[0047] Secondly, to determine the optimal sampling window length M, an objective analysis method based on information entropy is adopted. Using 1 to N standard monitoring time windows as the acquisition window length, N differential flight log sets are iteratively obtained, where N is a positive integer greater than M. The N differential flight log sets are traversed, and the information entropy value of each dataset is calculated. Information entropy is an information-theoretic metric describing the randomness and uncertainty of UAV flight data distribution in a three-dimensional grid. It is calculated by statistically analyzing the probability of flight events occurring in each partition unit and applying the Shannon entropy formula. Information entropy quantifies the uncertainty of flight activity distribution in the spatiotemporal dimensions.

[0048] For example, suppose a flight log set contains K flight events, and the low-altitude 3D partition model has G partition cells. First, calculate the frequency pi of flight events in each partition cell i, which is the ratio of the number of flight events in that partition cell to the total number of events K. Then, the formula for calculating the information entropy H of this dataset is: H=-Σ(pi*log2(pi))(i from 1 to G). The more uniform the distribution of flight events in a certain partition unit, that is, the closer the pi values ​​of each unit are, the larger the entropy value, indicating that the flight activity distribution is more dispersed and the uncertainty is higher; when the flight events are concentrated in a few partition units, that is, some pi is close to 1 and the rest is close to 0, the smaller the entropy value, the more concentrated the flight activity distribution is and the lower the uncertainty.

[0049] Furthermore, an entropy-window length curve is plotted with the sampling window length as the horizontal axis and the information entropy value as the vertical axis. This entropy-window length curve is a characteristic curve used to characterize the trend of information entropy changing with the sampling window length, indicating the changing law of historical data information as the sampling time increases, and usually exhibits obvious marginal effect characteristics.

[0050] Furthermore, by applying an elbow recognition algorithm to the entropy-window length curve, the inflection point where the entropy growth rate changes significantly is accurately extracted. The sampling window length corresponding to this inflection point is the optimal sampling window length M. Elbow recognition refers to an algorithm that determines the optimal parameters by identifying the point of maximum change in the curve curvature. By selecting the critical point where information gain tends to saturate but data noise has not yet increased significantly, it ensures that the historical data used contains sufficient spatial activity pattern information while avoiding noise interference introduced by excessively long historical data, thus providing a high-quality data foundation for subsequent heat map calculations.

[0051] It is worth noting that the optimal sampling window length M determined by elbow recognition must satisfy M≥2. This condition ensures that the sampling data covers at least two complete standard monitoring time windows, thereby capturing the basic periodic changes in UAV activity, such as the differences between daytime and nighttime patterns, and avoiding the problem of incomplete pattern recognition caused by single-window data. Simultaneously, a relatively large upper limit N (N>M≥2) is preset during the analysis process to generate multiple datasets with window lengths from 1 to N. This provides the necessary number of data points for constructing the entropy-window length curve, enabling the elbow recognition algorithm to accurately capture the inflection point characteristics of information entropy changes.

[0052] Furthermore, based on the determined optimal sampling window length M, flight log data from the most recent M standard monitoring time windows within the target area are extracted from the UAV monitoring database to form the original flight dataset. This dataset is then cleaned based on the second activity space boundary, such as removing out-of-bounds points, abnormal elevations, and invalid state data, ultimately yielding high-quality historical flight data that conforms to airspace rules. All flight trajectory points in this historical flight data are within the valid airspace activity range, satisfying spatial consistency, and cover continuous standard time windows without interruptions, satisfying temporal continuity.

[0053] Furthermore, a multi-factor weighted evaluation model is used to calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model, and a three-dimensional heat map is generated based on the comprehensive heat value, including:

[0054] The multi-factor weighted evaluation model is defined as follows: the multi-factor weighted evaluation model includes at least a time decay factor, an event weight factor, an aircraft type weight factor, and a frequency factor.

[0055] Based on the historical flight data, the grid flight data corresponding to each partition unit is extracted. The grid flight data includes multiple flight records, and each flight record includes at least the associated stored flight timestamp, flight event marker, and UAV model marker.

[0056] According to the multi-factor weighted evaluation model, each flight record of the grid flight data is evaluated for heat intensity, and the heat intensity evaluation results are accumulated to the corresponding partition unit to obtain the grid accumulated heat intensity.

[0057] Calculate the ratio of the cumulative heat intensity of each grid to the cell volume of the corresponding partition cell to obtain the comprehensive heat intensity value;

[0058] The three-dimensional heat map is obtained by smoothing the combined heat values ​​of multiple partition units using Bézier curves.

[0059] After collecting and preprocessing historical flight data, the raw data, consisting of discrete flight records, is insufficient to intuitively represent the overall airspace usage characteristics. Therefore, it needs to be transformed into quantifiable spatial thermal distribution information. A multi-factor weighted evaluation model is used to spatially aggregate and weight the discrete data, generating a three-dimensional heat map reflecting airspace usage intensity. This provides a visualization and analytical basis for subsequent monitoring point optimization. The specific steps are as follows:

[0060] First, a multi-factor weighted evaluation model is defined. This model is a spatial heat quantification algorithm based on multi-dimensional feature fusion. By fusing multi-dimensional features to quantify the heat contribution value of partition units, it can comprehensively reflect the spatiotemporal distribution characteristics and operational importance of airspace use. The multi-factor weighted evaluation model includes at least four core factors: a time decay factor, which dynamically calculates weights based on flight record timestamps, with recent events having significantly higher weights than historical events to ensure the timeliness of airspace use heat; an event weight factor, which distinguishes risk levels based on flight event markers, adding extra weight to areas where high-risk events such as "unauthorized flights" or "air traffic disruptions" have occurred; an aircraft type weight factor, which distinguishes risk differences based on drone type markers, with large drones and commercial drones having higher risk weights than small recreational drones; and a frequency factor, which directly counts the number of flights within the grid, with areas having higher flight frequencies having higher base heat values.

[0061] Secondly, grid flight data is extracted based on historical flight data. Specifically, for each partition unit in the low-altitude 3D partition model, all flight records within its spatial range are aggregated to form a grid flight dataset. Each flight record must contain three key fields: a flight timestamp, accurately recording the time when the drone entered the partition unit, with an accuracy of at least the second; a flight event marker, using classification codes to identify the flight mission type, such as aerial photography code 01, inspection code 02, transportation code 03, etc.; and a drone model marker, using standard model codes to identify the drone specifications, such as DJI M300 code DJI-M300, Zongheng drone code ZH-280, etc. These fields are stored together using a unified data structure to ensure a complete correspondence between spatiotemporal information and attribute information.

[0062] Furthermore, the heat accumulation is performed by traversing the grid flight data. Heat accumulation refers to the process of summing the heat contribution values ​​of all flight records within a partition cell. Specifically, for each partition cell, each flight record is processed one by one: the time weight coefficient of the record is calculated based on the time decay factor (e.g., using an exponential decay function); the feature weight coefficient is obtained by combining the event weight factor and the aircraft type weight factor; the base frequency (usually 1) is multiplied by the three weight coefficients to obtain the heat contribution value of the flight record; the heat contribution values ​​of all records are accumulated to obtain the grid accumulated heat.

[0063] Next, the heat index of the grid flight data is accumulated. Heat index accumulation refers to the process of weighting and summing the heat index contributions of all flight records within a partition unit to form the overall heat index score for that unit. Specifically, for each partition unit, each flight record is processed individually:

[0064] First, the time weighting coefficient of the record is calculated based on the time decay factor. For example, the exponential decay function Wt=e-λΔt is used, where Δt is the time difference between the current time and the event occurrence time, and λ is the decay rate constant. Based on the event type and aircraft type characteristics, the event weighting coefficient We is obtained, such as setting the weight of "black flight" events to 2.0 and normal flight to 1.0; the aircraft type weighting coefficient Wm is also obtained, such as setting the weight of large commercial drones to 1.8 and small recreational drones to 1.0. The base frequency, with a value of 1, is multiplied by the three weighting coefficients to obtain the heat contribution value of a single flight record: 1×Wt×We×Wm. Finally, the heat contribution values ​​of all flight records within the partition unit are summed to obtain the grid cumulative heat intensity, which is the sum of the heat contribution values ​​of all individual flight records. By quantifying and superimposing the risk contribution of each flight record, the overall heat intensity of the partition unit can be accurately reflected.

[0065] Next, volume standardization is performed. The ratio of the cumulative heat intensity to the physical volume of each partition cell is calculated to generate a comprehensive heat intensity value. Comprehensive heat intensity value = cumulative heat intensity of the grid / physical volume of the partition cell. This calculation eliminates heat intensity deviations caused by differences in partition cell size, ensuring the comparability of partition cells of different sizes. The cell volume is calculated based on the three-dimensional dimensions of the grid, specifically using the formula: Physical volume of partition cell = ΔX × ΔY × ΔZ, where ΔX, ΔY, and ΔZ represent the side lengths of the grid in three-dimensional space.

[0066] Finally, spatial smoothing is achieved using Bézier curves. Bézier curves are a parametric curve mathematical tool that generates smooth, continuous curves using control points and are widely used in computer graphics and data fitting. Specifically, a 3D Bézier curve fitting algorithm is applied to the comprehensive heat values ​​of all partitioned units. A control point matrix is ​​constructed based on the grid spatial coordinates and the comprehensive heat values. A continuous surface is generated through recursive interpolation, eliminating data abrupt changes caused by grid discretization and producing a continuous and smooth 3D heat map. This 3D heat map is visualized using gradient colors or isosurfaces, intuitively displaying the thermal distribution characteristics of different regions in the spatial domain, providing high-precision spatial data support for subsequent monitoring point optimization.

[0067] S30: Define the deployable area in the low-altitude three-dimensional zoning model, with the optimization objective of maximizing the weighted total coverage of the three-dimensional heat map and minimizing the total number of monitoring points, and use the monitoring point deployment scheme as the optimization variable to perform multi-objective optimization solution;

[0068] First, define the deployable area in the low-altitude three-dimensional zoning model. The deployable area refers to the spatial range in the three-dimensional airspace model that meets the physical conditions for equipment installation, is permitted by relevant regulations, and has infrastructure support. By defining the deployable area in the low-altitude three-dimensional zoning model, the feasible solution space of the optimization problem can be clearly defined, which makes it easier to limit the deployment scheme of monitoring points to the actual feasible range and avoid the optimization results from deviating from the actual engineering situation.

[0069] Furthermore, with the optimization objectives of maximizing the weighted total coverage of the three-dimensional heat map and minimizing the total number of monitoring points, and using the monitoring point deployment scheme as the optimization variable, a multi-objective optimization solution is performed, including:

[0070] Using the three-dimensional heat map as the weight spectrum of the three-dimensional space and the number of times the monitoring point is covered as the weighting object, a weighted total coverage operator is constructed, wherein the weighted total coverage operator is positively correlated with the optimization objective;

[0071] A monitoring point operator is constructed by combining the total number of monitoring points contained in the three-dimensional space, wherein the monitoring point operator is positively correlated with the optimization objective, and the total number of monitoring points is negatively correlated with the optimization objective;

[0072] By combining the monitoring point operator and the weighted total coverage operator, an optimization objective function is obtained;

[0073] Based on the monitoring task information of the target area, extract constraint information and regularize it to obtain the constraint set of the deployment scheme;

[0074] Based on a multi-objective optimization algorithm, the monitoring point deployment scheme is iteratively optimized by combining the deployment scheme constraint set and the optimization objective function to obtain the multi-objective optimization solution result. The monitoring point deployment scheme is generated by randomly selecting from the deployable area.

[0075] Specifically, firstly, a weighted total coverage operator is constructed using a three-dimensional heat map as the spatial weighting basis. The weighted total coverage operator is a quantitative evaluation index that comprehensively considers spatial heat distribution and equipment coverage effectiveness. By calculating the sum of the products of the heat value of each zone unit and the number of times the monitored point is covered, the overall coverage benefit can be quantified. Its value is positively correlated with the optimization objective; that is, the larger the value, the better the coverage effect in high-heat areas. Weighted total coverage = Σ(zone unit heat value × number of times the unit is covered), where the number of times a zone unit is covered is determined by the spatial sensing range of the monitored point and the relative positional relationship of the zone unit.

[0076] Secondly, a monitoring point operator is constructed based on the total number of monitoring points in three-dimensional space. This operator is an evaluation function used to quantify the resource investment cost of monitoring equipment. While negatively correlated with the number of devices, it is mathematically transformed into a positively correlated objective function, characterizing equipment utilization efficiency. Specifically, the more monitoring points there are, the higher the equipment cost and the lower the optimization objective value. In practical optimization models, this is typically transformed into a positively correlated objective using its reciprocal form: equipment utilization efficiency = 1 / total number of monitoring points. This transformation ensures the consistency of the optimization objective function's direction; a larger objective value indicates that fewer devices are needed to achieve the same coverage effect, resulting in higher resource utilization efficiency.

[0077] Furthermore, the monitoring point operator and the weighted total coverage operator are fused together to form an optimization objective function. This optimization objective function includes two optimization objectives: weighted total coverage and equipment efficiency, aiming to maximize both weighted coverage and equipment efficiency simultaneously.

[0078] Furthermore, based on the monitoring requirements of the target area, constraints such as equipment performance constraints, cooperative positioning requirements, and deployment area limitations are extracted and transformed into mathematical constraints using regularization methods, constructing a complete set of deployment scheme constraints. Finally, based on multi-objective optimization algorithms, such as NSGA-II, and combining the constraint set and the optimization objective function, the deployment scheme of initial monitoring points randomly generated within the deployable area is iteratively optimized. Through non-dominated sorting, crowding calculation, and genetic operations, the population is continuously evolved, ultimately outputting the multi-objective optimization solution. This solution provides a series of feasible deployment schemes that achieve the best trade-off between coverage effectiveness and economy.

[0079] Specifically, the set of constraints for the deployment scheme includes at least:

[0080] Hardware performance constraints, including the maximum viewing distance and field of view for each monitoring point;

[0081] The collaborative positioning constraint is used to define that any point within the low-altitude three-dimensional partition model is covered by at least K different monitoring points, where K is an integer greater than 1.

[0082] The location feasibility constraint is used to restrict the monitoring points to be deployed only in the deployable areas of the low-altitude three-dimensional partition model.

[0083] Specifically, the deployment scheme constraint set includes three types of core constraints, which together ensure that the optimization results simultaneously meet the requirements of technical feasibility and task reliability.

[0084] First, the deployment scheme constraint set includes hardware performance constraints. These constraints are set based on the physical characteristics of the monitoring equipment and cover the maximum visible distance and viewing angle of each monitoring point. The maximum visible distance limits the effective detection radius of the monitoring point, while the viewing angle limits the horizontal and vertical field of view of the equipment. This constraint calculates the actual coverage area of ​​each point through a geometric model to ensure that the coverage assessment meets the physical limits of the equipment.

[0085] Secondly, the deployment scheme constraint set includes collaborative positioning constraints. These constraints are set to ensure positioning accuracy and reliability, requiring that any point in the low-altitude three-dimensional partition model be covered by at least K different monitoring points, where K is an integer greater than 1. By determining the intersection of the multi-point field of view through spatial analytical geometry, it can ensure that there are no blind spots and effectively avoid target loss caused by single-point failure, thus significantly improving system redundancy.

[0086] Finally, the deployment scheme constraint set also includes point location feasibility constraints. These constraints are set based on actual deployment conditions, restricting monitoring points to be deployed only within predefined deployable areas in the low-altitude 3D zoning model. By eliminating infeasible areas and verifying spatial coordinates, the final scheme is ensured to be engineering-feasible. During the optimization process, penalty function methods or constraint dominance principles are used to handle non-compliant solutions, ensuring that the final scheme fully meets practical application requirements.

[0087] S40: Based on the multi-objective optimization solution results, obtain the monitoring point layout scheme and perform three-dimensional monitoring of the target area accordingly.

[0088] After obtaining the multi-objective optimization solution, the solution set contains multiple monitoring point deployment schemes. Each scheme represents a different trade-off between coverage effectiveness and the number of devices. The final deployment scheme can be selected from the solution set based on actual monitoring needs and resource budget. The final deployment scheme should clearly specify parameters such as the three-dimensional spatial coordinates and device orientation angle of each monitoring point.

[0089] Furthermore, based on the determined deployment plan, surveillance equipment is deployed within the target area, and a three-dimensional surveillance network is constructed. Through the collaborative work of the equipment, all-weather, multi-angle monitoring of the airspace is achieved. Specifically, firstly, the spatial distribution relationship of the surveillance points is utilized to establish a three-dimensional perception network covering the target area; secondly, through data fusion technology, the surveillance data from each point is integrated to generate real-time three-dimensional trajectories of UAV activities; finally, combined with air traffic control rules and heat maps, abnormal flight behavior is automatically identified and warned, forming a complete closed-loop management system for airspace surveillance. Through this step, the theoretical solution output by the optimized algorithm is transformed into a practically operable surveillance system, which not only fully leverages the technical advantages of the previous optimization calculations but also ensures the effective execution of airspace surveillance tasks, ultimately achieving refined and intelligent supervision of UAV activities.

[0090] In summary, the embodiments of this application have at least the following technical effects:

[0091] Compared to existing technologies, this application first constructs a low-altitude three-dimensional partition model that accurately describes the permitted activity space of UAVs by integrating three-dimensional map information and airspace control rules. This establishes a digital and structured foundation for airspace management, overcoming the shortcomings of traditional two-dimensional monitoring methods that cannot accurately reflect spatial altitude information, and providing a reliable spatial benchmark for UAV surveillance. Secondly, based on a multi-factor weighted evaluation model, it comprehensively analyzes multi-dimensional features such as time, events, and aircraft type in historical flight data to generate a three-dimensional map reflecting airspace usage intensity. This enables a quantitative assessment of UAV activity patterns, providing a scientific basis for surveillance resource allocation and solving the problem of blind deployment caused by reliance on experience in existing technologies.

[0092] Furthermore, intelligent deployment of monitoring points is achieved through multi-objective optimization algorithms, minimizing equipment investment costs while ensuring coverage. Finally, three-dimensional dynamic monitoring is implemented based on the optimization results, significantly improving the comprehensiveness and effectiveness of UAV monitoring in complex airspace environments and solving problems such as numerous blind spots, unreasonable resource allocation, and low monitoring efficiency in existing technologies.

[0093] Example 2, as Figure 2 As shown, based on the same inventive concept as the UAV flight dynamic monitoring method based on a 3D map provided in Embodiment 1, this embodiment of the invention also provides a UAV flight dynamic monitoring system based on a 3D map, including:

[0094] The airspace grid management module 11 is used to acquire three-dimensional map information of the target area and construct a low-altitude three-dimensional partition model in combination with airspace control rules. The low-altitude three-dimensional partition model is used to describe the space in which UAVs are allowed to operate.

[0095] The flight heat analysis module 12 is used to collect historical flight data of UAVs in the target area, calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model using a multi-factor weighted evaluation model, and generate a three-dimensional heat map based on the comprehensive heat value.

[0096] The optimization deployment decision module 13 is used to define the deployable area in the low-altitude three-dimensional partition model. The optimization objective is to maximize the weighted total coverage of the three-dimensional heat map and minimize the total number of monitoring points. The multi-objective optimization solution is performed with the monitoring point deployment scheme as the optimization variable.

[0097] The three-dimensional dynamic monitoring module 14 is used to obtain the monitoring point layout scheme based on the multi-objective optimization solution results, and to perform three-dimensional monitoring of the target area accordingly.

[0098] Specifically, the airspace grid management module 11 is used for:

[0099] A three-dimensional map of the target area is acquired, and a low-altitude three-dimensional partition model is constructed in conjunction with airspace control rules. This low-altitude three-dimensional partition model describes the permitted activity space for UAVs, including:

[0100] The boundary of the first activity space is determined based on the airspace control rules and the three-dimensional map information;

[0101] Based on the historical air traffic records and UAV registration information of the target area, the intrinsic parameter information of the UAV is extracted accordingly.

[0102] Traverse the intrinsic parameter information of the UAV, extract the minimum airspace distance constraint, and spatially shrink the first activity space boundary according to the minimum airspace distance constraint to obtain the second activity space boundary;

[0103] The low-altitude three-dimensional partition model is constructed by combining the three-dimensional map information with the boundary of the second activity space.

[0104] Specifically, the flight thermal analysis module 12 is used for:

[0105] Collect historical flight data of UAVs within the target area, including:

[0106] Based on a preset standard monitoring time window, determine the optimal sampling window length M, where M≥2;

[0107] Obtain the UAV flight logs for the first M standard monitoring time windows within the target area as raw flight data;

[0108] Define the data space boundary according to the second activity space boundary, and clean the original flight data accordingly to obtain the historical flight data.

[0109] Specifically, based on a preset standard monitoring time window, the optimal sampling window length M is determined, including:

[0110] Combining the standard monitoring time window, with 1 to N standard monitoring time windows as the acquisition window length, N differential flight log sets are iteratively obtained, where N > M;

[0111] Calculate the corresponding data entropy by traversing N sets of the differential flight logs, and plot the entropy-window length curve with the acquisition window length and the data entropy as the coordinate axes;

[0112] Elbow identification is performed based on the entropy-window length curve, and the corresponding sampling window length is extracted and determined as the optimal sampling window length M.

[0113] Furthermore, a multi-factor weighted evaluation model is used to calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model, and a three-dimensional heat map is generated based on the comprehensive heat value, including:

[0114] The multi-factor weighted evaluation model is defined as follows: the multi-factor weighted evaluation model includes at least a time decay factor, an event weight factor, an aircraft type weight factor, and a frequency factor.

[0115] Based on the historical flight data, the grid flight data corresponding to each partition unit is extracted. The grid flight data includes multiple flight records, and each flight record includes at least the associated stored flight timestamp, flight event marker, and UAV model marker.

[0116] According to the multi-factor weighted evaluation model, each flight record of the grid flight data is evaluated for heat intensity, and the heat intensity evaluation results are accumulated to the corresponding partition unit to obtain the grid accumulated heat intensity.

[0117] Calculate the ratio of the cumulative heat intensity of each grid to the cell volume of the corresponding partition cell to obtain the comprehensive heat intensity value;

[0118] The three-dimensional heat map is obtained by smoothing the combined heat values ​​of multiple partition units using Bézier curves.

[0119] The optimized deployment decision module 13 is specifically used for:

[0120] The optimization objective is to maximize the weighted total coverage of the three-dimensional heat map and minimize the total number of monitoring points. A multi-objective optimization solution is performed, using the monitoring point deployment scheme as the optimization variable. This includes:

[0121] Using the three-dimensional heat map as the weight spectrum of the three-dimensional space and the number of times the monitoring point is covered as the weighting object, a weighted total coverage operator is constructed, wherein the weighted total coverage operator is positively correlated with the optimization objective;

[0122] A monitoring point operator is constructed by combining the total number of monitoring points contained in the three-dimensional space, wherein the monitoring point operator is positively correlated with the optimization objective, and the total number of monitoring points is negatively correlated with the optimization objective;

[0123] By combining the monitoring point operator and the weighted total coverage operator, an optimization objective function is obtained;

[0124] Based on the monitoring task information of the target area, extract constraint information and regularize it to obtain the constraint set of the deployment scheme;

[0125] Based on a multi-objective optimization algorithm, the monitoring point deployment scheme is iteratively optimized by combining the deployment scheme constraint set and the optimization objective function to obtain the multi-objective optimization solution result. The monitoring point deployment scheme is generated by randomly selecting from the deployable area.

[0126] Specifically, the set of constraints for the deployment scheme includes at least:

[0127] Hardware performance constraints, including the maximum viewing distance and field of view for each monitoring point;

[0128] The collaborative positioning constraint is used to define that any point within the low-altitude three-dimensional partition model is covered by at least K different monitoring points, where K is an integer greater than 1.

[0129] The location feasibility constraint is used to restrict the monitoring points to be deployed only in the deployable areas of the low-altitude three-dimensional partition model.

[0130] The three-dimensional dynamic monitoring module 14 is specifically used for:

[0131] Based on the results of the multi-objective optimization, the layout scheme of the monitoring points is obtained, and the three-dimensional monitoring of the target area is carried out accordingly.

[0132] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0133] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0134] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A method for monitoring the flight dynamics of unmanned aerial vehicles (UAVs) based on a 3D map, characterized in that, include: Acquire 3D map information of the target area and construct a low-altitude 3D partition model in combination with airspace control rules. The low-altitude 3D partition model is used to describe the permitted activity space of the UAV. Historical flight data of UAVs within the target area are collected, and a multi-factor weighted evaluation model is used to calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model. A three-dimensional heat map is then generated based on the comprehensive heat value. Define the deployable area in the low-altitude three-dimensional zoning model, with the optimization objectives of maximizing the weighted total coverage of the three-dimensional heat map and minimizing the total number of monitoring points, and use the monitoring point deployment scheme as the optimization variable to solve the multi-objective optimization problem; Based on the results of the multi-objective optimization, the layout scheme of the monitoring points is obtained, and the three-dimensional monitoring of the target area is carried out accordingly.

2. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 1, characterized in that, A three-dimensional map of the target area is acquired, and a low-altitude three-dimensional partition model is constructed in conjunction with airspace control rules. This low-altitude three-dimensional partition model describes the permitted activity space for UAVs, including: The boundary of the first activity space is determined based on the airspace control rules and the three-dimensional map information; Based on the historical air traffic records and UAV registration information of the target area, the intrinsic parameter information of the UAV is extracted accordingly. Traverse the intrinsic parameter information of the UAV, extract the minimum airspace distance constraint, and spatially shrink the first activity space boundary according to the minimum airspace distance constraint to obtain the second activity space boundary; The low-altitude three-dimensional partition model is constructed by combining the three-dimensional map information with the boundary of the second activity space.

3. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 2, characterized in that, Collect historical flight data of UAVs within the target area, including: Based on a preset standard monitoring time window, determine the optimal sampling window length M, where M≥2; Obtain the UAV flight logs for the first M standard monitoring time windows within the target area as raw flight data; Define the data space boundary according to the second activity space boundary, and clean the original flight data accordingly to obtain the historical flight data.

4. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 3, characterized in that, Based on a preset standard monitoring time window, the optimal sampling window length M is determined, including: Combining the standard monitoring time window, with 1 to N standard monitoring time windows as the acquisition window length, N differential flight log sets are iteratively obtained, where N > M; Calculate the corresponding data entropy by traversing N sets of the differential flight logs, and plot the entropy-window length curve with the acquisition window length and the data entropy as the coordinate axes; Elbow identification is performed based on the entropy-window length curve, and the corresponding sampling window length is extracted and determined as the optimal sampling window length M.

5. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 3, characterized in that, A multi-factor weighted evaluation model is used to calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model, and a three-dimensional heat map is generated based on the comprehensive heat value, including: The multi-factor weighted evaluation model is defined as follows: the multi-factor weighted evaluation model includes at least a time decay factor, an event weight factor, an aircraft type weight factor, and a frequency factor. Based on the historical flight data, the grid flight data corresponding to each partition unit is extracted. The grid flight data includes multiple flight records, and each flight record includes at least the associated stored flight timestamp, flight event marker, and UAV model marker. According to the multi-factor weighted evaluation model, each flight record of the grid flight data is evaluated for heat intensity, and the heat intensity evaluation results are accumulated to the corresponding partition unit to obtain the grid accumulated heat intensity. Calculate the ratio of the cumulative heat intensity of each grid to the cell volume of the corresponding partition cell to obtain the comprehensive heat intensity value; The three-dimensional heat map is obtained by smoothing the combined heat values ​​of multiple partition units using Bézier curves.

6. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 5, characterized in that, The optimization objective is to maximize the weighted total coverage of the three-dimensional heat map and minimize the total number of monitoring points. A multi-objective optimization solution is performed, using the monitoring point deployment scheme as the optimization variable. This includes: Using the three-dimensional heat map as the weight spectrum of the three-dimensional space and the number of times the monitoring point is covered as the weighting object, a weighted total coverage operator is constructed, wherein the weighted total coverage operator is positively correlated with the optimization objective; By combining the total number of monitoring points in the three-dimensional space, a monitoring point operator is constructed. The monitoring point operator is an evaluation function used to quantify the resource investment cost of monitoring equipment. The monitoring point operator is positively correlated with the optimization objective and negatively correlated with the number of equipment, representing the equipment utilization efficiency. The total number of monitoring points is negatively correlated with the optimization objective. By combining the monitoring point operator and the weighted total coverage operator, an optimization objective function is obtained; Based on the monitoring task information of the target area, extract constraint information and regularize it to obtain the constraint set of the deployment scheme; Based on a multi-objective optimization algorithm, the monitoring point deployment scheme is iteratively optimized by combining the deployment scheme constraint set and the optimization objective function to obtain the multi-objective optimization solution result. The monitoring point deployment scheme is generated by randomly selecting from the deployable area.

7. The method for monitoring the flight dynamics of a UAV based on a three-dimensional map as described in claim 6, characterized in that, The deployment scheme constraint set includes at least: Hardware performance constraints, including the maximum viewing distance and field of view for each monitoring point; The collaborative positioning constraint is used to define that any point within the low-altitude three-dimensional partition model is covered by at least K different monitoring points, where K is an integer greater than 1. The location feasibility constraint is used to restrict the monitoring points to be deployed only in the deployable areas of the low-altitude three-dimensional partition model.

8. A UAV flight dynamic monitoring system based on a three-dimensional map, characterized in that, A method for performing UAV flight dynamic monitoring based on a three-dimensional map as described in any one of claims 1-7 includes: The airspace grid management module is used to acquire three-dimensional map information of the target area and construct a low-altitude three-dimensional partition model in combination with airspace control rules. The low-altitude three-dimensional partition model is used to describe the space in which UAVs are allowed to operate. The flight heat analysis module is used to collect historical flight data of UAVs in the target area, calculate the comprehensive heat value of each partition unit in the low-altitude three-dimensional partition model using a multi-factor weighted evaluation model, and generate a three-dimensional heat map based on the comprehensive heat value. The optimization deployment decision module is used to define the deployable area in the low-altitude three-dimensional zoning model. The optimization objectives are to maximize the weighted total coverage of the three-dimensional heat map and minimize the total number of monitoring points. The multi-objective optimization solution is performed with the monitoring point deployment scheme as the optimization variable. The three-dimensional dynamic monitoring module is used to obtain the monitoring point deployment scheme based on the multi-objective optimization solution results and perform three-dimensional monitoring of the target area accordingly.

Images