An adaptive and dynamically adjusted low-altitude aircraft take-off and landing point intelligent scheduling method

Abstract

This invention belongs to the field of low-altitude traffic operation scheduling and intelligent optimization technology, specifically involving an adaptive and dynamically adjusted intelligent scheduling method for low-altitude aircraft take-off and landing points. The method includes: Step 1: Multi-source dynamic data perception and feature library construction; Step 2: Definition of scheduling constraint system and scientific calculation of parameters; Step 3: Construction of multi-objective optimization model; Step 4: Solving with adaptive optimization algorithm; Step 5: Output of planning scheme and dynamic adjustment. The advantages of this invention are: focusing on market-based bidding and reservation of take-off and landing point time slots, personalized pilot adaptation, and dynamic opening and closing of take-off and landing point status; adding a take-off and landing point time slot option pricing system, a pilot human factors engineering adaptation system, and a multi-aircraft type differentiated scheduling system; realizing market-based dynamic pricing and optimal allocation of time slot resources; achieving personalized adaptation scheduling of take-off and landing points with pilot technical level, operating habits, and physiological state; and improving scheduling safety.

Description

Technical Field

[0001] This invention belongs to the field of low-altitude traffic operation scheduling and intelligent optimization technology. Specifically, it relates to an intelligent scheduling method and system for low-altitude aircraft take-off and landing points that integrates multi-source dynamic data, multi-aircraft model differentiated adaptation, human factors engineering and financial pricing mechanisms. It focuses on the characteristics of take-off and landing point time slot allocation and dynamic state scheduling, and is applicable to time slot bidding, aircraft model adaptation, personnel matching and operation scheduling management of fixed take-off and landing points in multiple scenarios such as urban low-altitude logistics distribution, emergency rescue, and short-distance commuting. Background Technology

[0002] With the low-altitude economy being incorporated into the national strategic emerging industries, the large-scale application of low-altitude aircraft (including eVTOL and manned aircraft) has become a trend. As the core hub of the low-altitude transportation network, the scheduling efficiency of take-off and landing points directly determines the safety, efficiency, and economy of low-altitude operations. However, existing technologies have significant shortcomings in the take-off and landing point scheduling process. Chinese patent CN116088585A discloses a multi-UAV take-off and landing sequence planning system and method based on the Hungarian algorithm. This scheme only plans the take-off and landing sequence based on static parameters such as UAV battery power, arrival time, and execution batches, without considering airspace dynamics, weather changes, pilot status, and market-based pricing of time slots, resulting in insufficient dynamic adaptability. Chinese patent CN118759975B discloses a vertical take-off and landing field control system. This system focuses on take-off and landing field environmental monitoring and path planning, but it lacks mechanisms for multi-aircraft type differentiation, pilot human factor matching, and time slot bidding, thus failing to achieve optimal resource market allocation. Chinese patent CN119090240B discloses an approach and departure route and time slot allocation method for UAV take-off and landing sites. This scheme only presets time slot windows based on take-off and landing site capacity and aircraft traffic, without introducing dynamic pricing models, human factor adaptation rules and multi-scale dynamic response mechanisms. The time slot allocation is rigid, which can easily lead to resource misallocation and cannot adapt to real-time fluctuations in weather, airspace and pilot status.

[0003] Paper 1: A hybrid task allocation approach for multi-UAV systems with complex constraints: a market-based bidding strategy and improved NSGA-III optimization. This approach adopts a market bidding strategy and improves NSGA-III to achieve multi-UAV task allocation. Although it introduces the concept of market-based bidding, it only allocates tasks at the task level and does not extend to the scheduling of take-off and landing time slots. It does not integrate human factors engineering, multi-model adaptation and multi-scale dynamic data. The bidding mechanism has poor adaptability to low-altitude take-off and landing scenarios and cannot solve core problems such as time slot dynamic pricing and pilot-take-off and landing point matching.

[0004] Paper 2: Incentive-Compatible Vertiport Reservation in Advanced Air Mobility: An Auction-Based Approach proposes an auction-based method for reservations at advanced air traffic take-off and landing sites, focusing on an incentive-compatible reservation mechanism. However, the auction mechanism differs fundamentally from the Black-Scholes option pricing model. Furthermore, this paper does not consider the coupled effects of pilot human factors, multi-aircraft type differentiation, and multi-scale dynamic factors (weather, airspace, fatigue). It also fails to construct a multi-objective optimization system that coordinates safety, efficiency, cost, human factors, and resource allocation, making the scheduling scheme prone to imbalance.

[0005] Paper 3: The Dynamic Operator Allocation / Multiple Attribute Decision-Making Methods for the Dynamic Operator Allocation Problem uses operator fatigue as a triggering factor, employs AHP to determine attribute weights, and combines fuzzy logic and TOPSIS methods for dynamic operator allocation. However, this technology only uses fatigue as a single triggering factor and does not establish a multi-dimensional quantitative matching mechanism for technical level, operating habits, and physiological state. Furthermore, it is only applicable to general manufacturing systems and cannot be adapted to low-altitude take-off and landing point scheduling scenarios.

[0006] The aforementioned existing technologies and standards all revolve around the allocation of takeoff and landing resources and basic control, but they have significant shortcomings in key areas such as scheduling mechanisms, human factor adaptation, dynamic collaboration, multi-objective optimization, and algorithm adaptation, making them unable to support high-density, multi-scenario, and high-safety-requirement urban low-altitude operations. Overall, existing technologies exhibit a four-layer progressive deficiency in the takeoff and landing point scheduling stage—mechanism layer, adaptation layer, collaboration layer, and algorithm layer, as detailed below: ① Defects at the mechanism level: rigid scheduling rules and lack of market-based allocation. Traditional scheduling is mainly based on "first come, first served" or "fixed priority preemption", without introducing a market-based resource allocation mechanism. This easily leads to waste of take-off and landing time slot resources and unfair allocation. The problem of no available time slots when aircraft arrive under high-density commercial operation is prominent.

[0007] ② Adaptation layer defects: Insufficient adaptation of both human factors and aircraft type; the scheduling scheme ignores the differences in the technical level, operating habits and physiological state of the pilots, which can easily lead to take-off and landing operation risks during manned or remote driving phases; at the same time, the precise matching logic between multiple aircraft types and take-off and landing sites and time slots has not been established, and resource mismatches such as large sites serving small aircraft types and small sites being unable to meet the needs of large aircraft types are common.

[0008] ③ Defects in the coordination layer: The dynamic response is lagging, the multi-objective optimization is unbalanced, and the scheduling only responds to basic dynamic data such as airspace and meteorology. There is no dynamic pricing and real-time adjustment mechanism for time slot resources, resulting in a lag in response to emergencies. Moreover, the focus is on single-objective optimization of time slot utilization, without forming an optimization system that coordinates safety, efficiency, cost, human factors, and resource allocation. This easily leads to imbalances such as "emphasizing efficiency over safety" and "emphasizing allocation over adaptation".

[0009] ④ Algorithm layer defects: Insufficient solution performance makes it difficult to support real-time scheduling. Existing scheduling algorithms are not optimized for low-altitude dynamics, human factors adaptability and market-based time slot scheduling characteristics. They have slow convergence speed and are prone to getting trapped in local optima, which cannot meet the real-time and efficient scheduling requirements in high-density take-off and landing scenarios.

[0010] Current technologies generally equate "dynamism" simply with real-time data updates, failing to recognize that the core challenge of low-altitude scheduling lies in the fact that multiple dynamic factors have completely different time scales and are coupled together, jointly influencing scheduling decisions. ① Airspace occupation and conflict risks: sudden changes on the order of seconds to minutes, requiring scheduling to have millisecond-level response; ② Fluctuations in weather and time slot demand: These changes occur on a minute-by-minute basis, requiring minute-by-minute rescheduling; ③ Driver fatigue and physiological state: gradual changes over hours, requiring hourly adaptation and calibration; ④ Status of take-off and landing facilities and site conditions: fluctuations on a minute to hour level, requiring minute-level constraint verification.

[0011] Dynamic factors at different time scales impose varying requirements on constraint rigidity, real-time performance, computational overhead, and optimization objective weights: second-level conflicts require rapid optimization, minute-level pricing requires continuous updates, and hourly human factors require stable adaptation. Existing technologies employ a single cycle, a single algorithm, and a single constraint weight, which cannot perform hierarchical processing, time-sharing response, graded penalties, or adaptive weight adjustment for multi-scale dynamic factors. They either miss the optimal scheduling window due to computational lag, cause system oscillations due to frequent recalculations, or sacrifice safety or efficiency due to a one-size-fits-all approach to constraints. Essentially, they lack the ability to handle multi-scale, strongly coupled dynamic environments.

[0012] Existing research on traditional takeoff and landing scheduling methods lacks market-based pricing design. Some intelligent scheduling studies only focus on hardware resource matching, failing to integrate human factors engineering and financial pricing models, and even more so, failing to address the core scientific problem of dynamic coupling across multiple time scales. This prevents the formation of a complete closed loop of "data perception—market-based pricing—human factor adaptation—dynamic scheduling—iterative optimization." Therefore, there is an urgent need to construct an intelligent scheduling method and system for low-altitude aircraft takeoff and landing points that focuses on dynamic scheduling of takeoff and landing slots, integrates financial pricing and personnel adaptation, is adaptable to multiple types of aircraft, and can uniformly handle multi-scale dynamic factors. Summary of the Invention

[0013] In view of the above problems, the purpose of this invention is to provide an adaptive and dynamically adjustable intelligent scheduling method for low-altitude aircraft take-off and landing points, which focuses on market-based bidding reservation of take-off and landing point slots, personalized pilot adaptation, dynamic opening and closing of take-off and landing point status, and adds a take-off and landing point slot option pricing system, a pilot human factors engineering adaptation system and a multi-aircraft type differentiated scheduling system to overcome the shortcomings of the above-mentioned prior art.

[0014] This invention provides an adaptive and dynamically adjustable intelligent scheduling method for the take-off and landing points of low-altitude aircraft, comprising the following steps: Step 1: Multi-source dynamic data perception and feature library construction. Construct a multi-dimensional perception network, collect multi-source dynamic data, and simultaneously establish feature libraries for multiple types of aircraft and pilots to complete the construction of a multi-source dynamic data feature library. Step 2: Definition of scheduling constraint system and scientific calculation of parameters. Construct a complete constraint system by establishing hard constraints, soft constraints, time slot pricing constraints, human factor adaptation constraints, and dynamic opening and closing constraints at take-off and landing points; Step 3: Constructing a multi-objective optimization model. Construct a collaborative weighted multi-objective optimization model that considers safety redundancy, operational efficiency, demand coverage, human factor adaptability, aircraft type adaptability, and cost optimization objectives. Combine this with a complete constraint system to deeply couple the option pricing mechanism with the scheduling objectives, thereby achieving multi-objective collaborative optimal matching of scheduling schemes under different low-altitude application scenarios. Step 4: Solve using the adaptive optimization algorithm. The design improves the adaptive NSGA-III optimization algorithm by incorporating scene weight adaptive adjustment, constraint penalty mechanism and local optimization strategy to achieve efficient solution of collaborative weighted multi-objective optimization model and output of optimal scheduling scheme, ensuring that the solution speed and global optimality are adapted to the low-altitude real-time scheduling requirements. Step 5: Output and Dynamic Adjustment of Planning Scheme Output the take-off and landing point scheduling plan, and at the same time build a closed-loop mechanism for real-time perception, dynamic analysis, and iterative adjustment. Combine multi-source dynamic data, scene changes, and constraint fluctuations to make real-time corrections to the take-off and landing point scheduling plan.

[0015] As a preferred embodiment of the present invention, step 1 further includes the following step: Step 1.1: Deployment of perception nodes and definition of data types. Construct a multi-dimensional perception network and deploy six types of perception nodes: airspace, environment, demand, facilities, aircraft status, and pilots. Collect static geographic data, dynamic operational data, facility status data, real-time data of various types of aircraft, and pilot status data. Step 1.2 Data Acquisition and Preprocessing Specifications Step 1.2.1, Data acquisition frequency, ① High-frequency dynamic data: The acquisition cycle is 3 minutes to ensure real-time performance; ②Medium frequency dynamic data: Acquisition cycle of 15 minutes, balancing real-time performance and computational cost; Step 1.2.2, Data Preprocessing Methods ①Outlier removal: using 3 The criteria remove outliers from meteorological and signal strength data; ② Missing value imputation: Linear interpolation is used to imput short-term missing data, and an LSTM-based time series prediction model is used to imput long-term missing data; ③ Data standardization: Normalize data of different dimensions; Step 1.3: Construction of a feature library for multiple types of aircraft. Establish an aircraft feature database including logistics drones, eVTOL, emergency rescue drones, police drones, and manned aircraft, and extract core performance parameters including: maximum wingspan, maximum liftoff weight, maximum wind resistance threshold, range, minimum takeoff and landing area, communication signal requirements, and charging power.

[0016] As a preferred embodiment of the present invention, step 2 further includes the following step: Step 2.1, Hard Constraints Step 2.1.1, Safety distance constraints, (1) Safety distance constraints between take-off and landing points and fixed obstacles, (2) Minimum horizontal distance constraint between the take-off and landing point and the moving obstacle. Step 2.1.2, Spatial Constraints, (1) Distance constraints between take-off and landing points and the boundary of controlled airspace, (2) Distance constraints between takeoff and landing points and the centerline of the flight path, Step 2.1.3, Meteorological constraints, (1) Wind speed constraint, (2) Visibility constraints, (3) Precipitation constraints, Step 2.1.4, Facility Constraints (1) Supply voltage fluctuation constraint, (2) Communication signal strength constraints, (3) Minimum available area constraint, Step 2.1.5, Human Factors-Based Adaptation Constraints (1) Fatigue state constraints, (2) Consecutive operation time constraints, (3) Safety constraints on physiological indicators, (4) No-fly zone restrictions, Step 2.2, Soft Constraints Step 2.2.1, Efficiency Constraints (1) Service radius constraints of take-off and landing points, (2) Average utilization rate constraint of take-off and landing time slots, (3) Average distance constraint from the demand point to the nearest take-off and landing point. Step 2.2.2, Cost Constraints (1) Real-time operating cost adaptation constraints at take-off and landing points. (2) Constraints on real-time connectivity between take-off and landing points and existing infrastructure. Step 2.2.3, Sustainability Constraints (1) Noise emission constraints at take-off and landing points, (2) Real-time avoidance constraints in the buffer zone of the ecological protection area. Step 2.2.4, Time Slot Utilization Constraints Step 2.3, Time Slot Pricing Constraints Based on the BS option pricing model, and combined with the real-time non-renewability of take-off and landing time slots and the dynamic optimization of scenarios, the option price and scheduling rules are dynamically adjusted. Step 2.3.1, Constraints of the core formula for BS option pricing. Market-based dynamic scheduling is achieved through time-slot options. ; ; ; In the formula, The real-time premium for takeoff and landing slot options is the fee that users need to pay to reserve a slot. for The real-time value of a time slot is determined by location, time period, airspace status, and demand intensity. and All are intermediate variables. It reflects the relative strength of the underlying asset price relative to the strike price, and takes into account the effects of time, volatility and risk-free interest rate; This represents the adjusted probability that the option will be exercised. The option exercise price is the final cost paid when the slot is actually used. This is the current risk-free interest rate, used for value discounting; The remaining time to expiration of the option is equal to the planned takeoff and landing time minus the current time; The slot value volatility reflects price fluctuations caused by low-altitude supply and demand and environmental changes. The cumulative distribution function is a standard normal distribution, representing the probability of time slot usage and the expected risk. Step 2.3.2, Time Slot Value Volatility Constraints Step 2.3.3, Option expiration time constraint, Step 2.3.4, Time-slot option price volatility constraints, Step 2.4 Human Factors Adaptation Constraints By combining real-time operation characteristics optimization for low-altitude driving, the intelligent scheduling system for low-altitude take-off and landing points collects driver data in real time to achieve dynamic and personalized matching of time slots between drivers and take-off and landing points. Step 2.4.1: Real-time adaptation constraints between the driver and the takeoff and landing points. Step 2.4.2, Real-time matching degree constraint of operating habits. Step 2.4.3: Real-time matching constraint between take-off and landing point difficulty and driver skill level. Step 2.4.4: Real-time matching degree constraint of operation level. Step 2.4.5: Real-time increase of safety redundancy constraints under fatigue conditions. Step 2.4.6: Real-time physiological state matching degree constraint. Step 2.5, Option Premium A linkage mechanism with scheduling decisions.

[0017] As a preferred embodiment of the present invention, the option premium in step 2.5 In addition to being used as a time slot reservation fee, it also participates in scheduling decisions through the following rules: Demand feedback linkage: The intelligent scheduling system for low-altitude take-off and landing points will calculate the option premium in real time. The information is presented to users through a display layer; users decide whether to book based on their budget and the urgency of the task; when the option premium... When the user's preset threshold is exceeded, the user will cancel or postpone the reservation. The low-altitude take-off and landing point intelligent scheduling system collects behavioral changes in real time through demand sensing nodes, updates demand distribution data, and thus indirectly affects the demand coverage target. and operational efficiency goals ; Dynamic adjustment trigger: The intelligent scheduling system for low-altitude take-off and landing points monitors option premiums. short-term rate of change and its deviation from the historical average; according to the dynamic adjustment mechanism of the scheduling plan, the short-term rate of change The value determines the level of adjustment triggered; Execution Priority Ranking: When generating the planning scheme output, the low-altitude take-off and landing point intelligent scheduling system will prioritize the reserved time slots according to option premiums. Sorted from highest to lowest value; high-value time slots are allocated preferentially to technology levels. High pilot and aircraft compatibility High-altitude aircraft; Operating cost adjustment: The operator periodically adjusts operating costs based on historical option premiums. The accumulated revenue will be used to dynamically adjust the upper limit of operating costs for order slots in different scenarios within the soft constraint. .

[0018] As a preferred embodiment of the present invention, step 3 further includes the following step: Step 3.1: Define the objective function. A multi-objective weighted summation objective function is constructed, and the weight coefficients are dynamically adjusted for different low-altitude application scenarios. The basic weights are determined by the analytic hierarchy process. The core objective function of the multi-objective optimization model : ; In the formula, For safety redundancy targets; For operational efficiency targets; To cover the target of demand; Human factors fit target; For model compatibility targets; To optimize costs; The target weighting coefficient for safety redundancy; Weighting coefficients for operational efficiency targets; Determine the target weighting coefficient for demand coverage; Human factors fit target weight coefficient; The target weighting coefficient for device compatibility; Optimize the target weight coefficients for cost.

[0019] As a preferred embodiment of the present invention, step 4 further includes the following step: Step 4.1: Improve the overall design of the adaptive NSGA-III optimization algorithm. Based on the NSGA-III framework and combined with the optimization and iteration mechanism for low-temperature scenarios, the process of initializing the population, calculating the constraint penalty, non-dominated sorting, adaptive weight update, crossover mutation, local elite search, and termination judgment is adopted. The solution strategy is adaptively switched for different scenarios such as emergency response, logistics, and commutation, ensuring rapid convergence to the Pareto optimal solution set. Step 4.2, Population initialization, Based on a multi-source dynamic data feature library, basic data such as takeoff and landing point layout, aircraft parameters, pilot information, and real-time demand are extracted to construct a decision variable matrix, which includes decision variables such as takeoff and landing point opening and closing status, time slot allocation scheme, service radius, pilot-takeoff and landing point matching relationship, and aircraft type-takeoff and landing point adaptation scheme. An initial population is randomly generated, and the population size is set to 100-200 according to the number of takeoff and landing points in the planning area to ensure population diversity. At the same time, invalid individuals that do not meet hard constraints are removed to improve the initial solution efficiency. Step 4.3: Constructing a constraint and punishment mechanism. To address the hard constraints, soft constraints, human factor fit, and machine model fit constraints in the collaborative weighted multi-objective optimization model, a hierarchical penalty function is designed to penalize individuals that violate the constraints based on their fitness, ensuring that the optimal solution meets the constraint requirements. Hard constraint penalty: Individuals that violate any hard constraint are given the maximum penalty value and the invalid individual is directly eliminated to uphold the bottom line of scheduling safety; Soft constraint penalties: In routine scenarios (daily logistics, commuting, and other normalized operational scenarios), individuals who violate soft constraints are assigned gradient penalty values ​​according to the degree of violation; in emergency rescue scenarios, the penalty intensity is reduced by relaxing the rules (in emergency scenarios, under the premise of ensuring that the safety bottom line of hard constraints is not breached, the penalty coefficients and compliance thresholds of soft constraints such as efficiency, cost, and noise are appropriately reduced to prioritize the execution of emergency tasks) to achieve flexible adjustment based on the scenario. Adaptation constraint penalty: Individuals that fail to meet the minimum adaptation thresholds for human factors and device models will be subject to the maximum penalty to ensure that core adaptation requirements are met; Step 4.4, Adaptive Adjustment and Local Search, (1) Adaptive adjustment of scenario-based weights. The improved adaptive NSGA-III optimization algorithm can identify the current low-altitude application scenario in real time and automatically match the target weights in the core objective function of the collaborative weighted multi-objective optimization model. When the scenario changes or the requirements change suddenly, the collaborative weighted multi-objective optimization model can automatically recalibrate the weight coefficients and update the core objective function according to the rule of increasing the weight of the core objective and decreasing the weight of the other objectives in proportion, so as to ensure that the optimization direction is in line with the requirements of the scenario. (2) Adaptive crossover mutation strategy, The crossover probability is dynamically adjusted based on the population evolution generation, constraint satisfaction, and objective function convergence. With the probability of mutation ,in, In the early stages of population iteration, increase the crossover probability and mutation probability to expand the search range and avoid getting trapped in local optima. In the later stages of population iteration, reduce the probability of crossover and mutation, accelerate the convergence speed, and accurately approximate the optimal solution; When the constraint satisfaction level is lower than a preset threshold, the mutation probability is increased by a fixed amount within a preset range to generate more individuals that satisfy the constraints; the crossover probability range is... Range of mutation probability values To achieve adaptive dynamic adjustment; (3) Local elite search optimization, For elite individuals after non-dominated sorting, a local search operator is introduced to refine the optimization of take-off and landing time slot allocation, pilot matching, and aircraft type adaptation, thereby improving the practicality of the optimal solution. At the same time, elite individuals are retained to accelerate algorithm convergence and ensure the uniformity and optimality of the Pareto solution set. Step 4.5: Solve for the termination condition using the algorithm. The algorithm terminates its iteration when any of the following conditions are met: It reaches the preset maximum number of iterations to meet the timeliness requirements of low-altitude real-time scheduling; The Pareto optimal solution set remained unchanged after 10 consecutive iterations, and the objective function value converged and stabilized. The generated scheduling scheme fully satisfies all hard constraints, and the compliance rate of soft constraints and adaptation constraints reaches the preset threshold. Finally, multiple Pareto optimal scheduling schemes are output, and the optimal execution scheme is determined by combining the scenario priority.

[0020] As a preferred embodiment of the present invention, step 5 further includes the following step: Step 5.1, Output of the planning scheme, (1) Basic static planning scheme, Landing point layout and configuration: Clearly define the fixed locations, construction standards, and supporting facility parameters of landing points within the planning area, and determine the difficulty level of the landing point sites; Aircraft compatibility list: Defines the types of aircraft compatible with each take-off and landing point, and marks the minimum take-off and landing area, wind resistance threshold, and charging power compatibility parameters; Basic constraint thresholds: solidify hard constraint thresholds to determine the benchmark parameters for soft constraints, time slot pricing, and human factor adaptation under normal operating scenarios (daily logistics, commuting, and other routine operating scenarios); (2) Scenario-based dynamic execution scheme, Take-off and landing point opening and closing scheme: clearly define the open / closed / warning status of each take-off and landing point, and mark the scheduling restriction rules for take-off and landing points in the warning status; Time slot allocation scheme: Plan the time slot scheduling plan for each take-off and landing point, determine the time slot utilization rate and service radius configuration, and avoid airspace conflicts; Demand Coverage Plan: Prioritize demand points according to emergency > logistics > commuting, formulate demand point coverage strategies, and clarify the guarantee mechanism for high-priority demands; Personnel and aircraft matching scheme: Generates an optimal matching list of pilots and takeoff / landing points, and aircraft types and takeoff / landing points, ensuring compatibility meets standards; simultaneously, the intelligent scheduling system for low-altitude takeoff / landing points allocates reserved time slots according to option premiums. The high-value time slots are allocated to drivers with high technical skills and aircraft with high suitability in order to optimize safety during high-risk periods. Cost control plan: Plan the construction and operation cost control targets for take-off and landing sites, and clarify the cost optimization path throughout the entire life cycle; (3) Visual output of the solution, Simultaneously output visual reports of the scheduling plan, including heat maps of take-off and landing point operation status, airspace conflict early warning maps, demand coverage distribution maps, adaptability level tables, and cost-benefit analysis tables, which facilitate intuitive management and execution. Step 5.2, Dynamic adjustment mechanism for scheduling scheme, (1) Real-time data monitoring, Real-time collection of take-off and landing point operation data, aircraft status, pilot physiological and skill data, airspace environment, demand distribution, and cost data; the dataset is updated every 1-5 minutes, and the constraint satisfaction, objective function achievement, and scenario changes are analyzed simultaneously. (2) Grading adjustment rules, Fine-tuning: When demand fluctuations, soft constraint deviations, slot utilization changes fall below a preset low threshold, or option premium changes occur... At that time, only local thresholds are fine-tuned for take-off and landing point service radius, time slot allocation, and aircraft type adaptation, without changing the overall scheduling scheme; Mid-adjustment: When low-altitude application scenarios are switched, or when the option premium is more than twice the historical average for three consecutive update cycles, the target weight is automatically recalibrated, the adaptive NSGA-III optimization algorithm is improved to solve the problem quickly through iteration, and the scenario-based scheduling scheme is updated. Re-adjustment: When extreme weather occurs, hard constraints are breached, emergency demand surges, key take-off and landing points fail, or option premiums drop by more than 50% and are accompanied by hard constraints being breached, global re-planning is immediately triggered, take-off and landing points that do not meet hard constraints are forcibly closed, high-priority needs are prioritized, and emergency dispatch plans are quickly generated. (3) Constraint dynamic adaptation adjustment, The hard constraints must always be kept inviolable. Once a deviation from the hard constraints occurs, the take-off and landing points will be forcibly shut down and relocated immediately. Soft constraints follow the principle of relaxing thresholds according to preset rules in emergency scenarios to ensure that emergency needs are met; Human factors and aircraft compatibility constraints are kept to the minimum threshold, pilot status and aircraft performance are updated in real time, and matching relationships are dynamically adjusted to ensure that the compatibility meets the standards. (4) Adjust the closed-loop process. A closed-loop adjustment process is formed, which includes data perception, deviation analysis, algorithm recalculation, scheme update, implementation, and effect feedback. The adjustment response time is controlled within 4 minutes to meet the needs of real-time, efficient, and safe scheduling of low-altitude take-off and landing points, and to always ensure the optimal synergy of the six objectives.

[0021] The beneficial effects of this invention are as follows: 1. This invention addresses the problems of rigid scheduling and inefficient resource allocation at take-off and landing points by introducing an option pricing model from the financial field to establish a bidding and reservation mechanism for take-off and landing point time slots. This enables market-based dynamic pricing and optimal allocation of time slot resources, replacing the traditional single scheduling rule.

[0022] 2. This invention addresses the problem of neglecting human factors in scheduling by integrating human factors engineering to construct a driver personality profile and status monitoring system. This enables personalized scheduling that adapts take-off and landing points to the driver's skill level, operating habits, and physiological state, thereby improving scheduling safety.

[0023] 3. To address the problem of insufficient dynamic adaptability in take-off and landing point scheduling, this invention establishes a dynamic scheduling mechanism for take-off and landing point time slots and status, enabling rapid response to multiple dynamic factors such as weather, airspace, time slot demand, and pilot status.

[0024] 4. This invention addresses the differentiated needs of different types of aircraft by establishing a scheduling system that adapts the performance parameters of multiple aircraft models to take-off and landing time slots and site conditions, thereby achieving precise matching and scheduling of take-off and landing points for eVTOL, large / small logistics drones, emergency rescue drones, and other aircraft models.

[0025] 5. This invention constructs a scientific multi-objective optimization scheduling model that coordinates and balances six core objectives: safety redundancy, demand coverage, operational efficiency, construction and operation costs, human factors adaptation, and optimal resource allocation, thereby avoiding the imbalance problem of single-objective optimization.

[0026] 6. This invention optimizes the adaptive solution algorithm to adapt to the characteristics of scenarios such as dynamic scheduling of take-off and landing time slots, personalized matching of human factors and market-based pricing, thereby improving the solution efficiency and global optimality of the scheduling scheme. Attached Figure Description

[0027] Figure 1 This is the core flowchart of the present invention; Figure 2 This is a diagram showing the relationship between the option premium and the objective function in this invention. Detailed Implementation

[0028] This embodiment provides an adaptive and dynamically adjusted intelligent scheduling method for low-altitude aircraft takeoff and landing points, comprising the following steps: Step 1: Multi-source dynamic data perception and feature library construction. Construct a multi-dimensional perception network, collect multi-source dynamic data, and establish feature libraries for multiple types of aircraft and pilots. Complete the construction of the multi-source dynamic data feature library to provide data support for subsequent pricing, adaptation and scheduling. Step 1.1: Deployment of perception nodes and definition of data types. Construct a multi-dimensional perception network and deploy six types of perception nodes: airspace, environment, demand, facilities, aircraft status, and pilots. Collect static geographic data, dynamic operational data, facility status data, real-time data of various types of aircraft, and pilot status data. ① Airspace awareness nodes: Based on miniature radar and ADS-B (Automatic Dependent Surveillance-Broadcast) equipment, the deployment density is one per 10km² on average, which monitors airspace occupancy rate, real-time aircraft position, temporary no-fly zone boundaries, and dynamic route adjustment information in real time; ② Environmental sensing nodes: integrate meteorological sensors (wind speed, visibility, precipitation, temperature) and terrain monitoring equipment, with an average deployment density of 1 per 5km², focusing on collecting key meteorological parameters and terrain deformation data that affect aircraft take-off and landing; ③ Demand perception node: Real-time acquisition of demand point coordinates, demand intensity, and demand priority (emergency > logistics > commuting) through logistics platform API interface, end-user location data, and emergency command center instruction interface. ④ Facility sensing nodes: Deployed around existing take-off and landing points, power supply stations, and communication base stations to monitor the dynamic status of infrastructure such as site occupancy, power supply load, communication signal strength, and the number of available charging piles.

[0029] ⑤ Aircraft Status Nodes: Aircraft status perception nodes are divided into fixed take-off and landing point types and airborne mobile types. They collaboratively collect core data such as aircraft identity, performance, take-off / landing / flight status, and transmit and verify the data with low latency.

[0030] ⑥ Driver perception node: Collect driver's operation data, physiological data (heart rate, fatigue level) and environmental interaction data through wearable devices and driver-end monitoring devices, and transmit and verify them with low latency; Step 1.2 Data Acquisition and Preprocessing Specifications Step 1.2.1, Data acquisition frequency, ① High-frequency dynamic data (weather, airspace occupancy rate, facility occupancy rate, physiological data of pilots or controllers): The collection cycle is 3 minutes to ensure real-time performance; ②Medium frequency dynamic data (demand distribution, communication signal strength): acquisition cycle of 15 minutes, balancing real-time performance and computational cost; Step 1.2.2, Data Preprocessing Methods ①Outlier removal: using 3 The criteria (three standard deviation criteria, Laida criteria) are used to remove outliers from meteorological and signal strength data (such as data where the instantaneous change in wind speed exceeds twice the maximum wind resistance threshold of the aircraft). ② Missing value imputation: Linear interpolation is used to imput short-term missing data, and LSTM (Long Short-Term Memory)-based time series prediction model is used to imput long-term missing data; ③ Data standardization: Normalize data of different dimensions (e.g., standardize communication signal strength [-100dBm~-50dBm] to the [0,1] interval); the formula is as follows: ; In the formula, This is the result after data standardization; This is the original data; 、 These are the theoretical minimum and maximum values ​​for this type of data, respectively. Step 1.3: Construction of a feature library for multiple types of aircraft. Establish a feature library for mainstream aircraft types, including logistics drones, eVTOL (passenger / freight) drones, emergency rescue drones, police drones, and manned aircraft, and extract core performance parameters for each type: maximum wingspan, maximum takeoff weight, maximum wind resistance threshold, range, minimum takeoff and landing area, communication signal requirements, and charging power, to provide a basis for subsequent constraint definition and adaptive scheduling.

[0031] Step 2: Definition of scheduling constraint system and scientific calculation of parameters. Based on low-altitude operation safety requirements, performance characteristics of various types of aircraft, and human factors characteristics of pilots, a complete constraint system is constructed, consisting of hard constraints, soft constraints, time slot pricing constraints, human factor adaptation constraints, and dynamic opening and closing constraints of take-off and landing points. The parameters of the complete constraint system adopt a scenario-based calculation method. For those without calculation conditions, the current regulations or technical manuals are referenced. Adaptation constraints for various types of aircraft and take-off and landing points are added. Principles for calculating constraint parameters: For different scenarios (core urban area / suburbs / remote suburbs, logistics / emergency / commuting), parameter values ​​are calculated based on data such as regional geographical characteristics, proportion of aircraft types, and demand density, through regression analysis / threshold derivation. Step 2.1, Hard constraints (must be met, no compromise). Step 2.1.1, Safety distance constraints, (1) Safety distance constraints between take-off and landing points and fixed obstacles, ; In the formula, For the first Real-time horizontal distance between each take-off and landing point and the nearest fixed obstacle (building, high-voltage line); The vertical height of the obstacle relative to the ground level at the take-off and landing point; This is the standard glide slope angle for aircraft (usually taken as 3° to 5°). For obstacle safety factors; slender and high-risk obstacles such as high-voltage lines and communication towers: =2.0~2.5; large fixed obstacles such as high-rise buildings and mountains: =1.5~2.0; Low-risk obstacles such as ordinary trees and low-rise structures: =1.2~1.5; This represents the maximum wingspan of the aircraft to be scheduled; (2) Minimum horizontal distance constraint between the take-off and landing point and the moving obstacle. ; ; In the formula, For the first The real-time minimum horizontal distance between each take-off and landing point and an aircraft operating within the airspace; The safe distance to be maintained when encountering different moving obstacles at the take-off and landing point; This is the minimum safe interval for the aircraft model; It is the relative velocity; For driver / automatic system reaction time; The distance required for an aircraft to brake in an emergency. Step 2.1.2, Spatial Constraints, (1) Distance constraints between take-off and landing points and the boundary of controlled airspace, ; ; In the formula, For the first The distance between each take-off and landing point and the boundary of the controlled airspace; Distances that should be maintained between takeoff and landing points and the boundaries of different types of controlled airspace; restricted areas and danger zones: ≥2000m; Controlled Restricted Area, Airport Control Zone: ≥1000m; airspace control sector, low-altitude control corridor: ≥500m; Horizontal distance between the boundaries of reported airspace and uncontrolled airspace: ≥200m; This refers to the vertical height difference between the takeoff and landing point and the boundary of the controlled airspace; The height of the take-off and landing point ground level; The minimum altitude at the boundary of the controlled airspace; Vertical safety separation for controlled airspace; restricted and danger zones: ≥300m; Controlled Restricted Area, Airport Control Zone: ≥150m; airspace control sector, low-altitude control corridor: ≥100m; Vertical safety separation at the boundaries of reported airspace and uncontrolled airspace: ≥50m; (2) Distance constraints between takeoff and landing points and the centerline of the flight path, ; In the formula, For the first Distance between each takeoff and landing point and the centerline of the flight path; The route safety factor is set between 1.5 and 3.0, and is determined by considering factors such as aircraft speed, navigation accuracy, and airspace density. Half-width of the aircraft's flight path protection zone; small unmanned aerial vehicles: =15~25m; Medium-sized logistics drones: =25~40m; eVTOL / Manned Spacecraft: =40~60m; Step 2.1.3, Meteorological constraints, (1) Wind speed constraint, ; In the formula, Wind speed; The wind speed threshold; (2) Visibility constraints, ; Visibility; (3) Precipitation constraints, ; This represents the hourly rainfall. Step 2.1.4, Facility Constraints (1) Supply voltage fluctuation constraint, ; In the formula, Real-time voltage of the power supply system at the take-off and landing points; The rated voltage of the power supply system at the take-off and landing points; For general thresholds, =5%; (2) Communication signal strength constraints, ; The real-time wireless signal strength between the take-off and landing point and the aircraft and control center; (3) Minimum available area constraint, ; In the formula, This refers to the actual usable takeoff and landing area at the takeoff and landing point; This is a general density coefficient; ; This refers to the maximum takeoff weight of the aircraft intended for takeoff and landing. Step 2.1.5, Human Factors-Based Adaptation Constraints (1) Fatigue state constraints, ; In the formula, Real-time driver fatigue level; The fatigue threshold; (2) Consecutive operation time constraints, ; In the formula, This refers to the driver's continuous working hours for the day; The maximum continuous operating time is as stipulated by regulations / standards (the pilot's continuous flight time shall not exceed 4 hours; the cumulative flight time per month shall not exceed 100 hours). (3) Safety constraints on physiological indicators, ; In the formula, Real-time heart rate; The minimum safe heart rate (generally 60 beats / minute); The maximum safe heart rate (generally 100 beats / minute); (4) No-fly zone restrictions, ; This means that the pilot cannot be in a state of temporary no-fly zone, abnormal health condition, or expired qualification that prohibits operation; Step 2.2, Soft constraints (prioritize meeting them, but can be appropriately relaxed in emergency scenarios). Step 2.2.1, Efficiency Constraints (1) Service radius constraints of take-off and landing points, ; In the formula, Provides a real-time scheduling service radius for take-off and landing points; This is to determine the real-time remaining range of the aircraft to be scheduled; (2) Average utilization rate constraint of take-off and landing time slots, ; In the formula, For the first Average utilization rate of time slots at each take-off and landing point in nearly one hour; For the first The first take-off and landing point Real-time occupancy status of each time slot (1 = occupied, 0 = idle); This represents the total number of time slots in approximately one hour. (3) Average distance constraint from the demand point to the nearest take-off and landing point. ; In the formula, The average distance from newly added demand points to the most recently activated take-off and landing points; To record the number of new demand points in real time; For the first The distance from each newly added demand point in real time to the most recently activated take-off and landing point; Step 2.2.2, Cost Constraints (1) Real-time operating cost adaptation constraints at take-off and landing points. ; In the formula, For the first Real-time operating cost per take-off and landing point per time slot; Upper limit on operating costs for order slots in different scenarios; (2) Constraints on real-time connectivity between take-off and landing points and existing infrastructure. ; In the formula, For the first The horizontal distance between each take-off and landing point and the real-time operating substation / communication base station; Step 2.2.3, Sustainability Constraints (1) Noise emission constraints at take-off and landing points, ; In the formula, The real-time equivalent continuous A-weighted sound level within 100 meters of the take-off and landing point; (2) Real-time avoidance constraints in the buffer zone of the ecological protection area. ; In the formula, For the first Real-time distance between each take-off and landing point and the boundary of the ecological protection zone buffer zone; The width of the buffer zone for the ecological protection area; Step 2.2.4, Time Slot Utilization Constraints ; In the formula, For the first The first take-off and landing point Real-time prediction of utilization rate for each time slot; These are the scheduled takeoff and landing times within the time slot; The duration of the time slot; Step 2.3, Time Slot Pricing Constraints Based on the Black-Scholes (BS) option pricing model, combined with the real-time non-renewability of take-off and landing slots and the dynamic optimization of scenarios, the parameter boundary values ​​refer to the pricing norms of financial futures options and the real-time scarcity characteristics of low-altitude slot resources, which are the core constraints of market-based scheduling. The low-altitude take-off and landing intelligent scheduling system needs to update pricing parameters every minute and dynamically adjust option prices and scheduling rules. Step 2.3.1, Constraints of the core formula for BS option pricing. This model transforms traditional fixed scheduling into market-based dynamic scheduling, achieved through time slot options: 1) High-value time slots are locked in advance to avoid no slots available during peak periods; 2) Prices reflect real-time scarcity, guiding demand to be staggered; 3) Supports minute-level dynamic price adjustments to adapt to multi-scale changes in weather, airspace, and demand; 4) Linked with adaptive optimization algorithms to achieve a closed loop of pricing, allocation, and adjustment. ; ; ; In the formula, the parameters for the low-altitude scene are defined as follows: The real-time premium for takeoff and landing slot options is the fee that users need to pay to reserve a slot. for The real-time value of a time slot is determined by location, time period, airspace status, and demand intensity. As an intermediate variable, it reflects the "relative strength" of the underlying asset price relative to the strike price, and takes into account the effects of time, volatility and risk-free interest rate; is an intermediate variable representing the adjusted probability of the option being exercised; The option exercise price is the final cost paid when the slot is actually used. This is the current risk-free interest rate, used for value discounting; The remaining time to expiration of the option is equal to the planned takeoff and landing time minus the current time; The slot value volatility reflects price fluctuations caused by low-altitude supply and demand and environmental changes. The cumulative distribution function is a standard normal distribution, representing the probability of time slot usage and the expected risk. Step 2.3.2, Time Slot Value Volatility Constraints ; Step 2.3.3, Option expiration time constraint, ; In the formula, This refers to the real-time exercise time of the option; The planned takeoff time for the aircraft is updated in real time. The exercise time window; Step 2.3.4, Time-slot option price volatility constraints, ; In the formula, for Time slot option price; for Time slot option price; Step 2.4 Human Factors Adaptation Constraints The core innovation of this scheduling patent is the human factors adaptation constraint. Based on the aviation human factors engineering SHEL model and the real-time operation big data of domestic UAV operators / pilots, the parameter boundary values ​​refer to the "Evaluation Standard of Human Factors Engineering for Aviation Pilots". Combined with the optimization of real-time operation characteristics of low-altitude driving, the intelligent scheduling system for low-altitude take-off and landing points collects pilot data in real time to achieve dynamic and personalized matching between pilots and take-off and landing point time slots. Step 2.4.1: Real-time adaptation constraints between the driver and the takeoff and landing points. ; ( and ); In the formula, Real-time compatibility between the driver and the takeoff and landing points; This represents the real-time matching degree between the driver's skill level and the difficulty of the take-off and landing points, with a value of [0,1]. This represents the real-time matching degree between the driver's operating habits and the take-off and landing site, with a value of [0,1]. This represents the real-time matching degree between the driver's physiological state and the operational requirements at the take-off and landing points, with a value of [0,1]. This is the weighting coefficient for the degree of technical matching. The weighting coefficient for the degree of matching with operating habits; The physiological state matching degree weighting coefficient; Step 2.4.2, Real-time matching degree constraint of operating habits. ; In the formula, The coefficient representing the driver's real-time operating habits; Real-time adaptation of operating parameters to the take-off and landing site; Pick ~ To avoid the denominator being 0; When the calculation result is less than 0, it is directly set to 0 to ensure that the matching degree is not negative. The closer the matching degree is to 1, the higher the matching degree. Step 2.4.3: Real-time matching constraint between take-off and landing point difficulty and driver skill level. ; In the formula, Real-time difficulty level of take-off and landing points (1-5, 1=easiest, 5=most complex). The driver's real-time skill level (1-5, 1=novice, 5=expert). Step 2.4.4: Real-time matching degree constraint of operation level. ; Step 2.4.5: Real-time increase of safety redundancy constraints under fatigue conditions. ; In the formula, The first under fatigue state Real-time safety redundancy standards for individual drivers; This serves as a standard for routine safety redundancy. To increase the coefficient; Real-time driver fatigue level; Step 2.4.6: Real-time physiological state matching degree constraint. ; In the formula, This represents the upper limit threshold for fatigue. Step 2.5, Option Premium The linkage mechanism with scheduling decisions, Option premium In addition to being used as a time slot reservation fee, it also participates in scheduling decisions through the following rules: Demand feedback linkage: The intelligent scheduling system for low-altitude take-off and landing points will calculate the option premium in real time. The information is presented to users through the presentation layer (step 6); users decide whether to book based on their budget and the urgency of the task; when the option premium... When the user's preset threshold is exceeded, the user will cancel or postpone the reservation. The low-altitude take-off and landing point intelligent scheduling system collects behavioral changes in real time through the demand perception node (step 1.1), updates the demand distribution data, and thus indirectly affects the demand coverage target. (Demand Coverage) and Operational Efficiency Goals (Operating efficiency); Dynamic adjustment trigger: The intelligent scheduling system for low-altitude take-off and landing points monitors option premiums. short-term rate of change Its deviation from the historical average; according to the dynamic adjustment mechanism of the scheduling plan (step 5.2), the short-term rate of change. The value determines the level of adjustment to be triggered (see step 5.2 for details); where, ; Execution Priority Ranking: When generating the planning scheme output (dynamic execution scheme) in step 5.1, the low-altitude take-off and landing point intelligent scheduling system will allocate the reserved time slots according to the option premium. Sorted from highest to lowest value; high-value time slots are allocated preferentially to technology levels. Higher-level drivers (meeting) and device compatibility High-altitude aircraft to reduce operational risks and ensure the safe execution of high-value missions; Operating cost adjustment: The operator periodically adjusts operating costs based on historical option premiums. The accumulated revenue will be used to dynamically adjust the upper limit of operating costs for order slots in different scenarios within the soft constraint. (Step 2.2). For example, when the average option premium over 30 consecutive days... When the value exceeds the baseline by 20%, the system may suggest or automatically adjust the settings. An increase of 5% allows for increased investment in facility maintenance; this adjustment can be performed manually or automatically via the human-machine interaction module in step 6. The aforementioned linkage mechanism will not include option premiums. Instead of using it as a direct variable in multi-objective optimization, the market signals are transmitted to scheduling decisions through data closure, rule triggering, and parameter adjustment, thus maintaining the simplicity of the optimization model.

[0032] Step 3: Constructing a multi-objective optimization model. This step uses the analytic hierarchy process (AHP) to determine the basic weights and constructs a collaborative weighted multi-objective optimization model that includes safety redundancy, operational efficiency, demand coverage, human factor adaptability, aircraft type adaptability, and cost optimization objectives. Combined with a complete constraint system, the option pricing mechanism is deeply coupled with the scheduling objectives to achieve multi-objective collaborative optimal matching of scheduling schemes under different low-altitude application scenarios. This provides a standardized mathematical model (six-objective collaborative weighted multi-objective optimization model) for subsequent adaptive optimization algorithm solutions. Step 3.1: Define the objective function (based on the analytic hierarchy process to determine weights). With maximizing the overall scheduling benefits as the core, a multi-objective weighted sum objective function is constructed. The weight coefficients are dynamically adjusted for different low-altitude application scenarios (emergency rescue / urban logistics / short-distance commuting), and the basic weights are determined by the analytic hierarchy process. Ensuring safety is the primary objective, human factors and aircraft model compatibility are the core adaptation objectives, and efficiency, coverage, and cost are the collaborative optimization objectives. The core objective function of the multi-objective optimization model : ; In the formula, For safety redundancy targets; For operational efficiency targets; To cover the target of demand; Human factors fit target; For model compatibility targets; To optimize costs; The target weighting coefficient for safety redundancy; Weighting coefficients for operational efficiency targets; Determine the target weighting coefficient for demand coverage; Human factors fit target weight coefficient; The target weighting coefficient for device compatibility; Optimize the target weight coefficients for cost.

[0033] Step 3.1.1, Security Redundancy Target , ; In the formula, ; The number of take-off and landing points within the planned area; For the first The actual distance between each take-off and landing point and the nearest obstacle; The hard constraint threshold for obstacle safety distance; This represents the ideal safe distance from obstacles. The probability of airspace conflict (calculated based on airspace occupancy rate and takeoff and landing point distribution density). , The number of aircraft in the region in real time. The total airspace area serving the take-off and landing points. (Total airspace area of ​​the planned region). The target weighting coefficient for safety redundancy; Step 3.1.2, Operational Efficiency Target , ; In the formula, For the first The average utilization rate of each take-off and landing point (calculated based on historical data and demand forecasts); 85% is the recommended upper limit for the average utilization rate of take-off and landing point time slots (to avoid excessive occupation leading to scheduling congestion and safety risks); 30% is the minimum reasonable utilization rate of take-off and landing points (to avoid resource waste). For the first Real-time dispatch service radius for each take-off and landing point; The maximum permissible service radius for take-off and landing points; Minimum necessary service radius for take-off and landing points; Weighting coefficients for operational efficiency targets; Step 3.1.3, Requirements Coverage Objectives , ; In the formula, The number of demand points (the number of newly added demand points in real time); For the first Demand weights for each demand point (emergency demand = 3, logistics demand = 2, commuting demand = 1). For the first The distance from each demand point to the nearest take-off and landing point; Service radius of take-off and landing points; This is an indicator function that takes the value 1 if the condition is met, and 0 otherwise. Determine the target weighting coefficient for demand coverage; Step 3.1.4, Human Factors Fit Goals , ; In the formula, For the first The driver and the first Real-time human factor adaptation at each take-off and landing point; The total number of drivers participating in dispatch within the planned area; For the first The difficulty level of each take-off and landing point (level 1-5); For the first The technical level of each driver (level 1 to 5); This is an indicator function; it takes the value 1 if the condition is met, and 0 otherwise. Human factors fit target weight coefficient; Step 3.1.5, Model Compatibility Target , ; In the formula, The number of aircraft types that need to be adapted within the planned area; For the first The scene weight of aircraft-like vehicles is determined by low-altitude application scenarios; For model compatibility, and The value range is [0,1]; For the first Minimum compatibility rate for similar aircraft; This is an indicator function; it takes a value of 1 if the target is met and 0 if the target is not met. The target weighting coefficient for device compatibility; Step 3.1.6, Cost Optimization Objective , ; In the formula, The number of take-off and landing points within the planned area; For the first Real-time operating cost per take-off and landing point per time slot; This represents the upper limit (constraint threshold) of operating costs per time slot. Optimize the target weighting coefficients for cost; In summary, through the construction of a six-objective multi-objective optimization model, the option premium... As a core market-based regulatory variable, it deeply couples core objectives such as demand coverage, cost optimization, and human factor adaptation through a multi-path transmission mechanism to achieve dynamic adaptive optimization of the scheduling scheme. The correlation and influence logic between option premiums and the multi-objective optimization model is as follows: Figure 2 As shown.

[0034] Step 3.2, Mathematical expression of constraints, Step 3.2.1, Hard Constraints ; Step 3.2.2, Soft Constraints ; Step 3.2.3: Dynamic opening and closing constraints of take-off and landing points. ; In the formula, For the first Status of each takeoff and landing point; This refers to compliance with hard constraints; For soft constraint compliance rate; Step 3.2.4, Human Factors Adaptation Core Constraints, .

[0035] Step 4: Solve using the adaptive optimization algorithm. This step addresses the six-objective collaborative weighted multi-objective optimization model constructed in step 3. Combining the dynamic, multi-constraint, and scenario-differentiated characteristics of low-altitude take-off and landing point scheduling, an improved adaptive NSGA-III optimization algorithm is designed. This algorithm incorporates adaptive adjustment of scenario weights, constraint penalty mechanisms, and local optimization strategies to solve problems such as multi-objective conflicts, dynamic data iteration, and constraint boundary adaptation. This enables efficient solution of the collaborative weighted multi-objective optimization model and output of the optimal scheduling scheme, ensuring that the solution speed and global optimality are adapted to the real-time scheduling requirements of low-altitude areas. Step 4.1: Improve the overall design of the adaptive NSGA-III optimization algorithm. Based on the NSGA-III (third-generation non-dominated sorting genetic algorithm) framework, combined with the optimization iteration mechanism for low-temperature scenarios, the algorithm adopts the process of initializing the population, calculating the constraint penalty, non-dominated sorting, adaptive weight update, crossover and mutation, local elite search, and termination judgment. It takes into account the core requirements of hard constraints that cannot be broken, soft constraints that are dynamically adjusted, and multi-objective weights that are adapted to different scenarios. It realizes adaptive switching of solution strategies for different scenarios such as emergency response, logistics, and commuting, and ensures rapid convergence to the Pareto optimal solution set. Step 4.2, Population initialization, Based on the multi-source dynamic data feature library from step 1, basic data on takeoff and landing point layout, aircraft parameters, pilot information, and real-time demand are extracted to construct a decision variable matrix, which includes decision variables for takeoff and landing point opening and closing status, time slot allocation scheme, service radius, pilot-takeoff and landing point matching relationship, and aircraft type-takeoff and landing point adaptation scheme. An initial population is randomly generated, and the population size is set to 100-200 according to the number of takeoff and landing points in the planning area to ensure population diversity. At the same time, invalid individuals that do not meet hard constraints are removed to improve the initial solution efficiency. Step 4.3: Constructing a constraint and punishment mechanism. To address the hard constraints, soft constraints, human factor fit, and machine model fit constraints in the collaborative weighted multi-objective optimization model, a hierarchical penalty function is designed to penalize individuals that violate the constraints based on their fitness, ensuring that the optimal solution meets the constraint requirements. Hard constraint penalty: Individuals that violate any hard constraint are given the maximum penalty value and the invalid individual is directly eliminated to uphold the bottom line of scheduling safety; Soft constraint penalties: In routine scenarios (daily logistics, commuting, and other normalized operational scenarios), individuals who violate soft constraints are assigned gradient penalty values ​​according to the degree of violation; in emergency rescue scenarios, the penalty intensity is reduced by relaxing the rules (in emergency scenarios, under the premise of ensuring that the safety bottom line of hard constraints is not breached, the penalty coefficients and compliance thresholds of soft constraints such as efficiency, cost, and noise are appropriately reduced to prioritize the execution of emergency tasks) to achieve flexible adjustment based on the scenario. Adaptation constraint penalty: Individuals that fail to meet the minimum adaptation thresholds for human factors and device models will be subject to the maximum penalty to ensure that core adaptation requirements are met; penalty function : ; In the formula, The penalty coefficient is... ; These are the violation rates of hard constraints, soft constraints, and adaptive constraints, respectively. Step 4.4, Adaptive Adjustment and Local Search, (1) Adaptive adjustment of scenario-based weights. The improved adaptive NSGA-III optimization algorithm identifies the current low-altitude application scenario (emergency / logistics / commuting) in real time and automatically matches the target weights in the core objective function of the collaborative weighted multi-objective optimization model in step 3.1 without manual intervention. When the scenario changes or the demand changes suddenly, the collaborative weighted multi-objective optimization model automatically recalibrates the weight coefficients and updates the core objective function according to the rule of increasing the weight of the core objective and decreasing the weight of the other objectives proportionally, so as to ensure that the optimization direction is in line with the needs of the scenario. Adaptive crossover mutation strategy The crossover probability is dynamically adjusted based on the population evolution generation, constraint satisfaction, and objective function convergence. With the probability of mutation ,in, In the early stages of population iteration, increase the crossover probability and moderately increase the mutation probability to expand the search range and avoid getting trapped in local optima; In the later stages of population iteration, reduce the probability of crossover and mutation, accelerate the convergence speed, and accurately approximate the optimal solution; When the constraint satisfaction level is lower than a preset threshold, the mutation probability is increased by a fixed amount within a preset range to generate more individuals that satisfy the constraints; the crossover probability range is... Range of mutation probability values To achieve adaptive dynamic adjustment; (3) Local elite search optimization, For elite individuals after non-dominated sorting, a local search operator is introduced to refine the optimization of key decision variables such as take-off and landing time slot allocation, pilot matching, and aircraft type adaptation, thereby improving the practicality of the optimal solution. At the same time, elite individuals are retained to accelerate algorithm convergence and ensure the uniformity and optimality of the Pareto solution set. Step 4.5: Solve for the termination condition using the algorithm. The algorithm terminates its iteration when any of the following conditions are met: It reaches the preset maximum number of iterations (set to 50-100 generations) to meet the timeliness requirements of low-altitude real-time scheduling; The Pareto optimal solution set did not change significantly after 10 consecutive iterations, and the objective function value converged and stabilized. The generated scheduling scheme fully satisfies all hard constraints, and the compliance rate of soft constraints and adaptation constraints reaches the preset threshold. Finally, multiple Pareto optimal scheduling schemes are output, and the optimal execution scheme is determined by combining the scenario priority.

[0036] Step 5: Output and Dynamic Adjustment of Planning Scheme This step, based on the results of the adaptive optimization algorithm, outputs a standardized and implementable take-off and landing point scheduling plan. At the same time, it constructs a closed-loop mechanism for real-time perception, dynamic analysis, and iterative adjustment. Combining multi-source dynamic data, scene changes, and constraint fluctuations, it makes real-time corrections to the take-off and landing point scheduling plan, realizing "static planning, dynamic operation, and full-process adaptation" of low-altitude take-off and landing points.

[0037] Step 5.1, Output of the planning scheme, After the algorithm is solved, it outputs a basic static planning scheme and a scenario-based dynamic execution scheme, covering all scheduling elements. The specific content is as follows: Basic static planning scheme, Landing point layout and configuration: Clearly define the fixed locations, construction standards, and supporting facility parameters of landing points within the planning area, and determine the difficulty level of the landing point sites; Aircraft compatibility list: Defines the types of aircraft compatible with each take-off and landing point, and marks the minimum take-off and landing area, wind resistance threshold, and charging power compatibility parameters; Basic constraint thresholds: solidify hard constraint thresholds to determine the benchmark parameters for soft constraints, time slot pricing, and human factor adaptation under normal operating scenarios (daily logistics, commuting, and other routine operating scenarios); Dynamic execution scheme for the scenario Take-off and landing point opening and closing scheme: clearly define the open / closed / warning status of each take-off and landing point, and mark the scheduling restriction rules for take-off and landing points in the warning status; Time slot allocation scheme: Plan the time slot scheduling plan for each take-off and landing point, determine the time slot utilization rate and service radius configuration, and avoid airspace conflicts; Demand Coverage Plan: Prioritize demand points according to emergency > logistics > commuting, formulate demand point coverage strategies, and clarify the guarantee mechanism for high-priority demands; Personnel and aircraft matching scheme: Generates an optimal matching list of pilots and takeoff / landing points, and aircraft types and takeoff / landing points, ensuring compatibility meets standards; simultaneously, the intelligent scheduling system for low-altitude takeoff / landing points allocates reserved time slots according to option premiums. Sorted from highest to lowest, drivers with higher technical skill levels are allocated higher-value time slots (those who meet the requirements). (and more compatible models) to optimize security during high-risk periods; Cost control plan: Plan the construction and operation cost control targets for take-off and landing sites, and clarify the cost optimization path throughout the entire life cycle; Visual output of the solution Simultaneously output visual reports of the scheduling plan, including heat maps of take-off and landing point operation status, airspace conflict early warning maps, demand coverage distribution maps, adaptability level tables, and cost-benefit analysis tables, which facilitate intuitive management and execution. Step 5.2, Dynamic adjustment mechanism for scheduling scheme, Based on the real-time acquisition of multi-source dynamic data in step 1, and combined with the fluctuation of constraints, scene switching, and sudden changes in demand, a hierarchical dynamic adjustment mechanism is constructed to ensure that the scheduling scheme always fits the actual operation scenario and achieves global optimization. (1) Real-time data monitoring, Real-time collection of take-off and landing point operation data, aircraft status, pilot physiological and skill data, airspace environment, demand distribution, and cost data; the dataset is updated every 1-5 minutes, and the constraint satisfaction, objective function achievement, and scenario changes are analyzed simultaneously. Tiered adjustment rules, Fine-tuning (normal scenarios): When demand fluctuations, soft constraint deviations, slot utilization changes fall below a preset low threshold, or option premium changes... At this time, only local threshold fine-tuning (setting fine-tuning thresholds) is performed on the service radius of take-off and landing points, time slot allocation, and aircraft type adaptation, without changing the overall scheduling scheme; Mid-adjustment (scenario switching): When the low-altitude application scenario switches (emergency - logistics - commuting), or the option premium is more than twice the historical average for three consecutive update cycles, the target weight is automatically recalibrated, the adaptive NSGA-III optimization algorithm is improved to solve the problem quickly and update the scenario-based scheduling scheme. Re-adjustment (extreme / sudden scenarios): When extreme weather, hard constraint breach, surge in emergency demand, failure of key take-off and landing points, or option premium drops by more than 50% accompanied by hard constraint breach, global re-planning is immediately triggered, take-off and landing points that do not meet hard constraints are forcibly closed, high priority needs are prioritized, and emergency dispatch plans are quickly generated. (3) Constraint dynamic adaptation adjustment, The hard constraints must always be kept inviolable. Once a deviation from the hard constraints occurs, the take-off and landing points will be forcibly shut down and relocated immediately. The soft constraints follow the principle of "strict enforcement in routine situations and appropriate relaxation in emergencies." In emergency scenarios, the thresholds are slightly relaxed according to preset rules to ensure that emergency needs are met. Human factors and aircraft compatibility constraints are kept to the minimum threshold, pilot status and aircraft performance are updated in real time, and matching relationships are dynamically adjusted to ensure that the compatibility meets the standards. (4) Adjust the closed-loop process. A closed-loop adjustment process is formed, which includes data perception, deviation analysis, algorithm recalculation, scheme update, implementation, and effect feedback. The adjustment response time is controlled within 4 minutes to meet the needs of real-time, efficient, and safe scheduling of low-altitude take-off and landing points, and to always ensure the optimal synergy of the six objectives.

[0038] Step 6: Construction of an intelligent scheduling system for low-altitude take-off and landing points. This step, based on a multi-source dynamic data feature library, a complete constraint system, a collaborative weighted multi-objective optimization model (six-objective optimization model), an adaptive optimization algorithm, and a dynamic adjustment mechanism, builds a layered architecture, modular collaboration, real-time interaction, and scenario-adaptive intelligent scheduling system for low-altitude take-off and landing points. It achieves integrated operation of the entire process, including data acquisition, model calculation, scheme generation, scheduling execution, and dynamic adjustment, while taking into account system stability, real-time performance, and scalability. It supports efficient, safe, and orderly scheduling of low-altitude take-off and landing points in multiple scenarios such as emergency response, logistics, and commuting.

[0039] An adaptive and dynamically adjustable intelligent scheduling system for low-altitude aircraft takeoff and landing points includes: (1) Construct a data perception module, including: perception layer, data layer, model layer, application layer, and presentation layer; Perception layer: The system data acquisition terminal is used for real-time perception and acquisition of multi-source dynamic data. The multi-source dynamic data includes data from take-off and landing point operation sensors, aircraft airborne equipment, environmental monitoring devices, user demand terminals, and human factors monitoring devices, and is the data source for the entire system. Data Layer: Build a multi-source dynamic data feature library and data management platform to complete data cleaning, fusion, storage, and standardization processing, and construct a scheduling database, constraint parameter library, scenario knowledge base, and model weight library to provide high-quality data support for upper-layer model operations; Model layer: Integrating a complete constraint system (full-dimensional scheduling constraint system), a collaborative weighted multi-objective optimization model (six-objective optimization model), an adaptive optimization algorithm, and a dynamic adjustment mechanism, it completes objective function calculation, constraint verification, optimal solution solving, and scheduling decision generation, and is the core computing unit of the system; Application Layer: Develop scenario-based scheduling applications to meet actual scheduling needs, enabling functions such as take-off and landing point opening and closing control, time slot allocation, demand matching, personnel and aircraft model adaptation, cost control, and emergency scheduling. It also supports model layer decisions and implements them. Presentation layer: Build a visual management and control platform to intuitively present the system's operating status, scheduling plan, constraint compliance, airspace situation, and demand distribution information. It supports manual intervention, parameter calibration, and command issuance to achieve human-machine collaborative management and control. (2) Construct the core functional modules of the system, including: multi-source data fusion management module, constraint control module, multi-objective optimization solution module, scheduling scheme execution and control module, dynamic adaptive adjustment module, and visualization and human-computer interaction module; Multi-source data fusion management module: Integrates various types of data collected from the perception layer, including airspace environment, takeoff and landing point status, aircraft parameters, pilot information, real-time requirements, and cost data. Through data cleaning, noise reduction, normalization, and spatiotemporal alignment, it eliminates data redundancy and conflicts; establishes a real-time data update and caching mechanism, refreshing dynamic data at a frequency of 1-5 minutes to ensure data timeliness and accuracy; and constructs a data access control system to achieve secure data storage, rapid retrieval, and historical traceability, providing stable data support for model solving. Constraint Management Module: This module solidifies the scheduling constraint system from Step 2 and is divided into four sub-modules: hard constraint management, soft constraint management, human factors and aircraft model adaptation constraints, and time slot pricing constraints. It enables standardized configuration and real-time verification of constraint parameters; supports visual modification and scenario-based preset of constraint thresholds; allows one-click switching of soft constraint relaxation parameters in emergency scenarios; and monitors the compliance status of constraints at each take-off and landing point in real time. It immediately issues warnings for situations such as hard constraint breaches or failure to meet adaptation constraints, triggering scheduling adjustment instructions to adhere to the bottom-line requirements of scheduling. Multi-objective optimization solution module: Equipped with the collaborative weighted multi-objective optimization model (six-objective collaborative optimization model) in step 3 and the improved adaptive NSGA-III optimization algorithm in step 4, it integrates scene recognition, adaptive weight adjustment, constraint penalty calculation, population iterative solution, and optimal solution selection functions; it automatically identifies the current low-altitude application scenario, matches the corresponding target weights, and quickly completes the model solution; it supports manual intervention in target weights and solution parameters, taking into account both automatic algorithm calculation and manual decision optimization, and outputs the Pareto optimal scheduling scheme that meets the actual operation requirements, with the solution response time controlled within 30 seconds, adapting to real-time scheduling requirements; The scheduling scheme execution and control module receives the scheduling schemes output by the optimization solution module, and realizes the decomposition, distribution, and execution control of the schemes. This includes functions such as dynamic opening and closing of take-off and landing points, intelligent time slot allocation, demand priority scheduling, driver-to-take-off and landing point matching, and aircraft-to-take-off and landing point adaptation. It automatically allocates scheduling resources according to the demand priority of emergency > logistics > commuting, and implements high-priority demand flow restriction control for take-off and landing points in early warning status. It synchronously records scheduling execution data and statistically analyzes indicators such as target achievement rate, constraint satisfaction rate, operational efficiency, and cost consumption, providing a basis for scheme optimization. Dynamic adaptive adjustment module: Based on real-time data monitoring results, a hierarchical dynamic adjustment mechanism is constructed to automatically execute micro-adjustment, medium-adjustment, and heavy-adjustment strategies; when there is a scene switch, sudden change in demand, or fluctuation in constraints, the model is automatically recalculated and the scheme is updated to realize real-time correction of the scheduling scheme; at the same time, an anomaly handling mechanism is established to quickly activate emergency scheduling plans in response to emergencies such as extreme weather, take-off and landing point failures, and airspace conflicts, so as to ensure the safety of low-altitude operations and the continuity of scheduling. Visualization and Human-Computer Interaction Module: This module establishes an integrated visual management interface, displaying heatmaps of takeoff and landing point operation status, airspace conflict situation maps, demand coverage distribution maps, scheduling scheme lists, constraint compliance reports, cost-benefit curves, and other content. It supports manual viewing, filtering, and modification of scheduling parameters, issuing start / stop commands, adjusting priorities, and calibrating constraint thresholds, enabling collaborative human-machine management. Simultaneously, it generates daily and weekly scheduling operation reports, providing data support for system operation and maintenance and decision optimization.

Claims

1. An adaptive and dynamically adjustable intelligent scheduling method for the take-off and landing points of low-altitude aircraft, characterized in that, Includes the following steps: Step 1: Multi-source dynamic data perception and feature library construction. Construct a multi-dimensional perception network, collect multi-source dynamic data, and simultaneously establish feature libraries for multiple types of aircraft and pilots to complete the construction of a multi-source dynamic data feature library. Step 2: Definition of scheduling constraint system and scientific calculation of parameters. Construct a complete constraint system by establishing hard constraints, soft constraints, time slot pricing constraints, human factor adaptation constraints, and dynamic opening and closing constraints at take-off and landing points; Step 3: Constructing a multi-objective optimization model. Construct a collaborative weighted multi-objective optimization model that considers safety redundancy, operational efficiency, demand coverage, human factor adaptability, aircraft type adaptability, and cost optimization objectives. Combine this with a complete constraint system to deeply couple the option pricing mechanism with the scheduling objectives, thereby achieving multi-objective collaborative optimal matching of scheduling schemes under different low-altitude application scenarios. Step 4: Solve using the adaptive optimization algorithm. The design improves the adaptive NSGA-III optimization algorithm by incorporating scene weight adaptive adjustment, constraint penalty mechanism and local optimization strategy to achieve efficient solution of collaborative weighted multi-objective optimization model and output of optimal scheduling scheme, ensuring that the solution speed and global optimality are adapted to the low-altitude real-time scheduling requirements. Step 5: Output and Dynamic Adjustment of Planning Scheme Output the take-off and landing point scheduling plan, and at the same time build a closed-loop mechanism for real-time perception, dynamic analysis, and iterative adjustment. Combine multi-source dynamic data, scene changes, and constraint fluctuations to make real-time corrections to the take-off and landing point scheduling plan.

2. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 1, characterized in that, Step 1 also includes the following steps: Step 1.1: Deployment of sensing nodes and definition of data types. Construct a multi-dimensional perception network and deploy six types of perception nodes: airspace, environment, demand, facilities, aircraft status, and pilots. Collect static geographic data, dynamic operational data, facility status data, real-time data of various types of aircraft, and pilot status data. Step 1.2 Data Acquisition and Preprocessing Specifications Step 1.2.1, Data acquisition frequency, ① High-frequency dynamic data: The acquisition cycle is 3 minutes to ensure real-time performance; ②Medium frequency dynamic data: Acquisition cycle of 15 minutes, balancing real-time performance and computational cost; Step 1.2.2, Data Preprocessing Methods ①Outlier removal: using 3 The criteria remove outliers from meteorological and signal strength data; ② Missing value imputation: Linear interpolation is used to imput short-term missing data, and an LSTM-based time series prediction model is used to imput long-term missing data; ③ Data standardization: Normalize data of different dimensions; Step 1.3: Construction of a feature library for multiple types of aircraft. Establish an aircraft feature database including logistics drones, eVTOL, emergency rescue drones, police drones, and manned aircraft, and extract core performance parameters including: maximum wingspan, maximum liftoff weight, maximum wind resistance threshold, range, minimum takeoff and landing area, communication signal requirements, and charging power.

3. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 1, characterized in that, Step 2 also includes the following steps: Step 2.1, Hard Constraints Step 2.1.1, Safety distance constraints, (1) Safety distance constraints between take-off and landing points and fixed obstacles, (2) Minimum horizontal distance constraint between the take-off and landing point and the moving obstacle. Step 2.1.2, Spatial Constraints (1) Distance constraints between take-off and landing points and the boundary of controlled airspace, (2) Distance constraints between takeoff and landing points and the centerline of the flight path, Step 2.1.3, Meteorological constraints, (1) Wind speed constraint, (2) Visibility constraints, (3) Precipitation constraints, Step 2.1.4, Facility Constraints (1) Supply voltage fluctuation constraint, (2) Communication signal strength constraints, (3) Minimum available area constraint, Step 2.1.5, Human Factors-Based Adaptation Constraints (1) Fatigue state constraints, (2) Consecutive operation time constraints, (3) Safety constraints on physiological indicators, (4) No-fly zone restrictions, Step 2.2, Soft Constraints Step 2.2.1, Efficiency Constraints (1) Service radius constraints of take-off and landing points, (2) Average utilization rate constraint of take-off and landing time slots, (3) Average distance constraint from the demand point to the nearest take-off and landing point. Step 2.2.2, Cost Constraints (1) Real-time operating cost adaptation constraints at take-off and landing points. (2) Constraints on real-time connectivity between take-off and landing points and existing infrastructure. Step 2.2.3, Sustainability Constraints (1) Noise emission constraints at take-off and landing points, (2) Real-time avoidance constraints in the buffer zone of the ecological protection area. Step 2.2.4, Time Slot Utilization Constraints Step 2.3, Time Slot Pricing Constraints Based on the BS option pricing model, and combined with the real-time non-renewability of take-off and landing time slots and the dynamic optimization of scenarios, the option price and scheduling rules are dynamically adjusted. Step 2.3.1, Constraints of the core formula for BS option pricing. Market-based dynamic scheduling is achieved through time-slot options. ； ； ； In the formula, The real-time premium for takeoff and landing slot options is the fee that users need to pay to reserve a slot. for The real-time value of a time slot is determined by location, time period, airspace status, and demand intensity. and All are intermediate variables; The option exercise price is the final cost paid when the slot is actually used. This is the current risk-free interest rate, used for value discounting; The remaining time to expiration of the option is equal to the planned takeoff and landing time minus the current time; The slot value volatility reflects price fluctuations caused by low-altitude supply and demand and environmental changes. The cumulative distribution function is a standard normal distribution, representing the probability of time slot usage and the expected risk. Step 2.3.2, Time Slot Value Volatility Constraints Step 2.3.3, Option expiration time constraint, Step 2.3.4, Time-slot option price volatility constraints, Step 2.4 Human Factors Adaptation Constraints By combining real-time operation characteristics optimization for low-altitude driving, the intelligent scheduling system for low-altitude take-off and landing points collects driver data in real time to achieve dynamic and personalized matching of time slots between drivers and take-off and landing points. Step 2.4.1: Real-time adaptation constraints between the driver and the takeoff and landing points. Step 2.4.2, Real-time matching degree constraint of operating habits. Step 2.4.3: Real-time matching constraint between take-off and landing point difficulty and driver skill level. Step 2.4.4: Real-time matching degree constraint of operation level. Step 2.4.5: Real-time increase of safety redundancy constraints under fatigue conditions. Step 2.4.6: Real-time physiological state matching degree constraint. Step 2.5, Option Premium A linkage mechanism with scheduling decisions.

4. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 3, characterized in that, Option premium in step 2.5 In addition to being used as a time slot reservation fee, it also participates in scheduling decisions through the following rules: Demand feedback linkage: The intelligent scheduling system for low-altitude take-off and landing points will calculate the option premium in real time. The information is presented to users through a display layer; users decide whether to book based on their budget and the urgency of the task; when the option premium... When the user's preset threshold is exceeded, the user will cancel or postpone the reservation. The low-altitude take-off and landing point intelligent scheduling system collects behavioral changes in real time through demand sensing nodes, updates demand distribution data, and thus indirectly affects the demand coverage target. and operational efficiency goals ; Dynamic adjustment trigger: The intelligent scheduling system for low-altitude take-off and landing points monitors option premiums. short-term rate of change and its deviation from the historical mean; According to the dynamic adjustment mechanism of the scheduling plan, the short-term rate of change The value determines the level of adjustment triggered; Execution Priority Ranking: When generating the planning scheme output, the low-altitude take-off and landing point intelligent scheduling system will prioritize the reserved time slots according to option premiums. Sorted from highest to lowest value; high-value time slots are allocated preferentially to technology levels. High pilot and aircraft compatibility High-altitude aircraft; Operating cost adjustment: The operator periodically adjusts operating costs based on historical option premiums. The cumulative revenue will be used to dynamically adjust the upper limit of the operating cost for order slots in different scenarios within the soft constraint. .

5. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 1, characterized in that, Step 3 also includes the following steps: Step 3.1: Define the objective function. A multi-objective weighted summation objective function is constructed, and the weight coefficients are dynamically adjusted for different low-altitude application scenarios. The basic weights are determined by the analytic hierarchy process. The core objective function of the multi-objective optimization model : ； In the formula, For safety redundancy targets; For operational efficiency targets; To cover the target of demand; Human factors fit target; For model compatibility targets; To optimize costs; The target weighting coefficient for safety redundancy; Weighting coefficients for operational efficiency targets; Determine the target weighting coefficient for demand coverage; Human factors fit target weight coefficient; The target weighting coefficient for device compatibility; Optimize the target weight coefficients for cost.

6. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 1, characterized in that, Step 4 also includes the following steps: Step 4.1: Improve the overall design of the adaptive NSGA-III optimization algorithm. Based on the NSGA-III framework and combined with the optimization and iteration mechanism for low-temperature scenarios, the process of initializing the population, calculating the constraint penalty, non-dominated sorting, adaptive weight update, crossover mutation, local elite search, and termination judgment is adopted. The solution strategy is adaptively switched for different scenarios such as emergency response, logistics, and commutation, ensuring rapid convergence to the Pareto optimal solution set. Step 4.2, Population initialization, Based on a multi-source dynamic data feature library, basic data such as takeoff and landing point layout, aircraft parameters, pilot information, and real-time demand are extracted to construct a decision variable matrix, which includes decision variables such as takeoff and landing point opening and closing status, time slot allocation scheme, service radius, pilot-takeoff and landing point matching relationship, and aircraft type-takeoff and landing point adaptation scheme. An initial population is randomly generated, and the population size is set to 100-200 according to the number of takeoff and landing points in the planning area to ensure population diversity. At the same time, invalid individuals that do not meet hard constraints are removed to improve the initial solution efficiency. Step 4.3: Constructing a constraint and punishment mechanism. To address the hard constraints, soft constraints, human factor fit, and machine model fit constraints in the collaborative weighted multi-objective optimization model, a hierarchical penalty function is designed to penalize individuals that violate the constraints based on their fitness, ensuring that the optimal solution meets the constraint requirements. Hard constraint penalty: Individuals that violate any hard constraint are given the maximum penalty value and the invalid individual is directly eliminated to uphold the bottom line of scheduling safety; Soft constraint penalty: Individuals who violate soft constraints in normal scenarios are assigned gradient penalty values ​​according to the degree of violation; in emergency rescue scenarios, the penalty intensity is reduced by relaxing the rules, so as to achieve flexible adjustment according to the scenario. Adaptation constraint penalty: Individuals that fail to meet the minimum adaptation thresholds for human factors and device models will be subject to the maximum penalty to ensure that core adaptation requirements are met; Step 4.4, Adaptive Adjustment and Local Search, (1) Adaptive adjustment of scenario-based weights. The improved adaptive NSGA-III optimization algorithm can identify the current low-altitude application scenario in real time and automatically match the target weights in the core objective function of the collaborative weighted multi-objective optimization model. When the scenario changes or the requirements change suddenly, the collaborative weighted multi-objective optimization model can automatically recalibrate the weight coefficients and update the core objective function according to the rule of increasing the weight of the core objective and decreasing the weight of the other objectives in proportion, so as to ensure that the optimization direction is in line with the requirements of the scenario. (2) Adaptive crossover mutation strategy, The crossover probability is dynamically adjusted based on the population evolution generation, constraint satisfaction, and objective function convergence. With the probability of mutation , When the constraint satisfaction level is lower than a preset threshold, the mutation probability is increased by a fixed amount within a preset range to generate more individuals that satisfy the constraints; the crossover probability range is... Range of mutation probability values To achieve adaptive dynamic adjustment; (3) Local elite search optimization, For elite individuals after non-dominated sorting, a local search operator is introduced to refine the optimization of take-off and landing time slot allocation, pilot matching, and aircraft type adaptation, thereby improving the practicality of the optimal solution. At the same time, elite individuals are retained to accelerate algorithm convergence and ensure the uniformity and optimality of the Pareto solution set. Step 4.5: Solve for the termination condition using the algorithm. The algorithm terminates its iteration when any of the following conditions are met: It reaches the preset maximum number of iterations to meet the timeliness requirements of low-altitude real-time scheduling; The Pareto optimal solution set remained unchanged after 10 consecutive iterations, and the objective function value converged and stabilized. The generated scheduling scheme fully satisfies all hard constraints, and the compliance rate of soft constraints and adaptation constraints reaches the preset threshold. Finally, multiple Pareto optimal scheduling schemes are output, and the optimal execution scheme is determined by combining the scenario priority.

7. The adaptive dynamic adjustment intelligent scheduling method for low-altitude aircraft take-off and landing points according to claim 1, characterized in that, Step 5 also includes the following steps: Step 5.1, Output of the planning scheme, (1) Basic static planning scheme, Landing point layout and configuration: Clearly define the fixed locations, construction standards, and supporting facility parameters of landing points within the planning area, and determine the difficulty level of the landing point sites; Aircraft compatibility list: Defines the types of aircraft compatible with each take-off and landing point, and marks the minimum take-off and landing area, wind resistance threshold, and charging power compatibility parameters; Basic constraint thresholds: solidify hard constraint thresholds to determine the benchmark parameters for soft constraints, time slot pricing, and human factor adaptation in typical scenarios; (2) Scenario-based dynamic execution scheme, Take-off and landing point opening and closing scheme: clearly define the open / closed / warning status of each take-off and landing point, and mark the scheduling restriction rules for take-off and landing points in the warning status; Time slot allocation scheme: Plan the time slot scheduling plan for each take-off and landing point, determine the time slot utilization rate and service radius configuration, and avoid airspace conflicts; Demand Coverage Plan: Prioritize demand points according to emergency > logistics > commuting, formulate demand point coverage strategies, and clarify the guarantee mechanism for high-priority demands; Personnel and aircraft matching scheme: Generates an optimal matching list of pilots and takeoff / landing points, and aircraft types and takeoff / landing points, ensuring compatibility meets standards; simultaneously, the intelligent scheduling system for low-altitude takeoff / landing points allocates reserved time slots according to option premiums. The high-value time slots are allocated to drivers with high technical skills and aircraft with high suitability in order to optimize safety during high-risk periods. Cost control plan: Plan the construction and operation cost control targets for take-off and landing sites, and clarify the cost optimization path throughout the entire life cycle; (3) Visual output of the solution, Simultaneously output visual reports of the scheduling plan, including heat maps of take-off and landing point operation status, airspace conflict early warning maps, demand coverage distribution maps, adaptability level tables, and cost-benefit analysis tables, which facilitate intuitive management and execution. Step 5.2, Dynamic adjustment mechanism for scheduling scheme, (1) Real-time data monitoring, Real-time collection of take-off and landing point operation data, aircraft status, pilot physiological and skill data, airspace environment, demand distribution, and cost data; the dataset is updated every 1-5 minutes, and the constraint satisfaction, objective function achievement, and scenario changes are analyzed simultaneously. (2) Grading adjustment rules, Fine-tuning: When demand fluctuations, soft constraint deviations, slot utilization changes fall below a preset low threshold, or option premium changes occur... At that time, only local thresholds are fine-tuned for take-off and landing point service radius, time slot allocation, and aircraft type adaptation, without changing the overall scheduling scheme; Mid-adjustment: When low-altitude application scenarios are switched, or when the option premium is more than twice the historical average for three consecutive update cycles, the target weight is automatically recalibrated, the adaptive NSGA-III optimization algorithm is improved to solve the problem quickly through iteration, and the scenario-based scheduling scheme is updated. Re-adjustment: When extreme weather occurs, hard constraints are breached, emergency demand surges, key take-off and landing points fail, or option premiums drop by more than 50% and are accompanied by hard constraints being breached, global re-planning is immediately triggered, take-off and landing points that do not meet hard constraints are forcibly closed, high-priority needs are prioritized, and emergency dispatch plans are quickly generated. (3) Constraint dynamic adaptation adjustment, The hard constraints must always be kept inviolable. Once a deviation from the hard constraints occurs, the take-off and landing points will be forcibly shut down and relocated immediately. Soft constraints follow the principle of relaxing thresholds according to preset rules in emergency scenarios to ensure that emergency needs are met; Human factors and aircraft compatibility constraints are kept to the minimum threshold, pilot status and aircraft performance are updated in real time, and matching relationships are dynamically adjusted to ensure that the compatibility meets the standards. (4) Adjust the closed-loop process. A closed-loop adjustment process is formed, which includes data perception, deviation analysis, algorithm recalculation, scheme update, implementation, and effect feedback. The adjustment response time is controlled within 4 minutes to meet the needs of real-time, efficient, and safe scheduling of low-altitude take-off and landing points, and to always ensure the optimal synergy of the six objectives.

Images