Dynamic environment self-adaptive low-altitude route planning method and system

Abstract

The invention provides a dynamic environment adaptive low-altitude route planning method and system, and the method comprises the steps: collecting multi-source environment information in real time, and constructing a flight environment model; according to flight historical data, in combination with the flight environment model, carrying out route planning; based on the battery capacity of the unmanned aerial vehicle, the endurance demand of the air route is checked, and the optimal air route is output; verifying whether the optimal route meets the requirements of safety and tasks or not, and controlling the unmanned aerial vehicle to fly according to the optimal route after verification is passed; and the unmanned aerial vehicle feeds back the flight state in real time, and dynamically optimizes the air route in combination with the flight environment model. The problems that in the prior art, the environment perception capability is insufficient, situation perception is missing, the optimization target is single, and the algorithm practicability is poor are solved. The problem of route planning of the unmanned aerial vehicle in a dynamic environment is solved.

Description

Technical Field

[0001] This invention belongs to the field of autonomous navigation technology for unmanned aerial vehicles (UAVs), and specifically relates to a dynamic environment adaptive low-altitude route planning method and system. Background Technology

[0002] The vertical scope of low-altitude airspace is defined as airspace below 1000m true altitude. Compared to traditional airspace planning, low-altitude airspace is at a lower altitude and is more affected by the ground environment. Airspace planning in low-altitude airspace requires consideration of complex and diverse limiting factors, but it also has high economic value. In the future, with a highly developed low-altitude economy, the number of drones operating in urban low-altitude airspace will increase significantly, making urban low-altitude airspace unprecedentedly busy and an indispensable part of the urban transportation system. Therefore, it is foreseeable that in the near future, the demand for low-altitude airspace management, flight control, flight planning, and conflict resolution services will increase significantly, becoming the "lifeline" of the low-altitude economy and a fundamental guarantee of low-altitude safety.

[0003] Current low-altitude airspace management methods are not yet perfect, often relying primarily on simple isolation flight. For example, my country's current "Interim Regulations on the Management of Unmanned Aerial Vehicle Flights" requires drones to primarily operate in isolated flight, while also considering the needs of integrated flight. The "Measures for the Management of Air Traffic for Civil Unmanned Aerial Vehicle Systems," promulgated in 2016, also requires the separate designation of isolated airspaces for drones. In principle, the horizontal distance between the boundary of the isolated airspace and the boundary of the airspace used by other aircraft should not be less than 10 kilometers, and the vertical distance between the upper and lower limits and the airspace used by other aircraft should not be less than 600 meters. While this is sufficient to handle the current small number of low-altitude drone applications, it falls short in the face of the potential future scenario of large-scale drone operations in urban low-altitude airspace.

[0004] Existing UAV route planning systems have the following technical problems:

[0005] 1. Insufficient environmental perception capability: Unable to acquire and process meteorological changes in real time, such as dynamic factors like sudden changes in wind speed, precipitation, and terrain changes;

[0006] 2. Lack of situational awareness: Limited ability to detect surrounding aircraft (especially non-cooperative targets), resulting in a high risk of collision;

[0007] 3. Single optimization objective: Only considers the shortest path or the optimal time, ignoring multi-dimensional factors such as safety, energy consumption, and task priority;

[0008] 4. Poor algorithm practicality: Laboratory models are difficult to adapt to the complex constraints of real flight environments. Summary of the Invention

[0009] Therefore, the technical problem to be solved by the present invention is to provide a dynamic environment adaptive low-altitude route planning method and system, which overcomes the problems of insufficient environmental perception, lack of situational awareness, single optimization target, and poor algorithm practicality in the prior art.

[0010] In a first aspect, the present invention provides a dynamic environment adaptive low-altitude flight path planning method, comprising:

[0011] Real-time acquisition of multi-source environmental information to construct a flight environment model;

[0012] Based on flight history data and the aforementioned flight environment model, route planning is carried out.

[0013] Based on the drone's battery level, the flight path's endurance requirements are examined, and the optimal flight path is output.

[0014] Verify whether the optimal flight path meets the safety and mission requirements. Once the verification is successful, control the UAV to fly along the optimal flight path.

[0015] The drone provides real-time feedback on its flight status and, combined with the flight environment model, dynamically optimizes its flight path.

[0016] Furthermore, the construction of the flight environment model includes:

[0017] Spatiotemporal registration: unifying the coordinate system and synchronizing the time for the multi-source environmental information;

[0018] Data correlation: correlation between obstacles and terrain, correlation of meteorological data impact, and integration of traffic situation;

[0019] Situation assessment: safety risk level classification, flight difficulty score;

[0020] Generate a four-dimensional environmental situation map: generate four-dimensional situation information including weather, terrain, obstacles, and traffic. Furthermore, the route planning employs a hybrid optimization engine to generate an initial route, including: generating candidate routes through a reinforcement learning model and performing multi-objective optimization using a genetic optimization algorithm.

[0021] Furthermore, the hybrid optimization engine employs a two-stage optimization approach:

[0022] Phase 1: The reinforcement learning model is input with a multi-dimensional state vector consisting of the UAV's position, speed, flight environment, and battery status, and the PPO algorithm is used to generate the initial flight path.

[0023] Phase Two: The genetic optimization algorithm uses the initial route as the population seed and performs multi-objective optimization through SBX crossover and Gaussian mutation. The optimization function is:

[0024] J = αS + β·C + γ·P

[0025] Where S is the safety factor, C is the flight cost, P is the mission priority factor, α, β, and γ are adaptive weights, and α+β+γ=1.

[0026] Furthermore, the genetic optimization algorithm parameters are configured as follows: population size is 50, number of generations is 100, tournament selection size is 5, SBX crossover probability is 0.85, and Gaussian mutation probability is 0.15.

[0027] Furthermore, the range requirements for the test route include:

[0028] Predict the current battery level of a drone by assessing battery degradation;

[0029] Energy consumption forecasting is based on route planning results;

[0030] Determine whether the drone's current battery level is sufficient for the flight route based on the remaining battery life.

[0031] Furthermore, the function for predicting battery degradation is:

[0032]

[0033] Where: Q0 is the initial charge, λ is the self-discharge coefficient, k is the aerodynamic drag coefficient, and V i For segmented velocities, Δt i This refers to the flight time for a flight segment.

[0034] Furthermore, the function for energy consumption prediction is:

[0035]

[0036] Where: Phover is hovering power, k is aerodynamic coefficient, v is flight speed, h is rate of climb, and η is power system efficiency.

[0037] Furthermore, the function for determining the remaining battery life is:

[0038]

[0039] Among them, E reserve To ensure safe energy reserves, Qcurrent represents the current energy level.

[0040] Secondly, the present invention provides a dynamic environment adaptive low-altitude route planning system, comprising:

[0041] Data Acquisition Module: Collects multi-source environmental information in real time and constructs a flight environment model;

[0042] Planning module: Based on flight history data and the aforementioned flight environment model, route planning is carried out;

[0043] Endurance testing module: Based on the drone's battery level, it tests the endurance requirements of the flight path and outputs the optimal flight path;

[0044] Verification module: Verifies whether the optimal flight path meets the safety and mission requirements. After successful verification, controls the UAV to fly along the optimal flight path.

[0045] Dynamic optimization module: The UAV provides real-time feedback on its flight status and, in conjunction with the flight environment model, dynamically optimizes its flight path.

[0046] Beneficial effects:

[0047] 1. Multi-dimensional perception: Breaking through the limitations of single environmental perception, it achieves four-dimensional fusion of meteorology, terrain, obstacles, and traffic;

[0048] 2. Intelligent decision-making closed loop: RL (reinforcement learning) + GA (genetic optimization algorithm) dual-stage optimization + dynamic battery verification, balancing speed and quality;

[0049] 3. Strong robustness: It maintains functional integrity even in extreme scenarios such as communication interruption and battery degradation;

[0050] 4. High scalability: Supports collaborative operation of multiple drones, providing solutions for dense airspace scenarios. Attached Figure Description

[0051] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings.

[0052] Figure 1 This is a system architecture diagram of a specific embodiment of the present invention;

[0053] Figure 2 This is a schematic diagram of a four-dimensional perception system according to a specific embodiment of the present invention;

[0054] Figure 3 This is a diagram illustrating the optimization algorithm architecture of a specific embodiment of the present invention;

[0055] Figure 4 This is a schematic diagram of the feature extraction layer process in a specific embodiment of the present invention;

[0056] Figure 5 This is a flowchart illustrating the logic of battery model verification in a specific embodiment of the present invention. Detailed Implementation

[0057] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. The principles and features of the present invention are described below with reference to the accompanying drawings. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other. The embodiments given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0058] This invention provides a dynamic environment adaptive low-altitude flight path planning method and system, which solves the problem of poor adaptability of UAVs in dynamic environments. This invention can achieve:

[0059] 1. Four-dimensional situational awareness encompassing weather, terrain, obstacles, and aircraft;

[0060] 2. Three-dimensional optimization of safety, cost, and task priority;

[0061] 3. End-to-end response delay ≤ 1.5 seconds;

[0062] 4. Success rate of tasks in complex environments ≥ 98%. Specific Implementation

[0064] like Figure 1 As shown, the system architecture of this embodiment is as follows:

[0065] I. Support Layer

[0066] It provides basic support for the system, including a historical database (stores past route planning data for reference and learning), a rule engine (with built-in route planning-related rules and constraints), and a case library (stores planning cases in various scenarios to assist decision-making), providing data and rule basis for core processing.

[0067] II. Input Layer

[0068] The system collects multi-source environmental information, including real-time meteorological data (such as wind speed, wind direction, and temperature, which affect flight safety and energy efficiency), topographic and geographic information (mountains, buildings, and other landforms, which are spatial constraints for flight route planning), and dynamic obstacles (moving objects in the air, which need to be avoided in real time). This data is fused from multiple sources and integrated into unified and usable information. At the same time, it integrates mission parameters (flight mission objectives and requirements), UAV status (battery power, flight performance, etc.), and airspace traffic situation (other flight activities in the airspace), providing comprehensive input for subsequent processing.

[0069] III. Core Processing Layer

[0070] 1. Hybrid Planning Engine: Relying on historical data, rules, and cases from the support layer, and combined with fused information from the input layer, the engine performs route planning calculations. The planning results are first validated by a battery model (based on the characteristics of the UAV battery, verifying the match between the route's power requirements and its range), and then the optimized route is output.

[0071] 2. Environmental Modeling Module: Based on the input layer data, this module constructs a flight environment model, providing a foundation for dynamic risk assessment. Dynamic risk assessment utilizes the environmental model, combined with the UAV's status, to analyze potential risks along the flight path (such as sudden weather changes and obstacle collision risks).

[0072] 3. Route Verification Module: Further verifies the optimized route to confirm whether it meets safety, mission, and other requirements. Once verification is successful, control commands are generated to guide the flight.

[0073] IV. Output Control Layer

[0074] The flight control system receives control commands and manipulates the UAV platform to execute flight missions. At the same time, the UAV platform feeds back the flight status to the system. The human-machine interface sends mission commands on the one hand and receives flight status feedback on the other hand for monitoring and display, forming a closed loop to ensure that the flight process is monitorable and adjustable, and realizes adaptive low-altitude route automatic planning and flight control in dynamic environments.

[0075] like Figure 2 As shown, the four-dimensional perception system in this embodiment specifically includes:

[0076] 1. Meteorological data collection can be divided into:

[0077] Wind speed sensor: Real-time measurement of three-dimensional wind speed;

[0078] Temperature and humidity sensor: monitors changes in atmospheric conditions;

[0079] Barometer: Detects trends in air pressure changes;

[0080] Millimeter-wave radar: Inverting the distribution of precipitation particles.

[0081] The processing flow includes: multi-sensor data fusion, construction of a three-dimensional wind field model, precipitation probability inversion, and generation of short-term weather forecasts.

[0082] The output shows the weather conditions for the next 5 minutes, including wind shear warnings.

[0083] 2. Topographic layer data acquisition includes:

[0084] LiDAR: High-precision terrain scanning (accuracy ±5cm);

[0085] RGB-D camera: Visual terrain recognition;

[0086] GNSS / RTK: Centimeter-level positioning;

[0087] IMU: High-precision attitude measurement.

[0088] The processing flow includes: point cloud data processing and filtering, real-time visual SLAM localization, multi-source data fusion, and construction of a 3D terrain model.

[0089] The output is a digital elevation map (DEM) that can be used for accessibility analysis.

[0090] 3. Data acquisition for the barrier layer includes:

[0091] 77GHz radar: detects dynamic obstacles (range 200m);

[0092] Binocular vision: obstacle recognition and classification;

[0093] Ultrasonic sensor: Near-range obstacle detection.

[0094] The processing flow includes: multimodal obstacle detection, obstacle classification (static / dynamic / biological), motion trajectory prediction, and safe distance calculation.

[0095] The output includes obstacle position / velocity / type and collision risk prediction.

[0096] 4. Data collection at the traffic level includes:

[0097] ADS-B receiver: Receives signals from cooperative aircraft;

[0098] RF localization: Detecting non-cooperative targets;

[0099] Self-organizing network communication: Cooperative perception among unmanned aerial vehicles;

[0100] The processing flow includes: cooperative target tracking, non-cooperative target identification, and multi-UAV collaborative perception to generate an airspace situation map.

[0101] The output includes the real-time location of surrounding aircraft and a heat map of airspace traffic density.

[0102] 5. The processing flow of the data fusion center includes:

[0103] Spatiotemporal registration: unified coordinate system (WGS84 / local), time synchronization (PTP protocol, error <1ms);

[0104] Multi-source data association: obstacle and terrain association, meteorological impact analysis, and traffic situation fusion;

[0105] Situation assessment: classification of security risk levels and scoring of passage difficulty;

[0106] Security risk analysis: comprehensive risk assessment matrix, threat level classification;

[0107] Generate a four-dimensional environmental situation map: comprehensively visualize the environmental status, with an update frequency of 10Hz.

[0108] 6. The decision-making process of the UAV decision-making system includes:

[0109] Route planning: Generate initial routes based on situational maps and optimize for multiple objectives (safety / efficiency / energy consumption);

[0110] Dynamic adjustment: Response to environmental changes (delay ≤ 1.2s), real-time route optimization;

[0111] Flight control: precise track tracking and emergency maneuver control.

[0112] As shown in Table 1, the implementation and performance indicators of the four-dimensional situational awareness system in this embodiment are presented.

[0113] Table 1

[0114]

[0115]

[0116] like Figure 3 As shown, the hybrid optimization algorithm architecture of this embodiment is as follows:

[0117] 1. The input layer includes:

[0118] Thinking perception data: four-dimensional situational information including weather, terrain, obstacles, and traffic;

[0119] Drone status: Real-time parameters such as position, speed, attitude, and remaining battery power;

[0120] Task parameters: target point coordinates, task priority, time constraints;

[0121] Historical flight data: past flight records, energy consumption models, performance indicators;

[0122] Rules engine: Built-in rules and constraints related to route planning.

[0123] 2. For example Figure 4 As shown, the feature extraction layer includes:

[0124] Data standardization: Min-Max normalization processing;

[0125] Feature engineering: Constructing spatiotemporal feature matrices;

[0126] Output: 128-dimensional normalized feature vector (location, environment, task, device status).

[0127] 3. Hybrid Optimization Engine:

[0128] 3.1 Reinforcement Learning Module (PPO Algorithm)

[0129]

[0130]

[0131] The input consists of 128-dimensional state features, and the output is an initial route plan (containing 5-10 waypoints). Key technologies include: policy gradient optimization, advantage function calculation, and importance sampling.

[0132] 3.2 Genetic Algorithm Module:

[0133]

[0134]

[0135] The key technical parameters are as follows:

[0136] Population size: 50; Generations: 100; Crossover probability: 0.85; Mutation probability: 0.15;

[0137] Fitness function:

[0138]

[0139] Where S = safety factor, C = flight cost, and P = mission priority.

[0140] 4. Battery model verification

[0141] Energy consumption prediction model:

[0142]

[0143] Where: Phover: hovering power, k: aerodynamic coefficient, v: flight speed Climb rate, η: Power system efficiency

[0144] Remaining battery life calculation:

[0145]

[0146] Ereserve is a safety reserve of energy (≥15% of total energy).

[0147] The logical flow of battery model verification, such as Figure 5 As shown.

[0148] 5. Output

[0149] Optimized route output: includes waypoint sequence, flight altitude, and speed profile;

[0150] Route adjustment mechanism: Automatic adjustment is triggered when battery verification fails; Output format:

[0151]

[0152] The multi-objective cost function in this embodiment includes:

[0153] Define the three-dimensional optimization objective:

[0154] J = α·S + β·C + γ·P

[0155] in:

[0156] S: Safety factor (normalized value of minimum distance to obstacles / aircraft); C: Flight cost (including energy consumption E + time T): C = k1E + k2T; P: Mission priority factor (dynamically weighted according to mission type);

[0157] α, β, γ: adaptive weight (α+β+γ=1);

[0158] The intelligent optimization algorithm in this embodiment includes:

[0159] Hybrid optimization model architecture:

[0160]

[0161]

[0162] Algorithm parameter configuration:

[0163] PPO reinforcement learning:

[0164] State space: 48 dimensions (position / velocity / environment / battery state);

[0165] Action space: 36 discrete actions (10° intervals for heading angle, 5 speed levels);

[0166] Reward function: R = 200·S - 0.2·C - 1000·δcollision;

[0167] Genetic Algorithm:

[0168] Population size: 50;

[0169] Selection operator: Tournament selection (size = 5);

[0170] Crossover rate: 0.85 (SBX crossover);

[0171] Mutation rate: 0.15 (Gaussian mutation);

[0172] The battery degradation model in this embodiment:

[0173]

[0174] Where: Q0: initial charge, λ: self-discharge coefficient (measured value 0.001 / h), k: aerodynamic drag coefficient (drone-specific parameter), V i : Piecewise velocity, Δt i Flight time for a flight segment.

[0175] This embodiment has already been applied in real-world scenarios, with the specific implementation scenarios as follows:

[0176] (I) Hardware Configuration

[0177] Drone side:

[0178]

[0179]

[0180] Ground station:

[0181] Edge computing nodes: Intel i9-13900K + NVIDIA RTX 6000Ada; Database: TimescaleDB (spatiotemporal data storage);

[0182] Communication: 5G private network + LoRaWAN dual-link redundancy;

[0183] (II) Software Implementation

[0184] Core code for route optimization:

[0185]

[0186]

[0187] (III) Work Process

[0188] Task initialization: Input target point, task priority (level 1-5), and safety threshold; Environmental awareness: Real-time acquisition of four-dimensional situational awareness data (update frequency 10Hz);

[0189] Route generation: RL generates 3 candidate paths, and GA optimization selects the Pareto optimal solution.

[0190] Flight control: Waypoints are executed in segments, environmental changes are detected every 500ms, and replanning response time is ≤1.2 seconds.

[0191] Mission terminated: Upon reaching the target point or triggering a safety mechanism.

[0192] (iv) Exception handling mechanism, as shown in Table 2:

[0193] Table 2

[0194] Exception types Processing strategy Response time sudden obstacles Speed ​​obstacle avoidance + route replanning ≤0.8s Battery warning Switch to energy-saving mode + nearest landing point planning ≤1.0s Communication interruption Local cache route execution + self-organizing network collaboration ≤0.5s Sudden weather changes Adaptive deceleration + altitude adjustment ≤1.2s

[0195] After implementation and experimental verification, this embodiment has been successfully implemented as follows:

[0196] (I) Test Environment:

[0197] Location: A 5km x 5km urban simulation environment (including 30 buildings);

[0198] Weather: Wind speed varies randomly from 0 to 15 m / s, with a 30% chance of precipitation;

[0199] Transportation: 20 background drones flying randomly;

[0200] Task: Medical supplies delivery (Priority Level 4);

[0201] (II) Performance comparison, as shown in Table 3:

[0202] Table 3

[0203]

[0204] (III) Typical scenario tests, as shown in Table 4:

[0205] Table 4

[0206]

[0207]

[0208] The specific experimental data for this embodiment are as follows:

[0209] (a) Sensor accuracy data, as shown in Table 5:

[0210] Table 5

[0211]

[0212] (ii) Algorithm performance data, as shown in Table 6:

[0213] Table 6

[0214]

[0215]

[0216] (III) Energy consumption model data, as shown in Table 7:

[0217] Table 7

[0218] Flight status Power consumption (W) Influencing factors Hover 1200 Altitude Cruise (8m / s) 1500 Speed ​​+ Wind Resistance Climb (5m / s) 2200 Climb rate + Gravity Energy saving mode 950 Slowing down + optimizing the path Emergency Obstacle Avoidance 1800 Acceleration + Attitude Change

[0219] (iv) Multi-objective optimization weights, as shown in Table 8:

[0220] Table 8

[0221] Task type Safety weight (α) Cost weight (β) Priority weight (γ) Medical emergency 0.75 0.15 0.1 Logistics and distribution 0.6 0.3 0.1 Agricultural spraying 0.55 0.35 0.1 Inspection and monitoring 0.65 0.25 0.1 Surveying and Exploration 0.5 0.4 0.1

[0222] (v) Historical training data, as shown in Table 9:

[0223] Table 9

[0224]

[0225]

[0226] The innovations and advantages of this embodiment are as follows:

[0227] 1. Four-dimensional situational awareness architecture: The first integrated perception solution for weather / terrain / obstacles / aircraft, with multi-sensor spatiotemporal synchronization technology (error ≤10ms);

[0228] 2. Hybrid Optimization Engine: A two-stage optimization mechanism combining RL (Reinforcement Learning) and GA (Genetic Optimization Algorithm), with a three-dimensional cost function of safety, cost, and priority;

[0229] 3. Engineering battery model:

[0230] Pavailable=Prated·(1-0.005·ΔT)·e-0.0012·t

[0231] Where ΔT is the temperature change and t is the flight time;

[0232] 4. Distributed decision-making architecture: Edge computing node processing latency ≤300ms, and basic functions can still be maintained when communication is interrupted.

[0233] The improvement effect of this implementation compared with the existing technology is shown in Table 10:

[0234] Table 10

[0235]

[0236]

[0237] This embodiment solves the route planning problem of UAVs in dynamic environments through a technical closed loop of multi-source perception, hybrid optimization, and dynamic compensation, and has significant technical innovation and engineering practical value.

[0238] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A dynamic environment adaptive low-altitude flight path planning method, characterized in that, include: Real-time acquisition of multi-source environmental information to construct a flight environment model; Based on flight history data and the aforementioned flight environment model, route planning is carried out. Based on the drone's battery level, the flight path's endurance requirements are examined, and the optimal flight path is output. Verify whether the optimal flight path meets the safety and mission requirements. Once the verification is successful, control the UAV to fly along the optimal flight path. The drone provides real-time feedback on its flight status and, combined with the flight environment model, dynamically optimizes its flight path.

2. The method according to claim 1, characterized in that, The constructed flight environment model includes: Spatiotemporal registration: unifying the coordinate system and synchronizing the time for the multi-source environmental information; Data correlation: correlation between obstacles and terrain, correlation of meteorological data impact, and integration of traffic situation; Situation assessment: safety risk level classification, flight difficulty score; Generate a four-dimensional environmental situation map: generate four-dimensional situation information including weather, terrain, obstacles, and traffic.

3. The method according to claim 1, characterized in that, The route planning uses a hybrid optimization engine to generate initial routes, including: generating candidate routes through a reinforcement learning model and performing multi-objective optimization using a genetic optimization algorithm.

4. The method according to claim 3, characterized in that, The hybrid optimization engine employs a two-stage optimization approach: Phase 1: The reinforcement learning model is input with a multi-dimensional state vector consisting of the UAV's position, speed, flight environment, and battery status, and the PPO algorithm is used to generate the initial flight path. Phase Two: The genetic optimization algorithm uses the initial route as the population seed and performs multi-objective optimization through SBX crossover and Gaussian mutation. The optimization function is: J = α·S + β·C + γ·P Where S is the safety factor, C is the flight cost, P is the mission priority factor, α, β, and γ are adaptive weights, and α+β+γ=1.

5. The method according to claim 4, characterized in that, The genetic optimization algorithm parameters are configured as follows: population size is 50, number of generations is 100, tournament selection size is 5, SBX crossover probability is 0.85, and Gaussian mutation probability is 0.

15.

6. The method according to claim 1, characterized in that, The range requirements for the test route include: Predict the drone's current battery level by assessing battery degradation; Energy consumption forecasting is based on route planning results; Determine whether the drone's current battery level is sufficient for the flight route based on the remaining battery life.

7. The method according to claim 6, characterized in that, The function for predicting battery degradation is: Where: Q0 is the initial charge, λ is the self-discharge coefficient, k is the aerodynamic drag coefficient, and V i Let Δt be the segmented velocity. i This refers to the flight time for a flight segment.

8. The method according to claim 6, characterized in that, The function for energy consumption prediction is: Where: Phover is hovering power, k is aerodynamic coefficient, and v is flight speed. Let η be the climb rate and η be the efficiency of the power system.

9. The method according to claim 6, characterized in that, The function for determining the remaining battery life is: Among them, E reserve To ensure safe energy reserves, Qcurrent represents the current energy level.

10. A dynamic environment adaptive low-altitude route planning system, characterized in that, include: Data Acquisition Module: Collects multi-source environmental information in real time and constructs a flight environment model; Planning module: Based on flight history data and the aforementioned flight environment model, route planning is carried out; Endurance testing module: Based on the drone's battery level, it tests the endurance requirements of the flight path and outputs the optimal flight path; Verification module: Verifies whether the optimal flight path meets the safety and mission requirements. After successful verification, controls the UAV to fly along the optimal flight path. Dynamic optimization module: The UAV provides real-time feedback on its flight status and, in conjunction with the flight environment model, dynamically optimizes its flight path.

Images