Spectrum trajectory joint optimization method and system for unmanned aerial vehicle task efficiency

Abstract

The application discloses a spectrum trajectory joint optimization method and system for unmanned aerial vehicle task execution efficiency, and relates to the technical field of unmanned aerial vehicle task time optimization. When all the sensing nodes are located in the communication range of the unmanned aerial vehicle, a minimum unmanned aerial vehicle task time optimization model is constructed. A convex processing is performed on the minimum unmanned aerial vehicle task time optimization model by using a continuous convex approximation and a block coordinate descent method. The minimum unmanned aerial vehicle task time optimization model after the convex processing comprises an unmanned aerial vehicle task time and unmanned aerial vehicle flight trajectory joint optimization model and an unmanned aerial vehicle task time and channel allocation factor joint optimization model. The unmanned aerial vehicle task time and unmanned aerial vehicle flight trajectory joint optimization model and the unmanned aerial vehicle task time and channel allocation factor joint optimization model are iteratively optimized to determine the unmanned aerial vehicle task time. The application further reduces the task time of the unmanned aerial vehicle and improves the execution efficiency of the unmanned aerial vehicle.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) mission time optimization technology, and in particular to a spectrum trajectory joint optimization method and system for improving the efficiency of UAV mission execution. Background Technology

[0002] Drones have found widespread application in wireless communication due to their excellent maneuverability, deployment flexibility, and low cost. In drone-assisted uplink sensor networks, drones collect sensor information from various sensor nodes and bring the collected information back to headquarters, reducing the complex wired deployment between headquarters and individual sensor nodes. In drone-assisted uplink sensor networks, drones are used to perform information collection tasks. To improve the efficiency of drone missions, information collection must be completed in the shortest possible time. The time required for a drone to perform a mission is related to its flight path and the planning of spectrum resources.

[0003] Existing literature largely treats UAV spectrum, trajectory, and other communication resources as decision variables for joint optimization, focusing on energy efficiency, spectral efficiency, and system throughput. Research specifically addressing the efficiency of UAV mission execution is relatively limited. Current technologies primarily optimize UAV flight speed and trajectory to minimize information acquisition time. However, during flight, the spectrum allocated to each sensor node remains constant, failing to consider the impact of time-varying spectrum on mission duration. Summary of the Invention

[0004] The purpose of this invention is to provide a spectrum trajectory joint optimization method and system for improving the efficiency of UAV mission execution. While considering the UAV's flight trajectory and speed, the amount of spectrum allocated to each sensor node by the UAV is used as an optimization variable, which can further reduce the UAV's mission time and improve the UAV's mission execution efficiency.

[0005] To achieve the above objectives, the present invention provides the following solution:

[0006] In a first aspect, the present invention provides a spectrum trajectory joint optimization method for improving the efficiency of UAV mission execution, characterized by comprising:

[0007] When all sensor nodes are within the UAV's communication range, a minimum UAV mission time optimization model is constructed. All sensor nodes are located within a circular sensing area, the radius of which is smaller than the radius of the UAV's communication range. When the UAV enters the circular sensing area, it allocates spectrum resources to each sensor node using frequency division multiple access (FDMA). Each sensor node uses its allocated spectrum resources to upload sensing information. After receiving all the sensing information uploaded by the sensor nodes, the UAV flies away from the circular sensing area. The decision variables of the minimum UAV mission time optimization model include the UAV mission time, the UAV flight trajectory, and the channel allocation factor.

[0008] The minimum UAV mission time optimization model is convexized using continuous convex approximation and block coordinate descent method. The convexized minimum UAV mission time optimization model includes a joint optimization model of UAV mission time and UAV flight trajectory, and a joint optimization model of UAV mission time and channel allocation factor.

[0009] The joint optimization model of UAV mission time and UAV flight trajectory, as well as the joint optimization model of UAV mission time and channel allocation factor, are iteratively optimized to determine the UAV mission time.

[0010] Secondly, this invention provides a spectrum trajectory joint optimization system for improving the efficiency of UAV mission execution, comprising:

[0011] A minimum UAV mission time optimization model construction module is used to construct a minimum UAV mission time optimization model when all sensor nodes are within the UAV's communication range. Specifically, all sensor nodes are located within a circular sensing area, the radius of which is smaller than the radius of the UAV's communication range. When the UAV enters the circular sensing area, it allocates spectrum resources to each sensor node using frequency division multiple access (FDMA). Each sensor node uses its allocated spectrum resources to upload sensing information. After receiving all the sensing information uploaded by the sensor nodes, the UAV flies out of the circular sensing area. The decision variables of the minimum UAV mission time optimization model include the UAV mission time, the UAV flight trajectory, and the channel allocation factor.

[0012] The convexity processing module is used to perform convexity processing on the minimum UAV mission time optimization model using continuous convex approximation and block coordinate descent method; the minimum UAV mission time optimization model after convexity processing includes a joint optimization model of UAV mission time and UAV flight trajectory, and a joint optimization model of UAV mission time and channel allocation factor.

[0013] The UAV mission time calculation module is used to iteratively optimize the joint optimization model of UAV mission time and UAV flight trajectory, as well as the joint optimization model of UAV mission time and channel allocation factor, to determine the UAV mission time.

[0014] According to specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0015] This invention is based on an UAV-assisted uplink wireless sensor network. By jointly optimizing the UAV flight trajectory and spectrum resources, it minimizes the UAV mission time while ensuring the upload volume of each sensor node. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 A flowchart of a spectrum trajectory joint optimization method for improving the efficiency of UAV mission execution provided in an embodiment of the present invention;

[0018] Figure 2 This is a schematic diagram of an information collection network assisted by unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention.

[0019] Figure 3 This is a diagram illustrating the process of determining the starting point and target point of the UAV in each circle according to an embodiment of the present invention.

[0020] Figure 4 This is a UAV trajectory diagram under the proposed algorithm 3 provided in the embodiments of the present invention;

[0021] Figure 5 This is a spectrum allocation diagram of each sensing node provided in an embodiment of the present invention;

[0022] Figure 6 This is a convergence graph of the upload volume of each sensor node as a function of the number of iterations, provided in an embodiment of the present invention.

[0023] Figure 7 These are drone trajectory diagrams under different algorithms provided in embodiments of the present invention;

[0024] Figure 8 A convergence graph of UAV mission time with the number of iterations provided in an embodiment of the present invention;

[0025] Figure 9 This is a time-consuming diagram of drones under different upload volumes and different algorithms provided in an embodiment of the present invention;

[0026] Figure 10 This is a time-travel diagram of a UAV under different signal-to-noise ratios and different algorithms, provided in an embodiment of the present invention.

[0027] Figure 11 The spectrum allocation diagram of sensing node 2 provided in the embodiment of the present invention;

[0028] Figure 12 This invention provides drone trajectories and time-based graphs under different algorithms in embodiments of the invention.

[0029] Figure 13 The diagram shows the structure of a spectrum trajectory joint optimization system for improving the efficiency of UAV mission execution, as provided in this embodiment of the invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] The main research focus of this invention is on a UAV-assisted uplink wireless sensor network. By jointly optimizing the UAV flight trajectory and spectrum resources, it minimizes the UAV mission time while ensuring the upload volume of each sensor node. The main contributions of this invention are summarized as follows:

[0033] 1) First, this invention establishes a minimum UAV mission time optimization model where all sensor nodes are within the UAV's communication range. To address the non-convexity of this model, this invention uses a successful convex approximation (SCA) to transform some non-convex constraints, thus ensuring the uplink volume of each sensor node while solving for the minimum UAV mission time.

[0034] 2) Secondly, this invention extends the proposed algorithm to general scenarios. In this case, the distribution of sensor nodes is no longer limited to the communication range of the UAV. By combining algorithms such as minimum circle coverage and TSP, the minimum task time of the UAV in general scenarios can be solved.

[0035] 3) Finally, simulation results show that, compared with other benchmark algorithms, the algorithm proposed in this invention can effectively reduce the mission time of the UAV while ensuring the communication upload volume of the sensor node.

[0036] Example 1

[0037] like Figure 1 As shown in the figure, the present invention provides a spectrum trajectory joint optimization method for improving the efficiency of UAV mission execution, which includes the following steps.

[0038] Step 100: When all sensor nodes are within the communication range of the UAV, construct a minimum UAV mission time optimization model.

[0039] The process for determining the minimum UAV mission time optimization model is as follows:

[0040] like Figure 2 As shown, K sensing nodes are distributed within a circular sensing region of radius R, where K = {1, 2, ..., k, ..., K}. The position of the k-th sensing node is represented by two-dimensional coordinates. k∈K represents the sensor nodes. The sensor nodes continuously perceive surrounding information and upload this information to the drones that periodically come to perform tasks. Assuming the drone is a rotary-wing drone, it flies into a circular sensing area and allocates spectrum resources to each sensor node using frequency division multiple access (FDMA). Each sensor node uses its allocated spectrum resources to upload sensing information. After receiving all the sensing information uploaded by all sensor nodes, the drone flies away from the circular sensing area.

[0041] Assuming the radius R of the circular sensing area is smaller than the radius D of the UAV's communication range, the UAV can simultaneously receive sensing information uploaded by all sensing nodes. When the UAV enters and leaves the circular sensing area, the coordinates of its projected points on the circumference are e and g, respectively. The amount of sensing information that the k-th sensing node needs to send is Q. k bits, k∈K, the time T taken by the drone from e to g is the drone's mission execution time, during which the drone always flies at a constant altitude H.

[0042] Since the UAV's mission execution time T is the optimization variable, a path discretization algorithm is used to divide the projected trajectory q of the UAV's flight path within the circular sensing area into M path segments, represented by M+1 discrete points, denoted as . This represents the set of ground-projected coordinates of the UAV along path segment M. Where q0 and q... M These represent the ground projection coordinates of the drone entering and leaving the circular sensing area, respectively. Therefore, we have

[0043] q0=e,q M =g (1).

[0044] The length of the (m+1)th path segment can be represented as ||q m+1 -q mWithin the same path segment, the drone's position is considered constant, and its speed is considered constant. To satisfy this condition, the length of each path segment should be sufficiently small and much smaller than the drone's flight altitude H. Therefore, the length of each path segment should satisfy the following condition:

[0045] ||q m+1 -q m ||≤Δ max (2).

[0046] Where, Δ max Δ represents the upper limit of the length of each path segment. max <<H. In addition, M and Δ max The maximum distance constraint must also be satisfied, namely:

[0047] M·Δ max ≥D u (3).

[0048] In the above formula, D u Let T be the total flight distance of the drone. m+1 Let v represent the flight time of the drone in the (m+1)th segment of the path, then the speed of the drone in that segment is v. m+1 It can be represented as:

[0049]

[0050] Considering the maximum speed constraint v of the rotary-wing UAV max ,have:

[0051] ||q m+1 -q m ||≤v max T m+1 (5).

[0052] The total flight time for a drone to perform a mission can be expressed as:

[0053]

[0054] Considering the good line-of-sight link between the high-altitude drone and the sensor nodes, a Line-of-Sight (LOS) model is used to describe the channel between the drone and each sensor node. When the drone is on the m-th path segment, its distance to sensor node k can be expressed as:

[0055]

[0056] W k This represents the coordinates of sensor node k.

[0057] At this time, the channel power gain h between the UAV and the sensor node mk Can be written as:

[0058]

[0059] Where β0 represents the channel power gain at a reference distance of 1m.

[0060] Assume the total spectrum held by the drone is B (MHz), and it is evenly divided into C (C≥K) sub-channels, each with a bandwidth of w (MHz). The binary variable θ k,c,m Channel allocation factors are defined as follows:

[0061]

[0062] Furthermore, assuming that the transmit power of each sensor node is P, the communication traffic of sensor node k when the UAV is on the m-th path segment can be obtained as follows:

[0063]

[0064] in, This represents the additive white Gaussian noise power at the drone receiver. To ensure successful uploading of sensing information from all sensor nodes, this represents the total communication upload volume of sensor node k during the drone's flight. It should be greater than the perceived information Q k ,have:

[0065]

[0066] Considering the continuity of communication between the UAV and the sensor nodes, it is assumed that each sensor node obtains at least one channel in each segment of the UAV's path. Furthermore, the same channel is assigned to at most one sensor node at any given time, with a channel allocation factor θ. k,c,m It should meet the following requirements:

[0067]

[0068]

[0069] During the mission, the drone always flew within the circular sensing area, therefore:

[0070] ||q m+1 -o xy ||≤R (14).

[0071] In the above formula, o xy This indicates the center of the circular sensing area.

[0072] Based on the above analysis, the problem of minimizing the execution time of a UAV mission based on spectrum resource allocation and flight trajectory optimization can be written as the following optimization problem:

[0073]

[0074] In the above formula, C2 can be obtained from formulas (2) and (5), and C3 can be obtained from formulas (10) and (11), where, Furthermore, C3 is a non-convex constraint and involves the multiplication of multiple optimization variables; C3 to C6 are integer programming problems.

[0075] in, The objective function is the sum of the times for all path segments of the UAV; T m {q} represents the flight time of the UAV in the m-th path segment; M represents the total number of paths; m} represents the set of ground-projected coordinates of the UAV along path segment M; θ k,c,m This represents the probability that the UAV will assign the c-th channel to the k-th sensor node when it is on the m-th path segment.

[0076] C1 represents the starting and ending constraints of the UAV; q0 represents the ground projection coordinates of the UAV entering the circular sensing area, e represents the ground projection coordinate value of the UAV entering the circular sensing area, and q M q represents the ground projection coordinates of the drone as it leaves the circular sensing area. M The ground projection coordinates of the UAV after it leaves the circular sensing area; g represents the ground projection coordinate value of the UAV after it leaves the circular sensing area;

[0077] C2 represents the path segment length constraint condition for the UAV; Δ max This represents the upper limit of the length of each path segment; v max q represents the maximum speed of the drone; m+1 q represents the ground projection coordinates of the UAV on the (m+1)th segment of its path; m This represents the ground projection coordinates of the UAV on the m-th segment of its path;

[0078] C3 represents the upload constraints for each sensor node; P represents the transmit power of the sensor node, and β0 represents the channel power gain at a reference distance of 1m. The summative white Gaussian noise power at the UAV receiver is represented by w; the channel bandwidth is w; the UAV's flight altitude is h; and Q represents the UAV's altitude. k Represents perceived information;

[0079] C4 represents the constraint condition for the range of channel allocation factor values;

[0080] C5 represents the constraint that each sensing node must obtain at least one channel;

[0081] C6 represents the constraint that a channel can be allocated to at most one sensor node, and K represents the total number of sensor nodes;

[0082] C7 represents the constraint that the drone's flight trajectory lies within the circular sensing area. xy R represents the center of the circular sensing area, and R represents the radius of the circular sensing area.

[0083] C8 represents a constraint condition where the task time is a non-negative value.

[0084] Step 200: Using the continuous convex approximation and block coordinate descent method, the minimum UAV mission time optimization model is convexized; the minimum UAV mission time optimization model after convexization includes a joint optimization model of UAV mission time and UAV flight trajectory, and a joint optimization model of UAV mission time and channel allocation factor.

[0085] Since (P1) is a mixed integer nonconvex problem, it is difficult to find the global optimum directly. Therefore, the following method is used to solve it.

[0086] When all sensor nodes are within the communication range of the drone

[0087] First, the communication upload constraint C3 in (P1) is made convex. Then, a slack variable A is introduced. k,c,m ,make

[0088]

[0089] Then C3 can be written as:

[0090]

[0091] The above equation remains non-convex, but the quadratic part on the left-hand side can be bounded by the function at a specified iteration point using SCA. For the left-hand side of the above equation at the l-th iteration point... Performing a second-order Taylor expansion at that point, we obtain... At point The lower bound:

[0092]

[0093] Substituting the lower bound in the above formula into formula (17) yields:

[0094]

[0095] Next, we address the right-hand side of equation (16). The equal sign in the equation can be rewritten as less than or equal to:

[0096]

[0097] This does not affect the equivalence of the two, because as long as the less than sign holds, T on the right side of the equation... mWe can continue to optimize and reduce until the two are equal. Similarly, we use SCA to adjust the logarithmic part on the right-hand side of the above equation at the l-th iteration point. Performing a second-order Taylor expansion at this point yields a lower bound for the logarithmic part:

[0098]

[0099] in,

[0100]

[0101]

[0102] Substituting formula (20) into formula (21) yields:

[0103]

[0104] Due to θ k,c,m ||q m -w k || 2 Since the product of decision variables is non-convex, we introduce the Block Coordinate Descent (BCD) algorithm. This algorithm divides the decision variables into different subsets and iterates alternately over the decision variables in different subsets. When iterating over the decision variables of a certain subset, the decision variables of other subsets are treated as constants, thus transforming the original problem into multiple convex subproblems. Because the decision variables {T}... m} Simultaneously affected by {q m} and {θ k,c,m Due to the influence of}, all decision variables are divided into two subsets, namely {q}. m ,T m} and {θ k,c,m ,T m By iterating alternately over the two sets of variables, a suboptimal solution to the original problem can be obtained.

[0105] Step 300: Iteratively optimize the joint optimization model of UAV mission time and UAV flight trajectory, and the joint optimization model of UAV mission time and channel allocation factor to determine the UAV mission time.

[0106] For the first time, {θ k,c,m Treat} as a constant, for the subset {q m ,T m By iterating, the original optimization problem (P1) can be written as:

[0107]

[0108] Replacing constraint C3 with equations (19) and (24), we obtain...

[0109]

[0110] At this point, (P1.2) has been transformed into a convex optimization, and therefore can be solved using the CVX solver. The detailed solution process of (P1.1) is summarized as Algorithm 1.

[0111] Table 1. Solution based on SCA iterative algorithm (P1.1)

[0112]

[0113]

[0114] Next, {q m Treat} as a constant, for {θ k,c,m ,T m By iterating, the original optimization problem (P1) can be written as:

[0115]

[0116] Divide both sides of equation (20) by wT simultaneously. m Can be written

[0117]

[0118] In the above formula, the left side is about {A} m,k,c} and {T m The convex function of}, the right side of which is about {θ} k,c,m The affine function of}, therefore the above expression is about {θ} k,c,m} and {T m The convex function of}. Finally, the integer variable θ in (P1) k,c,m Relaxing the variables as continuous variables yields the following results:

[0119] 0≤θ k,c,m ≤1 (29).

[0120] Therefore, the original optimization problem (P1) can be rewritten as follows:

[0121]

[0122] (P1.4) is a convex optimization, so it can be solved efficiently using CVX. The detailed solution process for (P1.3) is summarized in Algorithm 2.

[0123] Table 2. Solution based on SCA iterative algorithm (P1.3)

[0124]

[0125]

[0126] By iterating over the subproblems (P1.1) and (P1.3) alternately, the solution to the original optimization problem (P1) can be obtained. The detailed process is summarized as Algorithm 3.

[0127] Table 3. Solution based on SCA iterative algorithm (P1)

[0128]

[0129]

[0130] Furthermore, the method provided in this embodiment of the invention also includes:

[0131] (1) When some sensor nodes are within the communication range of the UAV and other sensor nodes are not within the communication range of the UAV, a spiral algorithm is used to divide the sensor nodes and obtain multiple circular sensing areas.

[0132] (2) By constructing a minimum UAV mission time optimization model, the mission time of the UAV in each circular sensing area is determined.

[0133] (3) Use the minimization function of the external path to determine the mission time when the UAV is not in the circular sensing area.

[0134] (4) The task time of the UAV in each circular sensing area is added together with the task time of the UAV outside the circular sensing area to obtain the final task time of the UAV.

[0135] When the distribution range of sensor nodes exceeds the communication range of the UAV, to improve the applicability of the algorithm, it is combined with helical algorithms, TSP, etc., thus making it applicable to scenarios with a wide range of sensor nodes. Assume the UAV takes off from headquarters, collects the perception information from each sensor node, and returns to headquarters after completion.

[0136] First, a spiral algorithm is used to cover all sensor nodes with a minimum number of circles of diameter D, thus dividing all sensor nodes into several parts. Since all nodes within the same circle can communicate with the drone simultaneously, the proposed algorithm can be used to find the minimum time for the drone within each circle. Therefore, the minimum time for the drone in a typical scenario can be considered as consisting of two parts: the time inside the circle and the time outside the circle. By minimizing the time inside the circle and the time outside the circle, the drone's mission time can be minimized.

[0137] For out-of-circle time, since the UAV does not communicate with the sensor nodes while flying outside the circles, it can fly at its maximum speed. Therefore, minimizing out-of-circle time can be transformed into minimizing out-of-circle distance. Treating the centers of each circle and the headquarters as city nodes in the Time-of-Sight (TSP) problem, and using the TSP algorithm, the minimum out-of-circle flight distance of the UAV is obtained by determining the order in which the UAV visits each circle. Assuming all sensor nodes are covered by G circles, the minimum distance obtained from TSP is L, and the minimum flight distance and time of the UAV are L-GD and (L-GD) / V, respectively. max , where GD represents the distance within the circle.

[0138] For time within a circle, the intersection of the shortest straight path determined by the TSP with each circle can be considered as the starting and target points of the UAV in each circle, such as... Figure 3 As shown in the diagram, Algorithm 3 is applied within each circle to obtain the minimum time for the drone in each circle. The process of finding the minimum time for the drone in a general scenario is summarized as Algorithm 4.

[0139] Table 4. Minimum Drone Time Calculation Table for General Scenarios

[0140]

[0141]

[0142] The technical solution protected by this invention will be verified through simulation below.

[0143] This section verifies the proposed algorithm through simulation; some simulation parameters are shown in Table 5.

[0144] Table 5 shows some simulation parameters.

[0145]

[0146] First, consider the scenario where all sensor nodes are within the UAV's communication range. Let the radius of the circle be D0 = 800m, and the number of sensor nodes distributed within it be K = 3. The amount of sensing information generated by each sensor node is Q. k =500 MBits. The positions of each sensor node, as well as the coordinates of the UAV's starting and target points, are given in Table 6.

[0147] Table 6: Coordinates of each point

[0148]

[0149] Based on the simulation parameters above, the proposed algorithm 3 is used to solve problem (P1). First, the initial conditions are set. The line connecting the starting point e and the target point g is sampled at equal intervals. In this simulation, the number of paths M is assumed to be 203, meaning 203 points are sampled as the initial trajectory of the UAV. Furthermore, assuming the UAV spectrum is evenly distributed among the three sensor nodes, the initial spectrum allocation is obtained. Initial duration of each path segment Through T max / M is obtained, while the drone is in T max The task will definitely be completed within the time limit, during simulation T max It can be obtained from the following formula

[0150]

[0151] In the formula, it is assumed that the drone always maintains the maximum distance from each node, which is the diameter of the circle. Δ t It is an additional time item.

[0152] In the simulation, Δ t The value is set to 50. Furthermore, the parameter γ in Algorithm 3 is set to 0.5. The UAV trajectory and spectrum allocation obtained using the above simulation parameters and Algorithm 3 are as follows: Figures 4-12 As shown. Figure 4 The mission time of the drone is marked in the text. Figure 6 The convergence curves of the upload volume of each sensor node as a function of the number of iterations are presented. Figure 7 The proposed algorithm and the trajectories of the UAV under different algorithms are presented. Figure 8 Convergence curves of UAV mission time versus iteration number are presented for the proposed algorithm and other algorithms. Figure 9 The time taken by the proposed algorithm and other algorithms for drone missions under different target upload volumes is presented. Figure 10 The time taken for UAV missions using the proposed algorithm and other algorithms is presented under different signal-to-noise ratios at the UAV receiver. Figure 11 The spectral distribution curves of sensor node SN2 under the proposed algorithm and other algorithms are presented. Figure 12 The drone trajectories and corresponding task times of the proposed algorithm and other algorithms in a general scenario are presented.

[0153] Compared to other algorithms, the proposed algorithm reduces drone mission time by 6.4%, 11.9%, 27.7%, and 32.9%, respectively.

[0154] Example 2

[0155] In order to implement the method corresponding to Embodiment 1 above and achieve the corresponding functions and technical effects, a spectrum trajectory joint optimization system for improving the efficiency of UAV mission execution is provided below.

[0156] like Figure 13 As shown, a spectrum trajectory joint optimization system for improving the efficiency of UAV mission execution includes:

[0157] Minimum UAV mission time optimization model construction module 1 is used to construct a minimum UAV mission time optimization model when all sensor nodes are within the UAV's communication range. Specifically, all sensor nodes are located within a circular sensing area, the radius of which is smaller than the radius of the UAV's communication range. When the UAV enters the circular sensing area, it allocates spectrum resources to each sensor node using frequency division multiple access (FDMA). Each sensor node uses the allocated spectrum resources to upload sensing information. After receiving all the sensing information uploaded by the sensor nodes, the UAV flies away from the circular sensing area. The decision variables of the minimum UAV mission time optimization model include UAV mission time, UAV flight trajectory, and channel allocation factor.

[0158] The convexity processing module 2 is used to perform convexity processing on the minimum UAV mission time optimization model using continuous convex approximation and block coordinate descent method; the minimum UAV mission time optimization model after convexity processing includes a joint optimization model of UAV mission time and UAV flight trajectory, and a joint optimization model of UAV mission time and channel allocation factor.

[0159] The UAV mission time calculation module 3 is used to iteratively optimize the joint optimization model of UAV mission time and UAV flight trajectory, as well as the joint optimization model of UAV mission time and channel allocation factor, to determine the UAV mission time.

[0160] This invention addresses the problem of minimizing the mission completion time of unmanned aerial vehicles (UAVs) under limited spectrum resources, and studies a joint optimization method for spectrum resources and trajectory based on block coordinate descent and continuous convex approximation. First, a minimum time optimization model is studied and established with all sensor nodes within the UAV's communication range. Second, since the established model is a non-convex optimization, a convergent iterative algorithm based on block coordinate descent and continuous convex approximation is proposed to transform the model into a convex optimization problem for solution. Finally, the algorithm is applied to a general scenario for simulation analysis. Simulation results show that, while ensuring the upload capacity of each sensor node, the proposed algorithm can effectively reduce the UAV's mission time compared to other benchmark algorithms.

[0161] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0162] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for jointly optimizing spectrum trajectory for unmanned aerial vehicle (UAV) to perform a task efficiently, characterized in that, The application relates to an unmanned aerial vehicle (UAV) task time optimization method. When all the sensor nodes are located in the communication range of the UAV, a minimum UAV task time optimization model is constructed; all the sensor nodes are located in a circular sensing area, and the radius of the circular sensing area is smaller than the radius of the communication range of the UAV; when the UAV flies into the circular sensing area, the UAV uses frequency division multiple access to allocate frequency spectrum resources to the sensor nodes; the sensor nodes upload sensing information by using the obtained frequency spectrum resources; after the UAV receives all the sensing information uploaded by the sensor nodes, the UAV flies away from the circular sensing area; the decision variables of the minimum UAV task time optimization model include a UAV task time, a UAV flight trajectory and a channel allocation factor; The minimum UAV task time optimization model is convexly processed by using a continuous convex approximation and a block coordinate descent method; the convexly processed minimum UAV task time optimization model includes a UAV task time and UAV flight trajectory joint optimization model and a UAV task time and channel allocation factor joint optimization model; The UAV task time and UAV flight trajectory joint optimization model and the UAV task time and channel allocation factor joint optimization model are iteratively optimized to determine the UAV task time. 2.The method of claim 1, wherein, The minimum UAV task time optimization model is as follows: wherein, is an objective function; the objective function is the time sum of all path segments of the UAV; T m represents the flight time of the UAV in the mth path segment; M represents the total number of paths; {q m} represents the set of ground projection coordinates of the UAV in the M path segments; θ k,c,m represents the probability value of assigning the cth channel to the kth sensor node when the UAV is in the mth path segment; C1 represents the start constraint condition and the end constraint condition of the UAV; q0 represents the ground projection coordinate of the UAV entering the circular sensing area, e represents the ground projection coordinate value of the UAV entering the circular sensing area, q M represents the ground projection coordinate of the UAV leaving the circular sensing area, q M represents the ground projection coordinate of the UAV leaving the circular sensing area; g represents the ground projection coordinate value of the UAV leaving the circular sensing area; C2 represents a path segment length constraint condition of the UAV; Δ max represents an upper limit value of the length of each path segment; v max represents the maximum speed of the UAV; q m+1 represents the ground projection coordinates of the UAV on the (m+1)th path segment; q m represents the ground projection coordinates of the UAV on the mth path segment; C3 represents the upload constraint condition of each sensor node; P is the transmit power of the sensor node, β0 represents the channel power gain at the reference distance 1m, represents the additive white Gaussian noise power at the receiver end of the UAV; w is the channel bandwidth; H represents the UAV flight height value; Q k represents the perception information; C4 represents a channel allocation factor value range constraint condition; C5 represents a constraint condition that each sensor node obtains at least one channel; C6 represents a constraint condition that one channel is allocated to at most one sensor node, and K represents the total number of sensor nodes; C7 represents a constraint that the flight trajectory of the UAV is located within the circular sensing region, o xy o represents the center of the circular sensing region, and R represents the radius of the circular sensing region. C8 represents a constraint condition that the task time is a non-negative value. 3.The method of claim 1, wherein, The UAV task time and UAV flight trajectory joint optimization model and the UAV task time and channel allocation factor joint optimization model are iteratively optimized to determine the UAV task time, and the method specifically comprises the following steps. The UAV task time, the UAV flight trajectory and the channel allocation factor corresponding to the current iteration number are determined; the UAV flight trajectory corresponding to the current iteration number is determined by taking the channel allocation factor corresponding to the last iteration number as a fixed variable and solving the UAV task time and UAV flight trajectory joint optimization model; the channel allocation factor and the UAV task time corresponding to the current iteration number are determined by taking the UAV flight trajectory corresponding to the current iteration number as a fixed variable and solving the UAV task time and channel allocation factor joint optimization model; It is judged whether the difference between the UAV task time corresponding to the current iteration number and the UAV task time corresponding to the last iteration number is greater than a first set threshold value; If yes, the UAV task time corresponding to the current iteration number is determined as the final UAV task time; If no, the channel allocation factor corresponding to the last iteration number is updated as the channel allocation factor corresponding to the current iteration number, the UAV flight trajectory corresponding to the last iteration number is updated as the UAV flight trajectory corresponding to the current iteration number, and then the step of determining the UAV task time, the UAV flight trajectory and the channel allocation factor corresponding to the current iteration number is returned.

4. The method of claim 3, wherein, The UAV flight trajectory corresponding to the current iteration number is determined by taking the channel allocation factor corresponding to the last iteration number as a fixed variable, and solving the UAV task time and UAV flight trajectory joint optimization model by using a CVX solver.

5. The method of claim 4, wherein, The channel allocation factor and the UAV task time corresponding to the current iteration number are determined by taking the UAV flight trajectory corresponding to the current iteration number as a fixed variable, and solving the UAV task time and channel allocation factor joint optimization model by using a CVX solver and the preliminary task time corresponding to the current iteration number. The preliminary task time corresponding to the current iteration number is determined by taking the channel allocation factor corresponding to the last iteration number as a fixed variable, and solving the UAV task time and UAV flight trajectory joint optimization model by using a CVX solver.

6. The method of claim 1, wherein, Further comprising: When a part of the sensor nodes are located in the UAV communication range and another part of the sensor nodes are not located in the UAV communication range, the sensor nodes are divided by using a spiral algorithm to obtain a plurality of circular sensing areas; wherein the sensor nodes are located in the circular sensing areas; The task time of the UAV in each circular sensing area is determined by using the minimum UAV task time optimization model; The task time of the UAV not in the circular sensing area is determined by using the minimization function of the out-of-circle distance; The task time of the UAV in each circular sensing area and the task time of the UAV not in the circular sensing area are superimposed to finally obtain the task time of the UAV.

7. The method of claim 6, wherein, The task time of the UAV not in the circular sensing area is determined by using the minimization function of the out-of-circle distance, specifically comprising: The task time of the UAV not in the circular sensing area is determined by solving the minimization function of the out-of-circle distance by using a TSP algorithm.

8. A spectrum trajectory joint optimization system for improving the efficiency of UAV mission execution, characterized in that, Comprising: A minimum UAV task time optimization model construction module is configured to construct a minimum UAV task time optimization model when all the sensor nodes are located in the UAV communication range; wherein all the sensor nodes are located in the circular sensing areas, and the radius of the circular sensing area is smaller than the radius of the UAV communication range, and when the UAV flies into the circular sensing area, the UAV allocates frequency spectrum resources to each sensor node in a frequency division multiple access manner, each sensor node uploads sensing information by using the obtained frequency spectrum resources, and after the UAV receives all the sensing information uploaded by the sensor nodes, the UAV flies away from the circular sensing area; the decision variables of the minimum UAV task time optimization model include the UAV task time, the UAV flight trajectory, and the channel allocation factor; A convex processing module is configured to perform convex processing on the minimum UAV task time optimization model by using a continuous convex approximation and a block coordinate descent method; the convexly processed minimum UAV task time optimization model includes a UAV task time and UAV flight trajectory joint optimization model, and a UAV task time and channel allocation factor joint optimization model; A UAV task time calculation module is configured to iteratively optimize the UAV task time and UAV flight trajectory joint optimization model, and the UAV task time and channel allocation factor joint optimization model to determine the UAV task time.

Images