Method for optimizing the number and task completion time of multi-UAV-assisted IoT networks
By building a multi-UAV mission allocation model and ant colony algorithm to optimize the task allocation and trajectory of drones, the optimization of the number of drones and task completion time in the Internet of Things network is solved, and efficient data collection and cost reduction are achieved .
Patent Information
- Application Number
- CN202211033595.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-26
- Publication Date
- 2024-11-05
- Estimated Expiration
- 2042-08-26
AI Technical Summary
In the Internet of Things network, it is difficult for the existing technology to effectively optimize the number of drones and task completion time, resulting in the inability to meet the needs of emergency data collection tasks. At the same time, the purchase and maintenance costs of drones are high, and the existing path planning methods are available The time impact of drones under different data collection modes is not fully considered.
By building a multi-UAV mission allocation model, combining ant colony algorithm to optimize the task allocation and trajectory of the UAV, using a data collection model that combines the flying-hover-collecting mode to optimize the UAV's Flight path and hover position to minimize drone count and maximum mission completion time.
It realizes the need to shorten the maximum mission time of the drone while reducing network cost and overhead, and improves the efficiency of IoT data collection by optimizing the number and trajectory of the drone.
Smart Images

Figure CN115268498B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of mobile communication technology, and in particular relates to a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network. Background Art
[0002] With the rapid development of mobile communication technology, the outstanding performance of IoT application technology in the modern information industry has attracted widespread attention. IoT plays an important role in mobile detection application scenarios. In IoT networks, many ground IoT devices (GIDs) are deployed in areas with dense points of interest (PoI) to monitor information. Global controllers or gateways can be used to connect a large number of ground IoT devices and collect data from them for further analysis. However, the wide distribution of IoT devices and their diverse needs have brought new challenges to data collection in IoT networks.
[0003] As a new type of mobile platform, unmanned aerial vehicles (UAVs) can serve as a supplement to traditional ground communication systems, improve the communication connection, coverage and transmission quality of ground wireless devices, and play an important role in the communication network associated with IoT devices due to their mobility and flexibility. Due to the significant improvement of IoT performance brought by UAVs, UAV-assisted IoT systems have also received increasing attention, and UAV-assisted IoT data collection systems are one of the important applications. In certain emergency situations, such as disaster relief, industrial disasters and military surveillance, there are strict deadlines for the completion of the task of collecting data by UAVs. In addition, limited energy and cache capacity also pose new challenges to UAV data collection.
[0004] Although drones have great potential in assisting IoT data collection, there is still the problem of how to set the right number of drones in the IoT network. Due to the uncertainty of the task volume and the limited power and cache capacity of drones, a fixed number of drones may not be able to complete the data collection task while achieving the expected goal. At the same time, the purchase and maintenance costs of drones account for a large proportion of the entire network, and too many drones will bring unnecessary network overhead.
[0005] The application of drones in IoT networks, especially data collection, has received widespread attention. However, due to the urgency of the task and the limited network cost, the number of drones and the time to complete the task are greatly limited. In addition, with regard to the time to complete the task, the existing technical solutions mainly study the path planning of drones to shorten the time to complete the task. The analysis process does not consider the time of drones in different data collection modes. In fact, the time for drones to collect data will also greatly affect the time to complete the task. That is, the existing drone-assisted IoT path planning considerations are not comprehensive. Summary of the invention
[0006] To address the deficiencies of the above-mentioned prior art, the present invention provides a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network. By jointly optimizing the task allocation, hovering position and trajectory of the UAVs, the number of UAVs and the maximum task completion time are minimized, thereby achieving the goal of shortening the maximum task time of the UAVs while reducing network cost overhead.
[0007] In a first aspect, the present disclosure provides a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network, comprising the following steps:
[0008] According to the data collection problem of multi-drone assisted IoT, a multi-drone task allocation model is constructed, and the number of required drones is preliminarily determined in combination with the number of IoT devices in the actual IoT; the drone task allocation model includes objective functions and constraints to minimize the number of drones and the maximum task completion time;
[0009] The improved ant colony algorithm is used to obtain the best service order between IoT devices, which is then used to jointly optimize the task allocation and number of drones, determine the minimum number of drones, and obtain the trajectory profile of the shortest flight path for drones.
[0010] A drone data collection model that combines the flying-and-collecting mode with the flying-hovering-collecting mode is constructed. The joint optimization algorithm of the flight entry point, hovering point, and flight exit point in the IoT device transmission area is used to optimize the trajectory of the drone's shortest flight path, thereby minimizing the maximum mission completion time of the drone.
[0011] A further technical solution, wherein the improved ant colony algorithm is used to obtain the best service sequence between IoT devices, includes:
[0012] Initialize the pheromone concentration between every two preset hovering points of IoT devices according to the service time of the drone to the ground IoT devices and the influence of the service time on the distance between the preset hovering points of the two IoT devices;
[0013] Based on the pheromone concentration between the preset hovering points of every two IoT devices, the optimal service order between IoT devices is obtained through cyclic iteration.
[0014] In a further technical solution, the loop iteration comprises the following steps:
[0015] Step 1.1: If the current drone detects that there is no IoT device in the IoT that meets the signal-to-noise ratio and energy requirements, the initial point is added to the end of the current trajectory to determine the trajectory profile of the current drone, otherwise, step 1.2 is executed;
[0016] Step 1.2, based on the pheromone on the current trajectory profile, calculate the heuristic information and the transfer probability of each IoT device that meets the signal-to-noise ratio and energy conditions;
[0017] Step 1.3: Select the next IoT device according to the roulette method, and then re-execute step 1.1 for loop iteration to finally obtain the trajectory profile of the current drone.
[0018] A further technical solution is to jointly optimize the task allocation and number of drones, determine the number of minimized drones, and obtain the trajectory profile of the shortest flight path of the drones, including the following steps:
[0019] Step 2.1: Each drone searches for the current drone's track in order. If there is no IoT device that has not been visited in the IoT, go to step 2.3; otherwise, go to step 2.2.
[0020] Step 2.2: If there are any ground IoT devices that have not been visited in the IoT, add a new drone to the IoT devices, recalculate the transmission rate of each IoT device, and then go to step 2.1;
[0021] Step 2.3, obtain the current solution based on the set of paths that the drone traverses the hovering points, measure the quality of the current solution based on the fitness evaluation function, select the flight path with the smallest target value as the optimal path for the current iteration, then update the global optimal solution and the value of the pheromone;
[0022] Step 2.4: Re-execute step 2.1 and repeat the loop until the number of iterations is reached.
[0023] A further technical solution is that in the objective function of the drone task allocation model, the maximum task completion time of the drone includes the flight time and collection time of the drone for each IoT device; the collection time includes the drone's flying-in collection time, flying-out collection time, and hovering collection time.
[0024] A further technical solution is to use a joint optimization algorithm of the flight-in point, hovering point and flight-out point in the transmission area of the IoT device based on the drone data collection model, divide the transmission area of the IoT device into overlapping and non-overlapping cases, and optimize the flight-in point, flight-out point and hovering point of the drone respectively, so as to obtain the collection time of the drone for each IoT device and determine the minimum maximum task completion time of the drone.
[0025] A further technical solution is that when the transmission areas of IoT devices do not overlap, the drone can m The data collection time at is expressed as:
[0026]
[0027] Among them, the hovering point is projected onto the two-dimensional plane, and l is the distance from the hovering point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, t hov (l) is the drone at the hovering point OH m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m.
[0028] A further technical solution is that when the transmission areas of IoT devices overlap, the drone can m The data collection time at is expressed as:
[0029]
[0030] Among them, the hover point OH m Projected onto the two-dimensional plane, l is the distance from the hovering point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, Indicates that the drone has moved from the hovering point OH m To the FOP m The collection time, t hov (l) is the drone at the hovering point OH m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m.
[0031] In a second aspect, the present disclosure further provides an electronic device, comprising a memory and a processor, and computer instructions stored in the memory and executed on the processor, wherein when the computer instructions are executed by the processor, the steps of the method described in the first aspect are completed.
[0032] In a third aspect, the present disclosure further provides a computer-readable storage medium for storing computer instructions, which, when executed by a processor, complete the steps of the method described in the first aspect.
[0033] One or more of the above technical solutions have the following beneficial effects:
[0034] 1. The present invention provides a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network. On the basis of considering the UAV data collection mode and data collection time, the method minimizes the number of UAVs and the maximum task completion time by jointly optimizing the task allocation, hovering position and flight trajectory of the UAVs, thereby achieving the goal of shortening the maximum task time of the UAVs while reducing network cost overhead.
[0035] 2. After optimizing and determining the number of drones and their trajectory profiles, the optimization method described in the present invention divides the transmission areas of the Internet of Things devices into overlapping and non-overlapping situations for analysis. Based on minimizing the collection time, the entry point, exit point and hovering point of the drone for each Internet of Things device transmission area are optimized and determined respectively, and then the final flight trajectory of the drone is determined, at which time the maximum task completion time of the drone is minimized. BRIEF DESCRIPTION OF THE DRAWINGS
[0036] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.
[0037] Figure 1 This is a schematic diagram of a drone data collection scenario in Embodiment 1 of the present invention;
[0038] Figure 2 This is a flow chart of the optimization method described in Example 1 of the present invention. DETAILED DESCRIPTION
[0039] It should be noted that the following detailed descriptions are exemplary and are intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meanings as those commonly understood by those skilled in the art to which the present invention belongs.
[0040] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is also intended to include the plural form. In addition, it should be understood that when the terms "comprising" and / or "including" are used in this specification, it indicates the presence of features, steps, operations, devices, components and / or combinations thereof.
[0041] Embodiment 1
[0042] The application of drones in IoT networks, especially data collection, has attracted widespread attention. Due to the limited network cost and the urgency of the task, the number of drones and the task completion time are greatly limited. Existing technical solutions usually only focus on the path planning of drones to shorten the task completion time. However, in different data collection modes, the collection time will also greatly affect the task completion time. Therefore, for the data collection of ground IoT devices in an IoT network supported by multiple drones, this embodiment discloses a method for optimizing the number of drones and task completion time in a multi-drone assisted IoT network, which minimizes the number of drones and the maximum task completion time of drones by jointly optimizing the allocation, hovering position and trajectory of drones.
[0043] This embodiment provides a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network. Figure 2 As shown, the following steps are included:
[0044] According to the data collection problem of multi-drone assisted IoT, a multi-drone task allocation model is constructed, and the number of required drones is preliminarily determined in combination with the number of IoT devices in the actual IoT; the drone task allocation model includes objective functions and constraints to minimize the number of drones and the maximum task completion time;
[0045] The improved ant colony algorithm is used to obtain the best service order between IoT devices, which is then used to jointly optimize the task allocation and number of drones, determine the minimum number of drones, and obtain the trajectory profile of the shortest flight path for drones.
[0046] A drone data collection model that combines the flying-and-collecting mode with the flying-hovering-collecting mode is constructed. The joint optimization algorithm of the flight entry point, hovering point, and flight exit point in the IoT device transmission area is used to optimize the trajectory of the drone's shortest flight path, thereby minimizing the maximum mission completion time of the drone.
[0047] In this embodiment, if Figure 1As shown in the figure, for some scenarios where ground IoT devices are relatively scattered, a multi-UAV task allocation model is first constructed based on the data collection problem of multi-UAV-assisted IoT. In the data collection of ground IoT devices in the IoT network supported by multiple UAVs, it is necessary to consider the number of UAVs and the time it takes for UAVs to complete the task at the same time, to ensure that the time it takes for UAVs to complete the data collection task is minimized while minimizing the number of UAVs. Therefore, a multi-UAV-assisted data collection problem is proposed for the IoT network supported by UAVs. With the minimization of the number of UAVs and the maximum task completion time as the objective function, a multi-UAV task allocation model is constructed. By jointly optimizing the allocation, hovering position and trajectory of UAVs, the number of UAVs and the maximum task completion time of UAVs are minimized. The maximum task time here refers to the maximum task completion time of each UAV performing the data collection task in the process of multi-UAV-assisted IoT data collection. Multiple UAVs are responsible for completing the data collection tasks of all ground IoT devices in the area.
[0048] set up Represents the number of drones, the number of ground IoT devices, and the number of drones. n The objective function of the multi-UAV task allocation model is to minimize the number of UAVs and the maximum task completion time of UAVs, as follows:
[0049]
[0050] In the above formula, Nu represents the number of drones, U n (t) represents the trajectory of UAV n, M n represents the number of ground IoT devices served by drone n, α m,k [n] represents the association and scheduling status of drone n between adjacent ground IoT devices m and k, α m,k [n] = 1 means that drone n flies from hovering point m to hovering point k, is the set of associated states and sequences in the ground IoT devices, α m,0 [n] represents the association and scheduling status of drone n between ground IoT device m and the initial point (i.e., data center), and the objective function P1 represents the minimum number of drones required to collect data from all ground IoT devices; represents the flight time of the drone from IoT device m to IoT device k, represents the flight time of the drone from the last IoT device m back to the data center, represents the data collection time of the UAV at the IoT device m, and the objective function P2 represents minimizing the maximum task completion time of each UAV in the multi-UAV to complete the data collection of the ground IoT device.
[0051] The multi-UAV task allocation model also includes multiple constraints, specifically:
[0052] C1: γ mn (t)≥γ th
[0053]
[0054]
[0055]
[0056] C5:E n ≤E th
[0057]
[0058]
[0059]
[0060]
[0061]
[0062]
[0063] Among them, constraint C1 represents the signal-to-noise ratio γ between drone n and ground IoT device m. mn Equal to the signal-to-noise ratio threshold γ th When the IoT device has the largest transmission radius R, within the transmission range, the signal-to-noise ratio γ mn Greater than or equal to the signal-to-noise ratio threshold γ th , drones can continuously collect data from IoT devices.
[0064] Constraint C2 indicates that the hover point is within the transmission area of the IoT device, where OH m represents the coordinates of the hovering point OH in the transmission area of the IoT device m, R represents the maximum transmission radius of the IoT device m, and h represents the height of the hovering point OH from the ground.
[0065] Constraint C3 indicates that the data collection tasks of M IoT devices are assigned to Nu drones, where Nu represents the number of drones.
[0066] Constraint C4 represents the system bandwidth limitation, where represents the data collection time of the drone at IoT device m, R mn represents the transmission rate between drone n and ground IoT device m, Cm It represents the amount of data that needs to be collected at IoT device m.
[0067] Constraint C5 represents the upper limit of the energy required for the drone to complete the data collection task. The battery energy E carried by each drone is n The upper limit is E th .
[0068] Constraint C6 represents the upper limit of the number of drones that can be served within the drone’s scheduled trajectory.
[0069] Constraint C7 indicates that the mission completion time of all UAVs cannot exceed the given total mission completion time delay, S n Indicates the total task completion delay.
[0070] Constraint C8 represents α m,k [n] and α m,0 [n] is a binary variable.
[0071] Constraint C9 means that each IoT device transmits data to only one drone.
[0072] Constraint C 10 and C 11 It means that the drone is both the starting point and the end point. The drone starts from the data center and returns to the data center. The drone has no data collection time with respect to the data center. m,k [n] represents the flow variables associated between all ground IoT devices accessed by drone n, and M represents the number of ground IoT devices.
[0073] Based on the above multi-UAV task allocation model, the multi-UAV task allocation model is solved by jointly optimizing the allocation, flight entry point, flight exit point, hovering point and flight trajectory of the UAVs.
[0074] Through the above constraints, combined with the number of IoT devices in the actual IoT, and the physical parameters of the drone such as power and its cache capacity, the number of drones required to assist the IoT is preliminarily determined, that is, the number of drones required to collect ground IoT data is preset.
[0075] After the above-mentioned drone task allocation model is constructed, the model is solved. In this embodiment, an improved ant colony algorithm is used to obtain the optimal service order between IoT devices, thereby jointly optimizing the task allocation and number of drones, determining the minimized number of drones, and obtaining the trajectory profile of the shortest flight path of the drone.
[0076] Since the trajectory of the UAV is affected by the number of UAVs, the order of task assignment and the hovering point of the UAV, the number of UAVs and the order of task assignment are optimized by an improved ant colony algorithm to obtain the minimized number of UAVs and the approximate trajectory of the shortest flight path of the UAV, that is, the trajectory profile; then, based on the trajectory profile, the trajectory of the UAV is further optimized by jointly optimizing the hovering point of the UAV, so as to achieve the purpose of minimizing the maximum task completion time of the UAV and the number of UAVs.
[0077] First, initialize the pheromone concentration between every two preset hovering points of the Internet of Things. The initialization of pheromones in the traditional ant colony optimization algorithm is a simple and easy-to-operate uniform distribution, but in the initial stage of the algorithm, the same pheromone between any two hovering points makes the search range of the solution large and the convergence low. Therefore, in this embodiment, the traditional ant colony optimization algorithm is improved by distributing pheromones unevenly between the two hovering points according to the influencing factors of the problem solution to guide the global directional search of the ant colony in the iterative process. It should be clear that the position of the hovering point is preset at this time, and the position of the preset hovering point is directly above the Internet of Things device. The preset hovering point needs to be optimized later to determine the final flight trajectory of the drone.
[0078] The path between hovering point m and hovering point k is represented by d m,k , then the sum of the paths between any two hovering points is expressed as . Define the sub-path ratio Then, set the initial pheromone concentration to:
[0079]
[0080] Among them, T k represents the service time of the UAV to the ground IoT device k, w1 and w2 represent the influence of the service time of the UAV on the ground IoT device and the distance between the two hovering points respectively, w1, w2∈[0, 1] and w1+w2=1.
[0081] After completing the initialization of the hovering point pheromone, the optimal service order between IoT devices is obtained through loop iteration. The specific steps include:
[0082] Step 1-1: If the current drone detects that there is no IoT device in the IoT that meets the signal-to-noise ratio and energy requirements, the initial point (i.e., the data center) is added to the end of the current trajectory to determine the trajectory profile of the current drone. The trajectory profile here refers to the approximate trajectory, that is, to determine the ground IoT device that the current drone flies over; otherwise, execute Step 1-2.
[0083] Step 1-2, based on the pheromone on the current trajectory profile, calculate the heuristic information and the transfer probability of each IoT device that meets the signal-to-noise ratio and energy conditions.
[0084] Heuristic information Φ mk (t+1) reflects the subpath α in the hovering point search process at the (t+1)th iteration m,k The heuristic information is defined using the number of drones Nu, the length of the shortest path between hovering points, and the total energy consumption from the start to the current hovering point, that is, the heuristic information in the (t+1)th iteration is:
[0085]
[0086] in, Used to adjust the heuristic information Φ mk (t+1), dis(m, k) represents the shortest path length between two hovering points, Represents the total energy consumption from the initial point to the next hovering point.
[0087] The transition probability is the use of R mk (t+1) describes the probability of selecting a hovering point k as the next hovering point of the current hovering point m during the hovering point search process of the (t+1)th iteration, where α and β respectively describe the pheromone concentration τ mk (t+1) and heuristic information Φ mk The importance of (t+1), the transition probability is:
[0088]
[0089] Step 1-3, select the next IoT device according to the roulette method, and then re-execute Step 1-1 for loop iteration. Since each ground IoT device corresponds to a hovering point, the transfer probability is determined by the pheromone concentration and heuristic information, and then the path set of the drone traversing the hovering point is obtained, thereby determining the optimal order of the ground IoT devices that the drone flies over.
[0090] Afterwards, since the determination of the flight path of the above-mentioned drone is carried out on the basis of the preliminary determination of the number of drones required to assist the Internet of Things, after obtaining the path set of each drone traversing the hovering point for the preset number of drones through the above scheme, the number of drones and the trajectory profile are optimized, which specifically includes the following steps:
[0091] Step 2-1, each drone will search the trajectory of the current drone in order. If there is no IoT device that has not been visited in the IoT, go to Step 2-3, otherwise go to Step 2-2.
[0092] Step 2-2: If there are any ground IoT devices that have not been visited, add a new drone to the IoT devices, recalculate the transmission rate of each IoT device, and then go to Step 2-1.
[0093] Step2-3, obtain the current solution cur (i.e. the path of the optimal order) based on the set of paths traversed by the drone through the hovering points, and use the current solution as the global optimal solution. After each iteration of Step2-1 and Step2-2, measure the quality of the current solution according to the fitness evaluation function, and select the flight path with the smallest target value as the optimal path of the current iteration, that is, further optimize the optimal order obtained above, and then update the global optimal solution with the optimal path, and at the same time update the value of pheromone, so that the path set obtained in the next iteration is better.
[0094] Specifically, the objective value minimization of the fitness evaluation function is expressed as:
[0095]
[0096] Among them, S represents the set of all drone collection sequences, S n represents the collection sequence of drone n, They represent the nominal values of the number of drones and the task completion time, ω T Reflecting the relative importance of the number of drones and the maximum task completion time of drones, respectively. In the two cases of overlapping and non-overlapping transmission areas, the service time of drone n with respect to ground IoT device m is expressed as:
[0097]
[0098] Among them, a and b represent the weight coefficients of the flight distance and the collection distance respectively, and R is the radius of the transmission area of the ground IoT device.
[0099] The flight path with the smallest target value is selected as the optimal path of the current iteration, and then the global optimal solution is updated with the optimal path, and the value of the pheromone is updated at the same time. Specifically, the first current solution obtained is used as the initial value of the global optimal solution. The rule for updating pheromones is defined as:
[0100] τ mk (t+1)=(1-ρ)τ mk (t)+Δτ mk (t)
[0101]
[0102]
[0103]
[0104]
[0105]
[0106] Among them, E cur (t) represents the total energy consumption from the initial point to the current solution, and the subpath between the hovering point m and the hovering point k is represented by e m,k , S n (UAV mk , v0) represents the set of paths that do not have anyone traveling to and from data center v0, and defines the path sharing degree psd(S n (UAV mk , v0), e m,k ) indicates S n (UAV mk , v0) through e m,k The number of paths, ε mk (t) represents UAV mk The sum of deviations from the target value when traveling to and from data center v0.
[0107] Step 2-4: Until the number of iterations is reached, otherwise update the global pheromone and then re-execute Step 2-1 for loop iteration.
[0108] Through the improved ant colony algorithm mentioned above, the optimal service order between IoT devices is obtained. Through multiple cycles of iteration, the number of minimized drones is determined, and the trajectory profile of the shortest flight path of each drone is obtained.
[0109] Finally, after determining the optimal order for each drone to fly over the corresponding ground IoT devices, it is necessary to solve the problem of minimizing the maximum task completion time of the drone. At this time, it is necessary to consider the task completion time of a single drone. The most important thing to determine the task completion time is to determine the data collection time. Therefore, in order to further optimize the maximum task completion time of the drone, in this embodiment, considering the interaction between the flight service time and collection service time of the drone for each IoT device, a drone data collection model combining the flying-collecting mode and the flying-hovering-collecting mode is constructed. The joint optimization algorithm of the flight entry point, hovering point and flight exit point in the transmission area of the IoT device is adopted to optimize the trajectory of the shortest flight path of the drone, thereby determining the minimum maximum task completion time of the drone.
[0110] In order to minimize the mission completion time of the drone, the fly-in point FIP, fly-out point FOP and hovering point OH of the drone in the transmission area of each IoT device will be jointly optimized. Since the fly-out point of the previous IoT device is also the fly-in point of the next IoT device, the fly-in point and fly-out point of an IoT device are determined by the flight trajectory of the drone between the adjacent IoT devices, that is, the service time of each IoT device is coupled with each other.
[0111] Therefore, the minimized maximum task completion time of drone n can be reformulated as minimizing the sum of the drone’s flight time and collection time for each IoT device:
[0112]
[0113] stC1, C2, C4, C5, C7
[0114] d 1min ≤d m1 ≤d 1max
[0115]
[0116]
[0117] Among them, d m1 Indicates the flight distance from the flight exit point of the previous ground IoT device transmission area to the flight entry point of the next IoT device transmission area, d 1min The shortest distance from the flight-out point of the last ground IoT device transmission area to the flight-in point of the next ground IoT device transmission area, d 1max Indicates the longest distance from the flight-out point of the previous ground IoT device transmission area to the flight-in point of the next ground IoT device transmission area. When the transmission areas of the former and the latter overlap, the flight-out point of the previous ground IoT device transmission area and the flight-in point of the next ground IoT device transmission area may overlap. At this time, d m1 is equal to 0.
[0118] The flight-in point and flight-out point of the transmission area are relatively easy to determine. When the transmission area of the previous ground IoT device overlaps with the transmission area of the next ground IoT device, the flight-out point of the transmission area of the previous ground IoT device is also the flight-in point of the transmission area of the next ground IoT device, and the data center is the first flight-out point and the last flight-in point.
[0119] When the transmission areas of ground IoT devices do not overlap, 1min , d 1max ] into very small subintervals and calculate The value of [d 2min , d 2max ] into very small subintervals and calculate , according to the objective function value of this part, the flying point is obtained.
[0120] In the above objective function, the most important thing is the data collection time of the drone at the IoT device m ,For the collection time of each IoT device by the drone, a combination of flying and collecting and flying-hovering-collecting schemes is used for data collection. For the convenience of analysis and discussion, in this embodiment, the time spent by the drone in the transmission area of the IoT device is classified as data collection time. Drone n flies into the transmission area of the ground IoT device m with a radius of R. When the entry and exit points of the transmission area are determined, optimizing the hovering point of the drone becomes the key to optimizing the data collection time.
[0121] In the above analysis, only the optimal speed v is considered. opt The data collection time during flight, but in fact the drone needs to decelerate and accelerate from flying in to hovering and from hovering to flying out. For simplicity, it is assumed that the acceleration of the drone in the acceleration and deceleration phases are the same, and ε represents the acceleration of the drone from 0 to v opt time, , a is the acceleration of the drone. The data collection time of the IoT device is the sum of the flying-in collection time, the flying-out collection time, and the hovering collection time. According to the collection time problem of the drone for a single IoT device, a drone data collection model combining the flying-collection mode and the flying-hovering-collection mode is constructed:
[0122]
[0123] in,
[0124] In the above formula, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, Indicates that the drone has moved from the hovering point OH m To the FOP m The collection time, t hov (OH m ) is the drone in the hovering position OH m collection time.
[0125] The constraints of this drone data collection model are:
[0126]
[0127] τ hov (OH m )≥0 (2)
[0128]
[0129]
[0130]
[0131] Constraint (1) indicates that OH m Located within the transmission area of the IoT device, and R h (t,OH m ) represent the transfer rates of flying in, flying out, and hovering, respectively.
[0132] Constraint (2) indicates that the collection time of the drone in the flight-hover-collection mode (HM) is greater than or equal to 0. Depending on the amount of data in the IoT device, the drone may not have data collection time in the flight-hover-collection mode (HM). The HM mode is suitable for scenarios with large amounts of data in IoT devices.
[0133] Constraint (3) means that if the amount of data from the UAV is small, data collection can be completed in the flying and collecting mode (FM), and there is no need to consider the acceleration and deceleration of the UAV.
[0134] Constraint (4) limits the amount of data from IoT devices, further simplifying the calculation.
[0135] Constraint (5) indicates that the total energy consumed by UAV n serving ground IoT devices is less than the energy upper limit of each UAV.
[0136] Combined with the drone data collection model, it can be seen that the shorter the distance between the hovering point and the ground IoT device, the higher the transmission rate, and the collection time of the drone in the flight-hovering-collection mode is shortened, but the overall collection time may be extended. Therefore, in order to better optimize the hovering position of the drone, the transmission area of the IoT device is divided into overlapping and non-overlapping cases for discussion.
[0137] Lemma 1: The optimal hovering point when the transmission areas between IoT devices do not overlap (OH m ) is located in FIP m and FOP m The trajectory of the drone in the transmission area is a symmetrical trajectory.
[0138] Lemma 1 is applied to minimize the data collection time of IoT devices. Due to the symmetry of the trajectory, the amount of data collected by the drone during the flight-in and flight-out phases is the same. Therefore, the drone is in GID m The data collection time at can be expressed as:
[0139]
[0140] Among them, the hovering point is projected onto the two-dimensional plane, l(0≤l≤l max ) is from the hover point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, t hov (l) is the drone at the hovering point OH m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m. When l = 0, the hovering point OHm is directly above the ground IoT device m. When l = l max When hovering, click OH m Located in Q m .Q m FIP m and FOP m The midpoint between
[0141] When the transmission areas of IoT devices do not overlap, the collection trajectory of the drone is established based on the Fermat point in the triangle. When the drone is in the transmission area of the IoT device, the collection time of the drone under (flying and collecting) FM is proportional to the trajectory length, and the collection time of the drone under (flying-hovering-collecting) HM is proportional to the transmission distance. According to the properties of the isosceles triangle, the Fermat point of the isosceles triangle is on the perpendicular bisector of the triangle with the three vertices of the flight-in point, the flight-out point and the hovering point. Therefore, when the hovering point OH m When the drone is directly above the IoT device m, the trajectory of the drone in the IoT device transmission area is the longest. m Located in Q m When the drone has the shortest trajectory in the transmission area of the IoT device, the hovering point OH m The optimization problem is transformed into a convex problem that can be solved by convex optimization techniques.
[0142] When the transmission areas of IoT devices overlap, the trajectory of the drone in the outgoing area is asymmetric. Therefore, the drone in GIDm The data collection time at can be expressed as:
[0143]
[0144] In the above formula, the hovering point OH m Projected onto the two-dimensional plane, l is the distance from the hovering point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, Indicates that the drone has moved from the hovering point OH m To the FOP m The collection time, It's the drone at the hovering point. m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m.
[0145] When the transmission areas of IoT devices overlap, the GID m 、FIP m and FOP m Draw an isosceles triangle with three points as vertices, l=||OH m -FIP m ||+||OH m -FOP m ||(l mmin ≤l≤l max ). When l=l min , OH m Located in Q m , Q m Yes FIP m and FOP m The midpoint of max , OH m Located in GID m Directly above. min , l max ] into very small subintervals and calculate The value of the hovering point OH is obtained according to the objective function value of this part. m .
[0146] Given the flight entry point FIP and flight exit point FOP of the drone, the data collection time of a single IoT device is minimized by optimizing the hovering point of the drone, thereby minimizing the data collection time on the entire drone trajectory.
[0147] Through the above scheme, after determining the approximate number of drones and the approximate number of ground IoT devices served by each drone, the number of drones and the maximum mission completion time of each drone are further optimized by jointly optimizing the flight trajectory and hovering position of the drones. Finally, the optimal number of drones and the flight path of the drones (including the ground IoT devices flown over and their flight sequence, as well as the entry point, exit point and hovering point of each ground IoT device) are solved, so as to meet the demand of shortening the maximum mission time of drones while reducing network cost overhead.
[0148] Embodiment 2
[0149] This embodiment provides an electronic device, including a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps in the method for optimizing the number and task completion time of multi-UAV-assisted Internet of Things networks are completed as described above.
[0150] Embodiment 3
[0151] This embodiment also provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps in the method for optimizing the number and task completion time of multi-UAV-assisted Internet of Things networks as described above are completed.
[0152] The steps involved in the above embodiments 2 to 3 correspond to the method embodiment 1. For the specific implementation, please refer to the relevant description part of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.
[0153] Those skilled in the art should understand that the modules or steps of the present invention described above can be implemented by a general-purpose computer device, or alternatively, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. The present invention is not limited to any specific combination of hardware and software.
[0154] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.
[0155] Although the above describes the specific implementation mode of the present invention in conjunction with the accompanying drawings, it is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art on the basis of the technical solution of the present invention without creative work are still within the scope of protection of the present invention.
Claims
1. A method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network, characterized in that: include: According to the data collection problem of multi-drone assisted Internet of Things, a multi-drone task allocation model is constructed, and the number of required drones is preliminarily determined in combination with the number of IoT devices in the actual Internet of Things; the drone task allocation model includes an objective function and constraints to minimize the number of drones and the maximum task completion time; in the objective function of the drone task allocation model, the maximum task completion time of the drone includes the flight time and collection time of the drone for each IoT device; the collection time includes the drone's flying-in collection time, flying-out collection time and hovering collection time; The improved ant colony algorithm is used to obtain the best service order between IoT devices, which is then used to jointly optimize the task allocation and number of drones, determine the minimum number of drones, and obtain the trajectory profile of the shortest flight path for drones. The improved ant colony algorithm is used to obtain the best service sequence between IoT devices, including: Initialize the pheromone concentration between every two preset hovering points of IoT devices according to the service time of the drone to the ground IoT devices and the influence of the service time on the distance between the preset hovering points of the two IoT devices; Based on the pheromone concentration between each two preset hovering points of the initialized IoT devices, the optimal service order between the IoT devices is obtained through loop iteration; the loop iteration includes the following steps: Step 1.1: If the current drone detects that there is no IoT device in the IoT that meets the signal-to-noise ratio and energy requirements, the initial point is added to the end of the current trajectory to determine the trajectory profile of the current drone, otherwise, step 1.2 is executed; Step 1.2, based on the pheromone on the current trajectory profile, calculate the heuristic information and the transfer probability of each IoT device that meets the signal-to-noise ratio and energy conditions; Step 1.3: Select the next IoT device according to the roulette method, and then re-execute step 1.1 for loop iteration to finally obtain the trajectory profile of the current drone; The joint optimization of the task allocation and number of drones, determining the minimized number of drones, and obtaining the trajectory profile of the shortest flight path of the drones includes the following steps: Step 2.1: Each drone searches for the current drone's track in order. If there is no IoT device that has not been visited in the IoT, go to step 2.3; otherwise, go to step 2.
2. Step 2.2: If there are any ground IoT devices that have not been visited in the IoT, add a new drone to the IoT devices, recalculate the transmission rate of each IoT device, and then go to step 2.1; Step 2.3, obtain the current solution based on the set of paths that the drone traverses the hovering points, measure the quality of the current solution based on the fitness evaluation function, select the flight path with the smallest target value as the optimal path for the current iteration, then update the global optimal solution and the value of the pheromone; Step 2.4: Re-execute step 2.1 to iterate until the number of iterations is reached; A drone data collection model that combines the flying-and-collecting mode with the flying-hovering-collecting mode is constructed. The joint optimization algorithm of the flight entry point, hovering point, and flight exit point in the IoT device transmission area is used to optimize the trajectory of the drone's shortest flight path, thereby minimizing the maximum mission completion time of the drone.
2. The method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network as claimed in claim 1, characterized in that: Based on the UAV data collection model, a joint optimization algorithm of the flight-in point, hovering point and flight-out point in the transmission area of the IoT device is adopted. The transmission area of the IoT device is divided into overlapping and non-overlapping cases. The flight-in point, flight-out point and hovering point of the drone are optimized and determined respectively. Then the collection time of the drone for each IoT device is obtained, and the maximum task completion time of the drone is minimized.
3. The method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network as claimed in claim 2, characterized in that: When the transmission areas of IoT devices do not overlap, drones in GID m The data collection time at is expressed as: Among them, the hovering point is projected onto the two-dimensional plane, and l is the distance from the hovering point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, t hov (l) is the drone at the hovering point OH m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m.
4. The method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network as claimed in claim 2, characterized in that: When the transmission areas of IoT devices overlap, drones in GID m The data collection time at is expressed as: Among them, the hover point OH m Projected onto the two-dimensional plane, l is the distance from the hovering point OH m The distance to the ground IoT device m, Indicates that the drone flies from the entry point FIP m To the hover point OH m The collection time, Indicates that the drone has moved from the hovering point OH m To the FOP m The collection time, t hov (l) is the drone at the hovering point OH m The collection time, and R h (t, l) represent the transmission rates of flying in, flying out and hovering, respectively, C m It represents the amount of data that needs to be collected at IoT device m.
5. An electronic device, characterized in that: The invention comprises a memory and a processor, and computer instructions stored in the memory and executed on the processor. When the computer instructions are executed by the processor, the steps of a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network as claimed in any one of claims 1 to 4 are completed.
6. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of a method for optimizing the number and task completion time of a multi-UAV-assisted Internet of Things network as described in any one of claims 1-4.
