Improved hybrid nonlinear optimization unmanned aerial vehicle base station positioning method

Abstract

The invention discloses an improved hybrid nonlinear optimization unmanned aerial vehicle base station positioning method. The method comprises the following steps: step 1, establishing a communication network model for unmanned aerial vehicle base station positioning; 2, setting target correlation constraints for the communication network model, wherein the target correlation constraints comprise mobility constraints, power consumption constraints and path loss constraints; 3, solving the communication network model by adopting an improved Java method to obtain an initial position of the unmanned aerial vehicle base station; and step 4, optimizing the initial position by adopting an improved random hill climbing method to obtain a final unmanned aerial vehicle base station position. According to the invention, the accuracy and efficiency of unmanned aerial vehicle base station positioning are significantly improved.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV communication positioning optimization, and particularly to a method for positioning a UAV base station by improving hybrid non-linear optimization. Background Art

[0002] With the rapid development of wireless communication technology, in recent years, people's demand for location-based services has been continuously increasing, especially for three-dimensional location-based services. In outdoor scenarios, multiple global satellite navigation systems around the world, such as the Global Positioning System in the United States, Galileo in Europe, GLONASS in Russia, and the Beidou satellite navigation system in China, can provide high-precision positioning information. However, in cities with high-rise buildings and forest areas with complex terrain, due to the occlusion of obstacles such as walls and trees, signal propagation is attenuated and scattered, making it particularly difficult to achieve precise positioning under the environmental conditions where these global positioning and navigation satellite systems are strongly interfered. Therefore, base station positioning has become one of the key facilities for positioning in these environments.

[0003] With the large-scale coverage of 5G networks, base stations have gradually been upgraded to 5G base stations to ensure communication quality. Using the communication signals between users and ground 5G fixed base stations for positioning has become a current research hotspot. In recent years, with the continuous development of 5G communication networks, UAVs have played an important role in fields such as navigation, emergency communication, and geological exploration due to their advantages such as controllable mobility and low deployment cost, and are expected to become an important part of the future fifth-generation air communication platform. When communication is blocked due to sudden emergencies, communication problems can be solved by relying on airborne satellites, ground emergency communication vehicles, and UAVs. However, airborne satellites have high deployment costs and long communication delays, making it difficult to be widely applied; emergency communication vehicles have unstable signal transmission and are difficult to deploy when traffic is blocked. In contrast, UAVs are small in size and highly flexible. Using them to carry base stations as airborne base stations has become an important means to achieve local capacity enhancement.

[0004] Currently, UAV base station positioning is mainly applied in two aspects: One is in hot spots, such as sports events or celebration activities, where sudden traffic surges occur due to crowd gathering. Existing deployed ground base stations cannot accommodate all communication traffic, and temporary communication measures such as emergency support vehicles have low mobility. Therefore, after considering the distribution of ground users, additional capacity enhancement solutions are needed, and using UAV base station positioning is an effective supplementary method. The other is during natural disasters. In affected areas, there are often large-scale power outages, road breaks, and network outages, and it is difficult to quickly restore communication relying on traditional emergency means, which brings inconvenience to subsequent rescue work. UAV base station positioning can achieve rapid coverage of the disaster area and improve rescue efficiency to a certain extent. However, in the existing technology, there is still a problem of relatively low solution efficiency in solving the position of UAV base station positioning. Summary of the Invention

[0005] In view of this, the present invention proposes a method for positioning an unmanned aerial vehicle (UAV) base station by improving hybrid non-linear optimization to solve the problem of low efficiency in solving UAV base station positioning in the prior art.

[0006] The specific technical solution of the present invention is as follows:

[0007] A method for positioning an unmanned aerial vehicle (UAV) base station by improving hybrid non-linear optimization, comprising:

[0008] Step 1, establish a communication network model for UAV base station positioning;

[0009] Step 2, set target correlation constraints for the communication network model, including mobility constraints, power consumption constraints, and path loss constraints;

[0010] Step 3, use an improved Java method to solve the communication network model to obtain the preliminary position of the UAV base station;

[0011] Step 4, use an improved random hill climbing method to optimize the preliminary position to obtain the final position of the UAV base station.

[0012] Further, in Step 1, the core components of the communication network model include a UAV base station set and a mobile positioning device set. The UAV base station is used to provide stable and efficient communication services, and the mobile positioning device is used to receive communication signals from the UAV base station to ensure the accuracy and real-time update of positioning information; the time is divided into several fixed time steps for precisely controlling the position layout and communication power allocation of the UAV base station.

[0013] Further, Step 1 further includes: determining a dual optimization objective. On the one hand, the total power consumed by the movement and communication of the UAV base station is reduced to the lowest level, and on the other hand, the number of uncovered positioning devices is reduced; use a weighted normalization objective function to combine the dual optimization objectives, and achieve the overall optimization effect by assigning weights to each objective to ensure that the two objectives have comparable scales.

[0014] Further, the formula representation of Step 1 is: Wherein, represents the total power consumed by communication; represents the number of uncovered positioning devices; i represents the i-th UAV; j represents the j-th mobile positioning device; t represents the current time; the function f1 calculates the sum of the total power consumption of the UAV base station within all time steps, reflecting the objective of minimizing the overall power consumption of the UAV base station within the entire specified time range; the function f2 represents the total number of uncovered positioning devices, and the negative sign before the summation symbol indicates that the objective of this function is to minimize this index and reduce the number of uncovered positioning devices The function minf combines two objective functions f1 and f2 into an objective function, achieving the overall optimization effect by assigning weights w1 and w2 to each objective, where w1 + w2 = 1.

[0015] Further, in step two, the mobility constraints include restrictions on the vertical speed, horizontal speed, altitude change, and position change of the UAV; the power consumption constraints include restrictions on the power consumption of the UAV's motor, communication power consumption, and processing power consumption of the computing unit; the path loss constraint is used to accurately calculate the attenuation degree of the signal during transmission, and accordingly optimize and adjust parameters such as the flight altitude and transmission power of the UAV to ensure the stability and reliability of signal transmission.

[0016] Further, in step two, the path loss constraint of the UAV is expressed by the formula where represents the path loss of the UAV; B ij and q ij are the subchannel bandwidth and the transmission power allocated by the UAV base station to the positioning device respectively, ε 2 is the Gaussian white noise, and ∈ is the average path loss.

[0017] Further, in step three, the process of solving using the improved Java method includes: using the k - means clustering algorithm to process the data collected by the positioning device to initially determine the approximate position of the UAV base station, where the data collected by the positioning device includes the initial position of the UAV, signal strength, and factors affecting the base station position such as geographical information; through the iterative update process based on the current solution, optimal solution, and worst solution, continuously adjust the coordinates of the UAV base station until a certain convergence condition is met or the preset number of iterations is reached.

[0018] Further, step three is expressed by the formula:

[0019]

[0020] where represents the coordinate of the UAV base station on the x - axis; represents the coordinate of the UAV base station on the y - axis; represents the height of the UAV base station; and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and r is the assigned weight.

[0021] Furthermore, in step four, the process of using the improved stochastic hill climbing method for position optimization includes: by combining the advantages of randomness and local search, gradually approaching the global optimal solution, and introducing random elements and a penalty mechanism during the search process to avoid being trapped in a local optimum; using a fitness function combined with a penalty mechanism to evaluate candidate solutions, and the penalty mechanism adjusts the fitness value according to the difference between the maximum and minimum hill climbing travel lengths; designing a penalty factor that gradually decreases over time.

[0022] Furthermore, step four is expressed by the formula where Imbalance Penalty is the fitness function; max(Length) and min(Length) represent the maximum and minimum hill climbing travels respectively, and f is the penalty factor; f t+1 = max(f min , f t ×γ), where f t+1 = max(f min , f t ×γ) is the penalty factor at time t + 1; f t+1 = max(f min , f t ×γ) is the penalty factor at time t; f t+1 = max(f min , f t ×γ) is the minimum penalty factor; γ is the rate of decrease over time..

[0023] The beneficial effects of the present invention are as follows:

[0024] (1) Establish a communication network model, divide time into several time periods with a time step T = 10, and accurately control the base station layout and power allocation. Its dual objectives are to reduce the power consumption of the base station, extend the battery life, and at the same time reduce uncovered devices, improve the system coverage rate and reliability. The weighted normalized objective function makes the optimization more fair and accurate, avoiding being dominated by a single objective.

[0025] (2) Set target correlation constraints for the communication network model, covering mobility, power consumption, and path loss. The mobility constraint guarantees the ability of the unmanned aerial vehicle to perform tasks, the power consumption constraint ensures the battery life and efficiency, and the path loss constraint can accurately calculate the signal attenuation, optimize the flight altitude and transmission power, and ensure stable and reliable signal transmission.

[0026] (3) Use the improved Java method combined with the k - means clustering algorithm to solve. First, roughly determine the locations of the base stations through clustering, and then iteratively update the coordinates. This method enhances the convergence stability and speed, ensures the robust and efficient deployment of the base stations, can adapt to adverse conditions such as signal interference, and dynamically adjusts the position optimization performance.

[0027] (4) Optimize the position using an improved random hill - climbing method, combining random and local search. Set the hill - climbing location, introduce random elements and a penalty mechanism to avoid local optima. Balance the workload initially and then shift to optimizing the overall position in the later stage to enhance fitness, generate an ideal layout, and improve the overall performance and efficiency of the system. Description of the Drawings

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0029] Figure 1 It is a schematic flowchart of the method for positioning an unmanned aerial vehicle (UAV) base station with improved hybrid non - linear optimization of the present invention. Detailed Embodiments

[0030] In order to make the technical problems, technical solutions and beneficial effects to be solved by the present invention clearer, the following further details the present invention with reference to the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0031] The present invention proposes a method for positioning an unmanned aerial vehicle (UAV) base station with improved hybrid non - linear optimization. First, establish a communication network model for UAV base station positioning. Then, set the target correlation constraints according to the established communication network model for UAV base station positioning. Next, use an improved Java method to solve it to obtain the position of the UAV. Finally, use an improved random hill - climbing method for position optimization. The specific steps of the method of the present invention are as follows:

[0032] Step 1: Establish a communication network model for UAV base station positioning. The present invention first constructs and optimizes a communication network model for UAV base station positioning. The core components of this model cover multiple UAV base stations, which are uniformly identified as set I, and each base station is responsible for providing stable and efficient communication services. At the same time, there is a group of mobile positioning devices that need to continuously receive communication signals from the UAV base stations to ensure the accuracy and real - time update of the positioning information. For distinction, these devices are also given specific symbols for representation, denoted by J.

[0033] During the construction and optimization of the model, time is finely divided into several time steps, with the time step T = 10. Each time step represents a crucial optimization decision moment. Such a time segmentation strategy enables us to more precisely control the location layout of the drone base station and the allocation of communication power, thereby more effectively achieving the optimization goal.

[0034] The optimization goal of this network communication model has distinct duality. On the one hand, it aims to minimize the total power consumed by the movement and communication of the drone base station to the lowest possible level. This can not only significantly extend the flight time of the drone but also greatly improve the energy efficiency performance of the entire system. On the other hand, it also endeavors to reduce the number of uncovered positioning devices to ensure that more devices can stably and reliably receive communication signals from the drone base station, thereby further enhancing the coverage and reliability of the system.

[0035] The specific formula is expressed as where, represents the total power consumed by communication; represents the number of uncovered positioning devices; i represents the i-th drone; j represents the j-th mobile positioning device; t represents the current time; the function f1 calculates the sum of the total power consumption of the drone base station over all time steps, and this function reflects the goal of minimizing the overall power consumption of the drone base station within the entire specified time range; the function f2 represents the total number of uncovered positioning devices, and the negative sign before the summation symbol indicates that the goal of this function is to minimize this indicator, that is, to reduce the number of uncovered positioning devices The function minf combines the two objective functions f1 and f2 into an objective function, and achieves the overall optimization effect by assigning weights w1 and w2 to each objective, where w1 + w2 = 1.

[0036] The weighted normalized objective functions f1 and f2 adopted in this step have significant advantages. It ensures that the two objective functions have comparable scales, thereby making the optimization process more interpretable and fair. At the same time, it also makes each term in the combined objective function f contribute fairly and evenly to the optimization process, effectively preventing the situation where a certain objective disproportionately dominates due to scale differences.

[0037] Step 2: Establish target relevance constraints for the communication network model of drone base station positioning established in Step 1. In Step 1, we have successfully constructed the communication network model for drone base station positioning. To further optimize this model and ensure that the drone base station can meet the established performance and limitation requirements during actual deployment and operation, the core task of Step 2 is to set target relevance constraints for this model. These constraints specifically cover mobility constraints, power consumption constraints, and path loss constraints.

[0038] The mobility of the drone is a core element when it performs tasks. To ensure that the drone can perform tasks efficiently and accurately, strict constraints need to be imposed on its mobility. The vertical and horizontal speeds of the drone have a crucial impact on its maneuverability and overall performance. These speed parameters not only determine how the drone navigates and adapts to different flight scenarios but also are directly related to its ability to perform tasks. The mobility constraints of the drone are expressed by the formula

[0039] covering multiple key aspects such as the speed, altitude, and position change of the drone. Among them, V i，t represents the speed of the i-th drone at time t; represents the vertical speed of the i-th drone at time t, and h i，t and h i，t-1 represent the altitudes of the i-th drone at times t and t - 1, represents the speed of the i-th drone on the X-axis at time t, and x i，t and x i，t-1 represent the positions of the i-th drone on the X-axis at times t and t - 1, represents the speed of the i-th drone on the Y-axis at time t, and y i，t and y i，t-1 represent the positions of the i-th drone on the Y-axis at times t and t - 1, represents the horizontal speed of the i-th drone at time t, and represent the wind speed and heading angle of the i-th drone at time t, respectively.

[0040] The power consumption of the drone is directly related to its endurance and task execution efficiency. The power consumption constraints of the drone can be divided into three parts: motor power consumption, communication power consumption, and computing unit processing power consumption. To ensure that the drone can meet the power consumption limit requirements during actual operation, it is expressed by the formula Among them, and are the motor power consumption and communication power consumption consumed by the drone during horizontal and vertical flights, is the computing unit processing power consumption, P0 is the power consumption of the drone blade profile, is the angular velocity of the motor rotor, R is the rotor radius, and P i is the motor induction power consumption of the i-th drone, V0 is the induced velocity during hovering, ρ is the air density, and S FP is the frontal area of the drone, and W iis the induced velocity coefficient, A is the area of the drone exposed to the air, P c is the circuit power consumption, τ t , σ and P are the normalized load power consumption, amplifier power consumption, and transmission power respectively. Therefore, the power consumption constraint of the drone is expressed by the formula where represents the power consumption of the drone.

[0041] Path loss is an important parameter in wireless communication, which determines the attenuation degree of the signal during transmission. For the communication network model of drone base station positioning, the path loss constraint is equally crucial. The path loss constraint of the drone is expressed by the formula where represents the path loss of the drone; B ij and q ij are the sub-channel bandwidth and the transmission power allocated by the drone base station to the positioning device respectively, ε 2 is the Gaussian white noise, and ∈ is the average path loss. Through this formula, the attenuation degree of the signal during transmission can be accurately calculated, and parameters such as the flight altitude and transmission power of the drone can be optimized and adjusted accordingly to ensure the stability and reliability of signal transmission.

[0042] Step 3, solve the communication network model established in Step 1 and Step 2 using an improved Java method to obtain the optimal position of the drone base station.

[0043] First, based on the rich data collected by the positioning device, these data cover the initial position of the drone, signal strength, geographical information, and any other factors that may affect the base station position. Subsequently, apply the k-means clustering algorithm to process these data. K-means clustering is an unsupervised learning algorithm that can divide data points into k clusters, and the center point of each cluster represents the average position of the data points within the cluster. Here, the value of k represents the number of drone base stations expected to be deployed. Through cluster analysis, the approximate position of the drone base station can be initially determined.

[0044] Next, the parameters are adjusted to balance the quality of the solution and the computational efficiency. An improved Java method is adopted to iteratively update the coordinates of the UAV base station. After determining the initial position, this method starts to iterate based on the current solution (i.e., the current position of the UAV base station), the optimal solution (the best position found so far), and the worst solution (the worst position found so far). The iterative update process of the improved Java method is strictly controlled by the optimization equation, continuously adjusting the coordinates of the UAV base station until a certain convergence condition is met (such as the change in the coordinates of the UAV base station is less than a certain threshold) or the preset number of iterations is reached. At convergence, it can be considered that a sufficiently good solution has been found, or at least an optimal solution under the given conditions.

[0045] The specific formula is expressed as

[0046] where represents the coordinate of the UAV base station on the x-axis; represents the coordinate of the UAV base station on the y-axis; represents the height of the UAV base station; and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and r is the assigned weight.

[0047] This step adopts an improved Java method to enhance the stability and speed of convergence, ensuring the robustness and efficiency of the UAV base station deployment. At the same time, this method can adapt to potential adverse conditions, such as signal interference, geographical obstacles, etc., thus optimizing the performance of the UAV base station in the actual scenario. By dynamically adjusting the position of the UAV base station, this method can maintain efficient and robust communication network performance in different environments, providing strong support for the deployment and optimization of the UAV base station.

[0048] Step Four, the improved random hill climbing method is used to optimize the position of the UAV obtained in Step Three. In the process of optimizing the UAV position in the present invention, for the UAV position preliminarily determined in Step Three, an improved random hill climbing method is adopted for further optimization. The core of this method lies in combining the advantages of randomness and local search, and gradually approaching the global optimal solution through iteration.

[0049] Specifically, starting from the UAV position generated in Step 3, the present invention takes it as the starting point of the search. During the search process, whenever a new position is observed to bring an improvement in fitness, that is, the new position is superior to the current position under a certain evaluation criterion, the UAV is iteratively moved to this better position. To increase the diversity of the search and avoid falling into local optima, the present invention arranges two hill-climbing locations among the UAV positions and will consider swapping positions between two different UAVs during the search process to generate new candidate solutions from adjacent UAV positions. At the same time, to jump out of local optima and explore a broader solution space, the present invention randomly selects random elements of the search direction and step size during the search process, or uses different methods to generate candidate (or optimized) positions of the UAV. The introduction of such random elements greatly increases the diversity and variability of the search.

[0050] When evaluating the advantages and disadvantages of each candidate solution, the present invention uses a fitness function and incorporates a penalty mechanism. This penalty mechanism adjusts the fitness value according to the difference between the maximum and minimum hill-climbing travel lengths, such that solutions with larger differences in travel lengths are penalized during evaluation, thereby promoting the generation of more balanced solutions. The specific formula is expressed as where ImbalancePenalty is the fitness function; max(Length) and min(Length) represent the maximum and minimum hill-climbing travels respectively, and f is the penalty factor.

[0051] To shift the focus of the random hill-climbing method from workload balance to optimizing the total UAV position in subsequent iterations, the present invention designs a penalty factor that gradually decreases over time. This means that at the initial stage of the search, the penalty factor is larger, which helps to balance the workload; as the search progresses, the penalty factor gradually decreases, causing the search focus to gradually shift to optimizing the total UAV position. The specific formula is expressed as f t+1 = max(f min , f t ×γ), where f t+1 = max(f min , f t ×γ) is the penalty factor at time t + 1; f t+1 = max(f min , f t ×γ) is the penalty factor at time t; f t+1 = max(f min , f t ×γ) is the minimum penalty factor; γ is the rate of decrease over time.

[0052] Finally, through an improved random hill climbing method, the optimized UAV positions are obtained. These positions have a significant improvement in fitness, and due to the introduction of random elements and penalty mechanisms, the deterministic trap is successfully avoided, thus generating a more ideal and efficient UAV position layout. This step not only increases the diversity and variability of the search but also avoids falling into local optima, ultimately generating a globally optimal or near-optimal UAV position layout, significantly enhancing the overall performance and efficiency of the UAV system.

[0053] The beneficial effects of the present invention are as follows:

[0054] (1) A communication network model is established, and time is divided into several time periods with a step size T = 10 to accurately control the base station layout and power allocation. Its dual objectives are to reduce the power consumption of the base station and extend the battery life, while reducing uncovered devices and improving the system coverage rate and reliability. The weighted normalization objective function makes the optimization more fair and accurate, avoiding the dominance of a single objective.

[0055] (2) Target correlation constraints are set for the communication network model, covering mobility, power consumption, and path loss. The mobility constraint guarantees the UAV's task execution ability, the power consumption constraint ensures the battery life and efficiency, and the path loss constraint can accurately calculate the signal attenuation, optimizing the flight altitude and transmission power to ensure stable and reliable signal transmission.

[0056] (3) An improved Java method combined with the k-means clustering algorithm is used for solution. First, the approximate positions of the base stations are initially determined through clustering, and then the coordinates are iteratively updated. This method enhances the convergence stability and speed, ensuring a robust and efficient base station deployment, being able to adapt to adverse conditions such as signal interference, and dynamically adjusting the positions to optimize the performance.

[0057] (4) The improved random hill climbing method is used to optimize the positions, combining random and local searches. The climbing locations are set, random elements and penalty mechanisms are introduced to avoid local optima. The workload is balanced in the initial stage, and then it turns to optimizing the total positions, improving the fitness, generating an ideal layout, and enhancing the overall performance and efficiency of the system.

[0058] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. An improved method for positioning an unmanned aerial vehicle base station by hybrid non - linear optimization, characterized in that, Including: Step 1, establish a communication network model for UAV base station positioning; Step 2, set target relevance constraints for the communication network model, including mobility constraints, power consumption constraints, and path loss constraints; Step 3, use an improved Java method to solve the communication network model to obtain the preliminary position of the UAV base station; Step 4, use an improved random hill climbing method to optimize the preliminary position to obtain the final position of the UAV base station.

2. The improved UAV base station positioning method for hybrid non-linear optimization according to claim 1, wherein In Step 1, the core components of the communication network model include a UAV base station set and a mobile positioning device set. The UAV base station is used to provide stable and efficient communication services, and the mobile positioning device is used to receive communication signals from the UAV base station to ensure the accuracy and real-time update of positioning information; the time is divided into several fixed time steps for accurately controlling the position layout and communication power allocation of the UAV base station.

3. The improved UAV base station positioning method with hybrid non-linear optimization according to claim 2, wherein Step 1 also includes: determining a dual optimization goal. On the one hand, reduce the total power consumed by the movement and communication of the UAV base station to the lowest level. On the other hand, reduce the number of uncovered positioning devices; use a weighted normalization objective function to combine the dual optimization goals, and achieve the overall optimization effect by assigning weights to each goal to ensure that the two goals have comparable scales.

4. The improved UAV base station positioning method for hybrid non-linear optimization according to claim 3, characterized in that, The formula of the first step is expressed as: Wherein, represents the total power consumed by communication; represents the number of un-covered positioning devices; i represents the i-th drone; j represents the j-th mobile positioning device; t represents the current time; the function f1 calculates the sum of the total power consumption of the drone base stations within all time steps, reflecting the goal of minimizing the overall power consumption of the drone base stations within the entire specified time range; the function f2 represents the total number of un-covered positioning devices, and the negative sign before the summation symbol indicates that the goal of this function is to minimize this indicator and reduce the number of un-covered positioning devices The function minf combines the two objective functions f1 and f2 into an objective function, and realizes the overall optimization effect by assigning weights w1 and w2 to each objective, where w1 + w2 = 1.

5. The method for positioning a UAV base station with improved hybrid non-linear optimization according to claim 1, wherein, In Step 2, the mobility constraints include restrictions on the vertical speed, horizontal speed, altitude change, and position change of the UAV; the power consumption constraints include restrictions on the power consumption of the UAV motor, communication power consumption, and processing power consumption of the computing unit; the path loss constraint is used to accurately calculate the attenuation degree of the signal during transmission, and accordingly optimize and adjust parameters such as the flight altitude and transmission power of the UAV to ensure the stability and reliability of signal transmission.

6. The improved UAV base station positioning method for hybrid non - linear optimization according to claim 5, characterized in that, In the second step, the path loss constraint of the UAV is expressed by the formula where represents the path loss of the UAV; B ij and q ij are the sub-channel bandwidth and the transmission power allocated by the UAV base station to the positioning device respectively, ε 2 is the Gaussian white noise, and ∈ is the average path loss.

7. The improved UAV base station positioning method for hybrid non-linear optimization according to claim 1, wherein, In Step 3, the process of using an improved Java method for solving includes: using the k-means clustering algorithm to process the data collected by the positioning device to initially determine the approximate position of the UAV base station, where the data collected by the positioning device includes factors affecting the base station position such as the initial position of the UAV, signal strength, and geographical information; through an iterative update process based on the current solution, optimal solution, and worst solution, continuously adjust the coordinates of the UAV base station until a certain convergence condition is met or the preset number of iterations is reached.

8. The improved UAV base station positioning method for hybrid non-linear optimization according to claim 7, characterized in that, Step 3 is expressed by the formula: Among them, represents the coordinate of the UAV base station on the x-axis; represents the coordinate of the UAV base station on the y-axis; represents the height of the UAV base station; and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and are the current, optimal, and worst solutions respectively, and r is the assigned weight.

9. The improved UAV base station positioning method for hybrid non - linear optimization according to claim 1, characterized in that, In Step 4, the process of using an improved random hill climbing method for position optimization includes: by combining the advantages of randomness and local search, gradually approaching the global optimal solution, and introducing random elements and a penalty mechanism during the search process to avoid falling into a local optimum; using a fitness function combined with a penalty mechanism to evaluate candidate solutions, and the penalty mechanism adjusts the fitness value according to the difference between the maximum and minimum hill climbing travel lengths; design a penalty factor that gradually decreases over time.

10. The improved UAV base station positioning method for hybrid non-linear optimization according to claim 9, wherein, The fourth step is expressed by the formula Imbalance Penalty = [max(Length) - min(Length)] * f, where Imbalance Penalty is the fitness function; max(Length) and min(Length) represent the maximum and minimum climbing distances respectively, and f is the penalty factor; f t+1 = max(f min , f t ×γ), where f t+1 = max(f min , f t ×γ) is the penalty factor at time t + 1; f t+1 = max(f min , f t ×γ) is the penalty factor at time t; f t+1 = max(f min , f t ×γ) is the minimum penalty factor; γ is the rate of decrease over time..

Images