Improved multi-unmanned aerial vehicle maximum coverage deployment method

By combining the iterative optimization method of Kalman filter, Weiszfeld algorithm and the fastest descent method and Goldstein rules, the local optimization and energy efficiency problems in multi-UAV coverage deployment are solved, and efficient coverage and energy consumption are minimized, which is enhanced. System adaptability and reliability.

CN119071795APending Publication Date: 2024-12-03GUANGXI TEACHERS EDUCATION UNIV
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202411090840.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-09
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

The prior art has local optimal problems and uneven convergence speeds of energy efficiency and path planning in multi-UAV coverage deployment, especially in terms of maximizing coverage and minimizing system energy consumption.

Method used

Adopt the improved multi-UAV maximizing coverage deployment method, combines the Kalman filter to obtain target position information, decentralize based on the Weiszfeld algorithm and adds disturbance to prevent local optimization, use the fastest descent method and Goldstein rules to iterate and optimize energy efficiency, and plan Deploy the path and realize path optimization.

Benefits of technology

It achieves stable and efficient coverage of the target area, reduces the energy consumption of the drone, enhances the durability and reliability of the system, and has the ability to quickly respond to environmental changes, and effectively responds to challenges in complex environments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119071795A_ABST
    Figure CN119071795A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of target coverage, in particular to an improved multi-unmanned aerial vehicle maximized coverage deployment method, which comprises the following steps: firstly, analyzing initial target information, adding disturbance to prevent local optimum to obtain an optimal coverage set after decentralization based on a Weiszfeld algorithm, and optimizing energy efficiency and planning a path by means of an SDG algorithm after the initial position of an unmanned aerial vehicle is positioned; the energy efficiency of the unmanned aerial vehicle is optimized by considering the energy consumption, speed, moving distance and other factors of the unmanned aerial vehicle, so that the aim of minimizing the energy consumption of the system is achieved. According to the invention, through accurate positioning and energy efficiency optimization, stable and efficient coverage of the target area is realized, the energy consumption of the unmanned aerial vehicle is greatly reduced, and the durability and reliability of the system are enhanced. In addition, the method also has the capabilities of quickly responding to environment change and re-planning a coverage path, and effectively deals with challenges in a complex environment.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of target coverage technology, and in particular to an improved multi-UAV maximum coverage deployment method. Background Art

[0002] With the continuous advancement of communication technology, multi-drone coverage deployment technology plays a vital role in disaster response and emergency communication support. This technology significantly improves adaptability and service efficiency in complex environments by providing continuous communication and monitoring services in a specific area. However, existing technologies face challenges in achieving stable and effective coverage of multiple targets in a wide area, especially in terms of maximizing coverage, minimizing system energy consumption, and optimizing drone movement paths.

[0003] The basic idea of ​​the Weiszfeld algorithm is to gradually approach the geometric median through an iterative process. The geometric median, also known as the median point, is the geometric center of a set of points in Euclidean space, which minimizes the total distance to these points. However, using only the Weiszfeld algorithm sometimes results in a local optimum.

[0004] The steepest descent method (SD) is a combination of the gradient descent method and one-dimensional search. Its search process is zigzag, and the direction (negative gradient) obtained in each step is orthogonal to the previous step. It converges quickly when far away from the extreme point, but converges very slowly when approaching the extreme point. Summary of the invention

[0005] The purpose of the present invention is to provide an improved multi-UAV maximization coverage deployment method, aiming to solve the local optimal problem of the Weiszfeld algorithm in maximizing coverage and the technical problem of uneven convergence speed of the steepest descent method in energy efficiency and path planning, so as to achieve efficient deployment of UAVs.

[0006] To achieve the above object, the present invention provides an improved multi-UAV maximum coverage deployment method, comprising the following steps:

[0007] Step 1: Use Kalman filter to obtain target position information;

[0008] Step 2: Based on the Weiszfeld algorithm, after decentralization, disturbance is added to prevent local optimality and find the maximum coverage set of drones;

[0009] Step 3: Locate the drone’s location information;

[0010] Step 4: SD algorithm combined with Goldstein rule iteratively optimizes energy efficiency;

[0011] Step 5: Plan the deployment path according to the iterative value to achieve path optimization.

[0012] Optionally, the update formula of the Kalman filter in step 1 is

[0013] x k =x k-1 +K k (z k -Hx k-1 )

[0014] Among them, x k is the state estimate at time k, x k-1 is the predicted state based on the previous estimate, K k is the Kalman gain, z k is the observed value, and H is the observation model.

[0015] Optionally, the execution process of adding disturbance to prevent local optimality after decentralization based on the Weiszfeld algorithm in step 2 includes the following steps:

[0016] Step 2.1: Use the Weiszfeld algorithm to perform decentralization and find the optimal center position by calculating the weighted average center point;

[0017] Step 2.2: Set a disturbance range for the center of the circle and perform a disturbance test;

[0018] Step 2.3: If the number of targets covered increases after the disturbance, update the position of the circle center and evaluate the coverage effect;

[0019] If the target covered does not change or decreases after the disturbance, further adjust or maintain the original position;

[0020] Step 2.4: After determining the optimal circle center position, remove the target position covered by the corresponding circle center from the target set;

[0021] Step 2.5: Check whether all target points have been covered or whether the preset number of drones have been deployed; if either condition is met, the optimization process ends; if the target has not been reached, adjustment and optimization will continue.

[0022] Optionally, the formula for the relative distance between the drone and the target in step 3 is as follows:

[0023]

[0024] where x i ,y i is the horizontal and vertical coordinates of the i-th UAV, x j ,y j are the horizontal and vertical coordinates of the j-th target, U is the drone group, ui is the i-th drone, O is the target user, and Oj is the j-th user; P(ui ) is the spatial position of the UAV u i ; P(o i ) is the spatial position of the target user o i ; H is the altitude of the UAV relative to the ground.

[0025] Optionally, the objective function expression for quantifying and evaluating the energy efficiency of the UAV in step 4 is as follows:

[0026] L(D, N) = μ d l d + μ c l c = μ d D + μ c N

[0027] μ d + μ c = 1

[0028] min f(x) = L(D(C * ), N(C * ))

[0029] where L represents the loss of the UAV, D represents the total distance between the position points and the UAV, N represents the total number of UAV resources around each task point, μ d and μ c are the weight coefficients for balancing the distance loss and the information loss, and are positive real numbers. At the same time, two key performance indicators are introduced: D(C * ), N(C * ). D(C * ) refers to the sum of the straight-line distances between all users and their corresponding UAVs when adopting the optimal coverage set strategy. N(C * ) represents the sum of the number of UAVs deployed around each user in the optimal coverage set.

[0030] Optionally, the SD algorithm in step 4 combines with the Goldstein rule to select an appropriate step size α. The determination of whether the step size α is too large or too small is determined by the following formula:

[0031]

[0032] where the parameters c 1 , c 2 are the parameters for controlling the step size selection, satisfying 0 < c < 0.5. The right side is the Armijo condition to ensure that the step size α k is not too large; while the left side is to avoid finding an α k that is too small, resulting in too slow algorithm operation speed. The positive scalar α k and the vector p kare the direction of step size and increment decrease respectively;

[0033] The step size α is enlarged or reduced according to the following formula:

[0034]

[0035] where θ 1 ,θ 2 is the zoom range, θ 1 >1,0<θ 2 <1.

[0036] The present invention provides an improved multi-UAV maximization coverage deployment method, firstly, by analyzing the initial target information, based on the Weiszfeld algorithm decentralized after adding disturbance to prevent local optimality to obtain the preferred coverage set, after locating the initial position of the UAV, the SDG algorithm is used to optimize the energy efficiency and plan the path, specifically by considering the energy consumption, speed, moving distance and other factors of the UAV, the energy efficiency of the UAV is optimized to achieve the goal of minimizing the energy consumption of the system, that is, the SW algorithm (Strong Weiszfeld Algorithm, SW) is first proposed to solve the coverage problem, and then the SDG algorithm (Steepest Descent with Goldstein Algorithm, SDG) is proposed to solve the energy consumption problem. The present invention not only achieves stable and efficient coverage of the target area through precise positioning and energy efficiency optimization, but also greatly reduces the energy consumption of the UAV, and enhances the durability and reliability of the system. In addition, the present invention also has the ability to quickly respond to environmental changes and re-plan coverage paths, effectively responding to challenges in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 It is a schematic flow chart of the steps of an improved multi-UAV maximum coverage deployment method of the present invention.

[0039] Figure 2 It is a schematic diagram of the improved SW algorithm flow of the present invention.

[0040] Figure 3 It is a schematic diagram of the improved SDG algorithm flow of the present invention. DETAILED DESCRIPTION

[0041] Embodiments of the present invention are described in detail below, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and should not be construed as limiting the present invention.

[0042] The following is an explanation of the English abbreviations mentioned in the article:

[0043] SW algorithm (Strong Weiszfeld Algorithm, SW): SW algorithm inherits the global search advantages of the traditional Weiszfeld algorithm and introduces a multiple perturbation mechanism;

[0044] The SDG algorithm (Steepest Descent with Goldstein Algorithm, SDG) combines the inexact search method--Goldstein rule with the Steepest Descent Algorithm (SD).

[0045] See also Figure 1 The present invention provides an improved multi-UAV maximum coverage deployment method, comprising the following steps:

[0046] S1: Use Kalman filter to obtain target position information;

[0047] S2: Based on the Weiszfeld algorithm, disturbances are added after decentralization to prevent local optimality and find the maximum coverage set of drones;

[0048] S3: Locate the drone’s location information;

[0049] S4: SD algorithm combined with Goldstein rule to iteratively optimize energy efficiency;

[0050] S5: Plan the deployment path according to the iterative numerical value to achieve path optimization.

[0051] The following is a further explanation based on the specific implementation steps:

[0052] The purpose of step S1 is to obtain target location information.

[0053] Obtaining target location information is the first step in using the Kalman filter for target coverage deployment. The Kalman filter is an effective recursive filter that can estimate the state of a dynamic system from noisy measurement data. In target coverage, the Kalman filter can be used to estimate the location of the target. The core of the Kalman filter lies in its update formula:

[0054] x k =xk-1 +K k (z k -Hx k-1 ) (1)

[0055] Among them, x k is the state estimate at time k, x k-1 is the predicted state based on the previous estimate, K k is the Kalman gain, z k is the observed value, and H is the observation model.

[0056] Through this iterative process, the Kalman filter can not only handle the noise in the sensor data, but also adapt to the rapid changes in the target's motion state, providing an efficient and accurate solution for obtaining target position information.

[0057] The execution process of step S2 is to calculate the maximum covering set.

[0058] In the task of covering the target area, calculating the maximum coverage set is a crucial task. The purpose of this process is to ensure that every point in the target area can be effectively monitored or served through optimal resource allocation. In step S2, based on the improvement of Weiszfeld, a corresponding SW algorithm is proposed. The algorithm flow is described in detail in Figure 2 shown.

[0059] First, the invention starts with defining the target area and available drone resources. Then, the center of the circle located in the target dense area is identified and retained, which are potential locations for drone deployment and can provide the largest initial coverage.

[0060] In order to further optimize the positions of these circle centers, the Weiszfeld algorithm is used for decentralization, as shown in formulas (2) and (3). This is an iterative algorithm that finds the optimal circle center position by calculating the weighted average center point, thereby achieving more efficient coverage.

[0061]

[0062] Then, a perturbation range is set for the center of the circle, and a perturbation test is performed. This means that a small adjustment is made to the current center position to observe changes in the coverage effect. If the number of targets covered increases after the perturbation, the center position is updated, see formula (4).

[0063] BEST i n+1 =BESTp i n +δ*r i (4)

[0064] p i is the iterative deployment position of the i-th UAV. Each UAV will finally iterate to an optimal position BESTpi*. BESTpi* belongs to C*, that is, C* is the “optimal coverage set” of the UAV swarm. δ is the preset small disturbance range, r i is a random vector drawn from some probability distribution (such as the normal distribution).

[0065] After updating the center position, you need to evaluate whether the coverage effect has changed. If the disturbance causes the number of covered targets to increase, keep the new position; if there is no change or a decrease, you may need to make further adjustments or keep the original position.

[0066] Once the optimal center positions are determined, these centers are marked as key centers BESTp i * , and remove the target positions covered by these circle centers from the target set. This helps to further optimize the coverage strategy and ensure efficient use of resources.

[0067] Finally, it is necessary to check whether all target points have been covered or whether the preset number of drones have been deployed. If either condition is met, the optimization process ends. If the target has not been reached, adjustments and optimizations will continue until maximum coverage is achieved.

[0068] Step S3: Locate the drone location information:

[0069] The system locates each drone through the GPS signaler carried by the drone to ensure accurate navigation and monitoring when performing tasks. As a key component of the drone, the GPS module receives signals from satellites in the earth's orbit and calculates the three-dimensional coordinates of the drone in real time: latitude, longitude and altitude, as shown in formula (5). In this regard, the relative distance of the drone to the target is shown in formula (6).

[0070]

[0071] where x i ,y i is the horizontal and vertical coordinates of the i-th UAV, x j ,y j are the horizontal and vertical coordinates of the j-th target.

[0072] Step S4 is an iterative optimization process for energy efficiency:

[0073] In terms of improving the energy efficiency of UAV missions, iterative optimization methods play a key role. To address this problem, the present invention proposes an improved SDG algorithm, which can accurately select an appropriate step size in combination with the Goldstein rule, and then iteratively update the position of the UAV to ensure that each step of the update moves towards the direction of optimizing energy efficiency. The specific process is detailed in Figure 3 .

[0074] First, an objective function is defined, as shown in formulas (7), (8), and (9). This function comprehensively considers factors such as flight path, speed, and energy consumption to quantitatively evaluate the energy efficiency performance of the UAV.

[0075] L(D,N)=μ d l d +μ c l c =μ d D+μ c N (7)

[0076] μ d +μ c =1 (8)

[0077] min f(x)=L(D(C * ),N(C * )) (9)

[0078] Among them, L represents the loss of the drone, D represents the total distance between the location point and the drone, N represents the total number of drone resources around each mission point, and μ d and μ c is the weight coefficient for balancing distance loss and information loss, which is a positive real number. At the same time, two key performance indicators are introduced: D(C * ),N(C * ). D(C * ) refers to the sum of the straight-line distances between all users and their corresponding drones when the optimal coverage set strategy is adopted. N(C * ) represents the sum of the number of drones deployed around each user in the optimal coverage set.

[0079] Next, the system calculates the gradient of the objective function. This step reveals the location with the worst energy efficiency, that is, the direction that needs to be optimized. Subsequently, the system must select an appropriate step size α, which is a key parameter in the iterative search algorithm and determines the amplitude of the drone position update in each iteration. The judgment of whether the step size α is too large or too small is determined by formula (10), and the amount of enlargement and reduction is handled according to formula (11).

[0080]

[0081] Among them, parameter c1 , c 2 is a parameter for controlling step size selection, satisfying 0 < c < 0.5. The right side is the Armijo condition to ensure that the step size α k will not be too large; while the left side is to avoid finding an α k that is too small, resulting in an overly slow algorithm operation speed. Among them, the positive scalar α k and the vector p k are the step size and the direction of incremental descent respectively.

[0082]

[0083] where θ 1 , θ 2 are the scaling ranges, θ 1 > 1, 0 < θ 2 < 1.

[0084] Once the step size is adjusted to an appropriate size, the system will output this parameter and use it to update the position of the UAV, while saving the new position information. Subsequently, the system will check the current optimization progress, evaluate whether the objective function value has dropped below the preset limit value, or whether the number of iterations has reached the preset upper limit.

[0085] If the optimization goal has been achieved, that is, the energy efficiency reaches the expectation or the number of iterations is exhausted, the system will end the iterative optimization process. Otherwise, the system will return to the gradient calculation step and continue the next round of iterative optimization. Through this process, the UAV can minimize energy consumption while ensuring the task execution efficiency, achieving the optimization of energy efficiency, which is of great significance for extending the flight time of the UAV and improving the task execution efficiency.

[0086] Step S5: Plan and deploy the path according to the iterative values to achieve path optimization.

[0087] In the iterative optimization of energy efficiency in step S4, the present invention not only updates and saves the position of the UAV in each iteration, but also based on these data, can carefully plan an efficient moving path for the UAV. This path is the result of careful calculation, which takes into account the current position of the UAV, the gradient information of the objective function, and the optimized step size, ensuring that the UAV can achieve the maximum coverage effect with the minimum energy consumption when performing tasks.

[0088] By this method, the flight path of the UAV is no longer a single straight line or a preset trajectory, but a dynamically optimized path that can respond to environmental changes and task requirements in real time. Such path planning not only improves the energy efficiency of the UAV, but also enhances its adaptability and flexibility in complex environments.

[0089] Moreover, this optimized movement path can also serve as an important reference for the deployment of drones, providing decision support for drone operators or automatic control systems, thereby achieving more accurate and efficient task allocation and resource management. In this way, it can ensure that drones can achieve optimal performance in terms of energy consumption and time efficiency when performing tasks.

[0090] What is disclosed above is only a preferred embodiment of the present invention, and it certainly cannot be used to limit the scope of rights of the present invention. Ordinary technicians in this field can understand that all or part of the processes of the above embodiment and equivalent changes made according to the claims of the present invention still fall within the scope of the invention.

Claims

1. An improved multi-UAV maximum coverage deployment method, characterized in that: The following steps are involved: Step 1: Use Kalman filter to obtain target position information; Step 2: Based on the Weiszfeld algorithm, after decentralization, disturbance is added to prevent local optimality and find the maximum coverage set of drones; Step 3: Locate the drone’s location information; Step 4: SD algorithm combined with Goldstein rule iteratively optimizes energy efficiency; Step 5: Plan the deployment path according to the iterative value to achieve path optimization.

2. The improved multi-UAV maximum coverage deployment method as claimed in claim 1, characterized in that: The update formula of the Kalman filter in step 1 is x k =x k-1 +K k (z k -Hx k-1 ) Among them, x k is the state estimate at time k, x k-1 is the predicted state based on the previous estimate, K k is the Kalman gain, z k is the observed value, and H is the observation model.

3. The improved multi-UAV maximum coverage deployment method as claimed in claim 2, characterized in that: In step 2, the execution process of adding disturbance to prevent local optimality after decentralization based on the Weiszfeld algorithm includes the following steps: Step 2.1: Use the Weiszfeld algorithm to perform decentralization and find the optimal center position by calculating the weighted average center point; Step 2.2: Set a disturbance range for the center of the circle and perform a disturbance test; Step 2.3: If the number of targets covered increases after the disturbance, update the position of the circle center and evaluate the coverage effect; If the target covered does not change or decreases after the disturbance, further adjust or maintain the original position; Step 2.4: After determining the optimal circle center position, remove the target position covered by the corresponding circle center from the target set; Step 2.5: Check whether all target points have been covered or whether the preset number of drones have been deployed; if either condition is met, the optimization process ends; if the target has not been reached, adjustment and optimization will continue.

4. The improved multi-UAV maximum coverage deployment method as claimed in claim 3, characterized in that: The formula for the relative distance between the drone and the target in step 3 is as follows: where x i ,y i is the horizontal and vertical coordinates of the i-th UAV, x j ,y j are the horizontal and vertical coordinates of the j-th target, U is the drone group, ui is the i-th drone, O is the target user, and Oj is the j-th user; P(u i ) is the drone u i The spatial position of P(o i ) is the target user o i The spatial position of the UAV; H is the altitude of the UAV relative to the ground.

5. The improved multi-UAV maximum coverage deployment method as claimed in claim 4, characterized in that: The objective function expression used to quantitatively evaluate the energy efficiency of the drone in step 4 is as follows: L(D,N)=μ d l d +m c l c =μ d D+μ c N m d +m c =1 min f(x)=L(D(C * ),N(C * )) Among them, L represents the loss of the drone, D represents the total distance between the location point and the drone, N represents the total number of drone resources around each mission point, and μ d and μ c is the weight coefficient for balancing distance loss and information loss, which is a positive real number. * ) refers to the sum of the straight-line distances between all users and their corresponding drones when the optimal coverage set strategy is adopted, N(C * ) represents the sum of the number of drones deployed around each user in the optimal coverage set.

6. The improved multi-UAV maximum coverage deployment method as claimed in claim 5, characterized in that: The SD algorithm in step 4 is used to determine the search direction, and the Goldstein rule selects an appropriate step size α. The judgment of whether the step size α is too large or too small is determined by the following formula: Among them, parameters c1 and c2 are parameters for controlling step size selection, satisfying 0 < c < 0.

5. The right side is the Armijo condition to ensure that the step size α k will not be too large; while the left side is to avoid finding an α k that is too small, resulting in an overly slow algorithm operation speed. The positive scalar α k and the vector p k are the step size and the direction of incremental descent, respectively; The step size α is enlarged or reduced according to the following formula: Among them, θ1 and θ2 are the scaling ranges, θ1>1, 0<θ2<1.

Citation Information

Cited By

  • Task scheduling-oriented unmanned aerial vehicle positioning deployment method based on scale-free attribute

    CN121486971A